Determination of Royalty Rates and Terms for Making and Distributing Phonorecords (Phonorecords III)
| Court | Copyright Royalty Board,Library Of Congress |
| Citation | 88 FR 54406 |
| Published date | 10 August 2023 |
| Record Number | 2023-14925 |
| Section | Rules and Regulations |
Federal Register, Volume 88 Issue 153 (Thursday, August 10, 2023)
[Federal Register Volume 88, Number 153 (Thursday, August 10, 2023)]
[Rules and Regulations]
[Pages 54406-54486]
From the Federal Register Online via the Government Publishing Office [www.gpo.gov]
[FR Doc No: 2023-14925]
[[Page 54405]]
Vol. 88
Thursday,
No. 153
August 10, 2023
Part IILibrary of Congress-----------------------------------------------------------------------Copyright Royalty Board-----------------------------------------------------------------------37 CFR Part 385Determination of Royalty Rates and Terms for Making and Distributing
Phonorecords (Phonorecords III); Final Rule
Federal Register / Vol. 88 , No. 153 / Thursday, August 10, 2023 /
Rules and Regulations
[[Page 54406]]
-----------------------------------------------------------------------
LIBRARY OF CONGRESS
Copyright Royalty Board
37 CFR Part 385
[Docket No. 16-CRB-0003-PR (2018-2022) (Remand)]
Determination of Royalty Rates and Terms for Making and
Distributing Phonorecords (Phonorecords III)
AGENCY: Copyright Royalty Board, Library of Congress.
ACTION: Final rule and order.
-----------------------------------------------------------------------
SUMMARY: The Copyright Royalty Judges announce their final
determination after remand of the rates and terms for making and
distributing phonorecords for the period beginning January 1, 2018, and
ending on December 31, 2022.
DATES:
Effective date: August 10, 2023.
Applicability date: The regulations apply to the license period
beginning January 1, 2018, and ending December 31, 2022.
ADDRESSES: The final determination after remand is posted in eCRB at
https://app.crb.gov/. For access to the docket to read the final
determination after remand and submitted background documents, go to
eCRB and search for docket number 16-CRB-0003-PR (2018-2022) (Remand).
FOR FURTHER INFORMATION CONTACT: Anita Brown, CRB Program Assistant,
(202) 707-7658, [email protected].
SUPPLEMENTARY INFORMATION:
Final Determination After Remand
On October 26, 2020, the United States Court of Appeals for the
D.C. Circuit (D.C. Circuit) issued its mandate vacating and remanding
in part the original Determination \1\ issued by the Copyright Royalty
Judges (Judges) in the captioned proceeding. See Johnson v. Copyright
Royalty Board, 969 F.3d 363 (D.C. Cir. 2020). In its ruling on appeal,
the D.C. Circuit found that in the original Determination, the Judges
(1) failed to give adequate notice to participants of their overhaul of
the royalty rate structure combined with significantly increased and
uncapped rates for section 115 licenses; (2) failed to explain why they
rejected a benchmark based on a past settlement agreement \2\ in lieu
of overhauling of the rate structure and significantly increasing
rates; and (3) failed to identify their legal authority to redefine a
material term after they promulgated a definition of that term in the
original Initial Determination circulated to the participants. See
Johnson, 969 F.3d at 367, 381; Initial Determination, Determination of
Royalty Rates and Terms for Making and Distributing Phonorecords
(Phonorecords III), 16-CRB-0003-PR (2018-2022) (Jan. 27, 2018).
---------------------------------------------------------------------------
\1\ Determination of Royalty Rates and Terms for Making and
Distributing Phonorecords (Phonorecords III), 84 FR 1918 (Feb. 5,
2019) (final rule and order) (original Determination); see also
Final Determination, 16-CRB-0003-PR (2018-2022) (Nov. 5, 2018). The
original Determination was issued by two of the Judges (Majority)
and was accompanied by a dissenting opinion (Dissent) authored by
the third Judge. The Dissent is appended to and part of the same
document as the original Determination.
\2\ The referenced settlement agreement formed the basis for
regulatory terms relating to section 115 musical works royalties and
was adopted as a final rule in Adjustment [or] Determination of
Compulsory License Rates for Mechanical and Digital Phonorecords,
Docket No. 2011-3 CRB Phonorecords II, 78 FR 67938 (Nov. 13, 2013).
See also Technical Amendment at 78 FR 76987 (Dec. 20, 2013).
---------------------------------------------------------------------------
After receipt of the D.C. Circuit's ruling and mandate, the Judges
consulted with the parties to the appeal and established procedures for
the remand proceeding. See Order Adopting Schedule for . . . Remand
(Dec. 23, 2020).\3\ Each side offered opening submissions, responsive
submissions, additional evidentiary filings, and further supplemental
briefing requested by the Judges. The parties' submissions included
legal briefing and incorporated evidence from the original proceeding
as well as evidence newly developed for the remand proceeding. After
preliminary deliberations, the Judges asked for supplemental briefing
from the parties responsive to a proposed alternative rate structure.
See Notice and Sua Sponte Order Directing the Parties to Provide
Additional Materials (Dec. 9, 2021). With respect to redefinition of
the material term Bundled Revenue, the Judges also sought legal
analysis from the parties relating to the D.C. Circuit's directive that
the Judges either provide ``a fuller explanation of the agency's
reasoning at the time . . .'' or take ``new agency action accompanied
by the appropriate procedures.'' See Johnson, 969 F.3d at 392 (citing
Department of Homeland Security v. Regents of the Univ. of Cal., 140 S.
Ct. 1891, 1908). On February 9, 2022, the Judges invited additional
briefing on the Bundled Revenue definition issue, specifically
permitting the parties to offer additional analysis of possible
characterization of the Copyright Owners' motion for clarification
following the Determination as a motion for rehearing under the
Copyright Act, title 17, United States Code at sec. 803(c)(2). See Sua
Sponte Order Regarding Additional Briefing (Feb. 9, 2022).
---------------------------------------------------------------------------
\3\ Following the original remand scheduling order, the Judges
amended the remand proceeding schedule by, e.g., permitting
additional briefing, changing due dates, and seeking additional
input with regard to specific issues. See, e.g., Order . . .
Modifying Scheduling Orders (Dec. 13, 2021) (eCRB no. 225973).
---------------------------------------------------------------------------
At the request of the parties, the Judges agreed to forego live
testimony. On March 8, 2022, all parties were afforded an opportunity
to present oral argument on all remand issues.\4\ On July 1, 2022, the
Judges issued an Initial Ruling and Order after Remand (Initial Ruling)
\5\--applying Johnson and considering the entire record developed pre-
remand and post-remand.
---------------------------------------------------------------------------
\4\ Copyright Owners and Services divided the time for oral
argument. George Johnson dba GEO Music Group waived oral argument.
\5\ The Initial Ruling (eCRB no. 26938) is included in Related
Rulings and Orders as section A. The findings and conclusions in the
Initial Ruling were adopted by a majority of the Judges, but two
Judges filed separate opinions. See Initial Ruling at 2 n.5. One
Judge, former Chief Judge Suzanne Barnett, dissented from the
Majority's conclusion in the Initial Ruling regarding the
Phonorecords II rate structure (section II of the Initial Ruling),
though not from the exception to that benchmark with regard to the
headline rate of 15.1% and the imposition of a cap on the TCC rate
prong. See Dissent in Part re Benchmark (July 1, 2022) (eCRB no.
26943). The other opinion was issued by Judge Strickler, who
dissented from the reasoning relating to the adoption of the
definition of Service Revenue (section V), but concurred in the
adoption of that definition. See Dissent in Part as to Section IV of
the Initial Ruling and Order after Remand . . . (July 1, 2022) (eCRB
no. 26965).
---------------------------------------------------------------------------
In the Initial Ruling, the Judges directed the parties to attempt
to submit jointly agreed-upon regulatory provisions implementing the
Initial Ruling for the Judges to consider. The Judges further ruled
that, if the parties could not agree on all the regulatory language,
they should make separate submissions regarding regulatory provisions
in dispute. See Initial Ruling at 114.
The parties agreed to many regulatory provisions but disagreed as
to several such provisions. Accordingly, they filed separate
submissions and respective replies regarding the regulatory provisions.
Services' Joint Submission of Regulatory Provisions (July 18, 2022);
Copyright Owners' Submission of Regulatory Provisions to Implement the
Initial Ruling (July 18, 2022); Services' Joint Response to Copyright
Owners' Submission of Regulatory Provisions (Aug. 5, 2022); Copyright
Owners' Response to Judges' July 27, 2022 Order Soliciting Responses
Regarding Regulatory Provisions (Aug. 5, 2022).
The Judges considered those submissions and entered an order
addressing the disputed regulatory provisions. See Corrected Order
regarding Regulatory Provisions
[[Page 54407]]
Following Initial Ruling and Order (after Remand) (Nov. 10, 2022)
(November 10th Order).\6\
---------------------------------------------------------------------------
\6\ The November 10th Order corrected an otherwise substantively
identical order issued two days earlier, on November 8, 2022, which
had inadvertently included a small amount of text. See November 10th
Order at 1 (eCRB no. 27312).
---------------------------------------------------------------------------
On November 30, 2022, the parties filed a Joint Submission in which
they provided joint regulatory language no longer in dispute that
applied the binding rulings of the Judges and the D.C. Circuit.\7\
However, the parties identified the single issue in dispute that
relates to the ``Total Content Cost'' (``TCC'') rates for nine
offerings made by interactive streaming services. Joint Submission . .
. Regarding Regulatory Provisions Following Initial Ruling and Order
(after Remand) (Nov. 30, 2022) (Joint Submission) (eCRB no. 27337).
---------------------------------------------------------------------------
\7\ The Judges largely adopt the regulations in the Joint
Submission, which reflect the substance of the Judges' post-remand
rulings, the substance and formatting that the Judges had adopted in
the pre-remand Final Determination that were not raised as issues on
appeal, and updates to references to subparagraphs of Section 115 to
conform to statutory amendments made pursuant to the Music
Modernization Act in 2018. Any differences in language or style are
made for ease of reference, consistent with the parties' post-remand
joint filings.
---------------------------------------------------------------------------
Having considered the parties' submissions (including the Joint
Submission), the Initial Ruling, and all other pertinent material, the
Judges adopted the several TCC rates set forth in the Phonorecords II-
based benchmark as proposed by the Services. See Order 43 on
Phonorecords III Regulatory Provisions (eCRB no. 28210).\8\
---------------------------------------------------------------------------
\8\ The Judges also found good cause to adopt a joint proposal
for modified language regarding late fees, in 37 CFR 385.3. Order 43
on Phonorecords III Regulatory Provisions at 9.
---------------------------------------------------------------------------
Based on the entirety of the record, the Judges adopt in toto \9\
the Initial Ruling and the Order 43 on Phonorecords III Regulatory
Provisions which are set out in this document. Accordingly, those two
documents are adopted by reference in this Final Determination After
Remand. Additionally, the regulatory terms that will codify this Final
Determination After Remand are set out in this document.\10\
---------------------------------------------------------------------------
\9\ But see Judge Strickler's Dissent, cited at n.5 supra, in
which--although he agrees with the Majority as to the definition of
a Service Revenue Bundle--he disagrees as to the legal reasoning
supporting that conclusion.
\10\ The documents are: Initial Ruling and Order After Remand,
designated as Related Rulings and Orders, section A; Order 43 on
Phonorecords III Regulatory Provisions, designated as Related
Rulings and Orders, section B; Dissent in Part as to Section IV of
the Initial Ruling and Order after Remand by Judge David R.
Strickler, designated as Related Rulings and Orders, section C; and
Dissent in Part re Benchmark, designated as Related Rulings and
Orders, section D.
---------------------------------------------------------------------------
On the basis of the foregoing, the Judges propound the rates and
terms described in this Final Determination After Remand for the period
January 1, 2018, through December 31, 2022.\11\ No participant having
filed a timely petition for rehearing, the Judges have made no
substantive alterations to the body of the Initial Determination After
Remand. The Register of Copyrights reviewed the Judges' Final
Determination After Remand for legal error in resolving a material
issue of substantive law under title 17, United States Code, and has
closed her review. Non-substantive typos have been corrected and non-
substantive formatting changes have been made to the version reviewed
by the Register in order to accommodate the Federal Register's
formatting standards. The Librarian shall cause the Judges' Final
Determination After Remand, and any correction thereto by the Register,
to be published in the Federal Register no later than the conclusion of
the Register's 60-day review period.
---------------------------------------------------------------------------
\11\ The regulations applicable to the period 2018 through 2022,
as set forth following this SUPPLEMENTARY INFORMATION section, will
appear in the CFR as appendix A to the current regulations. Although
these Phonorecords III regulations adopt the substance of the
Phonorecords II-based benchmark where the Judges so require, in
Sec. Sec. 385.21 and 385.22, these Phonorecords III regulations are
structured, consistent with the parties' Joint Submission, in the
same consolidated manner as set forth in the pre-remand Phonorecords
III regulations (a structure as to which no party appealed). See
Exhibit A to the Joint Submission at 16, n. 47; see also Exhibit B
to the Joint Submission at n.17 (red-lined version of Exhibit A,
supra).
---------------------------------------------------------------------------
Related Rulings and Orders
A. Initial Ruling and Order After Remand (Redacted Version With Federal
Register Naming and Formatting Conventions)
On October 26, 2020, the United States Court of Appeals for the
D.C. Circuit (D.C. Circuit) issued its mandate vacating and remanding
in part the Determination \12\ issued by the Copyright Royalty Judges
(Judges) in the captioned proceeding. See Johnson v. Copyright Royalty
Board, 969 F.3d 363 (D.C. Cir. 2020). In its ruling on appeal, the D.C.
Circuit found that in the Determination, the Judges (1) failed to give
adequate notice to participants of their overhaul of the royalty rate
structure combined with significantly increased and uncapped rates for
section 115 licenses; (2) failed to explain why they rejected a
benchmark based on a past settlement agreement \13\ in lieu of
overhauling of the rate structure and significantly increasing rates;
and (3) failed to identify their legal authority to redefine a material
term after they promulgated a definition of that term in the Initial
Determination circulated to the participants. See Johnson, 969 F.3d at
367, 381; Initial Determination, Determination of Royalty Rates and
Terms for Making and Distributing Phonorecords (Phonorecords III), 16-
CRB-0003-PR (2018-2022) (Jan. 27, 2018).
---------------------------------------------------------------------------
\12\ Determination of Royalty Rates and Terms for Making and
Distributing Phonorecords (Phonorecords III), 84 FR 1918 (Copyright
Royalty Board Feb. 5, 2019) (final rule and order)
(``Determination''); See also Final Determination, 16-CRB-0003-PR
(2018-2022) (Nov. 5, 2018) (citations to the Determination and to
the Dissent in this Initial Ruling and Order after Remand (Initial
Ruling) are found in this document). The Determination was issued by
two of the Judges (Majority) and was accompanied by a dissenting
opinion (Dissent) authored by the third Judge. The Dissent is
appended to and part of the same document as the Determination.
\13\ The referenced settlement agreement formed the basis for
regulatory terms relating to section 115 musical works royalties and
was adopted as a final rule in Adjustment of Determination of
Compulsory License Rates for Mechanical and Digital Phonorecords,
Docket No. 2011-3 CRB Phonorecords II, 78 FR 67938 (Nov. 13, 2013),
Technical Amendment at 78 FR 76987 (Dec. 20, 2013). In this Initial
Ruling, references to Phonorecords II, PR II, and PR II-based
benchmark are references to this final rule.
---------------------------------------------------------------------------
After receipt of the D.C. Circuit's ruling and mandate, the Judges
consulted with the parties to the appeal and established procedures for
the remand proceeding. See Order Adopting Schedule for . . . Remand
(Dec. 23, 2020).\14\ Each side offered opening submissions, responsive
submissions, additional evidentiary filings and further supplemental
briefing requested by the Judges. The parties' submissions included
legal briefing and incorporated evidence from the original proceeding
as well as evidence newly developed for the remand proceeding. After
preliminary deliberations, the Judges asked for supplemental briefing
from the parties responsive to a proposed alternative rate structure.
See Notice and Sua Sponte Order Directing the Parties to Provide
Additional Materials (Dec. 9 Order). The Judges also sought legal
analysis from the parties relating to the D.C. Circuit's directive that
the Judges either provide ``a fuller explanation of the agency's
reasoning at the time . . .'' or take ``new agency action accompanied
by the appropriate procedures.'' See Johnson, 969 F.3d at 392 (citing
Dep't of Homeland Sec. v. Regents of the Univ. of Cal., 140 S. Ct.
1891, 1908 (Regents)). On February 9, the Judges invited additional
briefing on
[[Page 54408]]
the service bundle definition issue, specifically permitting the
parties to offer additional analysis of possible characterization of
the Copyright Owners' motion for clarification following the
Determination as a motion for rehearing under the Copyright Act, title
17, United States Code (Act) at sec. 803(c)(2).
---------------------------------------------------------------------------
\14\ Following the original remand scheduling order, at the
request of parties or on their own motion, the Judges amended the
remand proceeding schedule by, e.g., permitting additional briefing,
changing due dates, and seeking additional input with regard to
specific issues. See, e.g., Order . . . Modifying Scheduling Orders
(Dec. 13, 2021).
---------------------------------------------------------------------------
At the request of the parties, the Judges agreed to forego live
testimony. On March 8, 2022, all parties were afforded an opportunity
to present oral argument on all remand issues.\15\ Following oral
argument, the Judges deliberated and now issue this Initial Ruling
after Remand.
---------------------------------------------------------------------------
\15\ Copyright Owners and Services divided the time for oral
argument. George Johnson dba GEO Music Group waived oral argument.
---------------------------------------------------------------------------
After due consideration of all of the evidence and oral argument of
counsel, the Judges \16\ determine: \17\
---------------------------------------------------------------------------
\16\ The findings and conclusions in this Initial Ruling are
adopted by a majority of the Judges. One Judge dissents from the
adoption of the entirety of the Phonorecords II rate structure
(section II), though not from the exception to that benchmark with
regard to the headline rate of 15.1% and the imposition of a cap on
the TCC rate prong. One Judge dissents in part from the reasoning
relating to adoption of the definition of Service Revenue (section
V), but not from the adoption of that definition.
\17\ As addressed infra, the Judges also order that the
participants in this remand proceeding prepare and submit regulatory
provisions consistent with this ruling. See Footnote 174.
---------------------------------------------------------------------------
(1) With regard to the applicable rates and rate structure, the
percent-of-revenue all-in headline royalty rate for the mechanical
license shall be set at 15.1%, phased-in, as set forth below:
2018-2022 All-In Headline Royalty Rates
----------------------------------------------------------------------------------------------------------------
2018 2019 2020 2021 2022
----------------------------------------------------------------------------------------------------------------
Percent of Revenue........................ 11.4% 12.3% 13.3% 14.2% 15.1%
----------------------------------------------------------------------------------------------------------------
In all other respects, the rates and rate structure of the
Phonorecords II-based benchmark proposed by the Services (as that
benchmark is defined herein) shall constitute the rates and rate
structure for the Phonorecords III period.\18\
---------------------------------------------------------------------------
\18\ The Services include in their Joint Rate Proposal a chart
summarizing the proposed rates for their offerings. That chart is
attached as an Addendum to this Initial Ruling.
---------------------------------------------------------------------------
To be clear: the 15.1% headline percentage rate substitutes for the
headline percentage rates in subparts B and C of the Services
Phonorecords II-based benchmark, and the definition of ``Service
Revenue'' for bundles shall be the definition contained in 37 CFR
385.11 (paragraph (5) for the ``Service Revenue'' definition) as
proposed in the Services' Phonorecords II-based benchmark.
(2) The Services' Phonorecords II-based benchmark is the better of
the benchmarks proposed by the parties and satisfies the requirements
of 17 U.S.C. 801(b)(1) in all respects. However, as noted supra, to be
consistent with this statutory section and the decision in Johnson, the
royalty rate of 10.5% in that benchmark shall be replaced with the
15.1% rate set forth in paragraph (1) above.
(3) To reiterate for clarity, consistent with the adoption of the
Phonorecords II-based benchmark, and for the reasons more fully
developed herein, the Judges adopt the definition of ``Service Revenue
for Bundled Services'' as it appeared in the Initial Determination in
the underlying proceeding. Following are the Judges' analysis and
ruling after remand.
I. Preliminary Issue: Burden of Proof
As a preliminary matter, the Judges address the issue of burden of
proof raised by both parties. Pursuant to the Administrative Procedure
Act (APA), ``the proponent of a rule or order has the burden of
proof.'' 5 U.S.C. 556(d). See also Initial Remand Submission of
Copyright Owners at 48 (Apr. 1, 2021) (``CO Initial Submission'')
(citing section 556(d) of the APA as setting forth ``a basic rule of
these rate-setting proceedings that a participant is required to
provide evidence establishing the propriety of all aspects of its own
proposed rates and terms, including all aspects of the participant's
proposed rate structure.''). Accordingly, it is clear to the Judges
that the Services should continue to bear the burden of proof regarding
the sufficiency of their proffered Phonorecords II-based benchmark in
this remand proceeding. And, in like fashion, because on remand
Copyright Owners have assumed the mantle of pursuing the vacated rate
structure and rates, they bear the burden of proof with regard to their
proposal.
However, Copyright Owners assert that it is the Services who bear
the burden of proof as to Copyright Owners' proposal regarding the
appropriateness, vel non, of an uncapped TCC rate prong. According to
Copyright Owners, this burden falls on the Services because ``only the
Services . . . proposed TCC prongs at the hearing,'' in the form of the
mix of capped and uncapped TCC prongs contained in the Services'
Phonorecords II benchmark. Id. at 47. The Judges find that the fact
that the Phonorecords II-based benchmark advanced by the Services
contains this mix of capped and uncapped TCC prongs does not bear on
Copyright Owners' duty, under 5 U.S.C. 556(d), to satisfy the burden of
proof with regard to the rates and rate structure they are advancing on
this remand. Moreover, the D.C. Circuit has already held that the fact
that some of the Streaming Services' proposals contemplated continued
use of an uncapped total content cost prong for some categories ``does
not mean they anticipated that the [Judges] would uncap the total
content cost prong across the board . . . [which] is quite different.''
Johnson, 369 F.3d at 382. The difference, according to Johnson, is that
``[u]ncapping the total content cost prong across all categories leaves
the Streaming Services exposed to potentially large hikes in the
mechanical license royalties they must pay.'' Id.
Accordingly, the Judges find that Copyright Owners indeed do bear
the burden of proof with regard to the appropriateness of uncapped rate
structure and rates they are proposing on remand and the Services bear
the burden of proof with regard to the appropriateness of the
Phonorecords II-based benchmark they are continuing to advance on
remand.
II. Rate Structure and Rates
A. Relevant Rulings in Johnson
In establishing a royalty rate structure and the rates within it in
the context of this remand proceeding, the Judges are guided by the
rulings in Johnson.
1. Percent of Revenue Prong
The D.C. Circuit noted that the Judges found the royalties in the
Phonorecords II period were too low and that record companies were
receiving a disproportionate share of the sum of the mechanical and
sound recording royalties. Johnson, 969 F.3d at 384-85. The D.C.
Circuit acknowledged that ``[t]he Judges . . . then carefully
[[Page 54409]]
analyzed the competing testimony and drew from it rates that were
grounded in the record and supported by reasoned analysis.'' Id. at
385. The D.C. Circuit found that the Judges acted well within their
discretion and not arbitrarily, relying on substantial evidence in
establishing the ``zone of reasonableness'' for the rates. Id. As the
D.C. Circuit noted, the Judges' process was ``the type of line-drawing
and reasoned weighing of the evidence [that] falls squarely within the
[Judges'] wheelhouse as an expert administrative agency.'' Id. at 385-
86 (emphasis added).
2. Uncapped TCC Prong
The D.C. Circuit found fault, however, in the Judges' determination
to establish an uncapped and increased percentage-based total content
cost (TCC).\19\ Id. at 380. This approach ``removed the only structural
limitation on how high the [TCC] . . . can climb.'' Id. The D.C.
Circuit reasoned that uncapping the TCC alternative rate prong across
all categories of service exposed the Services to potentially large
hikes in the overall mechanical royalties they must pay. Id. at 382.
The D.C. Circuit noted: ``As the [Judges] acknowledge, sound recording
rightsholders have considerable market power vis-[agrave]-vis
interactive streaming service providers . . . . The interactive
streaming services are . . . exposed to the labels' market power and
record companies could, if they so chose, put those services out of
business entirely . . . . [B]y virtue of their oligopoly power, the
sound recording copyright holders have extracted `inflated' royalties.
. . .'' Id. (cleaned up).
---------------------------------------------------------------------------
\19\ ``TCC'' refers to ``Total Content Cost,'' and is defined as
``a percentage of the royalties paid by the service . . . to sound
recording copyright holders.'' Johnson, 969 F.3d at 370; see also
Determination at 13 n.38 (``TCC'' is an industry acronym for ``Total
Content Cost'', a shorthand reference to the extant regulatory
language describing generally the amount paid by a service to a
record company for the section 114 right to perform digitally a
sound recording.'').
---------------------------------------------------------------------------
While the Services had advocated uncapping the TCC alternative rate
prong for some categories of service, that ``does not mean they
anticipated that the [Judges] would uncap the total content cost prong
across the board. That is quite different.'' Id. at 382. The D.C.
Circuit found that the Judges ``failed to provide adequate notice of
the drastically modified rate structure [they] ultimately adopted.''
Id. at 381. The D.C. Circuit emphasized that the failure to provide
adequate notice of their intentions ``is no mere formality [because]
[i]nterested parties' ability to provide evidence and argument . . .
not only protects the parties' interests, it also helps ensure that the
[Judges'] ultimate decision is well-reasoned and grounded in
substantial evidence.'' Id. at 381-82.
To support their adoption of an uncapped TCC rate prong, the Judges
``predicted that the sound recording copyright owners' royalty rates
would naturally decline in the course of their negotiations with
interactive streaming services.'' Id. at 372. The Judges found
persuasive the rebuttal testimony of one of Copyright Owners' economic
expert witnesses, Professor Watt, that an increase in mechanical
royalties payable by the Services would lead to a corresponding
decrease in the Services' sound recording royalty obligations. See
Determination at 73-74 (``[S]ound recording royalty rates in the
unregulated market will decline in response to an increase in the
compulsory license rate for musical works [and] Professor Watt's
bargaining model predicts that the total of musical works and sound
recordings royalties would stay ``almost the same'' in response to an
increase in the statutory royalty.''). The Services painstakingly
criticized this ``see-saw'' theory.
The D.C. Circuit concluded that, on remand, if and when the Judges
consider the ``uncapped'' rate structure, they shall address all
substantive challenges to that approach raised by the Services,
including the issue of whether ``an increase in mechanical license
royalties would lead to a decrease in sound recording royalties.'' Id.
at 383.
Thus, the D.C. Circuit held, the Judges erred procedurally in
adopting an uncapped TCC alternative rate prong. The D. C. Circuit
therefore instructed the Judges to provide the parties with the
opportunity to fully address the issues regarding the uncapped TCC
prong, and for the Judges to address the ``substantive challenges''
raised by the Services.
3. Four Itemized Statutory Objectives
The statutory standard found in section 801(b)(1) instructs the
Judges to set rates that are not only ``reasonable,'' but also
reflective of four itemized objectives, or factors, which, as the D.C.
Circuit stated, set forth ``competing priorities.'' 17 U.S.C.
801(b)(1)(A)-(D); Johnson, 969 F.3d at 387.\20\ With regard to these
four priorities, the D.C. Circuit found that the Judges properly
analyzed and applied the first objective (Factor A). Id. at 387-88. In
particular, the D.C. Circuit did not disturb the Judges' ruling that an
increase in the royalty rates for mechanical licenses was necessary in
order to satisfy Factor A. Johnson, 369 F.3d at 387-88. According to
Johnson, in making this finding, the Judges had engaged in a
``reasonable reading of the record'' and had relied on ``substantial
evidence.'' Id. at 388. Thus, Factor A (when considered without regard
to the other three objectives) indicated that the statutory rate needed
to be higher than it was during the Phonorecords II period.\21\
---------------------------------------------------------------------------
\20\ These competing objectives are: (A) To maximize the
availability of creative works to the public; (B) To afford the
copyright owner a fair return for his or her creative work and the
copyright user a fair income under existing economic conditions; (C)
To reflect the relative roles of the copyright owner and the
copyright user in the product made available to the public with
respect to relative creative contribution, technological
contribution, capital investment, cost, risk, and contribution to
the opening of new markets for creative expression and media for
their communication; and (D) To minimize any disruptive impact on
the structure of the industries involved and on generally prevailing
industry practices. Id.
\21\ However, as the D.C. Circuit also noted, because the four
section 801(b)(1) objectives reflect ``competing priorities, id'' at
387, the holding that Factor A militates toward a higher rate is not
ultumately dispositive. Rather, it must be weighed with the other
statutory factors.
---------------------------------------------------------------------------
With regard to the other three objectives, Johnson stated that
``[t]he question whether the [Judges] adequately addressed factors B
through D . . . is intertwined with the nature of the rate structure
ultimately imposed by the [Judges].'' Id. at 389. Accordingly, the D.C.
Circuit concluded that it ``need not . . . address whether the [Judges]
adequately considered these remaining factors.'' Id.\22\
---------------------------------------------------------------------------
\22\ The phrase ``intertwined with the nature of the rate
structure'' requires emphasis because the Majority independently
considered how to weigh Factors B and C specifically as to the 15.1%
revenue rate, without regard to the overall rate structure, as
discussed infra.
---------------------------------------------------------------------------
Within the parameters of the holdings in Johnson, the Judges
consider the record facts and the arguments made in this remand
proceeding, together with the pertinent facts and arguments made in the
original proceeding.
B. Rate Evidence for the 33-Months From January 2018 Through September
2020
After the Determination was issued, from its effective inception on
January 1, 2018, through September 30, 2020--a 33-month period--the
parties operated under the rates and rate structure set forth in that
ruling. In light of the D.C. Circuit's decision in Johnson, as of
October 1, 2020, the parties reverted to the Phonorecords II rates. The
Services have asserted in this remand proceeding that, during the 33-
month period when the Majority's new and higher
[[Page 54410]]
Phonorecords III rates were in effect, [REDACTED]. By contrast,
Copyright Owners, on remand, looking at the same data over this 33-
month period, aver that they prove the existence of the seesaw theory.
1. Services' Position
According to the Services, [REDACTED]. [REDACTED]. Moreover,
according to the Services, [REDACTED]. The Services further maintain
that, [REDACTED].
The Services make the [REDACTED]. And, [REDACTED]. Id. ]] 5, 9-13,
16-19, 22-23, 26-27.
The Services claim that [REDACTED]. More particularly, [REDACTED].
[REDACTED]. [REDACTED].
The Services' economic experts rushed to judgment upon learning of
these facts, claiming that they disproved the seesaw theory. See Katz
WDRT ]] 25-27 (relying on testimonies cited supra and concluding that
seesaw theory was disproved, based on [REDACTED]); Marx WDRT ]] 48-51
(relying on same testimonies and likewise finding because [REDACTED]);
Leonard WDRT ] 17. ([REDACTED]).
2. Copyright Owners' Position
Copyright Owners analyzed the royalty data over the same 33-month
period (January 2018 through September 2020) and reach the opposite
conclusion. One of their economic expert witnesses, Dr. Jeffrey
Eisenach, testified that [REDACTED]. Moreover, he opined that
[REDACTED]. See Eisenach RWRT sec. 2(A) & appx. C.
Based on this analysis, Professor Watt declares empirical
vindication of his seesaw theory. Watt RWRT ]] 41-42, 46 (``The
[Judges'] bargaining theory insights about the relationship between
royalty rates were correct. . . . [REDACTED]. . . .'').
3. Analysis and Decision Regarding Evidence of Post-Determination Rates
The Judges are perplexed by the willingness of the expert economic
witnesses on both sides to opine that the rate changes from January
2018 through September 2020 can serve as confirmation of their clients'
respective positions. The issue to be considered empirically was
whether the sound recording rate would decrease in response to the
increase in the mechanical rate. That is, if the record labels had
previously set royalties at a level that would allow the Services
merely to survive, would the record labels agree to lower their sound
recording rate if more of the Services' surplus were acquired by
Copyright Owners? To answer this question, the economists on both sides
applied sophisticated bargaining models and critiques to explain the
nature of the negotiations that would ensue.
In the process, the economists lost track of an obvious, elementary
point: The Phonorecords III rates were being challenged by the
Services' appeal, and might not persist. Indeed, the rates were
ultimately vacated and the parties returned in October 2020 to the
Phonorecords II rates.\23\ Now, the rates will be changed again by this
post-remand Determination, and going forward may be subject to further
potential change, consistent with the provisions of title 17. In light
of such ongoing fundamental uncertainty, why would any economist or
businessman assume that the sound recording companies would agree to
adjust their rates in response to a change in the mechanical rate? The
Judges are amazed that the economic experts neglected even to raise
this uncertainty as a complicating issue, let alone a dispositive
one.\24\
---------------------------------------------------------------------------
\23\ There also was uncertainty as to the effective inception
date of the Phonorecords III rate period, because the Services had
appealed (ultimately unsuccessfully) the CRB Judges' finding that
the period commenced, retroactively, as of January 1, 2018.
\24\ To place this point in the economic context of this
proceeding, the Judges characterize the ongoing ``legal
uncertainty'' as another ``independent variable'' to add to the
economic experts' list of such variables, discussed infra, that
affect the ``dependent variable,'' viz., the sound recording rate.
---------------------------------------------------------------------------
Moreover, no party called as a witness any representatives of the
Majors, or subpoenaed their testimony or documents, to provide the
Judges with evidence of how these record companies perceived the seesaw
issue, whether as a permanent phenomenon or as an uncertain matter,
given the pendency of the legal proceedings regarding the ultimate
mechanical rate. Any of the parties could have requested that the
Judges subpoena a sound recording industry witness to give testimony
and produce documents as to this issue, pursuant to 17 U.S.C.
803(b)(6)(C)(ix), but none did so. Further, Copyright Owners, who are
representing the music publishing interests of inter alios, Sony,
Universal, Warner, and Merlin, likely could have produced such sound
recording witnesses without the need for a subpoena. Witnesses from
these entities who negotiated with the Services after the Phonorecords
III rates and rate structure became effective certainly would have
knowledge relevant to the testimony of the Services' witnesses
[REDACTED] who claimed that [REDACTED].
Simply put, the period from period from January 2018 through
September 2020 was a time the Judges construe as ``33-months of
uncertainty,'' see 3/8/22 Tr. 87, 91 (Closing Argument) when no party
could ascertain with any assuredness the ultimate Phonorecords III
rates and rate structure. Thus, for the economists and the parties to
claim vindication for their arguments by reliance on how the record
labels did or did not respond to the challenged and ever-shifting rates
during this ``33 months of uncertainty'' reflects the elevation of
adversarial zeal over objective judgment.
Accordingly, the Judges place no weight on the purported changes or
stability of the sound recording rates during the Phonorecords III rate
period.
C. Percent-of-Revenue Rate Prong
1. Copyright Owners' Position
In their initial remand submission, Copyright Owners provided no
new evidence to support any aspect of the 15.1% revenue-based rate (or
for that matter, any new evidence to support the rates or rate
structure in the Determination), and elected to rely on the pre-remand
record. In fact, in their initial remand submission, Copyright Owners
do not so much as mention the 15.1% revenue rate derived by the Judges.
However, in their reply remand submission (which the Judges found also
to constitute, in part, a substantive initial submission \25\)
Copyright Owners do address the 15.1% revenue rate. In the reply
submission, Copyright Owners simply stated: ``[T]he Circuit affirmed
the Board's derivation of rate percentages, including raising the
revenue rate to 15.1%.'' Copyright Owners' Reply Brief on Remand (in
Reply Remand Submission of Copyright Owners, Vol. 1) at 64, n.48 (July
2, 2021) (``CO Reply''). In a subsequent submission, Copyright Owners
added that ``[t]he narrow mandate on this Remand does not allow for
reopening the rate percentage determination in the [ ]Determination.''
Copyright Owners' Motion for Reconsideration or Clarification at 15 &
n.10 (Dec. 17, 2021) (emphasis added) (Dec. 17th Motion).
---------------------------------------------------------------------------
\25\ See Order Denying in Part and Granting in Part Services'
Motion to Strike Copyright Owners' Expert Testimony and Granting
Services' Request to File Supplemental Testimony and Briefing at 11
(Oct. 1, 2021) (Oct. 1st Order) (The Judges found that ''with one
exception . . . the challenged testimonial evidence of Copyright
Owners' economic expert witnesses serve the dual purposes of direct
and rebuttal statements'' and, as a consequence, ``provide[d] the
Services an opportunity to file supplemental testimony and briefing
in opposition.
---------------------------------------------------------------------------
Thereafter, Copyright Owners asserted that the D.C. Circuit's
affirmance of the
[[Page 54411]]
[Judges'] revenue percentage rate calculation was ``strong[ ]'' and
``detailed.'' Copyright Owners' Reply in Further Support of Motion for
Reconsideration or Clarification at 4 (January 5, 2022). Moreover,
Copyright Owners took note that the Services had relied on
substantively identical language in Johnson to support their argument
that other statements in that D.C. Circuit decision should be deemed
affirmed. See id. at 4-5 (noting Services' reliance on Johnson's
description of the Judges' rulings regarding student and family
discounts (``grounded in substantial record evidence . . . based on the
weight and credibility of the evidence [and] squarely within the
Judges' expertise)'' as demonstrating that the D.C. Circuit had
affirmed those rulings) (emphasis added); see also Copyright Owners'
Brief in Response to the Additional Materials Orders at 2, 6-7 (Jan.
24, 2022) (``CO Additional Submission'') (again asserting that ``the
15.1% revenue rate . . . was specifically affirmed in detail by
Johnson.'').
2. Services' Position
In their initial submission after the remand, the Services objected
to any continued application by the Judges of the 15.1% revenue rate
because, ``as the Majority acknowledged, this particular division of
revenues will never happen in the real world because of the
complementary oligopoly power of the record labels.'' Services' Joint
Opening Brief (in Services' Joint Written Direct Remand Submission at
Tab D) at 52 (``Services' Initial Submission'') (Apr. 1, 2021). More
particularly in this regard, the Services note that Professor Marx's
Shapley Value Model,\26\ which served as an input for the generation of
the 15.1% revenue rate, also indicated that only [REDACTED]% of the
interactive streaming revenue should be paid out as royalties to the
sound recording rightsholders, with the remaining [REDACTED]% of these
revenues retained by the interactive streaming services. Id. (``Both
Professor Marx's and Professor Watt's models show lower combined
royalties being paid by the services than are currently paid in the
marketplace. . . The discrepancy in total royalties between the models
and the real world is explained, in part, by the absence of supranormal
complementary oligopoly profits in the Shapley model, and the presence
of those profits in the actual market.''). Id. (quoting Phonorecords
III, 84 FR 1952).
---------------------------------------------------------------------------
\26\ Generally, a Shapley Value Model is a game theory analysis.
It models a hypothetical bargain that assigns each ``player'' the
average marginal value it contributes to the bargain and (after
accounting for the costs that each ``player'' would need to recover)
the remaining ``surplus'' is allocated among the players according
to their relative contributions. See Johnson, 969 F.3d at 372. For
the reasons discussed infra, in the present case, the Shapley
surplus from the streaming revenue is split essentially equally by
the owners of the sound recording and musical works owners inter se,
but the royalty rates themselves that would result from their
bargaining would be different as between these two inputs, because
of their differing costs. See, e.g., Gans WDT ] 73.
---------------------------------------------------------------------------
By this approach, the Services maintain, ``the Majority awarded the
Copyright Owners the full 15.1% of revenue dictated by its model
(phased in over time), and left it up to the Services to convince the
complementary oligopolist major labels to dramatically lower sound
recording rates.'' Id. at 54-55. The Services argue that, instead, the
Majority should have applied to Professor Marx's [REDACTED]% total
royalty obligation what they characterize as ``any of the[ ] real-world
ratios in place of the [REDACTED] ratio taken from ``Professor Gans'
``Shapley-inspired'' model. Id. at 54. According to the Services, these
lower ratios would have reduced the revenue percentage rate well below
15.1%. Id.
Alternatively, the Services propose, through Professor Marx's post-
remand written testimony, that the Judges now adopt ``a more balanced,
burden-sharing approach'' to address what she described as the
Majority's ``imbalance'' problem. Id. at 57; see also Marx WDRT ]] 52-
63.\27\ Essentially, her proposal begins with an assumption, based on
record evidence, that labels typically take specific shares of service
revenue, including shares of [REDACTED]%, [REDACTED]% and
[REDACTED]%.\28\ These shares are significantly higher than the
[REDACTED]% that Professor Marx generated from her Shapley model. Next,
Professor Marx's post-remand burden-sharing approach uses as inputs the
15.1% of service revenue and the [REDACTED]% of service revenue that
would be retained by the musical works owners and the Services
respectively.\29\ Putting these two factors together, she sets forth
the basic math: Using her [REDACTED]% sound recording share as an
example, she notes that there is not enough revenue for the labels to
take this [REDACTED]% share, if the musical works owners also receive
15.1% and the Services also retain the [REDACTED]% derived from her
model ([REDACTED]% + 15.1% + [REDACTED]% = [REDACTED]%, an irrational
result). See Services' Joint Opening Brief at 57.
---------------------------------------------------------------------------
\27\ Claiming consistency with the Majority's analysis,
Professor Marx appears to maintain that her ``burden-sharing''
approach generates the statutorily-required ``reasonable'' rate as
well as a rate that satisfies the ``fair return''/``fair income''
objectives of statutory Factor B. See Marx WDRT ] 52 (introducing
her correction of the alleged ``imbalance'' problem by noting that
``the ``right'' mechanical royalty rate is one that is
``reasonable'' and achieves the four objectives laid out in Section
801(b)(1).''
\28\ See Marx WDRT, fig. 7 ([REDACTED]).
\29\ The [REDACTED]% of revenue that the services would retain
is based on one of Professor Marx's ``Shapley Value Models.''
Shapley Value modeling is discussed infra.
---------------------------------------------------------------------------
Professor Marx engages in an analysis based on the following math
and logic (again, using the [REDACTED]% sound recording rate as an
example of the fixed amount taken by the labels): (1) [REDACTED]% of
the streaming revenues remain available to be split between the
services and the musical works copyright owners; (2) adding the 15.1%
revenue rate and her [REDACTED]% revenue retention percentage equals
[REDACTED]%; and (3) the 15.1% revenue rate, as a percent of this
[REDACTED]%, is [REDACTED]%; and (4) [REDACTED]% of the [REDACTED]%
available for splitting between the services and the musical works
copyright owners is [REDACTED]% (rounded). Id. at fig.8.
Thus, she identifies her version of a ``fair'' result: The Services
and Copyright Owners would split the residual revenue remaining after
the labels have exercised their complementary oligopoly power to take
an outsized fixed share--with the split proportional to the 15.1%-to-
[REDACTED]% revenue amounts calculated respectively by the Judges (the
15.1% musical works rate) and Professor Marx (the [REDACTED]% service
revenue retention). Id. 59, table. 8.\30\
---------------------------------------------------------------------------
\30\ Using the same logic and calculation method, Professor Marx
finds that the services would retain [verbar][REDACTED]% /
[REDACTED]%, which equals [verbar][REDACTED]%. Assuming again that
[REDACTED]% of the steaming revenue is available to split (because
the labels have appropriated [REDACTED]%), the services would retain
[REDACTED]% [REDACTED]% rounded) of the streaming revenue. Id.
---------------------------------------------------------------------------
In their final post-remand submission, the Services also flatly
state: ``[T]he D.C. Circuit did not ``affirm'' the 15.1% rate--it
vacated that rate.'' Services' Joint Rebuttal Brief Addressing the
Judges' Working Proposal at 2 (Feb. 24, 2022) (``Services' Additional
Submission''). However, the Services do not support that quoted
statement with any citation to Johnson. See id. Further, the Services
assert that the 15.1% revenue rate is not immune from post-remand
review and reduction because ``the D.C. Circuit withheld judgment ``on
whether that final percentage satisfies factors B through D of Section
801(b)(1). . . .'' Id. at 3.
[[Page 54412]]
3. Analysis and Decision Regarding 15.1% Revenue Rate Prong
The Judges determine that they are clearly bound by the D.C.
Circuit's decision in Johnson to maintain the 15.1% revenue rate, as
phased-in by the Determination. Several reasons support this decision.
First, the Judges conclude that the D.C. Circuit's decision in
Johnson is conclusive and unambiguous regarding the revenue percentage
rate. The D.C. Circuit rejected the Services' assertion that the Judges
acted ``arbitrarily'' as to this particular issue, noting that the
Services had misstated the relevant facts. Johnson, 969 F.3d at 385-86
(responding to Services' misdescription of Judges' analysis and
explaining what Services described as ``not what happened.'').
Moreover, the D.C. Circuit held that with regard to the construction of
the 15.1% revenue rate, the Judges had ``engaged in the type of line-
drawing and reasoned weighing of the evidence [which] falls squarely
within the [Judges'] wheelhouse as an expert administrative agency.''
Id. at 386. The D.C. Circuit further noted that the Judges
``proceed[ed] cautiously'' to set the 15.1% revenue rate by
establishing a ``zone of reasonableness'' for the revenue rate. Id. at
385. Indeed, with regard to each aspect of this revenue rate analysis,
the D.C. Circuit found that the Judges' decision making was ``grounded
in the record and supported by reasoned analysis'' and that
``[s]ubstantial evidence supports [their] judgment.'' Id. at 385.
Second, when the D.C. Circuit reviewed the Determination, it
applied ``the same standards set forth in the Administrative Procedure
Act, 5 U.S.C. 706.'' Id. at 375 (noting that 17 U.S.C. 803(d)(3) cross-
references 5 U.S.C. 706); see also id. (``[W]e will set aside the [ ]
Determination `only if it is arbitrary, capricious, an abuse of
discretion, or otherwise not in accordance with law, or if the facts
relied upon by the agency have no basis in the record.'').
Here, the D.C. Circuit explicitly found that the Judges' analysis
and findings in connection with the 15.1% revenue rate are not
arbitrary and capricious, and that the facts relied upon by the Judges
have a sufficient basis in (are ``grounded in'') the record. It seems
beyond dispute that the D.C. Circuit affirmed the Judges in their
setting of the 15.1% revenue rate as a rate that is reasonable, and
thus satisfies that aspect of the section 801(b)(1) standard.\31\
Indeed, it would border on the Orwellian to misconstrue the D.C.
Circuit's unequivocal and obvious affirmance of the reasonableness of
the 15.1% revenue rate as a vacating of that finding.
---------------------------------------------------------------------------
\31\ The CRB Judges intentionally distinguish between the
``reasonable'' rate standard in the initial body of section
801(b)(1) and the objectives set forth as Factors A-D of section
801(b)(1). A rate can satisfy the statutory ``reasonable rate''
requirement yet require adjustment (higher or lower) to reflect the
balancing of the four additional factors. Accordingly, the Judges
defer to a subsequent section, infra, a discussion of how Factors A-
D should be addressed on this remand.
---------------------------------------------------------------------------
Third, the Judges note that Johnson conspicuously declines to
identify the Judges' setting of the 15.1% percent-of-revenue rate as
one of the findings to be revisited on remand. Rather, Johnson states
that the three overarching issues for resolution on remanded are the
Majority's failure: (1) ``to provide adequate notice of the rate
structure it adopted,'' (2) ``to explain its rejection of a past
settlement agreement as a benchmark for rates going forward; and (3)
``[to] identif[y] the source of its asserted authority to substantively
redefine a material term after publishing its Initial Determination.''
Johnson, 369 F.3d at 367. The Majority's finding that the 15.1% royalty
rate is ``reasonable'' was not identified by the D.C. Circuit as a
finding that was vacated and subject to further review and, indeed, as
noted supra, the appellate panel credited what it characterized as the
Majority's careful analysis and line-drawing in arriving at that
finding.
The clarity of the D.C. Circuit's affirmance of the royalty rate of
15.1% for the percent-of-revenue prong moots the issue of whether
Professor Marx's attempt, described supra, to correct the so-called
``imbalance'' problem has merit. However, the Judges note that, even if
this issue had not been conclusively decided in Johnson, they would
reject her approach as futile. That is, Professor Marx fails to
acknowledge that any surplus that her approach would appear to provide
to the Services would be siphoned off by the Majors, given their
complementary oligopoly power.
More particularly, the sound recording royalty rates she posits
([REDACTED]%, [REDACTED]% and [REDACTED]%) are all functions of the
sound recording companies' understanding of the Services' non-content
costs (costs that the Services must recover out of retained revenues in
order to remain in operation, i.e., to ``survive'') and the then-
existing musical works content (royalty) costs (comprised of the
mechanical rate and the performance rate). If, as Professor Marx
contemplates, the mechanical rate is reduced so that Copyright Owners
``share the burden'' of the complementary oligopoly effect on sound
recording rates, that ``burden sharing'' would increase the revenues
retained by the Services (that is the purpose of Professor Marx's
approach!). But such an increase would raise the Services' revenue
above their ``survival'' rate, as understood by the record labels.
Thus, the record labels, given their complementary oligopoly power,
would increase the Services' royalty rate above what it otherwise would
have been.
Alternately stated, when Professor Marx hypothesizes a given sound
recording royalty rate in column 1 of Figure 8 in her WDRT, that rate
is assumed, by the logic of the complementary oligopoly theory, to have
already allowed the services to cover only their non-content costs and
musical works royalties, as understood by the record labels. So, her
assumed rate in column 1 is not a fixed parameter, but rather an
independent variable, which is a function of, inter alia, the costs
incurred by the services, i.e., their non-content costs plus their
musical works royalty costs.\32\ If those service costs decreased (for
example, in an attempt to reduce the services' burden of bearing the
full brunt of the labels' complementary oligopoly power as in Professor
Marx's attempt to correct the imbalance problem), the percentage in
column 1 of Figure 8 would increase, as the labels siphoned off that
surplus over the services' survival revenue requirements. To find
otherwise would be to refute the logic of the dynamics of the
complementary oligopoly effect.\33\
---------------------------------------------------------------------------
\32\ The interactive services also pay a separate royalty for
the performance license necessary to transmit a song. However, under
the Judges' ``All-In'' royalty structure, that performance royalty
is deducted from the ``All-In'' calculation to determine the
mechanical royalty. Also, the performance royalty paid to the
largest Performing Rights Organization (PROs) are subject to
determination by federal judges in the Southern District of New York
(the so-called ``rate court'').
\33\ To be clear, the Judges are not stating that the Services'
retention of only enough revenue to allow them to cover their
noncontent costs and thus merely ``survive'' is indicia of an
effectively competitive (or even healthy) market--but are merely
acknowledging the state of affairs given the unregulated nature of
the sound recording royalties and the complementary oligopoly power
that exists in that market.
---------------------------------------------------------------------------
Moreover, the defect in Professor Marx's attempt to remedy the so-
called ``imbalance'' problem is a consequence of the statutory
licensing and royalty scheme. To recap, the licensing of content used
by the interactive services is bifurcated. The sound recording
royalties paid by the interactive services to the record labels are not
regulated, and complementary oligopoly power exists in that market,
inflating sound recording royalty rates above an effectively
competitive level. See
[[Page 54413]]
Determination at 73 (``[T]he existence of complementary oligopoly
conditions in the market for sound recordings'' is the basis for ``the
record companies' ability to obtain most of the available surplus''
generated by interactive streaming.) \34\ However (and to state the
obvious), the mechanical rate paid by the interactive services for
musical works is regulated, pursuant to 17 U.S.C. 115 and, until the
2018 enactment of the Music Modernization Act,\35\ according to the
rate standards in 17 U.S.C. 801(b)(1). Thus, there is no statutory or
regulatory impediment to prevent record labels from responding to a
decrease in the mechanical rate by increasing the unregulated sound
recording rate if such an increase is in their economic interest.\36\
---------------------------------------------------------------------------
\34\ As the Judges have consistently noted, this complementary
oligopoly power is generated by the concentration of ownership of
sound recording licenses for ``Must Have'' repertoires among the
three Majors (Sony Music Group, Warner Music Group and Universal
Music Group), plus Merlin (a consortium of Indies sometimes referred
to as ``the fourth Major''), as indicated by their reported
collective 85% share of Spotify's streams in 2018, the first year of
the rate period at issue here. See https://www.midiaresearch.com/blog/smaller-independents-and-artists-direct-grew-fastest-in-2020.
\35\ In subsequent rate periods, the rate remains regulated, but
is subject to a different standard--the ``willing buyer-willing
seller marketplace standard,'' for shorthand) under 17 U.S.C. 115.
\36\ The inverse relationship between changes in the mechanical
royalty rate and changes in the sound recording royalty rate has
been characterized as the ``seesaw'' effect, which is discussed in
further detail infra, with regard to the uncapped TCC rate prong.
---------------------------------------------------------------------------
Accordingly, any attempt by the Judges to reduce the mechanical
royalty rate in order to allow the Services to retain more of the
surplus would fail; it would be like pouring water into a bucket with a
siphon at its base. More water would not remain in the bucket, but
rather would accumulate wherever the siphon leads--in this case, to the
record labels. The Judges could keep mechanical royalty rates depressed
and allow this to occur, but that would harm Copyright Owners while
providing no relief to the Services. And despite the old adage that
``misery loves company,'' the Judges detect no directive under section
801(b)(1) that they harm Copyright Owners without providing a gain for
the interactive streaming services--and that they provide a windfall
for the record labels, to boot.
Although Professor Marx's attempt to reduce the Services'
``misery'' by sharing it with Copyright Owners is unavailing, the
statutory scheme and market forces do appear to combine to mitigate the
burden created by the complementary oligopoly power of the sound
recording companies. If interactive streaming revenue were to grow over
the rate period,\37\ then the phase-in to the 15.1% rate will reflect
fixed annual percentages of a larger base, allowing services to retain
a higher dollar level of the interactive streaming revenues.\38\
[REDACTED]. See, e.g., Diab WDRT ]] 10-11 (Google agreements);
Mirchandani WDRT ]] 16-17 (Amazon agreements); Bonavia WDRT ]] 8; 14-19
(Spotify agreements); White WDRT ]] 6; 8-14; 19; 24; 27-28 (Pandora
agreements). Additionally, the Services' headline sound recording rates
[REDACTED]. Services' Joint Remand Reply Brief at 40 (and record
citations therein). Thus, assuming no increase in non-content costs (or
increases smaller than the increases in streaming revenue), the
Services will realize increased revenue above and beyond what they
needed to survive.
---------------------------------------------------------------------------
\37\ Because this proceeding was appealed and remanded, the
Judges have the benefit of knowing the ``future'' (beyond 2017),
during which U.S. interactive streaming revenues have continued to
grow, a fact that is undisputed, and as to which the Judges take
administrative notice. See, e.g., RIAA 2018 Year-End Music Industry
Revenue Report (available at https://www.riaa.com/wp-content/uploads/2019/02/RIAA-2018-Year-End-Music-Industry-Revenue-Report.pdf; RIAA 2020 Year-End Music Industry Revenue Report
(available at https://www.riaa.com/wp-content/uploads/2021/02/2020-Year-End-Music-Industry-Revenue-Report.pdf (interactive streaming
revenue increased within this rate period from (approximately) $1.6
billion in 2018 to $7.7 billion in 2019 and $8.8 billion in 2020).
\38\ For example, if a royalty is set at a flat rate of 15.1%
when a revenue base is $1,000, then the royalty is $151, leaving
$849 in revenues to cover other costs which, for this example, are
held constant. If the revenue base doubles to $2,000, the same flat
15.1% royalty rate generates $302 in royalties, leaving $1,698 in
revenues to cover other costs which, if constant, allow for the
additional revenue ($1,698-$849 = $849) to generate profits.
---------------------------------------------------------------------------
The Services and Copyright Owners recognize the mitigation of harm
to the Services generated by these facts (although they may well
disagree with the Judges' application of these facts). During colloquy
with counsel for Pandora and Spotify during closing arguments on
remand, the Judges asked why they should in essence apply the ``misery
loves company'' adage:
[JUDGE STRICKLER] [T]he problem is . . . the sound recording
[rates] are unregulated in the interactive market . . . . Congress
did not want that to be controlled at all. So every time I see . . .
the services' argument about how we have [to] set a rate that's fair
even though there's this ability of the sound recording [companies]
to take more, my margin note is always this: ``Are they arguing that
`misery loves company?' '' [W]hy shouldn't that misery be shared
with Copyright Owners? . . . Isn't that really Professor Marx's
argument in her proposed split . . . using the 15.1 percent figure .
. . ?
[COUNSEL] [Regarding] Judge Strickler['s] . . . ``misery loves
company'' issue. . . . I think . . . the way [Judge Strickler] put
it during the trial was, even if I thought rates needed to come
down, how would that help you; wouldn't the labels just take all
that surplus for themselves based on their complementary oligopoly
power? . . . . I want[ ] to address it right off the bat . . . . in
open session.
Relat[ed] to . . . the seesaw . . . our point is that these
label rates are sticky in both directions. If you see an increase in
musical works rates, you do not see a quick decrease in label rates,
and the opposite is true. These rates are sticky.
. . .
There's a lot of friction with respect to the ability of label
rates to change quickly in response to the dynamic marketplace or
the dynamic for business reasons or because of regulatory changes in
musical works rate. These are multi-year contracts. They take a long
time to negotiate. They are complex, et cetera.
So, I do think it's right that at a minimum you can buy time
where the ratio is more aligned with the 801(b) factors. In other
words, you don't have to worry that the labels will take it all
right away, even if you believe they will ultimately take that.
[JUDGE STRICKLER] So you are saying we have something that
reduces misery for a period of time until the misery returns?
[COUNSEL] That's right. And I think that would have been true in
2018 when you were sitting drafting the decision. It's even more
true today in 2022 when the label rates, as I mentioned, are
effectively set, bought and paid for.
3/8/22 Tr. 29-30, 43-46 (Closing Argument) (emphasis added).
Similarly, on this topic, Copyright Owners' counsel accurately
characterized the Judges' adoption of the static 15.1% Shapley-based
rate as the inevitable consequence of ``regulatory lag,'' that requires
a regulator to keep a rate constant over the statutory term because
there is no sufficient data to project future rates. Id. at 273-75; see
generally A. Kahn, 2 The Economics of Regulation at 48 (1971) ``The
regulatory lag [is] the inevitable delay that regulation imposes in the
downward . . . [and] upward adjustments'' to rate levels, and ``thus is
to be regarded not as a deplorable imperfection of regulation but as a
positive advantage [because] companies can for a time keep the higher
profits they reap from a superior performance. . . .'').\39\
---------------------------------------------------------------------------
\39\ The Judges emphasize two points that mitigate any negative
impact on Copyright Owners from the static nature of the 15.1%
revenue rate. First, as a percent-of-revenue rate, it generates more
royalty revenue in a growing market, so the quantum of revenue is
not static. Second, Copyright Owners' own economic expert witness,
Professor Gans, testified that the data in the ``market
observations'' from the Goldman Sachs Report on which he relied were
the result of ``negotiated rates in the free market and thus
``presumed to . . . fully consider[ ] . . . expectations of future
costs and revenues . . . . incorporate[ing] expectations of future
values.'' Gans WRT ]] 37-38. On this issue, it is noteworthy that
both the Majority and the D.C. Circuit credited Professor Gans's
reliance on these projections. See Determination at 70 (``The Judges
. . . find Professor Gans' reliance on financial analysts'
projections for the respective industries to be reasonable.'');
Johnson, 969 F.3d at 386 (holding that ``[t]he CRB Judges' finding
that Gans's . . . reliance on Goldman Sachs' profit projections''
was ``reasonable'' and the] . . . type of line-drawing and reasoned
weighing of the evidence [that] falls squarely within the [Copyright
Royalty Board's] wheelhouse as an expert administrative agency.'')
Thus, dynamic changes going forward in the rate term are
embodied in the 15.1% revenue rate, and dynamic market expectations
are incorporated in the modeling data used to establish that rate.
---------------------------------------------------------------------------
[[Page 54414]]
4. Consideration of Factors A-D in Section 801(b)(1)
Finally, the Judges consider the impact of Factors A-D of section
801(b)(1) in connection with the setting of the revenue percentage rate
of 15.1%.\40\ Regarding Factor A, it cannot be gainsaid that the D.C.
Circuit has left this issue unresolved. Rather, Johnson unambiguously
affirmed the Majority's finding that an increase in the mechanical
royalty rate was warranted. Specifically, Johnson states that the
Majority's decision in this regard met the ``test'' that it be
``supported by substantial evidence [and] reflect a reasonable reading
of the record.'' Johnson, supra, at 388. Moreover, with regard to the
level of the increase, the D.C. Circuit did not disturb the finding by
the Majority that ``[t]he rates determined by the Judges represent a
44% increase over the current headline rate, and thus satisfies the
Factor A objective. . . .'' Determination at 85.\41\
---------------------------------------------------------------------------
\40\ The D.C. Circuit ruled, with regard to the ``nature of the
rate structure,'' that because it had ``vacat[ed] and remand[ed] . .
. for lack of notice'' ``[t]he question whether the [Judges]
adequately addressed factors B through D is bound up with the
[Judges'] analysis of sound recording rightsholders' likely
responses to the new rate structure.'' Johnson, supra, at 389.
However, the 15.1% revenue rate, viewed separately, is not bound up
in the ``rate structure'' issue, which relates to the uncapped TCC
prong and how the 15.1% revenue rate may be ``intertwined'' with
that second rate prong. As explained infra, the Judges are not
adopting an uncapped TCC rate prong, so the 15.1% rate is no longer
``bound up'' with the vacated and remanded ``rate structure'' issue,
making moot the argument that a new post-remand analysis of Factors
B through D is necessary or appropriate. However, on remand,
Copyright Owners have placed in issue the ``disruption'' element of
Factor D, claiming that the Services have not proven that the
uncapped TCC rates and rate prong have or will cause disruption.
\41\ The 44% figure cited by the Majority reflects the
percentage increase of the headline rate, from 10.5% to 15.1%.
---------------------------------------------------------------------------
With regard to Factors B and C,\42\ even if Johnson were construed
as permitting the Judges to revisit this issue, they would not adjust
the 15.1% revenue rate on the basis of these two factors. In this
regard, the Judges note that the Majority found that the 15.1% revenue
rate was not only ``reasonable,'' but also a ``fair allocation of
revenue between copyright owners and services.'' Determination at 87
(emphasis added). The Majority thus found explicitly that ``with regard
to Factors B and C . . . there is no basis to depart from [its]
determination of the reasonable . . . rate structure and rates as set
forth supra.'' Id. More particularly, the Majority calculated the 15.1%
rate by utilizing the total royalty percentage revenue of only
[REDACTED]% as calculated by Spotify's economic expert witness,
Professor Marx, whose economic modeling intentionally reflected a
conception of fairness by reducing the effect of the labels'
complementary oligopoly market power. See Determination at 67-68
(noting that Professor Marx testified that this aspect of her model
``represents a fair outcome in the absence of market power [and] . . .
eliminates . . . market power'' which . . . if left in the economic
analysis would ``render[ ] . . . the analysis incompatible with the
objectives of Factors B and C of section 801(b)(1).)'') (emphasis
added).\43\
---------------------------------------------------------------------------
\42\ Factors B and C are typically considered jointly, because
of the overlap in the objectives of providing a ``fair return'' and
a ``fair income'' to the licensors and licensees respectively (the
Factor B objectives) and reflecting their relative roles in making
the streamed music available to the public (the Factor C
objectives). See Johnson, 969 at 388 (noting without criticism the
joint consideration of Factors B and C; Determination at 85-86
(noting without criticism the several experts' joint consideration
of Factors B and C).
\43\ Additonal facts support the Majority's finding that the
15.1% revenue rate is fair. The record evidence indicates that the
headline percent-of-revenue sound recording rate was between
approximately [REDACTED]% to [REDACTED]% in 2017. See Marx WDRT ]
58, fig 7. When the 15.1% mechanical rate is added to that rate
range, the range of the total royalty obligation (based on headline
rates) is [REDACTED]% to [REDACTED]%. (Plus, given the phase-in of
the rates expressly to avoid disruption, the total royalty
obligation would be even lower before 2022, at current sound
recording rates.) The evdence pre-remand indicated that the Services
were ``surviving'' while incurring noncontent of costs of
approximately [REDACTED]% of revenue, leaving about [REDACTED]% of
revenue available to pay royalties while still remaining in
business. See Eisenach WRT ] 79 (Copyright Owners' expert economic
witness); McCarthy WDT ]] 28-29 (Spotify's Chief Financial Officer.)
Thus, even if the Judges were to engage in a de novo analysis of the
potential applicability of Factors B and C to the 15.1% rate, they
would not find any basis sufficient to warrant a downward rate
adjustment, beyond the phase-in adopted in the Determination.
---------------------------------------------------------------------------
Accordingly, the Judges find it would be substantively unwarranted
to engage in any new consideration on remand of the impact, if any, of
Factors B and C on the otherwise reasonable 15.1% revenue rate.\44\
---------------------------------------------------------------------------
\44\ However, the Judges take note of their further observation,
discussed supra, that the combined impact of ``sticky'' sound
recording royalty rates and the inevitable regulatory lag provide an
additional modicum of fairness with regard to the mechanical royalty
rate.
---------------------------------------------------------------------------
The final itemized statutory factor--Factor (D)--instructs the
Judges to consider the ``competing priority'' of ``minimiz[ing] any
disruptive impact on the structure of the industries involved and on
generally prevailing industry practices.'' 17 U.S.C. 801(b)(1)(D). As
with Factors B and C, even if Johnson were construed to allow the
Judges to revisit this issue on remand with respect to the 15.1%
revenue rate, the Judges would not change the Majority analysis or
findings. In the Determination, the Judges adopted the following
interpretation of this standard set forth in previous determinations:
[T]he Judges reiterated their understanding of Factor D,
concluding that a rate would need adjustment under Factor D if that
rate directly produces an adverse impact that is substantial,
immediate and irreversible in the short-run because there is
insufficient time for either [party] to adequately adapt to the
changed circumstance produced by the rate change and, as a
consequence, such adverse impacts threaten the viability of the
music delivery service currently offered to consumers under this
license.
Determination at 86 (emphasis added).
Also, in order to minimize any economic disturbance to the
Services' businesses, the Majority decided to phase-in the 15.1% rate
over the five-year rate term, setting annual percent-of-revenue rates
as follows: 11.4% in 2018; 12.3% in 2019; 13.3% in 2020; and 14.2% in
2021, before the full 15.1% rate became effective in 2022 the final
year of the rate term. Id. at 87-88.
On remand, the Services have not made any argument that the rate
structure or rates set by the Majority were ``disruptive under this
standard.'' \45\ In sum, there is insufficient basis for the Judges to
change the Majority's application of Factor (D) to the 15.1% revenue
rate finding by the Majority.\46\
---------------------------------------------------------------------------
\45\ The Judges further discuss the Factor D ``disruption issue
infra in connection with their analysis of the uncapped TCC prong.
\46\ Additional facts further support the Majority's finding
that the 15.1% revenue rate is would not be disruptive under Factor
D. The record evidence indicates that the headline percent-of-
revenue sound recording rate was between approximately [REDACTED]%
to [REDACTED]% in 2017. See Marx WDRT ]] 14, 19. When the 15.1%
mechanical rate is added to that rate range, the range of the total
royalty obligation (based on headline rates) is [REDACTED]% to
[REDACTED]%. (Plus, given the phase-in of the rates expressly to
avoid disruption, the total royalty obligation would be even lower
before 2022, at current sound recording rates.) The evidence pre-
remand indicated that the Services were ``surviving'' while
incurring noncontent costs of approximately [REDACTED]% of revenue,
leaving about [REDACTED]% of revenue available to pay royalties
while still remaining in business. See Eisenach WRT ] 79 (Copyright
Owners' expert economic witness); McCarthy WDT ]] 28-29 (Spotify's
Chief Financial Officer). Thus, even if the Judges were to engage on
remand in a de novo analysis of the potential applicability of
Factor D to the 15.1% rate, they would not find any disruption
sufficient to warrant a downward rate adjustment, beyond the phase-
in adopted in the Determination.
---------------------------------------------------------------------------
[[Page 54415]]
5. Conclusion Regarding the 15.1% Revenue Rate
For the forging reasons, the Judges do not disturb the Majority's
finding that the percent-of-revenue rate at 15.1%, phased-in annually
over the rate period, constitutes a ``reasonable'' rate under section
801(b)(1) to be used as the statutory rate for the 2018 to 2022
period.\47\
---------------------------------------------------------------------------
\47\ The Services' assert that the Judges previously found that
the reasonableness of the 15.1% rate was subject to revision on
remand. In support of this position, the Services cite to the
Judges' Order Granting in Part and Denying in Part Copyright Owners'
Motion for Reconsideration or, in the Alternative, Clarification at
3, 4 n.7 (January 6, 2022) (Jan. 6th Order). But the Judges said in
that interlocutory proposal merely that Copyright Owners were
incorrect in their extreme assertion that the Judges could not make
an ``alternative rate and rate structure finding . . . except for
the re-adoption of the vacated rate and rate structure approach in
the Phonorecords III Determination [because] . . . [t]hat . . .
would . . . be inconsistent with Johnson [and] . . . would render
the D.C. Circuit's vacating and remanding of the proceeding without
force or effect.'' Id. at 4, n.7. That did not mean that certain
elements of the D.C. Circuit's ruling could be ignored. Further,
when the Judges provided the parties with the Judges' explicitly
tentative ``Working Proposal,'' they did not declare that the 15.1%
revenue rate calculation could be revisited. Rather, the Judges
``express[ed] a concern, not that the foregoing calculations could
be overridden, but rather that this analysis . . . is `incomplete' .
. .'' Jan. 6th Order at 6 (emphasis added). The parties' submissions
in response to the Judges' ``Working Proposal'' demonstrated that
the 15.1% revenue rate calculation was not ``incomplete'' in the
manner that had raised the Judges' concern. Nothing the Judges said
in this interlocutory and tentative ``Working Proposal'' constituted
a definitive statement regarding the Judges' view of what was and
was not subject to review on remand. See generally merriam-webster.com (defining the adjective ``working'' in this context as
``assumed or adopted to permit or facilitate further work or
activity . . . a working draft.''). Indeed, a primary purpose of the
``Working Proposal'' was to allow the Judges and the parties to
address potential issues and resolutions, without prejudice going
forward.
---------------------------------------------------------------------------
D. Uncapped TCC Rate Prong
1. Two Post-Remand Rationales for Uncapped TCC Rate Prong
The Determination set forth the following two primary reasons for
adopting a ``greater-of' rate structure that also included an uncapped
TCC rate prong:
First, the use of an uncapped TCC metric is the most direct
means of implementing a key finding . . . by the experts for
participants on both sides in this proceeding: the ratio of sound
recording royalties to musical works royalties should be lower than
it is under the current rate structure. Incorporating an uncapped
TCC metric into the rate structure permits the Judges to influence
that ratio directly.
Second, an uncapped TCC rate prong effectively imports into the
rate structure the protections that record companies have negotiated
with services to avoid the diminution of revenue.
Determination at 35-36.\48\
---------------------------------------------------------------------------
\48\ The Majority added two other reasons that are not germane
to this remand. In particular, the Majority stated that, compared to
the Phonorecords II benchmark proposed by the Services, the
``greater-of'' structure with the uncapped TCC rate prong was
``simpler'' to understand than the ``Rube Goldberg-esque'' nature of
the Phonorecords II rate structure. Id. at 36. This issue apparently
was not raised on appeal, as it was not mentioned in Johnson, and
Copyright Owners have not raised the issue on remand. See CO Initial
Submission, supra. (However, the Judges do consider this issue in
their analysis of the PR II-based benchmark, infra.) The final
reason provided by the Majority was that its adoption of an uncapped
TCC rate prong was supported by evidence of Google's agreements with
labels that included an uncapped rate structure, on which Google had
relied to propose, post-hearing, the same greater-of rate structure.
Id. However, the D.C. Circuit found that Google's proposal was
distinguishable, as it was based on a far lower TCC rate (15%) as
well as a far lower percent-of-revenue rate (10.5%). The D.C.
Circuit thus declined to rely on the Google-based approach as
support for the uncapped TCC rate prong. Johnson, 969 F.3d at 383.
---------------------------------------------------------------------------
2. Copyright Owners' Position
Copyright Owners claim that the uncapped TCC prong should be
adopted. They contend that the D.C. Circuit remand was merely
``procedural'' rather than substantive, and the Judges thus are not
precluded from readopting the uncapped TCC prong in this remand
proceeding. CO Initial Submission at 35-38 (and record citations
therein).
They further contend that the uncapped TCC prong was adopted to
provide protection against revenue deferment and displacement
occasioned by the Services choosing to elevate the growth of
subscribers and other listeners over revenue maximization. Id. at 38-43
(and record citations therein). The uncapped TCC prong was first
proposed by Google to persuade the Judges to reject Copyright Owners'
proposed ``greater-of'' rate structure containing a per-play prong and
a per subscriber prong. Id. at 43-46 (and record citations therein).
Copyright Owners argue that the uncapped TCC prong should be
adopted because: (1) the Services have not shown any actual or
threatened ``disruption'' or other harm resulting from the uncapped TCC
prong during the 33-month period; (2) the Services actually experienced
``unprecedented growth and profit'' during this period; and (3) the
Services paid lower percentages of revenues in mechanical and total
royalties when the uncapped TCC prong was in effect. Copyright Owners'
Reply Brief on Remand at 34-48 (and record citations therein).
Relatedly, according to Copyright Owners the Services' argument
that the ``see-saw'' effect is unsupported by empirical evidence has
collapsed, given the evidence relating to market performance. Further
Copyright Owners maintain that this argument is irrelevant to the rate
structure issue. Id. at 48-50 (and record citations therein).
3. Services' Position
The Services argue on remand that the uncapped TCC rate prong must
be rejected. The Services reject the ``seesaw'' theory claiming it is
disproved by the experience of the parties during the 33-month period.
Services' Joint Opening Brief at 48-49; Services' Joint Supplemental
Brief at 7-13 (Nov. 15, 2021) (and record citations therein). The
Services further contend that Copyright Owners have disavowed the
``seesaw'' theory as understood by the Majority. The Services allege
that Copyright Owners now claim that the theory was nothing more than
``a nod'' to certain ``core principles'' of bargaining theory, rather
than a specific prediction of a commensurate inverse relationship
between increases in the mechanical royalty rate and decreases in the
sound recording royalty rate. Services' Joint Supplemental Brief at 2,
5-7 (and record citations therein).
With regard to the uncapped TCC rate prong, the Services assert
that Copyright Owners have not even attempted to demonstrate--nor could
they demonstrate--that the uncapped TCC rate prong is consistent with
all four statutory objectives set forth in section 801(b)(1). Services'
Joint Reply Brief at 1, 3-4, 33-34 36 (July 2, 2021) (``Services'
Reply''); see also Services' Joint Opening Brief at 44-64 (and record
citations therein). The Services claim that ``yoking'' the mechanical
rate to the ``complementary oligopoly rates extracted by the labels is
plainly unreasonable.'' Services' Joint Opening Brief at 44-46. The
Services argue that the existence, vel non, of any ``disruptive
impact'' arising from the uncapped TCC rate prong, is misguided and not
dispositive, because it is only one of the four separately itemized
factors and, as this factor relates to Copyright Owners' proposed
uncapped TCC prong, they bear the burden of proof. Services' Reply at
35-37.
Finally, the Services contend that Copyright Owners have failed to
explain their self-contradictory pre-remand argument that ``an uncapped
TCC prong `does nothing to protect Copyright Owners from the Services'
revenue
[[Page 54416]]
displacement and deferment.' '' Services' Reply at 43.
4. Application of Johnson Findings Regarding Uncapped TCC Rate Prong
The Judges conclude that the D.C. Circuit affirmed the Majority's
derivation and calculation of the 26.1% TCC rate, but vacated and
remanded the Judges' application and inclusion of that rate prong in
the rate structure. The D.C. Circuit noted that, on appeal, the
Services contended that ``it was arbitrary and capricious for the
[Judges] to rely on information drawn from different expert analyses in
calculating the mechanical royalty rates.'' Johnson, 969 F.3d at 384.
Thus, the Services were making the same ``information''-based argument
in opposition to the calculation of both aspects of the mechanical
royalty rates--the revenue percentage prong and the TCC prong. See also
id. (``the Streaming Services separately leveled objections to the
particular percentages adopted by the Copyright Royalty Board to
calculate the revenue and total content cost prongs.'') (emphasis
added)
In fact, both rate prongs were indeed derived from the same
analyses. See Determination at 75 (table) (showing that both 15.1%
revenue rate and 26.2% TCC rate derived from same data--Professor
Marx's model showing total royalties as high as [REDACTED]% [Majority's
lower bound] and Professor Gans's ``Shapley-inspired'' model showing
TCC percent should be [REDACTED]%.) \49\
---------------------------------------------------------------------------
\49\ The reciprocal of Professor Gans's [REDACTED]ratio of sound
recording:musical works royalties is [REDACTED], or [REDACTED]%.
---------------------------------------------------------------------------
It is also clear from Johnson that the D.C. Circuit found that the
Majority had reasonably derived and calculated the 26.2% TCC rate:
When it came to . . . the ratio of sound recording to musical
work royalties that Gans derived from his analysis the [CRB Judges]
specifically found . . . reasonable Gans' equal value assumption
[for dividing the Shapley surplus . . . between sound recording and
musical works owners] and his reliance on Goldman Sachs' profit
projections. That type of line-drawing and reasoned weighing of the
evidence falls squarely within the Board's wheelhouse as an expert
administrative agency.
See Johnson, 969 F.3d at 385-86 (cleaned up) (emphasis added).
Accordingly, because the identical analysis was performed by the Judges
to derive the 26.2% TCC rate as was done to derive the 15.1% revenue
rate, the Majority's finding with regard to the derivation and
calculation of the TCC rate likewise is not subject to further
consideration on remand by the Judges.
However, it is equally clear that the D.C. Circuit vacated and
remanded the Majority's application and inclusion of the 26.2% TCC rate
in a separate ``greater-of'' TCC prong. The defect that generated the
vacating on this issue was procedural-- ``the Streaming Services had no
notice that they needed to defend against and create a record
addressing such a significant, and significantly adverse, overhaul of
the mechanical license royalty scheme . . .'' Id. at 382. The
consequence of the D.C. Circuit's action, however, was substantive. The
D.C. Circuit stated:
This is no mere formality. Interested parties' ability to
provide evidence and argument bearing on the essential components
and contours of the [Judges'] ultimate decision not only protects
the parties' interests, it also helps ensure that the [Judges']
ultimate decision is well-reasoned and grounded in substantial
evidence. . . .
The Streaming Services separately challenge the uncapped rate
structure as arbitrary and capricious. In particular, they argue
that the rate structure formulated by the [Judges] failed to account
for the sound recordings rightsholders' market power. They also
object that the [Judges] failed to provide a `satisfactory
explanation, or root in substantial evidence, [their] conclusion
that an increase in mechanical license royalties would lead to a
decrease in sound recording royalties [the ``inverse relationship''
a/k/a the ``seesaw'' effect].
Id. at 381-83 (cleaned up) (emphasis added). Thus, the D.C. Circuit
explicitly declined to address these substantive issues, because of the
deficient procedure. Instead, the D.C. Circuit remanded these
substantive issues back to the Judges. Id. Simply put, Johnson found
that the absence of notice here could be outcome-determinative. Thus,
the Judges categorically reject Copyright Owners' assertion that the
remand as to the uncapped TCC rate structure was merely ``procedural.''
The Judges do not accept the notion that the Majority simply committed
some ministerial faux pas that could be summarily corrected so that the
uncapped TCC rate structure could be rubber-stamped on remand. Rather,
the Judges' error rendered it impossible for them to consider the pros
and cons of such a rate structure without the necessary input from the
Services (and, for that matter, Copyright Owners as well).
Because the procedural infirmity precluded the D.C. Circuit from
deciding whether the Majority's decision was ``well-reasoned and
grounded in substantial evidence,'' there also can be no substantive
presumption of the appropriateness of the uncapped TCC rate prong, as
suggested by Copyright Owners. To the contrary, the D.C. Circuit's
opinion makes it clear that on remand the Judges must engage in a fresh
consideration of the statutory appropriateness, vel non, of the
uncapped TCC rate prong, by weighing and contextualizing the competing
evidence and testimony entered into the record both before and after
the remand.
Accordingly, although Copyright Owners correctly assert that
Johnson did not find the uncapped TCC rate structure to be ``unfair,
unreasonable or inequitable,'' Johnson just as clearly did not find
that structure to be ``fair, reasonable or equitable.'' Rather, the
purpose of the remand was for the Judges to make these determinations.
Accordingly, the Judges next examine whether setting the statutory
mechanical rate as an uncapped TCC rate is ``reasonable,'' as required
by section 801(b)(1).\50\
---------------------------------------------------------------------------
\50\ The Judges consider infra whether any of the four itemized
statutory factors require an adjustment to this analysis.
---------------------------------------------------------------------------
5. Determining Whether Uncapped TCC Rate Prong is ``Reasonable''
a. Rejection of First Rationale for Including Uncapped TCC Rate
Two substantive issues are implicated raised with regard to the
issue of reasonableness: (1) whether the ``seesaw'' theory is valid;
and (2) if it is valid, whether there exist sufficient data to support
the phased-in 26.2% uncapped TCC rate.\51\ To demonstrate that this
uncapped TCC rate prong and the (phased-in) 26.2% rate are reasonable,
Copyright Owners rely on the combined application of two economic
models--the Shapley Value model and a Nash Bargaining Model.
Accordingly, it is necessary to consider how these two models relate to
each other and how these models and their interrelationship impact the
setting of the statutory rate.
---------------------------------------------------------------------------
\51\ As noted supra, in the Judges' recitation of the parties'
remand arguments regarding the uncapped TCC rate prong, they make
other arguments as well, specifically regarding: (1)) whether it
would be necessary and/or appropriate to adopt this uncapped TCC
rate prong to offset revenue deferral and/or displacement by the
Services; (2) whether this rate prong has caused, or would cause,
economic ``disruption'' to the Services (under Factor D of section
801(b)(1)); (3) whether the uncapped TCC rate prong would satisfy
Factors B and C of section 801(b)(1); and (4) whether this rate
prong improperly imports the complementary oligopoly power of sound
recording licensors. The Judges consider these issues after
addressing the issues relating to the ``seesaw'' theory.
---------------------------------------------------------------------------
The D.C. Circuit described the Shapley Value Model methodology:
The Shapley methodology is a game theory model that seeks to
assign to each market player the average marginal value that the
player contributes to the market. This methodology first determines
the costs that
[[Page 54417]]
each player should recover, then divides the ``surplus'' among the
players in proportion to the value of their contributions to the
worth of the hypothetical bargain that would be struck.
Johnson, 969 F.3d at 372. The Judges provided a consistent but more
detailed definition:
The Shapley value gives each player his average marginal
contribution to the players that precede him, where averages are
taken with respect to all potential orders of the players. The
Shapley value approach models bargaining processes in a free market
by considering all the ways each party to a bargain would add value
by agreeing to the bargain and then assigns to each party their
average contribution to the cooperative bargain. The idea of the
Shapley value is that each party should pay according to its average
contribution to cost or be paid according to its average
contribution to value. It embodies a notion of fairness. The Shapley
model is a game theory model that is ultimately designed to model
the outcome in a hypothetical `fair' market environment. It is
closely aligned to bargaining models, when all bargainers are on an
equal footing in the process.
Determination at 62-63 (cleaned up).
To apply a Shapley Value Model in a rate proceeding, the economic
modeler must obtain usable cost and revenue data to be inputted into
the model. More particularly for this proceeding, the modeler must
identify the parties' input costs, including the Services' non-content
costs, and the revenue derived from interactive streaming.\52\ The
difference between these revenues and the Services' noncontent costs
represents the Shapley ``surplus'' that can be shared among the
Services, the sound recording companies and Copyright Owners.
---------------------------------------------------------------------------
\52\ Identifying useful data is a vexing problem. As one of
Copyright Owners' expert economic witnesses, Professor Watt, has
written: ``[T]he main problem with the Shapley approach . . . a
particularly pressing problem [is] that of data availability.'' R.
Watt, Fair Copyright Remuneration: The Case of Music Radio, 7 Rev.
Econ. Rsch Copyright. Issues at 21, 27 (2010).
---------------------------------------------------------------------------
(i) The Shapley Approach of the Parties' Economic Expert Witnesses
(a) Professor Gans's ``Shapley-Inspired'' Model
Professor Gans, Copyright Owners' expert, utilized royalty and
profit interactive streaming data for record companies and music
publishers that he obtained from ``a [then] recent music industry
equity analysis report,'' namely, a Goldman, Sachs Equity Research
report dated October 4, 2016 entitled ``Music in the Air, Stairway to
Heaven.'' Gans WDT ] 76 & n.39. As the Majority summarized Professor
Gans's approach, ``[h]e found that, for the music publishers to recover
their costs and achieve profits commensurate with those of the record
companies under his approach, the ratio of sound recording royalties to
musical works royalties derived from his Shapley-inspired analysis was
[REDACTED] (which attributes equal profits to both classes of rights
holders and acknowledges the higher costs incurred by record companies
compared to music publishers).'' Determination at 69 (citing Gans WDT ]
77 tbl.3) (emphasis added).
Regarding Professor Gans's Shapley-inspired analysis, the Majority
stated:
[T]he Judges find the ratio of sound recording to musical work
royalties that Professor Gans derived from his analysis to be
informative. Professor Gans computed this ratio based on an
assumption of equal Shapley values between musical works and sound
recording copyright owners. The Judges find this assumption to be
reasonable . . . . [53]
---------------------------------------------------------------------------
\53\ The assumption of equal Shapley values is based on the
understanding that a sound recording license and a musical works
license are both necessary (i.e., perfect complements) in order for
a service to stream a song. Determination at 69 & n.122 therein.
Determination at 70. This is part and parcel of the ``line-drawing''
undertaken by the Majority that the D.C. Circuit affirmed. Thus, on
remand, the Judges do not find cause to reconsider the Majority's
limited adoption of Professor Gans's Shapley-inspired analysis.\54\
---------------------------------------------------------------------------
\54\ Because the ratio of sound recording to musical works
royalties that Professor Gans derived from the data and other
evidence was the only portion of his testimony on which the Majority
relied, and because that reliance was affirmed by the D.C. Circuit,
the criticisms of other aspects of Professor Gans's modeling are no
longer relevant.
---------------------------------------------------------------------------
(b) Professor Marx's Shapley Value Model
Professor Marx constructed two Shapley Value Models, one of which
was relied upon by the Majority. In the model credited by the Majority,
Professor Marx assumed one collective owner of sound recording
copyrights and one collective owner of musical works. She also assumed
the presence of a single interactive service. See Determination at 64-
68. That approach yielded a total royalty obligation for sound
recordings and musical works ranging between [REDACTED]% and
[REDACTED]% of the hypothetical service's revenue. Dissent at 133.
Copyright Owners criticized Professor Marx's decision to assume in
her model only one interactive streaming service, rather than the
multiple services that actually existed. They contend that assumption
reduced the market power of the licensors in her model. According to
Copyright Owners' economic experts, Professor Marx's approach was a
misuse of the Shapley Value Model. They aver that the Shapley Value
approach is intended only to eliminate from the rate derivation the
bargaining ability of a ``Must Have'' input supplier (like the sound
recording companies and Copyright Owners) to ``hold-out'' and thus
squeeze licensees for higher royalties. By modeling every possible
``arrival ordering,'' they contend, the ``hold-out'' problem is
avoided. They further contend that Professor Marx misconstrued the
purpose of the Shapley approach by wrongly modeling market participants
in a manner that significantly reduced the actual market power of these
``Must Have'' input suppliers. Determination at 66-67.
The Majority agreed with Professor Marx. The two Judges in the
Majority found that her modeling reasonably ``attempts to eliminate a
separate factor--market power--that she asserts renders a market-based
Shapley Analysis incompatible with the objectives of Factors B and C of
section 801(b)(1).'' Id. at 68.
Although the Majority ultimately relied upon Professor Marx's
modeling in this regard, the Majority found that her data inputs were
problematic. Determination at 65. Specifically, Professor Marx relied
on 2015 data from Warner/Chappell and Warner Music Group for music
publisher sound recording company noncontent costs, respectively. The
Majority found that 2015 data was less probative than 2016 data and
understated the percentage of revenue to be paid to the two classes of
content providers. However, the Majority ultimately found only that
this one-year older data served to ``understate'' the allocation of
surplus to the upstream content providers, and thus rejected only her
lower [REDACTED]% bound for total royalties, The Majority did decide to
adopt her upper bound of [REDACTED]% value for total royalties, which
could (and ultimately did) ``constitute a lower bound for total
royalties in computing a royalty rate,'' applied by the Majority in
order to make a downward adjustment to offset the complementary
oligopoly effect of ``Must Have'' inputs. Id. at 73, 75.
(c) Professor Watt's Criticisms of and Adjustments to Professor Marx's
Shapley Modeling
Professor Richard Watt was called by Copyright Owners as a rebuttal
witness at the hearing, for the purpose of reviewing Professor Marx's
WDT. Watt WRT ] 3. He concluded that Professor Marx's Shapley Value
Model contains important methodological and data
[[Page 54418]]
flaws which, in his opinion, caused her to significantly understate the
mechanical and overall (musical works + sound recording) royalty rates
to be paid by interactive services pursuant to a proper Shapley
analysis. Id. at ] 5.
Professor Watt also criticized her Shapley Value Model for failing
to incorporate the fact that ``the different interactive streaming
companies--Spotify, Apple Music, Rhapsody/Napster, Google Play Music,
Amazon, etc.--do all compete (and rather fiercely) among themselves,
offering (perhaps perfectly) substitutable services.'' Id. at ] 25.
Even more strongly in this vein, Professor Watt relied on the following
description of the substitutability of the streaming services, inter
se:
Each [interactive streaming] service in the increasingly crowded
field is working frantically to overcome the perception that the
main distinction among the uniformly priced $9.99 a month offering
is little more than font style, quirky playlist title and color
scheme. . . . [M]usic platforms have long fought against the
perception that they're . . . selling a nearly interchangeable
product . . . You're getting sold the same car [with] just got a
different lick of paint on it.'').
Id. at ] 32 n.19.
Professor Watt claimed that incorporating this downstream
competition into the model would reduce the Shapley values of the
Services and increase the Shapley values for the input suppliers, by
recognizing which players provide ``essential inputs'' and which are in
competition with other suppliers of substitutable inputs. Id.
He further criticized Professor Marx for including in her model
``other distributors'' who are not interactive streaming services. Id.
at ] 27. According to Professor Watt, these other distributors ``do not
belong in a properly constructed Shapley Value Model because their
presence would ``show up'' in the model as lower revenues for
interactive services as their subscribers or listeners left for these
other distributors (such as noninteractive services). Id.
Additionally, because he criticized Professor Marx's use of 2015
data (as noted supra), Professor Watt re-worked Professor Marx's model
by examining how the use of 2016 data, as opposed to her 2015 data,
would ``better reflect[ ] . . . the reality of the market. Id. at ] 37;
see also id. at ] 44. When using the (higher) 2016 revenues (and making
some relatively more minor adjustments he found necessary), Professor
Watt estimated that the share of streaming revenues that would be paid
out in total royalties (for musical works + sound recordings) in
Professor Marx's model would range from [REDACTED]% to [REDACTED]%. Id.
at ]] 50-52.\55\
---------------------------------------------------------------------------
\55\ As noted supra, when the Majority weighed and credited
Professor Watt's entire Shapley analysis, in which his estimate of
total royalties was [REDACTED]%, those Judges contextualized
Professor Marx's [REDACTED]% total royalty calculation as the lower
bound of a zone of reasonable rates, and applied it as a measure
that, in their analysis, would offset the complementary oligopoly
effect of real-world royalties. Determination at 75 (text and tbl.).
---------------------------------------------------------------------------
After analyzing these Shapley analyses,\56\ the Majority found that
the mechanical royalty rate needed to be increased in order to provide
Copyright Owners with a reasonable rate as required by section
801(b)(1). As a matter of arithmetic though, if the mechanical rate
increased and the sound recording rate did not decrease by a
corresponding amount, then the total royalties paid by the Services
would increase. That issue brings the Judges to consideration of
Professor Watt's bargaining model, on which the Majority relied to
posit an inverse relationship (the seesaw effect), by which an increase
in the mechanical rate would result in a commensurate reduction in the
sound recording rate.
---------------------------------------------------------------------------
\56\ Because his testimony was made in rebuttal, leaving the
Services no procedural right to file written testimony in
opposition, the Majority gave little weight to Professor Watt's
total royalty projections and no weight to his proffered ratios of
sound recordings-to-musical works royalties. Determination at 75.
---------------------------------------------------------------------------
(ii) Professor Watt's Bargaining Model
Professor Watt's Nash Bargaining Model is the linchpin that
connects: (a) the higher mechanical royalty rates generated by the
Shapley Value results relied upon by the Majority with (b) the assumed
lower sound recording rates--a connection that the Majority found to
render ``reasonable'' and ``fair'' its uncapped TCC prong. See
Determination at 73-74 (``As to the issue of applying a TCC percentage
to a sound recording royalty rate that is artificially high as a result
of musical works rates being held artificially low through regulation,
the Judges rely on Professor Watt's insight (demonstrated by his
bargaining model) that sound recording royalty rates in the unregulated
market will decline in response to an increase in the compulsory
license rate for musical works.''). Alternately stated, Professor
Watt's bargaining model result, i.e., the seesaw effect, if
sufficiently supported in the record, is the phenomenon that would
allow the Judges on remand to apply the Shapley results by increasing
the mechanical rate, without unduly exposing the Services to the risk
of higher total royalties.
More particularly, the Majority recognized a potential problem that
those Judges would have to resolve before utilizing the Shapley Value
approach to create an uncapped TCC prong: ``This is problematic because
the sound recording rate against which the TCC rate would be applied is
inflated . . . both by . . . complementary oligopoly [market]
conditions . . . and the record companies' ability to obtain most of
the available surplus due to the music publishers' absence from the
bargaining table.'' Determination at 73.\57\ But the Majority found
that Professor Watt had provided a rationale which permitted them to
resolve the second problem:
---------------------------------------------------------------------------
\57\ The other problem the Majority needed to resolve was how to
deflate the market-based sound recording royalty rates to mitigate
the complementary oligopoly effect in those rates. Id. As discussed
supra, the Judges resolved this problem by applying the low total
royalty payment sum, [REDACTED]%, from Professor Marx's Shapley
Value Model.
As to the issue of applying a TCC percentage to a sound
recording royalty rate that is artificially high as a result of
musical works rates being held artificially low through regulation,
the Judges rely on Professor Watt's insight . . . that sound
recording royalty rates in the unregulated market will decline in
response to an increase in the compulsory license rate for musical
works. 3/27/17 Tr. 3090 (Watt) (``[T]he reason why the sound
recording rate is so very high is because the statutory rate is very
low. And if you increase the statutory rate, the bargained sound
---------------------------------------------------------------------------
recording rate will go down.'').
Determination at 73-74; see also Watt WRT ] 23 n.13 (``[I]in my
Appendix 3, I show that . . . if the musical works rate is increased to
what would be a realistically fair and reasonable rate, then the
negotiated fee for sound recordings would decrease almost dollar for
dollar . . . .''); see also id. at ] 36 (``The statutory rate for
mechanical royalties . . . is significantly below the predicted fair
rate, and the statutory rate effectively removes the musical works
rightsholders from the bargaining table with the services. Since this
leaves the sound recording rightsholders as the only remaining
essential input, bargaining theory tells us that they will successfully
obtain most of the available surplus.'').\58\
---------------------------------------------------------------------------
\58\ In full detail, Professor Watt concluded: ``[F]or every
dollar that the statutory rate for musical works undercuts a fair
and reasonable rate, the freely negotiated rate for sound recordings
will increase by an estimated [REDACTED] cents. That is, if the
musical works rate is increased to what would be a realistically
fair and reasonable rate, then the negotiated fee for sound
recordings would decrease almost dollar for dollar, with only a
minor change in the total royalty rate for all copyrights
combined.'' Id. at ] 23, n.13; see also id., appx. 3 at 12.
---------------------------------------------------------------------------
[[Page 54419]]
To repeat: This inverse relationship is what has been described as
the ``seesaw'' effect. The question in this regard on remand is whether
the record proves that the seesaw theory is valid and measurable going
forward. Alternately stated, does the record prove that Professor
Watt's bargaining model serves as the linchpin that would allow the
Judges to apply the Shapley results by increasing the mechanical rate,
without unduly exposing the Services to the risk of higher total
royalties?
To resolve this issue, the Judges examine this bargaining model
dispute in detail, as it bears on whether the uncapped TCC rate
structure can be incorporated into the statutory rate.
(a) Bargaining Model Dispute
Professor Watt utilized a general Nash Bargaining Model.\59\ In his
particular application, Professor Watt modeled the streaming services
and the labels each as a ``single unit,'' asserting (as is common in
Shapley analyses) that this single-unit modeling was done ``for
simplicity.'' Watt WRT, appx. 3 at 10. Applying this and other modeling
assumptions, Professor Watt posited: ``If there were to be no
successful deal, then each of these two bargainers [the assumed
``single'' interactive service and ``single'' label] would earn 0,
since in that case the interactive streaming service could not
operate.'' Id.
---------------------------------------------------------------------------
\59\ The Nash Bargaining Model is one type of game-theoretic
approach used by economists to model the distribution of ``gains
from trade'' between two parties ``in a manner that reflects
`fairly' the bargaining strength of the different agents. Marx WDRT
] 28 n.33 (citing A. Mas-Colell, M. Whinston, and J. Green,
Microeconomic Theory 838 (1995)). To understand the parties'
modeling dispute, it is necessary to appreciate the essential
elements of the Nash Bargaining Model, as previously summarized by
the Judges: ``In the Nash Framework [for full quotation, see eCRB
no. 27063 n.48].'' SDARS III Final Determination, 83 FR 65210, 65215
& n.32 therein (Dec. 19, 2018).
---------------------------------------------------------------------------
In his oral testimony at the hearing, Professor Watt did not opine
as to whether changes in variables other than musical works royalties
would also have an impact on the level of sound recording royalty
rates, even as higher musical works rates would otherwise place
virtually 1:1 downward pressure on the sound recording rate. However,
in his written rebuttal hearing testimony, i.e., his WRT, Professor
Watt did make varying assumptions regarding the changes in the
Services' non-content costs, by which he did change the total revenue
share for content providers. Watt WRT ]] 50-52. He concluded from this
varying replication of Professor Marx's Shapley model ``that the
results that it delivers are very dependent upon the amount of total
interactive streaming revenue and the fraction of that revenue that is
taken up by downstream non-content costs.'' Id. at ] 53 (emphasis
added).\60\
---------------------------------------------------------------------------
\60\ The Judges take note here of Professor Watt's presentment
of alternative scenarios, because, as discussed infra, the Services
and their economists accuse Professor Watt of changing his
testimony, post-remand, by limiting the scenarios in which his
``seesaw'' argument would apply in order to salvage the credibility
of his bargaining model.
---------------------------------------------------------------------------
The Services had no procedural right under part 351 of the Judges'
regulations to proffer surrebuttal written testimony from economic
witnesses to challenge Professor Watt's assertion, made for the first
time in rebuttal, of the seesaw relationship between changes in the
musical works royalty rate and the sound recording royalty rate paid by
interactive services. Moreover, the Services and their economists also
had no opportunity to weigh in on the Majority's application of same
(which was not revealed until the Judges rendered their decision). See
Johnson, 969 F.3d at 381 (``Streaming Services had no notice that they
needed to defend against and create a record addressing such a
significant, and significantly adverse, overhaul of the mechanical
license royalty scheme.'').\61\ Now though, on this remand, the
Services have been afforded the opportunity to present these
criticisms, through their expert witnesses.
---------------------------------------------------------------------------
\61\ The Services could have sought leave to file surrebuttal
testimony, and could have challenged the Majority's understanding of
Professor Watt's testimony, after the Initial Determination, by
filing a Motion for Rehearing pursuant to 37 CFR 353.1. However, a
party is not required to engage in either of these procedural
approaches, but rather may challenge the Determination on appeal, as
has occurred here.
---------------------------------------------------------------------------
(b) Professor Katz's Principal Criticism
Pandora's economic expert, Professor Michael Katz, levied several
criticisms of the bargaining model proffered by Professor Watt and
applied by the Majority. The most important problem with Professor
Watt's analysis, according to Professor Katz, is that the former's
model assumes an ``extremely unrealistic'' zero payoff to the label in
the absence of an agreement with a streaming service--an assumption
which is ``far from . . . innocuous.'' Written Direct Remand Testimony
of Professor Michael Katz (Katz WDRT) ]] 16, 20.
Professor Katz opines that this zero payoff assumption is
equivalent to assuming, contrary to undisputed market facts, that: (1)
subscribers and listeners to an interactive service would not switch to
other interactive services if that service failed to reach an agreement
with the labels; and (2) the interactive service is a ``Must-Have''
input supplier. Katz WDRT ]] 17-18. In terms of Nash modeling,
according to Professor Katz, Professor Watt's assumption is thus
equivalent to ``assum[ing] that the sound recording copyright owners
have no outside option.'' Katz WDRT ] 127 (app. A) (emphasis added).
Moreover, not only does Professor Katz assert the indisputability
that such substitution would occur, he points out that Professor Watt
himself acknowledged in his own testimony that such substitution would
occur. Katz WDRT ] 19.\62\
---------------------------------------------------------------------------
\62\ The Judges have quoted Professor Watt's testimony in this
regard supra.
---------------------------------------------------------------------------
Beyond this purported inconsistency, Professor Katz finds Professor
Watt's no-substitution assumption to be a serious modeling error
because, in order to quantify accurately each Nash bargainer's
contribution to the net surplus to be divided, the extent of
substitutability on each side of the market must be captured by the
modeling. Katz WDRT ] 20. That is, he opines that ``Professor Watt's
assumption that there is no substitution dramatically biases his model
toward finding a large seesaw effect and renders his analysis
unreliable . . . lead[ing]to a prediction that the share of an increase
in musical works royalties that will fall on the streaming services is
approximately eight times larger than Professor Watt's prediction. Id.
at ] 21.
As a matter of music business dynamics, Professor Katz interprets
Professor Watt's substitutability error as follows.
The assumption that a label receives a zero payoff if it does
not reach agreement with a streaming service is equivalent to
assuming that, if a streaming service shut down, none of the
consumers who would otherwise have used that streaming service will
switch to alternative streaming services or other sources of
licensed music. The two forms of the assumption are equivalent
because, when the services are substitutes, failure to reach an
agreement with one service will not drive a label's payoffs from
interactive streaming to zero. It will not result in the loss of all
of the benefits that could be enjoyed by reaching an agreement.
Instead, many consumers would engage in substitution and choose
other streaming services, which will allow the label to earn profits
from the additional royalties that would be paid to it by those
other services.
Id. at ] 18.
Professor Katz attempts to adjust Professor Watt's Nash Bargaining
Model to account for this substitution effect. In his Appendix A,
Professor Katz--acknowledging the reality of multiple interactive
services--changes Professor Watt's assumed single label's payoff
[[Page 54420]]
(designated as parameter ``A'' in the Nash Bargaining Model) from a
value of zero to a value equal to ``the share of revenues that would be
diverted to other streaming services'' multiplied by ``the royalty rate
that the label receives from the other interactive streaming
services.'' Id. ]] 119, 127. Professor Katz asserts that the diversion
to other streaming services represents an ``outside option'' available
to a label. Id. ] 127. Professor Katz incorporates this ``outside
option'' in his revised version of Professor Watt's Nash Bargaining
Model.
In addition, Professor Katz asserts that Professor Watt's modeling
is unreliable because ``his prediction of the size of the see-saw
effect is very sensitive to the assumed values of various other
parameters.'' Id. at ] 23. For example, Professor Katz asserts that a
change in the royalty rate paid to the labels could materially affect
the balance or even the existence of the seesaw effect. Id. at ] 127.
As further support for his opinion, Professor Katz relies on the
testimony of one of Copyright Owners' own economic expert witnesses,
who gave testimony clearly indicating that the ``seesaw'' effect was
not at all likely to occur. Id. ] 24, n.16 (citing Gans WRT ] 32).\63\
---------------------------------------------------------------------------
\63\ In this regard, Professor Gans testified: ``[When
considering] the general distribution of profit when royalty rates
for musical works rightsholders are increased[,] [i]n principle,
those funds could come from a decrease in service profit, a decrease
in sound recording royalties, or an increase in consumer pricing . .
. . The general redistribution of profit in response to increased
musical works royalties is fundamentally an empirical question. . .
.'' Gans WRT ] 32.
---------------------------------------------------------------------------
In sum, Professor Katz finds Professor Watt's Nash Bargaining Model
to be unusable as a foundation to set royalty rates because, although
``there are theoretical reasons to believe that a see-saw effect may
occur, . . . there are complications and it is difficult to predict how
big the effect will be.'' Id. ] 24 (emphasis added).
(c) Professor Watt's Rebuttal to Professor Katz
In rebuttal to Professor Katz's criticisms, Professor Watt states
that ``the record needs to be straight on Nash bargaining theory,'' in
order to explain ``the foundational error'' committed by Professor
Katz. Watt RWRT ] 52. This basic mistake, according to Professor Watt,
is Professor Katz's erroneous assertion that the bargaining model must
account for a label's ``outside option.'' Id. ] 53. Relying on economic
authority regarding bargaining theory, Professor Watt defines an
``outside option'' as ``the best alternative that a player can command
if he withdraws unilaterally from the bargaining process.'' Id. ] 59
(emphasis added); see also id. ] 53 (``An outside option is a payoff
that the label would receive if negotiations with the service do not
result in an agreement.'') (emphasis added).\64\
---------------------------------------------------------------------------
\64\ The phrase ``outside option'' suggests the existence of an
``inside option.'' Indeed, a treatise cited by Professor Watt
identifies the ``inside option,'' defining it as ``[t]he payoff the
[bargainer] obtains while the parties temporarily disagree''--
contrasting it with the ``outside option'' as (consistent with
Professor Watt's testimony) ``the payoff [the bargainer] obtains if
she chooses to permanently stop bargaining, and chooses not to reach
an agreement with [the counterparty].'' A. Muthoo, Bargaining Theory
with Applications at 137 (1999).
---------------------------------------------------------------------------
Connecting this principle of bargaining theory to economic theory,
Professor Watt explains his understanding of the relationship of the
``outside option'' to the more familiar economic concept of
``opportunity cost'':
An outside option could also be referred to as an ``opportunity
cost,'' since it is the value of what would be foregone should a
deal with the service actually be struck. It is . . . useful to
recognize the equivalence between an outside option and an
opportunity cost, because economics in general has a very long
history of understanding how opportunity costs weigh in on economic
decision making.
Id.
Professor Watt then opines how Professor Katz confused the
``outside option'' with the disagreement (a/k/a threat) point in the
Nash Bargaining Model:
[Professor] Katz claim[s] that the outside option value that the
labels would enjoy should they not reach an agreement with the
services should be included as part of the ``disagreement point''
within the bargaining model and reimbursed like a cost prior to
bargaining. Doing this can dramatically alter the results of the
model. It is also definitively not how such an option should be
modelled. [Professor] Katz [is] guilty of misunderstanding the Nash
bargaining model, and concretely, the meaning of a ``disagreement
point,'' and the way that an outside option should be brought into
the model.
Id. ] 55.
More particularly, according to Professor Watt, these outside
options/opportunity costs do not belong in a Nash Bargaining Model,
because they are ``not the types of status quo actual financial
payments that may be modelled as disagreement points.'' Id. ] 57.
Rather, he asserts that, as Professor Katz essentially acknowledged,
they are ``payoffs from substitution, [i.e.,] an option instead of the
deal, and they are not actual financial payments, but opportunity
costs. Id.
Professor Watt then explains that an outside option/opportunity
that by definition exists as an alternative to a bargain between two
parties lies outside the two parties' bargain, and is thus out-of-place
within a proper Nash Bargaining Model:
In the case at hand, if the parties never stop negotiating and
never take up substitute options, then no joint enterprise is
offered and there is no surplus to share, so each necessarily gets a
payoff equal to 0, just as I assumed in my model.
. . .
[A]gainst this backdrop, an outside option (a potential payoff
that is not directly related to a share of the surplus that is being
negotiated) . . . comes in [to the model] as a constraint upon the
set of feasible deals that could be struck, exactly as an
opportunity cost would be treated.
Id. ]] 57-58.
(d) Dr. Leonard's Criticisms of Professor Watt's Bargaining Model
According to Google's economic expert witness, Dr. Gregory Leonard,
the Majority wrongly relied on Professor Watt's bargaining model
because it is ``highly stylized'' and theoretically ``simplified'' in
ways that make it unable to predict that ``an increase in the musical
works royalty would be offset nearly dollar-for-dollar by a decrease in
the sound recording royalties (the ``seesaw effect''), thus leaving the
services virtually unaffected by the proposed increase in musical works
royalties.'' Leonard WDRT 8.
Pointedly, Dr. Leonard criticizes Professor Watt's bargaining model
as comprised of a ``veneer of `complexity' . . . mathematical formulas
and [a] reference to John Nash,'' adopted to provide a rationalization
for adoption of his Shapley Value modeling that would significantly
increase the mechanical royalty rate.'' Id. ] 16. These modeling
deficiencies, Dr. Leonard asserts, are not merely ``simplifying
assumptions [that] better focus on the specific question the model is
meant to address,'' but rather ``simplify away economic characteristics
. . . entirely abstract[ing] away economic characteristics . . .
central to the question at hand.'' Id. ] 18.
In particular, Dr. Leonard avers that Professor Watt's bargaining
model materially abstracts away from, inter alia: (1) the nature of
consumer demand for streaming services and competing forms of music;
(2) how services decide to enter or exit the streaming market; (3) the
nature of the oligopolistic interaction among the labels; (4) the
nature and timing of the bargaining between each label and each
service; (5)
[[Page 54421]]
the potential for ``hold-up'' \65\ by labels that perceive the services
to be in a vulnerable bargaining position due to their previous
industry-specific investments made under their assumption that the pre-
existing statutory structure would be maintained; and (6) the failure
of Professor Watt's bargaining model to grapple with the complementary
oligopoly structure of the sound recording market. Id. ]] 18, 20.
---------------------------------------------------------------------------
\65\ A hold-up problem occurs when: (1) parties to a future
transaction must make specific investments prior to the transaction
in order to prepare for it; and (2) the exact form of the optimal
transaction (e.g., how many units if any, what quality level, the
time of delivery) cannot be specified with certainty ex ante. W.
Rogerson, Contractual Solutions to the Hold-Up Problem, Rev. Econ.
Stud. 777 (1992). Here, the interactive services may need to commit
to paying for long-term investments, even though they cannot know
the level of their largest costs (content royalties) beyond a single
rate term.
---------------------------------------------------------------------------
These factors, he posited, are ``important for determining how
sound recording royalties would actually change in response to a change
in the statutory musical works royalty.'' Id. Professor Leonard
concludes that, by not modeling these factors, Professor Watt's
``prediction of a virtual dollar for dollar decrease in sound recording
royalties is unreliable as a basis for formulating policy.'' Id. ] 20.
Regarding the complementary oligopoly structure of the market and
its impact on the bargaining process, Professor Leonard emphasizes that
an important ``real-world hurdle'' assumed away by Professor Watt's
modeling of a single label entity is that ``each label would prefer to
have the other labels lower their sound recording royalties while
maintaining its own royalties at pre-existing levels . . . .'' Id. ]
21. More particularly, Dr. Leonard explains that ``even if a label were
to recognize that it is more efficient for overall sound recording
royalties to be lower, the label may not be willing to lower its
royalty rate without assurance that the other labels will do the
same,'' a result which he asserts ``is unlikely to happen absent some
form of collusive behavior.'' Id. Thus, Dr. Leonard maintains that the
existence and size of any ``seesaw''-induced decrease in sound
recording royalties remains indeterminate, and it remains ``within the
realm of theoretical possibility that the labels do not agree to any
reduction in sound recording royalties even if a reduction in overall
royalties would be economically efficient. Id.
(e) Professor Watt's Rebuttal to Dr. Leonard's Criticisms
Professor Watt replies with a spirited defense of economic modeling
in general and his economic bargaining model in particular. He begins
by pointing out that models are not supposed to be ``perfect
representations of reality [but rather] are intended to isolate what is
important, in order to expose a useful insight on some issue of
relevance.'' Watt RWRT ] 105. He adds that economic models (not merely
his bargaining model) ``do not necessarily deliver predictions of
situations that are immune to changes in variables outside the model,
but rather the results inform conclusions about the relationships
between the variables and parameters within the model, [which is] by
nature a crude representation[ ] of reality, but the lessons and
insights that they provide can be very relevant to real-world
applications.'' Id. ]] 106-07 (emphasis added).
With particular regard to his bargaining model, Professor Watt
takes issue with Dr. Leonard's assertion that in the former's model the
surplus is a ``fixed constant.'' See Watt RWRT ]] 110-111. Rather,
Professor Watt avers that his bargaining model assume[s] that when the
surplus . . . whatever value it takes . . . is to be shared, the
parties understand that the amount to be shared is, at that moment,
given.'' Id. ] 111 (emphasis added).
Turning to Dr. Leonard's critique regarding the purported
distortionary effect of Professor Watt's modeling assumption of a
single label and a single interactive service, Professor Watt responds
by acknowledging that, if he had modeled multiple labels and services
in the bargaining process, that would be ``not particularly
enlightening vis-[agrave]-vis the single bargain setting, as it will
not lead to different insights than those distilled by the
[Majority].'' Id. ] 113.\66\ Further, Professor Watt characterizes this
criticism as ``empty,'' because under either his two-player Nash model
or Dr. Leonard's posited multi-player (Nash-in-Nash) model, the labels
will not respond to a musical works royalty increase ipso facto with a
reduction in the sound recording royalty (i.e., the seesaw effect will
not occur if there is ``a change in some other variable.''). Id. ] 114.
---------------------------------------------------------------------------
\66\ Professor Watt describes Dr. Leonard's multiple
simultaneous negotiations in a bargaining model as a ``Nash-in-
Nash'' model, but the former does not explain why he concludes that
this approach ``will not lead to different insights'' than those the
Majority distilled from his two-party Nash model.
---------------------------------------------------------------------------
(f) Professor Marx's Criticisms of Professor Watt's Bargaining Model
Professor Marx criticizes Professor Watt's application of the Nash
Bargaining Model because, in her opinion, its ``precise prediction'' of
the nearly one-to-one seesaw relationship ``depends critically on the
assumptions that he makes and the numerical inputs that he uses.'' Marx
WDRT ] 33. First, criticizing his modeling assumptions, like Professor
Katz, she criticizes his decision to abstract from reality by positing
a single label and a single interactive streaming service. She opines
that his one label/one service modeling assumption ineluctably leads to
his conclusion that each of these two parties ``has a `disagreement
payoff' of zero [meaning that] each party ends up with nothing in the
absence of a deal.'' Id. ] 34. But this zero ``disagreement payoff'' is
merely a product of Professor Watt's abstraction from reality,
according to Professor Marx, because ``[i]n reality, if interactive
streaming went away, a share of the music listening that had occurred
through interactive streaming services would migrate to other forms of
music distribution, generating revenues for the label . . . meaning
that the disagreement payoff would be positive for the label). Id.
(emphasis added).\67\ Consistent with Professor Katz, she maintains
that Professor Watt himself acknowledged the presence of this
substitution effect when he testified that ``[t]he existing interactive
streaming companies do not hold an essential input, as first they
compete with the non-interactive services . . . .'' Id. ] 35, n.43
(citing Watt WRT, app. 3).
---------------------------------------------------------------------------
\67\ Professor Marx's reference to a substitution from a
shutdown interactive service to ``other forms of music
distribution'' is different from, but analytically analogous to,
Professor Katz's assertion that the shutdown of any one interactive
service would result in migration of its subscribers and other users
to the remaining interactive services. These analogous critiques are
complementary. See Marx WDRT ] 37 (``One would expect the same
decrease in the estimated see-saw effect by including a second,
competing interactive streaming service in the market instead of
just the one that Professor Watt uses. In that case, if no deal is
reached, users would migrate to an even closer substitute--a
competing interactive streaming service--resulting in an even higher
degree of profit migration and thus an even lower estimated see-saw
effect'').
---------------------------------------------------------------------------
More particularly, Professor Marx maintains, a record label's
disagreement payoff must be considered realistically ``in any
accounting of what would happen if record labels and interactive
streaming services failed to reach an Agreement . . . .'' Marx RWDT ]
35. And, she opines, when this real-world substitution effect is taken
into account, the seesaw effect that Professor Watt estimates is
reduced dramatically, because ``[t]he greater . . . the substitution
between streaming and other forms of distribution, the greater is the
revenue that the record label can capture in the event of disagreement
[[Page 54422]]
and the lower is the estimated see-saw effect.'' Id.\68\
---------------------------------------------------------------------------
\68\ In the context of the bargaining model, Professor Marx
identifies Professor Watt's choice of ``a market structure that is
completely symmetric between record labels and services not
reflective of the real world'' as forcing his model ``to attribute[
] all the . . . surplus division to . . . bargaining power . . . and
none of it to the market structure.'' Id. ] 38.
---------------------------------------------------------------------------
Professor Marx opines that modeling the bargaining process without
these real-world particulars diminishes the value of Professor Watt's
Nash model in several significant ways. First, because his model fails
to incorporate the presence of three major record labels, ``each with
substantial complementary oligopoly power,'' it fails to capture the
fact that ``each record label does not fully internalize the impact of
its rates on the viability of the industry.'' Id. ] 39. She points to
the Judges' Final Determination in Web IV, where the Judges note how
this aspect of complementary oligopoly compromises the value of a rate
as a useful benchmark. Id. ] 39 n.45 (quoting Web IV Final
Determination). More particularly, she opines that when, as here,
``there are multiple negotiations between multiple record labels and
multiple services,'' sound recording rates can be affected ``by the
order of negotiations'' among the several label:service negotiating
pairs--a factor that Professor Watt's bargaining model fails to
capture. Marx WRDRT ] 41.
Next, Professor Marx avers that Professor Watt's bargaining model
``does not explain how or over what time frame the market would move to
a new equilibrium.'' Id. ] 40. More particularly, she testifies,
because interactive services' ``agreements with record labels often
contain multi-year terms and can take many years to negotiate . . .
there may be little incentive or practical ability for both sides to
move to a new rate before the contract expires''. Id. ] 41. She takes
note that this point was established at the hearing during questioning
of Professor Watt from the bench:
JUDGE STRICKLER: What of the situation . . . that the . . . time
period for the existing agreements between the . . . labels and the
interactive streamers is such that they've already locked in a
particular rate and then we set a rate that's higher for the
mechanical to reflect the fact that the sound recording royalty
should drop, but it's locked in for a period of time? Are we running
the risk, then, of disrupting the market by having a total royalty
that's greater than what is indicated by your Shapley testimony,
simply because of the disparity of times in which the rates are . .
. implemented?
PROFESSOR WATT: That's a very fair point. And I didn't even
think of that until you've mentioned it . . . [T]he model I have
done is . . . assuming that . . . the bargained thing happens at the
same time as the--or in the same general period of time as a change
in the statutory rate. You're absolutely correct.
3/27/17 Tr. 3091-92 (Watt); see Marx WRDRT ] 42, n.46
Third, Professor Marx points out that Professor Watt's Nash model
does not attempt to capture the effects of the heterogeneous and
asymmetric distribution of information relevant to the bargain
available to each party at the time of negotiation. Id. ] 41.
Lastly, Professor Marx avers that Professor Watt's Nash Bargaining
Model fails to address, on a more general basis beyond informational
issues, other ``asymmetries among record labels and among services.''
Marx WDRT ] 41.
In sum, Professor Marx concludes that these foregoing real-world
points all preclude the Judges from relying on Professor Watt's
testimony to identify a stable relationship between changes in the
mechanical royalty rate and the sound recording royalty rate because
they all share a common defect--they ``lie outside Professor Watt's
model.'' Marx WRDT ] 41.
To be clear, Professor Marx does not criticize Professor Watt for
neglecting to include these points in his bargaining model; rather, she
acknowledges that ``[t]hese are difficult features to capture in a
tractable equilibrium model.'' Id. Indeed, she urges the Judges to
appreciate that relying on such a necessarily limited model, as the
Majority did, can have ``dramatic effects'' on the royalty rates
derived. Id. Professor Marx emphasizes that all of these inherent
modeling deficiencies are especially pernicious, if the bargaining
model is applied yet again on remand, to set specific rates over a
five-year period, when other variables will have independent effect on
royalty rates. Id.
(g) Professor Watt's Rebuttal to Professor Marx
Because Professor Marx's criticisms are of a similar nature to
Professor Katz's criticisms, Professor Watt responds to Professor Marx
as he did to Professor Katz. To summarize, Professor Watt responds to
Professor Marx's points as follows:
Her criticism is centered on what he characterizes as her
``bogus'' argument that he supposedly had predicted almost a ``dollar
for dollar'' sound recording rate reduction in response to an increase
in the musical works rate (the seesaw effect). Watt RWRT ] 19.
Professor Watt finds this argument ``particularly disheartening,''
because Nash bargaining theory explains why the seesaw would apply to
the splitting of the surplus based on the available data, and that
``there are quite apparent reasons why available surplus may not
decrease even if the musical works rate increased, because of
simultaneous changes to other variables in the model.'' Id. ] 34
(emphasis added).
Professor Marx implicitly contradicts her own reliance on
the complementary oligopoly power of the Major labels by modifying his
bargaining model through the insertion of a lower value for their
bargaining power. Id. ]] 19, 22-24, 26.
Professor Marx misconstrues the purpose of his Nash model,
which was to serve ``as a reply'' to Professor Marx's direct testimony,
and ``to show bargaining insights that bore upon aspects of the case.''
Id. ] 29.
Professor Marx, like Professor Katz, improperly includes
in her bargaining model a potential payoff for the label arising from
an ``outside option,'' i.e., from an alternative that the label can
choose only if the Nash bargaining terminates. Id. ]] 53--68.
(h) Professor Marx's Reply to Professor Watt's Criticism \69\
---------------------------------------------------------------------------
\69\ The Judges found that Professor Watt's remand testimony,
denoted as ``rebuttal,'' also provided de facto ``direct''
testimony, to which the Services could respond with supplemental
testimony and argument. Oct. 1st Order at 11-12. Professor Marx's
response in the following text was set forth in Spotify's permitted
supplemental testimony.
---------------------------------------------------------------------------
In her supplemental remand testimony, Professor Marx challenged
several of Professor Watt's criticisms contained in his remand
testimony. First, she takes issue with what he identified as two
``core'' economic principles of bargaining: (1) that all of the
available net surplus will be shared; and (2) that neither of the two
bargainers will demand a share such that more than the total net
surplus is shared. Marx WSRT ]] 7-8.
As an initial matter, she disputes the notion that these are
``core'' principles of bargaining. Id. ] 8. More particularly, she
states that, in the present case, because ``the label does not know
with exactitude the precise maximum that a service would be willing to
pay (i.e., its ``survival'' rate), and the service likewise does not
know the exact minimum that the label would be willing to accept,'' the
simple bargaining model must be expanded to address ``the potential for
delay and/or bargaining breakdown.'' Id.
As a further criticism, Professor Marx avers that ``[i]n the real
world, the negotiated royalty outcomes do not involve just two parties,
but rather a sequence of overlapping, interrelated,
[[Page 54423]]
bilateral bargains involving multiple competing services and multiple
record labels with complementary oligopoly power.'' Id. ] 12.\70\ This
complication, she opines, exacerbates the informational deficit noted
in the immediately preceding paragraph, such that negotiations within
the several pairings of labels and services ``are affected by
uncertainty and private information and . . . Professor Watt's
discussion of bargaining theory [thus] does not support any particular
real-world see-saw outcome.'' Id.
---------------------------------------------------------------------------
\70\ In like manner, Professor Marx opines that Professor
Spulber's discussion of bargaining theory is irrelevant to any
assessment of ``the complexities affecting real-world negotiations''
and the presence, vel non, of a seesaw outcome. Id. ] 13.
---------------------------------------------------------------------------
(iii) Resolution of the Bargaining Dispute
(a) Professor Watt's Nash Bargaining Model Does Not Support Adoption of
Uncapped TCC Rate
The purpose of Professor Watt's Nash Bargaining Model was to allay
the Judges' concern that increasing the mechanical rate would lead to
higher total royalties for the Services. His bargaining model was
understood by the Majority to show that such higher total royalties
would not result, because the model demonstrated the ``seesaw'' effect,
whereby the sound recording rate would fall almost dollar-for-dollar
with the increase in the mechanical rate. See Determination at 73-74
(``[T]he Judges rely on Professor Watt's insight . . . demonstrated by
his bargaining model that sound recording royalty rates in the
unregulated market will decline in response to an increase in the
compulsory license rate for musical works. . . . Professor Watt's
bargaining model predicts that the total of musical works and sound
recordings royalties would stay `almost the same' in response to an
increase in the statutory royalty.'') (emphasis added).\71\
---------------------------------------------------------------------------
\71\ Copyright Owners note the Majority's recognition that,
regardless of the rate structure, i.e., uncapped TCC or otherwise,
Professor Watt's ``insight'' from ``bargaining theory'' would still
apply. See Determination at 74, n.138. That being the case, the
Majority's first rationale for adopting an uncapped TCC rate is
undermined.
---------------------------------------------------------------------------
On the surface, the economic experts on both sides appear to be at
loggerheads regarding the existence and applicability of the seesaw
relationship. However, as discussed below, on further analysis of their
respective positions, in light of Professor Watt's remand testimony
regarding a key assumption in his bargaining model, their disagreement
narrows considerably and--in an important respect--vanishes
completely.\72\
---------------------------------------------------------------------------
\72\ This is unsurprising. The difference of opinion among
economists often lies in their assumptions, which may be left
unstated or opaque (intentionally or not). Once those assumptions
are laid upon the table, their differences often evaporate. As the
esteemed economist Fritz Machlup noted more than sixty years ago:
``The most prolific source of disagreement lies in differences of
factual assumptions. It is not customary for experts to state all
the assumptions that underlie their conclusions; it would be much
too cumbersome. But when they have reached very different
conclusions, then we are forced to go back and find out what
implicit assumptions they have made.'' F. Machlup, Why Economists
Disagree, 109 Proceedings of the American Philosophical Society 1, 3
(1965). In the modern world of more formal economic modeling as
well, the obfuscation of assumptions continues to be an important
source of dispute, according to a book written by a leading game
theorist upon which Professor Watt relies in his testimony. A.
Rubinstein, Economic Fables at 20 (2012) (``[T]he model's formal
mantle enables economists . . . to conceal from the layman the
assumptions the model uses.''); see J. Schlefer, The Assumptions
Economists Make at 29 (2012) ([S]ome assumptions made by economists
capture important insights, others are insane. All you have to do is
decide which capture insights, which are insane, and in which
situations.'')
---------------------------------------------------------------------------
To recap: In his WRT, Professor Watt stated
[W]ith an appropriately modelled bargaining analysis . . . in my
Appendix 3 . . . I show that for every dollar that the statutory
rate for musical works undercuts a fair and reasonable rate, the
freely negotiated rate for sound recordings will increase by an
estimated [REDACTED] cents.
That is, if the musical works rate is increased to what would be
a realistically fair and reasonable rate, then the negotiated fee
for sound recordings would decrease almost dollar for dollar, with
only a minor change in the total royalty rate for all copyrights
combined.
Watt WRT ] 23 & n.13. But nowhere in his WRT did he qualify this
statement by explicitly acknowledging that in his bargaining model
there are certain assumptions lurking, i.e., that his ``concrete''
analysis is subject to the ``ceteris paribus'' constraint--that all
other things are held constant (i.e., equal before and after the change
in the musical works rate) other things being equal).\73\
---------------------------------------------------------------------------
\73\ In his oral testimony, Professor Watt likewise did not
qualify his opinion by taking note of his ceteris paribus
assumption. See 3/27/17 Tr. 3026 et seq. (Watt).
It is only in his later remand testimony--after the D.C. Circuit's
remand had compelled him to confront criticism from adverse
economists--that Professor Watt expresses this assumption overtly,
making explicit the ``understanding'' that he had theretofore only
---------------------------------------------------------------------------
tacitly assumed:
In other words, a model in which only the two copyright rates
are permitted to change . . . as was the understanding in my
original model, allows the system to derive a clear relationship
between those two rates, and that relationship is that an increase
in one leads to a decrease in the other, that is, the `see-saw
effect.' But if . . . something else changes along with the musical
works rate . . . then the net effect does not predict that the
negotiated rate of the labels will decrease.''
Watt RWRT ] 35 (emphasis added).
Indeed, as noted supra, Professor Watt did give a nod to the
relaxing of his implied ceteris paribus assumption in his WRT, by
identifying varying ``scenarios'' in which he considered the impact of
potential changes in service revenues and service non-content costs,
leading to different percentages of royalties paid to content
providers. Watt WRT ]] 45-52. Professor Watt then used these several
assumptions and scenarios to opine as follows: ``The message that
should be taken from this exercise . . . is that the results . . . are
very dependent upon the amount of total interactive streaming revenue
and the fraction of that revenue that is taken up by downstream non-
content costs.'' Id. ] 53.\74\
---------------------------------------------------------------------------
\74\ Further, in his remand testimony, Professor Watt points out
that Professor Katz made clear in his testimony that he applied the
``all else equal'' assumption expressly in his own Nash bargaining
analysis at the hearing. Watt RWRT ] 20 (quoting Katz WRT ] 67).
---------------------------------------------------------------------------
Professor Spulber, on behalf of Copyright Owners, likewise
emphasizes on remand the importance of the ceteris paribus assumption
in economic modeling:
[A]long with an increase in the compulsory license rate, all
other things being equal, we would expect to see a decrease in sound
recording royalty rates.
. . .
``All other things being equal'' (ceteris paribus in Latin), is
a central principle for economic modelling. This economic analysis
of bargaining highlights an important relationship between two
content cost variables. However, that relationship does not exist in
a vacuum. Many other variables affect the bargaining situation and,
for any given period, the net effect of all of the different
variables may be different than the effect of the modeled variable
alone. Thus, this economic analysis of bargaining will not assure
that a streaming service will not face disruption in the real world
for any reason.
. . .
Economic modeling is supposed to simplify the situation in order
to distill useful principles and teachings.
Spulber RWRT ]] 26-28 (emphasis added).
The Judges agree that the ceteris paribus principle \75\ is a
fundamental
[[Page 54424]]
principle in economic analysis and modeling. Professor Watt succinctly
makes this point, quoting the Nobel laureate economist James Buchanan,
for the following proposition:
---------------------------------------------------------------------------
\75\ The phrase is often translated into English as ``all other
things equal.'' However, that is somewhat ambiguous. Equal to what?
Not to other things. Rather, every ``thing'' (i.e., every other
independent variable) whose effects are not being measured remain
``constant,'' or ``controlled,'' i.e., ``equal'' to their measure
prior to the change of the independent variable being examined. See
W. Nicholson, Microeconomic Theory: Basic Principles and Extensions
at 649 (9th ed. 2005) (defining ``ceteris paribus'' as ``[t]he
assumption that all other relevant factors are held constant when
examining the influence of one particular variable in an economic
model'').
At the heart of any analytical process lies simplification or
abstraction, the whole purpose of which is that of making problems
scientifically manageable. In the economic system we recognize, of
course, that `everything depends on everything else,' and also that
---------------------------------------------------------------------------
`everything is always changing'.
Watt RWRT ] 32 (quoting J. Buchanan, Ceteris paribus: Some Notes on
Methodology, 24 So. Econ. J. 259, 259 (1958).
However, Professor Watt does not quote another portion of Professor
Buchanan's article that makes a point that looms large in the present
proceeding, to wit, the limitations inherent in applying the necessary
ceteris paribus condition:
Real problems require the construction of models, and the skill
of the scientist is reflected in the predictive or explanatory value
of the model chosen. We simplify reality to construct these models,
but the fundamental truth of interdependence must never be
forgotten. . . . [However,] [f]ew, if any, meaningful results may be
achieved by using ceteris paribus to eliminate the study of large
numbers of variables. If such variables are closely related, they
must be studied simultaneously; there is no escape route open.
Id. at 259-60 (emphasis added); see also A. Rubinstein, Comments on
Economic Models, Economics, and Economists: Remarks on Economics Rules
by D. Rodrik, 55 J. Econ. Lit.162, 167 (2017) ``[W]hat matters to the
empirical relevance of a model is the realism of its critical
assumptions'') (emphasis added).\76\
---------------------------------------------------------------------------
\76\ The Judges note now that Professor Watt did not claim that
his bargaining model generated any predictions, but rather that it
explained the splitting of the Shapley surplus by the sound
recording and musical works copyright owners, respectively, and the
impact of that split on royalty rates, given the assumptions and the
data in his model.
---------------------------------------------------------------------------
This is not to say that Professor Watt was unaware of this caveat.
As noted supra, he recognizes the difficulty of extrapolating from a
ceteris paribus world to the real world. The present panel of Judges
likewise recognizes this. However, the Majority missed this distinction
in the Determination when it applied Professor Watt's correct but
ceteris paribus ``insight'' for a constant real-world relationship
between sound recording and musical works royalty rates. Again, not a
single economist made this improper analytical leap or proposed an
uncapped TCC rate in order to set a TCC ratio across the entire rate
term. Indeed, on careful inspection, no economist states in his or her
remand testimony that Professor Watt's bargaining model provides
economic support for the uncapped TCC rate prong.
With the foregoing testimony in mind, the Judges see particularly
relevant several additional points in Professor Watt's remand rebuttal
testimony that pertain to the appropriateness, vel non, of a TCC rate
prong. Referring to the application of his bargaining model to the
present case, Professor Watt made these crucial statements regarding
the lack of a seesaw effect that would generate decreases in sound
recording rates when the mechanical rate is increased:
[T]he actual effects one would expect to see several years later
would be based on the actual data at that time. Moreover, I would
expect many other variables to have a larger effect on the bargains
than the relatively small changes in the musical works rate. . . .
[U]nderstanding actual market outcomes requires understanding these
variables.
. . .
[A]n attempt to capture all aspects of the real world is too
complex for a simple statistical exercise involving an econometric
regression. There is no obvious data to actually use for some of the
independent variables, such as consumer demand equations, costs of
entry and exit, a measure of oligopolistic interaction, different
timings of different rate bargains, and the actual values of outside
options.
Watt WRWT ]] 6(iv), 118.\77\
---------------------------------------------------------------------------
\77\ In the language of econometrics, Professor Watt describes
this problem as the ``almost sure[ ] impossibil[ity] of
``introduce[ing] a control variable for each and every possible
aspect that could potentially impinge upon the relationship [that]
could easily lead to such a low R\2\, and/or statistically
insignificant key coefficients, as to make the regression
meaningless.'' Id. ] 118.
---------------------------------------------------------------------------
Although Professor Watt was hardly transparent in disclosing his
ceteris paribus assumption in his original testimony, it seems clear
that he always understood its presence, and that, when this assumption
was relaxed, ``the actual effects . . . several years later would be
based on the actual data at that time [and] many other variables [with]
a larger effect on the bargains than the relatively small changes in
the musical works rate.'' Id. ] 6(iv) (emphasis added).
Professor Spulber likewise opined that the absence of an explicit
statement of these assumptions in Professor Watt's testimony was
unremarkable and appropriate:
[A]ll other things being equal'. . . should be generally read
into economic modeling conclusions or predictions, whether or not
the words are repeated in each instance. Economists do not typically
repeat these words in each place where they apply, since it would
lead to constant repetition.
Spulber RWRT ] 46, n.8.
Regardless of whether economists invariably identify the existence
of implicit assumptions lurking in each other's models, Professor Watt
overlooked a cardinal rule of communication: Know your audience. Here,
his audience is comprised of three Judges, only one of whom is also an
economist.\78\ Failing to appreciate Professor Watt's implied ceteris
paribus assumption, the Majority transformed his limited (albeit
important) ``insight'' regarding the equal split of the Shapley surplus
between the two classes of rights holders--and the seesaw effect that
would have if the mechanical rate were increased when the split was
imposed--into a justification for the imposition of an uncapped TCC
rate prong over the five-year rate term. The Majority's language
reveals this point clearly:
---------------------------------------------------------------------------
\78\ The dissenting Judge (the only economist on the panel)
warned that the seesaw effect was rife with assumptions that
rendered it too speculative to be relied upon to support the
uncapped TCC rate prong. See Dissent at 7-8.
As to the issue of applying a TCC percentage to a sound
recording royalty rate that is artificially high as a result of
musical works rates being held artificially low through regulation,
the Judges rely on Professor Watt's insight . . . demonstrated by
his bargaining model that sound recording royalty rates in the
unregulated market will decline in response to an increase in the
compulsory license rate for musical works. See 3/27/17 Tr. 3090
(Watt) (``[T]he reason why the sound recording rate is so very high
is because the statutory rate is very low. And if you increase the
statutory rate, the bargained sound recording rate will go down.'')
Professor Watt's bargaining model predicts that the total of
musical works and sound recordings royalties would stay ``almost the
same'' in response to an increase in the statutory royalty. Id. at
3091.
Determination at 73-74 (emphasis added).
Making the point ever so plainly, Professor Watt now expressly
acknowledges that his `` `see-saw effect' was never really a
`prediction' '' at all! Watt RWRT ] 117. Rather, he now cautions the
present panel of Judges, that, ``to make the jump from the model to the
actual real-world effects, one cannot ignore the words that are
omnipresent in all economic modeling,
[[Page 54425]]
that predictions about causal relationships are understood to be ``all
else equal.'' Id. ] 32.
Without the benefit of these caveats regarding an extrapolation of
the ``seesaw'' theory to the real-world, and with absence of an
explicit statement of the ceteris paribus assumption, the Majority
misapplied his testimony as a basis to adopt a fixed TCC rate, based
upon data from a snapshot in time (2016) to cement that rate
relationship for the entire five-year period.\79\ The Majority
misapplied Professor Watt's correct insight from bargaining theory
regarding the use of a fixed ratio for the equal division by two ``Must
Have'' input suppliers of the Shapley surplus to set royalty rates in a
period, by using that insight incorrectly to establish a fixed ratio of
royalty rates over the rate term.\80\
---------------------------------------------------------------------------
\79\ The importance of Professor Watt's failure to make explicit
the ceteris paribus assumption in his WRT is demonstrated by his
need to make it explicit in his RWRT. But even now, rather than
acknowledge that the Majority missed the point, he claims that the
Services' are wrongly blaming the Majority for failing to understand
this assumption: ``The Services' testimony on this remand seems
primarily focused on creating a ``straw man'' argument . . .
accus[ing] the [Majority] of something that the [Majority] did not
do--that is, rely on a guarantee of a particular decrease in sound
recording royalty rates--and the Services then attack the Board's
determination by claiming that the decrease did not occur.'' Watt
RWRT ] 5. As shown supra, however, this is precisely how the
Majority interpreted Professor Watt's ``insight.'' The Judges
understand that, as a matter of tact and tactics, Copyright Owners
may be reluctant to acknowledge that the error lies in the
combination of their witness's opaque testimony and the Majority's
lack of understanding of the assumptions economists make. Copyright
Owners might prefer to cast the Majority as the victims of the
Services' incorrect accusation. But the plain language of the
Determination belies Copyright Owners' characterization as to how
the confusion arose.
\80\ The forgoing analysis as applied to the uncapped TCC rate
needs to be contrasted with the application of Professor Watt's
bargaining model to increase the percent of-revenue rate to 15.1%.
That higher rate was set by the Majority after its consideration of
the same Shapley approaches, pursuant to the Judges' combination of
inputs from Professor Gans model (his [REDACTED] round recording-to-
musical works ratio) and the Shapley Value Model of Professor Marx
that adjusted for complementary oligopoly power by establishing a
lower total royalty level ([REDACTED]%). But the difference is that
the 15.1% revenue rate was set by applying the Shapley results based
on actual and projected market data, see Gans WRT ] 38, whereas the
uniform uncapped TCC rate (26.2%) was based on the ceteris paribus
assumption that held constant the actual data regarding the
aforementioned independent variables. As explained above though,
Professors Watt and Spulber make it clear that the ``insight'' from
bargaining theory did not have implications to allow for a
``prediction'' of rates in future periods.
Thus, when the Majority engaged in its analysis and ``line-
drawing'' to apply the data and market projections relied upon by
Dr. Gans's data, the Majority was operating--to use the D.C.
Circuit's phrase--in its ``wheelhouse,'' making a finding that
withstood appeal. Johnson, supra, 969 F.3d at 385-86; see also
Determination at 69-70 (``Professor Gans utilized data from
projections in a Goldman Sachs analysis to identify the aggregate
profits of the record companies and the music publishers,
respectively. . . . The Judges also find Professor Gans's reliance
on financial analysts' projections for the respective industries to
be reasonable.'').
---------------------------------------------------------------------------
Additionally, an examination of the expert economists' testimony
reveals that their facial disagreements vanish once the necessary
assumptions are laid bare. Professor Watt and the Services' three
economists all identify the following independent variables that will
impact the relative levels of sound recording and musical works rates
paid by interactive services:
(1) the level of downstream consumer demand;
(2) entry costs;
(3) exit costs;
(4) oligopolistic interaction;
(5) the timing of sound recording agreements vis-[agrave]-vis
statutory rate setting; and
Professor Watt and the three Service economists agree with regard
to the relevancy of these six independent variables. Compare Watt RWRT
]] 6(iv), 118 (identifying all five independent variables) with Leonard
WDRT ] 18 (identifying independent variables 1-4 above); Marx WDRT ]]
4-5, 42; (identifying independent variables 1-5 above); Katz WDRT ]]
127, 134 n.115 (identifying independent variables 4 and 6 above).
Accordingly, the remand record shows a consensus as to the lack of
modeling of independent variables that would be important to estimate
an uncapped TCC royalty ratio that could be utilized by the Judges to
lock-in a ratio over the rate term.
Indeed, as noted supra, a careful reading of the remand testimony
by Copyright Owners' economists, Professors Watt and Spulber, reveals
that neither of them actually testifies that there is sufficient
theoretical and empirical evidence to support the uncapped TCC rate
prong and the 26.2% TCC rate phased in on that prong. Rather, those two
witnesses testify to something far narrower: the alleged correctness of
Professor Watt's ``seesaw'' theory as demonstrating an equal splitting
of the surplus between the two ``Must Have'' input suppliers, and the
effect of that split when all other relevant independent variable are
held constant.
In this regard, it is noteworthy that none of Copyright Owners'
several economic experts in this proceeding (Dr. Eisenach, Professor
Gans, Dr. Rysman, or Professor Watt) ever proposed an uncapped TCC rate
prong in any form, let alone within a greater-of formulation. Such a
proposal would have been improper, because, as the expert testimony
described above makes clear, the ceteris paribus assumption, reasonable
for modeling purposes to provide insight as to the surplus split, lacks
the input of the omitted variables that the experts on both sides find
relevant to the application of economic modeling in this proceeding. A
further review of Copyright Owners' economic expert witness testimony
on remand--the first time any of them had occasion to weigh-in on the
appropriateness of the uncapped TCC prong--reveals that they also have
not endorsed the uncapped TCC rate prong as a proper form of rate
setting. To be sure, they strongly endorse the insight first described
by Professor Watt in his WRT that the Nash surplus would be split
essentially evenly between the two suppliers of essential content,
given his simplifying assumptions. But such endorsement is hardly the
same as endorsement of the uncapped rate prong itself.
For these reasons, the Judges find erroneous the Majority's
identification of a fixed relationship between the sound recording and
mechanical royalty rates that could serve as a basis for the Majority's
first rationale for yoking the mechanical rate to an uncapped TCC rate
prong.
(b) The Services Have Not Rebutted Copyright Owners' Prima Facie
Showing That Professor Watt's Model Demonstrates a More Limited
``Seesaw'' Effect
The foregoing analysis and decision related to the absence of a
fixed relationship between the sound recording and mechanical royalty
rates. A separate fixed relationship--the one Professor Watt has
clarified he was demonstrating all along--is that if the Judges
increase the mechanical royalty rate, the Shapley surplus realized by
the labels will decrease almost dollar-for-dollar with the increase in
the mechanical rate. The Services' economists aver that even this
version of the seesaw is defective.
According to Professors Katz and Marx, the Nash Bargaining Model
constructed by Professor Watt is deficient because it fails to properly
characterize the ``disagreement payoff'' to the sound recording company
when it and an interactive service fail to reach an agreement. More
particularly, as explained supra, they assert that Professor Watt's
model omits the value of ``outside options'' available to the sound
recording company. This criticism relates to the issue of whether the
seesaw effect would occur as posited in Professor Watt's model. That
is, the increase in the sound recording
[[Page 54426]]
company's ``disagreement payoff'' (a/k/a ``threat point'') would lead
to a higher royalty in the Nash bargain between the sound recording
company and the interactive service than needed to generate the seesaw
effect to offset the higher mechanical royalty rate.
As the several experts' positions in this regard, discussed supra,
make clear, however, each side has a different understanding of whether
an ``outside option'' is properly included in the definition and
calculation of the ``disagreement payoff.'' On the one hand, Professors
Katz and Marx claim that the existence and value of ``outside options''
should be included in the ``disagreement payoff.'' However, they
provide no economic authority for that assertion.
By contrast, Professor Watt cites to multiple economic game theory
publications and authorities for the proposition that the presence and
value of ``outside options'' are not to be included in the
``disagreement payoff'' contained in a Nash Bargaining Model. See A.
Muthoo, Bargaining Theory with Applications at 105 (1999) (``I thus
emphasize that the outside option point does not affect the
disagreement point.''); M. Osborne & A. Rubinstein, Bargaining and
Markets at 88 (1990) (``it is definitely not appropriate to take as the
disagreement point an outside option. . . .''); K. Binmore, A.
Rubinstein & A. Wolinsky, The Nash Bargaining Solution in Economic
Modeling, 17 RAND J. Econ. 176, 185 (1986) (``An outside option is
defined to be the best alternative that a player can command if he
withdraws unilaterally from the bargaining process.'').
According to Professor Watt and these authorities, the reason for
excluding ``outside options'' from the Nash Bargaining Model is
fundamental to the nature of the model itself. In the Nash approach,
the negotiating parties are bargaining with each other only over the
surplus their deal can generate, and they are attempting to agree upon
an allocation of that surplus that exists within the bounds of their
respective ``disagreement payoffs.'' Each may have ``inside options,''
which are alternatives available to them while bargaining is ongoing
and they temporarily disagree. See Muthoo, supra, at 137. However,
``outside options'' are available to a Nash bargaining party only in
lieu of continuing the Nash bargaining with the original counterparty
if it ``withdraws'' from the Nash bargaining process. See Binmore et
al., supra. Professor Watt characterizes the distinction as follows:
[T]he Nash bargaining model [is] designed as [a] self-contained
portrayal[ ] of negotiating behavior. . . . Given a surplus to
share, the Nash model . . . provide[s] allowance for financial
payments that a party is actually receiving, only while negotiations
are ongoing, without walking away for another option, and that would
cease as a result of the deal, to be factored into modelling as a
cost in some situations.'')
. . .
[A]n outside option (a potential payoff that is not directly
related to a share of the surplus that is being negotiated) . . .
comes in as a constraint upon the set of feasible deals that could
be struck. . . .''
Watt RWRT ]] 56, 58.\81\
---------------------------------------------------------------------------
\81\ Professor Marx in fact cites several of these authorities
(for other points), without noting the distinction they make between
the appropriate inclusion of ``inside options'' and exclusion of
``outside options'' in Nash modeling. See id. ] 59.
---------------------------------------------------------------------------
The Services never sought to introduce further testimony regarding
this important dispute. This is particularly striking because the
Services filed a motion to strike certain portions of the CO Reply, or
for leave to file supplemental testimony responsive to those itemized
portions. The portions the Services identified in their motion did not
include Professor Watt's criticisms as to the inclusion of ``outside
options'' in their experts' Nash modeling. Further, after the Judges
granted the Services' motion by providing them leave to file
supplemental testimony--consistent with the designations in their
motion--the supplemental testimonies did not address this ``outside
options'' issue.
In the course of discussions among the parties and the Judges
regarding remand procedures, the Judges invited the parties to produce
witnesses for a hearing, at which one or more of the Services' economic
expert witnesses could have addressed this ``outside options'' issue.
However, the Services (and Copyright Owners) waived the opportunity to
produce witnesses at a hearing. Rather, they offered, and the Judges
agreed, that they would stand on their written testimonies and proceed
to closing arguments by counsel.
In the closing arguments, each side argued numerous points of
controversy and provided the Judges with dozens of demonstrative aids
summarizing record evidence and the parties' arguments, but none of
those arguments or demonstrative aids so much as mentioned this
``outside options'' dispute. Moreover, when the Judges inquired during
closing arguments as to whether Services' counsel would be addressing
any of the experts' ``modeling disputes,'' counsel said that they were
resting on their papers. 3/8/22 Tr. 86-87 (Closing Argument).
Similarly, when the Judges inquired of Copyright Owners' counsel
whether he would be addressing the modeling ``dust-up'' between
Professors Watt and Katz, counsel demurred, stating that although he
would ``love to engage on it but . . . ``there would be too many
slides. . . .'' Id. at 262-64.
Simply put, the Services' economic experts made an assertion
regarding the need for Professor Watt to have included ``outside
options'' in his Nash Bargaining Model, but Professor Watt presented
authority clearly stating that such inclusions would be improper. Thus,
Copyright Owners made a prima facie showing that in a Nash Bargaining
Model, the surplus generated by the streaming surpluses acquired by the
content providers would be split equally as between the sound recording
licensors and musical works licensors, and that, ceteris paribus, an
increase in the mechanical rate to provide Copyright Owners more of the
surplus (per the Shapley-based results relied on by the Majority) would
be essentially offset through a nearly 1:1 reduction in the sound
recording rate. In response to Copyright Owners' prima facie case, the
Services stood mute in response to the rebuttal argument claiming that
their experts misapprehended the Nash modeling distinctions between
``inside options'' and ``outside options.'' \82\
---------------------------------------------------------------------------
\82\ The third economic expert for the Services, Dr. Leonard,
did not utilize the ``outside option'' phraseology to describe his
critiques. Rather, he first criticized Professor Watt for assuming
the existence of a ``fixed surplus.'' Leonard WDRT ] 16. However, as
discussed supra, that assumption came from the Majority's
extrapolation from Professor Watt's hearing testimony. His explicit
statement regarding the ceteris paribus assumption makes clear that
he was not assuming a ``fixed surplus.'' Watt RWRT ]] 110-11.
(Again, the only ``fixed'' surplus was not ``assumed,'' but rather
quantified, in order to establish the Majority's percent-of-revenue
prong royalty rate of 15.1%.)
Dr. Leonard next claims that Professor Watt's assumption that
the labels would bear virtually the entirety of an increase in the
statutory rate, because they previously ``have captured almost all''
[the] surplus,'' has been contradicted by the evidence.
Specifically, he refers to the 33-month period in which the
Phonorecords III rates were effective (January 2018 through
September 2020). Leonard WDRT ] 16. However, as the Judges find in
this Determination, that 33-month period was marked by significant
uncertainty with regard to the ultimate rates and rate structure
(and the rates were being phased-in), so no findings could reliably
be made based on sound recording rate changes during that period.
The remainder of Dr. Leonard's critique concerns issues that
would make a fixed TCC ratio inappropriate over the rate term. The
Judges agree with those criticisms as previously discussed, but they
do not pertain to this narrower issue of whether the surplus
generated by interactive streaming would be split in a manner
consistent with Professor Watt's Nash Bargaining Model.
---------------------------------------------------------------------------
Accordingly, the Judges find that the Services' criticisms in this
regard are insufficient to rebut Copyright Owners' prima facie showing
that Professor Watt's Nash Bargaining Model properly
[[Page 54427]]
identified and valued the ``disagreement payoff.'' 83 84
---------------------------------------------------------------------------
\83\ To be clear, the Judges' ruling is narrow; they make no
finding beyond crediting this prima facie showing and the failure of
the Services to rebut sufficiently that showing. It might be the
case that the existence and definition of ``outside options''--and
their relationship to ``inside options''--have other implications
vis-a-vis a Nash Bargaining Model applied in the context of a rate
setting proceeding. However, the Judges may not introduce and rely
on analytical approaches not developed by the parties. See Johnson,
969 F.3d at 381 (the Judges must not ``procedurally blindside[ ]''
the parties with an ``approach . . . first presented in the
determination and not advanced by any participant.''). See generally
P. Wald, Limits on the Use of Economic Analysis in Judicial
Decisionmaking, 50 J. L. & Contemporary Problems 225, 228 (1987) (''
judicial analysis, economic or otherwise, takes place only in the
context of lawsuits between two or more parties imposes a practical
constraint on the judge's ability to use economic analysis.'').
\84\ Professor Katz also criticizes Professor Watt's assumption
that ``a label's non-content costs are proportional to licensing
revenues.'' Katz WDRT ] 22. More particularly, Professor Katz claims
that this is not ``plausible'' because ``the royalty rate does not
directly affect the sound recording copyright owners' non-content
cost.'' Id. ] 133. The effect of eliminating this assumption,
according to Professor Katz, is to reduce the seesaw effect in
Professor Watt's model of [REDACTED] slightly further away from a
1:1 ratio, to .92. Id.
In rebuttal, Professor Watt says this criticism is inconsistent
with Professor Katz's own analysis, because the latter also ``sets
the cost equal to a fraction of revenue. . . .'' Watt ] 82 n.31
(referring apparently to a comparison of Katz WDRT ] 129 with id. ]
133). Professor Watt concludes that not only does ``[Professor]
Katz's own model contain the same feature that he is critical of in
my model,'' it is also ``not a flaw in the bargaining model.'' Watt
] 82. As a substantive matter, Professor Watt defends the assumption
that non-content costs would rise with royalty income, because
``[g]reater revenue should be directly equated with a larger scale
of business'' and ``the additional royalty income would have to be
managed (i.e., distributed to those who need to be paid from it,
such as artists), implying higher administration costs.'' Id. ] 79.
The Judges find that the common use by both experts of this
assumed proportionality of a label's non-content costs to licensing
revenues alone blunts Professor Katz's criticism of Professor Watt's
modeling. Further, Professor Watt reasonably posits that higher
revenue would imply a larger scale of business with associated
general cost increases. (But the Judges do no agree that it was
reasonable for Professor Watt to assume that distribution and
administrative costs in particular would increase merely because of
an increase in royalty rates; simply paying more money, ceteris
paribus, is not self-evidently associated with an increase in
costs.)
---------------------------------------------------------------------------
b. Rejection of Second Rationale for Including Uncapped TCC Rate Prong
In the Determination, as noted supra, the Majority also justified
the adoption of the uncapped TCC rate prong because it had the effect
of ``import[ing] into the rate structure the protections that record
companies have negotiated with services to avoid the undue diminution
of revenue through the practice of revenue deferral.'' Determination at
36; see also Johnson, 369 F.3d at 372 (``By pegging the mechanical
license royalties to an uncapped total content cost prong, the Board
sought to ensure that owners of musical works copyrights were neither
undercompensated relative to sound recording rightsholders, nor harmed
by the interactive streaming services' revenue deferral strategies. . .
.'') (emphasis added).
(i) Parties' More Specific Arguments
Copyright Owners likewise argue that the uncapped TCC rate
structure should be ``adopted to provide protection against revenue
deferment and displacement in a revenue-based rate structure.'' CO
Initial Submission at 38; see also id. at 40 (describing uncapped TCC
rate prong as ``critical backstop in a revenue-based rate
structure.'').
Whereas Copyright Owners echo the Majority, the Services adopt the
reasoning of the Dissent. They argue as follows:
[A] rate structure with a capped TCC prong, like the
Phonorecords II settlement, achieves the same goal of protecting the
Copyright Owners from any potential revenue deferral through a
``structure that provides alternate rate prongs and floors, below
which the royalty revenue cannot fall,'' . . . and does so without
allowing Copyright Owners to impermissibly share in the labels'
complementary oligopoly power. . . . [T]he streaming industry has
twice concluded, after extensive negotiations, that the appropriate
way to address any concerns regarding revenue deferral is to have a
rate structure that includes a capped TCC prong. Phono I, 74 FR
4510; Phono II, 78 FR 67938.
Services' Joint Opening Brief at 62 (quoting Dissent, 84 FR 1990)
(emphasis added).
In their Reply, Copyright Owners argue that the Majority maintained
the benefits of price discrimination contained in the prior
Phonorecords II framework, but balanced that goal with added protection
against Service revenue deferral and displacement. Copyright Owners'
Reply Brief on Remand at 49 (``In adopting a rate structure with [an
uncapped] TCC for all service offerings, the [Majority] balanced its
concerns about fostering price discrimination while also protecting
against proven revenue diminution by the Services.'').
The Services, in their Reply, take note that pre-remand, Copyright
Owners had strenuously objected to any yoking of the mechanical royalty
rate to the sound recording rate, maintaining that, although the
Copyright Owners now advocate for an uncapped TCC rate to protect
against revenue displacement and diminution:
[I]n their [pre-remand] reply proposed findings, the Copyright
Owners had expressed a very different view, arguing that an uncapped
TCC prong ``does nothing to protect Copyright Owners from the
Services' revenue displacement and deferment'' [and] Copyright
Owners have not even tried to explain away their complete about-face
on this issue.
Services' Reply at 43.
(ii) Analysis and Decision Regarding Revenue Diminution or Deferral
The Judges find that the second rationale put forth to support an
uncapped TCC rate does not justify the adoption of that rate prong.
Several reasons support this finding.
First, there is insufficient evidence to show how the sound
recording companies contractually structure their own royalty rates,
which would constitute the rate base for an uncapped TCC rate for the
mechanical royalty. The sound recording royalty rate, when proffered
for use as a mechanical royalty rate base, is analogous to pegging the
value of a foreign currency to the U.S. dollar. That is no mere
benchmark. The Judges must have the benefit of sufficient record
evidence to demonstrate that the pegging (or, to use the D.C. Circuit's
word in Johnson, ``yoking'') of a statutory rate to an unregulated rate
serves the statutory purposes for the rate at issue, here, the
mechanical rate.
But Copyright Owners presented virtually no evidence regarding how
the sound recording companies structure their interactive service
royalties. Indeed, in the hearing, Dr. Eisenach acknowledged that the
``relative value of sound recording [to] musical works licenses may
depend on a variety of factors,'' but he intentionally eschewed
unnecessary ``assumptions, complexities and uncertainties associated
with theoretical debates'' as to why the particular market ratios
existed. See Determination at 44. Indeed, the Majority found fault with
Dr. Eisenach's willful ignoring of these issues, agreeing with the
Services' criticism that Dr. Eisenach's ``use of sound recording
royalties paid by interactive services embeds within his analysis the
inefficiently high rates that arise in that unregulated market through
the complementary oligopoly structure of the sound recording industry
and the Cournot Complements inefficiencies that arise in such a market.
See Determination at 47. The uncapped TCC rate advocated now by
Copyright Owners suffers from the same affliction.
The only reference to such sound recording rate formulae in
Copyright Owners' voluminous PFF after the hearing was its statement
that the effective revenue calculations in two of the Major labels'
agreements with the
[[Page 54428]]
services was based on [REDACTED]. See Copyright Owners' PFF ]] 72, 91
(cited post-remand at Copyright Owners' Motion for Reconsideration or
Clarification at 25, n.14). On remand, the Services have provided a
further summary of the types of [REDACTED]. See White WDRT ]] 6-7, 14-
15, 20, 24-26, 28-29 ([REDACTED]); Bonavia WDRT ]] 15-17 ([REDACTED]);
Mirchandani WDRT ]] 16, 21-24 ([REDACTED]). Clearly, the levels of
[REDACTED] would have to be weighed and the impact of complementary
oligopoly power would need to be identified in order to adjust the rate
prongs to account for that power. But the record is devoid of such
details.
Second, compounding this problem, because the uncapped TCC rate is
embedded in a ``greater-of'' rate structure, the labels can exploit
their complementary oligopoly power when creating the switching points
that toggle royalty payments between and among rate prongs. As the
Judges have explained previously, in declining to import a ``greater
of'' structure from the unregulated interactive market, this
structure[it] is based on ``agreements [which] were all negotiated in a
market characterized by the lack of effective competition, and that the
lack of competition would affect the structure as well as the level of
rates.'' SDARS III, 83 FR 65210, 65228 (Dec. 19, 2018) (emphasis
added). Further, the Judges held therein that the ``advantageous''
nature of a ``greater-of'' structure to sound recording licensors ``may
well represent an example of what licensors can and would obtain when
they exploit their ``must have'' status for a special competitive
advantage.'' Id.; see also Dissent at 47 (in absence of testimony
explaining how greater-of structure is consonant with effective
competition, use by licensor suggests a game of ``heads I win tails you
lose.).''
Thus, there is insufficient evidence or testimony that would permit
the Judges to make any adjustment for the complementary oligopoly power
that may be built into each prong of the sound recording royalty rate
structures.
Third, as the Services note, Copyright Owners pre-remand, opposed
the identical rate structure--consisting of a percent-of-revenue prong
and an uncapped TCC prong--before Copyright Owners were in favor of it,
post-remand.\85\ Although Copyright Owners took a 180-degree turn on
this issue, they never stated they were wrong to oppose it previously.
Indeed, the Dissent relied upon Copyright Owners' strenuous objection
to an uncapped TCC rate, quoting it verbatim:
---------------------------------------------------------------------------
\85\ When Copyright Owners opposed the concept of an uncapped
TCC rate prong in a greater-of structure, the proposed uncapped TCC
rate was Google's 15% (and its proposed percent-of-revenue rate was
10.5%). Determination at 13. But after the Majority set the uncapped
TCC rate at 26.2%--a 75% increase over the 15% TCC rate--Copyright
Owners became zealous converts to the concept of an uncapped TCC
rate proper.
Copyright Owners rightly note that they obtain no legal
protection under such a TCC prong. In making this argument regarding
displacement and deferral of revenue, Copyright Owners lay out
comprehensively all the problems inherent in an uncapped TCC prong
set in a greater of rate structure, such as adopted in the majority
opinion:
The notion that [the] TCC prong will provide protection from
revenue gaming, deferral and displacement, and other revenue prong
problems is unsupported and speculative. Relying on just the TCC to
solve those admitted problems leaves the Copyright Owners'
protection from such problems entirely outside the statute. . . .
the per-user rates in the label deals are what protects the
Copyright Owners from price-slashing by the services. What is left
unanswered . . .is . . . how can it be reasonable to ask the Judges
to set a rate that does not itself provide for a fair return . . .
but simply puts the Copyright Owners' fair return in the hands of
the labels to negotiate terms that will adequately protect the
publishers and songwriters as well? The labels do not have a mandate
to ensure that the Services provide a fair return to the Copyright
Owners, and cannot be directed to ensure such. Indeed, labels may
not have the same incentives as songwriters and publishers to
negotiate such protections in their deals. To wit, a label could
make an agreement with a service that includes only a revenue prong
in exchange for equity or some other consideration that it may never
include in the applicable revenue subject to the TCC. . . . [W]hat
if Google purchased one or more record labels and did not have to
pay any label royalties? Or what if Spotify chose to avail itself of
the compulsory license to create its own master recordings embodying
musical works--which it is already doing . . . and chose to
compensate itself for its use of the master recordings on a
sweetheart basis (or not at all)? Or what if one or more labels
decided to enter the interactive streaming market and did not have
to pay themselves royalties? In each case, the Copyright Owners'
protection--the protection that the Services admit the Copyright
Owners need and is provided by the TCC--would be gone.
Dissent at 5-6 (quoting Copyright Owners' RPFF-Google at 39-41)
(emphasis added). To make the identical point post-remand, but from the
Services' perspective, Pandora's economic expert witness, Professor
Katz, simply utilizes Copyright Owners' verbatim language (bolded
above), but substitutes the word ``Services'' for ``Copyright Owners''
(and ``income'' for ``return'') to highlight how reliance on the sound
recording royalty rate is improper:
What is left unanswered . . . is . . . how can it be reasonable
to ask the Judges to set a rate that does not itself provide for a
fair income . . . but simply puts the Services' fair income in the
hands of the labels to negotiate terms that will adequately protect
the Services as well? The labels do not have a mandate to ensure
that the Copyright Owners provide a fair income to the Services, and
cannot be directed to ensure such.
Katz WDRT ] 71.
The Judges find this argument persuasive, both in its own right and
in the fact that it has been advanced by Copyright Owners and the
Services alike.\86\
---------------------------------------------------------------------------
\86\ At Closing Arguments on remand, Judge Strickler queried
counsel for Copyright Owners regarding their prior rejection of an
uncapped TCC prong within a ``greater-of'' rate structure. Counsel's
response was that an uncapped TCC doesn't provide enough protection
against revenue diminution: ``It provides more than the Phonorecords
II rates, but not as much as we want,'' although ``still better
than'' the negotiated Phonorecords II approach. 3/8/22 Tr. 240-
41(Closing Argument). But Copyright Owners have neither
distinguished nor disavowed their persuasive legal point quoted in
the text above, to wit that an uncapped TCC rate would be
unreasonable if the ``protection'' it affords lies ``entirely
outside the statute.'' Whether the ``protection'' relates to
Copyright Owners' concern over revenue diminution or to the
Services' concern over uncapped mechanical rates, the legal defect
is the same--the unreasonableness of leaving the purported
protection ``entirely outside the statute.''
---------------------------------------------------------------------------
Fourth, the Judges note that the Majority did not find that revenue
diminution, via displacement, deferral, or otherwise was pervasive, as
Copyright Owners aver. Compare CO Initial Submission at 40 (``The
record overwhelmingly established that the percent of revenue prong
often results in musical works royalties that are too low . . .
drive[n] [by] . . . . revenue deferral [and] revenue displacement'')
with Determination at 21 (``The Judges agree that there is no support
for any sweeping inference that cross-selling has diminished the
revenue base.'') (emphasis added) and 36 (``The Judges find that the
present record indicates that the Services do seek to engage to some
extent in revenue deferral in order to promote their long-term growth
strategy.'') (emphasis added).
Given that the Majority found revenue diminution through
displacement and/or deferral exists only ``to some extent'' and is not
a ``sweeping'' issue, the Judges on remand find that the uncapped TCC
rate structure creates the potential for unbalanced harm. As noted
supra, the only protection against runaway mechanical rates, the seesaw
hypothesis, cannot justify yoking the mechanical rate to a fixed ratio
with the
[[Page 54429]]
unregulated sound recording rate.\87\ By contrast, and as discussed
infra, the Phonorecords II-based benchmark approach, despite its own
imperfections, is superior in this regard, because its series of
alternate rate prongs and floors represents a negotiated compromise
(negotiated by trade associations with countervailing power) between
the potential for revenue diminution that would harm Copyright Owners,
on the one hand, and the potential for runaway mechanical rates (yoked
to the sound recording companies' complementary oligopoly power) that
would injure the Services, on the other.
---------------------------------------------------------------------------
\87\ Even Google, the party that, post-hearing, broached in its
PFF the idea of an uncapped TCC prong, candidly identified the risk
arising from an uncapped TCC: ``Having no cap on TCC . . . leaves
the services exposed to the labels' market power, and would warrant
close watching if adopted. . . .'' Google PFF ] 73 (emphasis added).
But as the Dissent noted, there is no satisfactory way to monitor an
uncapped TCC rate prong: ``Who would do the ``watching''? When would
such watching occur? Congress directed the Judges to be the
``watchers,'' and Congress instructed that the ``watching'' should
occur only through rate proceedings. . . .'' Dissent at 4 (emphasis
in original).
---------------------------------------------------------------------------
(iii) Distinction Between the ``Reasonable'' Rate Statutory Standard
and the Factor (D) Objective To Minimize ``Disruptive Impact''
The Judges next consider an issue emphasized by Copyright Owners:
whether the Services have demonstrated that the uncapped TCC rate prong
would cause a ``disruptive impact'' as set forth in Factor (D) of
section 801(b)(1).\88\
---------------------------------------------------------------------------
\88\ Separate and apart from the ``disruptive impact'' argument
made by Copyright Owners, there is no need to consider how this
prong would relate to Factor D, because the Judges find the uncapped
TCC rate prong with the (phased-in) 26.2% rate to be
``unreasonable.'' If it were necessary to separately consider the
four itemized factors, the Judges would confirm that Factor A is
satisfied, because, as the D.C. Circuit found, the Majority
reasonably found that rates should increase from the Phonorecords II
period, and the 15.1% revenue rate represents a 44% increase. The
Judges would also find Factors B and C to be satisfied without a
separate uncapped TCC rate prong. The reason is that, under the
section 801(b)(1) standard, the ``reasonableness'' standard filters
out more statutorily infirm rates than the fairness objectives. By
contrast, when a rate does satisfy the ``reasonableness'' standards
under section 801(b)(1), the Judges must also consider the rate
through the finer ``fairness'' filter. Cf. Determination at 68 &
n.120 (distinguishing between: (1) a Shapley Value analysis that
filters out unreasonable rates by reducing licensors' ability to
abuse market power by threatening or exercising their refusal to
license (``hold-out or ``hold-up'' power); and (2) a Shapley Value
analysis that further filters out unfair rates by going beyond
eliminating abuse of market power to also make a ``market power
adjustment'' explicitly to address Factors B and C). Finally, as the
text infra, explains, the Judges also find no basis under Factor D
to alter their analysis.
---------------------------------------------------------------------------
Section 801(b)(1) provides that one of the competing priorities of
the Judges in setting the mechanical rate is ``[t]o minimize any
disruptive impact on the structure of the industries involved and on
generally prevailing industry practices.'' 17 U.S.C. 801(b)(1)(D). In
Johnson, the D.C. Circuit did not identify any argument by the Services
that was predicated on a claim that this statutory form of
``disruption'' had occurred, or was likely to occur, as a consequence
of the Majority's rates and rate structure. Additionally, the D.C.
Circuit did not ground its decision to vacate and remand the Judges'
uncapped TCC rate and rate structure rulings based on the potential
that these rulings would be disruptive to the Services, let alone would
cause a statutory ``disruptive impact.''
After the D.C. Circuit's ruling, an argument regarding
``disruption'' was first made by Copyright Owners, not the Services.
Copyright Owners argued that the vacated rates should nonetheless be
maintained as interim rates, during the pendency of the remand
proceeding. Motion of Copyright Owners to Adopt Interim Rates and Terms
Pending the Remand Determination, passim (Nov. 2, 2020). Copyright
Owners argued that reverting to the rates that existed before the
Determination would constitute a ``disruption'' and self-servingly
predicted that the Services would attempt to argue that the uncapped
TCC rate and rate structure were themselves ``disruptive.'' Copyright
Owners opined that such an argument would be a ``hollow exercise.'' Id.
at 12, n.5; see id. at 2-3, 9 (claiming absence of disruption from
uncapped TCC rate and structure despite absence of such argument by
Services).
In response to that motion, the Services did not assert that the
Majority's uncapped TCC rates and rate structure would constitute
disruption or have disruptive impact, whether under statutory Factor D
or otherwise. See Services' Opposition to the National Music
Publishers' Association (NMPA) and Nashville Songwriters Association
International's (NSAI) ``Interim Rates Motion'' (Nov. 18, 2020). In
reply, Copyright Owners shifted from anticipating a ``disruption''
argument to misinterpreting Johnson, asserting, without citation:
``With respect to the TCC prong, the remand directs only that services
be given opportunity to offer evidence of disruption from rates that
have now been in effect for three years without any disruption.''
Copyright Owners' Reply in Support of Motion to Adopt Interim Rates at
7-8 (Nov. 25, 2020) (emphasis added).
On December 10, 2020, the Services submitted to the Judges their
Proposal for Remand Proceedings, in which they made no argument that
the uncapped TCC rates and rate structure (or, for that matter, any
aspect of the Determination) would cause disruption or have a
disruptive impact, whether under statutory Factor D or otherwise. By
contrast, in their remand proposal, Copyright Owners reference twelve
times that, for the Judges to reject the uncapped TCC rates and
structure, the Services must show the presence of ``disruption''
arising from the Majority's uncapped TCC rates and structure. Copyright
Owners made this argument notwithstanding that the ``reasonable'' rate
standard is separate from the ``disruptive impact'' issue, which is an
itemized objective (one of four) to be considered as an adjustment to
what would otherwise constitute a ``reasonable'' rate. See Proposal of
Copyright Owners for the Conduct and Schedule of the Resolution of the
Remand at 2, 7-8, 22-24 (Dec. 10, 2020).\89\
---------------------------------------------------------------------------
\89\ When Copyright Owners do address an argument that the
Services actually made (on appeal) regarding the uncapped TCC rates
and structure, they note not that the Services had made a
``disruption'' argument, but rather that ``the Services appealed for
the reversal of the TCC prong as substantively unreasonable.'' Id.
at 22 (emphasis added). But Copyright Owners then assert, coyly,
that ``this request was not granted by the Circuit'' (citing
Johnson, 969 F.3d at 383), when in actuality, the D.C. Circuit did
not rule against the Services on this point, but rather stated only
that it was not addressing substantive arguments made by the
Services ``[b]ecause we have vacated the rate structure devised by
the [Judges] for lack of notice. . . .'' Id.
---------------------------------------------------------------------------
In the CO Initial Submission, Copyright Owners assert, without
citation to any of the Services' filings: ``The Services contend that,
had they been given such an opportunity [at the hearing], they
supposedly could have established that an ``uncapped'' TCC is
disruptive because the market for sound recordings is not effectively
competitive.'' Id. at 5. Copyright Owners further aver that the
Services must ``provide evidence, consistent with the [CRB Judges']
well-established disruption standard, that because of the labels'
supposed market power, the TCC structure adopted by the Board has
actually, substantially, immediately and irreversibly threatened the
continued viability of the interactive streaming industry'' in a manner
that will ``threaten the viability of the music delivery service
currently offered to consumes under [the] license.'' Id. at 7, 56
(citations omitted).
Copyright Owners then assert that the Services bear the burden of
proving disruption under Factor D from the
[[Page 54430]]
uncapped rates and rate structure embodied within the rate proposal
(even though only Copyright Owners are pursuing this approach on
remand). Further, Copyright Owners assert that the Services' objection
to the uncapped rates and rate structure must fail unless they can show
that such a disruptive impact occurred during the 33-month period (from
January 2018 through September 2020) when the Phonorecords III rates
were in effect. Id. at 56.
In their initial substantive remand briefing, the Services once
more did not assert that the Determination's uncapped TCC rates and
structure would cause disruption pursuant to Factor D of section
801(b)(1), or even assert a non-statutory disruption arising therefrom.
Rather, the Services directly attacked this rate approach as
inconsistent with the statutory ``reasonable'' rate requirement,
maintaining that ``[t]ying the mechanical rates directly to the
complementary oligopoly sound recording rates in the manner of the
Majority's uncapped TCC rates and rate structure is plainly
unreasonable.'' Services' Joint Opening Brief at 46 (Apr. 1, 2021)
(emphasis added). The Services also asserted that the uncapped TCC
rates and rate structure are ``unreasonable'' because they do not
promote the statutory objectives of Factor B (``fair income'' to the
copyright user) and Factor C (reflecting the copyright users' itemized
role in making the musical works ``available to the public.''). Id. at
45, 50-51, 55.\90\
---------------------------------------------------------------------------
\90\ The Services' only references to the concept of
``disruption'' relate to their argument that their own benchmark
premised on the prior Phonorecords II rate structure and rates would
not be disruptive. Id. at 4, 24, 29-30. That argument is properly
made by Services in this context, because a party seeking to
persuade the Judges to adopt its proposal bears the burden of proof,
pursuant to section 556(d) of the APA, regarding the consonance of
its proposal with all the standards contained in section 801(b)(1).
The Judges do note that one of the Services' expert witnesses,
Professor Katz, found the Majority's attempt to avoid disruption by
phasing-in the new rate provisions insufficient ``to mitigate the
risk of short-term market disruption''. That testimony does not
constitute a direct reliance by the Services on the statutory
disruption objective in Factor D, but rather emphasizes the
Majority's own concern with such disruption and the witness's
concern that the phase-in did not prevent the disruptive effect that
the Majority itself had contemplated. In any event, Professor Katz,
as an economist, cannot make a legal argument regarding the
applicability of the Factor D objective, the Services did not rely
on his testimony in that regard and, as noted, the Services made no
legal Factor D ``disruption'' argument on remand. Thus, the Judges
do not give any weight to Professor Katz's testimony in this regard.
---------------------------------------------------------------------------
In the Services' Reply, the Services attack Copyright Owners'
``singular focus on the disruptive impact of the uncapped TCC prong.''
Services' Reply at 35. In particular, the Services argue:
1. they have maintained and demonstrated that Copyright Owners'
uncapped rates and rate structure are ``unreasonable,'' separate and
apart from demonstrating that this uncapped approach also fails to
satisfy the four itemized statutory factors;
2. the burden of proof with regard to Factor D disruption lies with
Copyright Owners, because they are the ones who are advocating for the
uncapped TCC rates and rate structure;
3. the presence of Factor D disruption, vel non, is not
dispositive, because section 801(b)(1) and Johnson require the Judges
to apply the entirety of the statutory standard (which consists of the
``reasonable rate'' requirement and consideration of all four itemized
Factors; and
4. the ``full extent of the disruption to the Services from an
uncapped TCC prong was never tested in the marketplace [because] [t]he
Majority set escalating rates, and the [ ] Determination was vacated
before the significant hike in rate levels was fully implemented.''
Id. at 35-36.
In their Remand Reply, with regard to the issue of ``disruption,''
Copyright Owners assert:
1. The Services have ``completely abandoned'' their appellate
argument asserting disruption, and admit to having no evidence that the
Board's adopted rate structure has any materially disruptive impact.
Copyright Owners' Reply Brief on Remand at 5 (July 2, 2021).
2. The Services have not even attempted to show any Factor D
related effect or other disruption from the adopted rates and
structure. Id. at 15, n.9.
3. The failure of the Services to provide evidence of disruption or
to pursue the argument that disruption had occurred was inconsistent
with their prior assertions that the uncapped TCC rates and rate
structure created ``a real risk of economic harm'' and the ``impact''
or ``harm' that the uncapped approach generated. Id. at 35.
4. Each of the Services, in response to Copyright Owners' discovery
requests, acknowledges that it was not offering new evidence regarding
the ``impact'' of the Phonorecords III rates and rate structure. Id. at
36-38.
5. The Services did not merely suffer no disruption, they
experienced unprecedented growth and profit under the uncapped TCC rate
prong. Id. at 45.\91\
---------------------------------------------------------------------------
\91\ The Judges allowed the Services to make a supplemental
filing in response to Copyright Owners' remand reply, because those
papers contained direct as well as reply materials. In their
supplemental filing, the Services argued that they had not
``thrived,'' that the financial data on which Copyright Owners'
relied did not isolate revenue attributable to interactive services,
was not limited to U.S. generated revenue, and used changes in the
market capitalization of Amazon and Alphabet (Google's parent
corporation) as a proxy for the economic fortunes of their
interactive services. Services' Joint Supplemental Brief at 13-15.
As explained supra, the Judges find the permanency of the
Phonorecords III rate structure during the 33-month period from
January 2018 through September 2020 to have been in question,
pending the appeal that resulted in the vacating and remanding of
the Determination and the reversion back to the Phonorecords II
rates and rate structure. Given that uncertainty, the Judges find it
wholly inappropriate to draw any conclusions from the change or
stasis in the sound recording rates or the total royalty payments by
a Service over that period.
---------------------------------------------------------------------------
6. The Services on remand have attempted to replace their prior
``disruption'' assertion with a claim of ``unreasonableness.'' Id. at
50, n.36.
(iv) Analysis and Decision Regarding ``Disruption'' Issue
The full Factor D ``disruption'' standard, as set forth by the
Judges, states that an adjustment is warranted by Factor D if the rate
analysis made by the Judges would otherwise:
directly produce[ ] an adverse impact that is substantial, immediate
and in the short-run because there is insufficient time for either
[party] to adequately adapt to the changed circumstance produced by
the rate change and, as a consequence, such adverse impacts threaten
the viability of the music delivery service currently offered to
consumers under this license.
Determination at 87. Factor D is not applicable, particularly as
proposed by Copyright Owners. Thus, the Judges reject Copyright Owners'
assertion that the uncapped TCC prong should be adopted because of the
absence of evidence of ``disruptive impact'' proffered by the Services.
This rejection is based on several findings of fact and conclusions of
law.
First, the issue of ``disruptive impact'' pertains here to the
proposal advanced by Copyright Owners, not the Services. Thus, the
burden of proving that this uncapped TCC rate prong proposal satisfies
the elements, including Factor D, of the section 801(b)(1) standard in
a sufficient manner lies with Copyright Owners, not the Services. See 5
U.S.C. 556(d). Accordingly, the fact that the Services did not
affirmatively assert an argument of ``disruptive impact'' is of no
consequence. Moreover, as the review of the Services' filing makes
clear, the Services never abandoned that argument, because they never
made it.
[[Page 54431]]
Rather, they have consistently argued that the uncapped TCC rate prong
was unreasonable, not that it was statutorily ``disruptive'' as that
standard has been applied by the Judges.
Second, Copyright Owners did not demonstrate with sufficient
evidence or testimony that the uncapped TCC rate would be consistent
with Factor D. To be clear, by this the Judges do not mean that
Copyright Owners were obliged to prove a negative. Rather, they needed
to prove, and indeed attempted to do so, that it was unlikely that
their rates would cause a ``disruptive impact.''
In this regard, as an empirical matter, Copyright Owners proffered
the testimony of an economic expert witness, Dr. Eisenach, who opined
that the Services' [REDACTED]. Eisenach WRT ]] 12-41 ([REDACTED]) CO
Reply at 40-41. However, as the Judges discuss supra, that period
reflected ``33 months of uncertainty,'' during which no one could
predict the final mechanical rate and structure that would be adopted
by the Judges and/or the D.C. Circuit after appeals. Accordingly, that
factual evidence is unpersuasive.
Further, as a theoretical matter, Copyright Owners rely on
Professor Watt's testimony regarding the ``seesaw'' effect. In that
regard, and as discussed supra, the Majority took comfort in what it
understood to be Professor Watt's ``prediction'' that increases in
mechanical royalties would be offset almost dollar-for-dollar by
reductions in the sound recording royalty. However, as also discussed
supra, Professor Watt has now clarified on remand that he never made
such a ``prediction,'' and that his testimony regarding the so-called
``seesaw'' was limited to shifts in the share of the surplus to
Copyright Owners and from sound recording companies as a consequence of
an increase in the mechanical rate, holding all other factors unchanged
(the ceteris paribus assumption).
Moreover, Professor Watt further explained that many other factors
would likely impact the sound recording rate together with an increase
in the mechanical rate, including ``a measure of oligopolistic
interaction, different timings of different rate bargains, and the
actual values of outside options.'' Watt RWRT ] 118. Professor Watt
candidly acknowledged that he has not modeled these independent
variables, and he further notes that the data may not exist to allow
for such modeling. Id. But the inability to model the impact of
independent variables does not mean that their potential to cause
disruption can be ignored.
In particular, the purpose of the ``seesaw'' contention was that it
prevented economic harm to the Services in connection with a rise in
the mechanical rate. Although not of Professor Watt's design, that
connection is intentionally built into the Majority's uncapped TCC
rate. See Determination at 35 (``Incorporating an uncapped TCC metric
into the rate structure permits the Judges to influence that ratio
directly.'') But the ``measure of oligopolistic interaction''
referenced by Professor Watt was the very concern expressed by the
Dissent, which cautioned that there was no evidence that the sound
recording companies would be compelled to maintain the same industry
structure and accept the loss of substantial royalty income. See
Dissent at 4 (``[T]he record companies may decide to keep their rates
high despite the increase in mechanical rates, or decide it is in their
interest to avoid a reduction in royalty revenue by creating a
completely different paradigm for streaming, by which the record
companies move the streaming service in-house and effectively destroy
the existing services.'').\92\
---------------------------------------------------------------------------
\92\ The Dissent noted that this risk was speculative in nature
because there was no evidence proffered at the hearing regarding the
reactions of the sound recording companies. But no such evidence was
forthcoming in the remand proceeding either, and, as noted supra,
the burden of proof in this regard falls on Copyright owners as the
proponents of the uncapped TCC rate prong. In fact, because the
major publishers who are members of the NMPA (a constituent of
Copyright Owners) are part of the same corporate structure as the
sound recording Majors, the burden of producing evidence would fall
on Copyright Owners as well regarding the sound recording companies'
reaction to the ``seesaw'' effect.
---------------------------------------------------------------------------
Also, the ``different timings of different rate bargains,'' another
independent variable identified in Professor Watt's remand testimony,
was an issue raised to him at the hearing by Judge Strickler. Professor
Watt candidly agreed that the Judge was ``absolutely correct'' that
there is a ``risk, then, of disrupting the market by having a total
royalty that's greater than what is indicated by your Shapley
testimony, simply because of the disparity of times in which the rates
are . . . implemented.'' 3/27/17 Tr. 3091-92 (Watt) (emphasis added).
However, this admitted risk of disruption was not addressed by
sufficient record evidence.\93\
---------------------------------------------------------------------------
\93\ As noted supra, Copyright Owners did not call any sound
recording industry witnesses, or provide evidence from sound
recording companies, indicating that labels would even be amenable
to considering such renegotiated rate reductions. Instead, at the
hearing, Professor Watt merely speculated that the sound recording
companies might renegotiate their rates downward to reflect the
seesaw effect when mechanical rates increased. Tr.3/27/17 3093-94
(Watt) (``I'm not able to comment on how, you know, how possible it
is to take an agreement that's in force and then change it.''). Not
only was that mere speculation, it was provided by an economist who
is neither a music industry executive nor an attorney, and the
witness did not testify that he had spoken to anyone who would have
industry knowledge regarding whether a label would even be amenable
to considering such rate reductions.
---------------------------------------------------------------------------
Third, disruption in the narrow sense of Factor D as applied by the
Judges previously is not relevant to the present problem. An increase
in total royalties is not a short-run immediate issue, but rather an
ever-present possibility that the seesaw analysis does not sufficiently
address. Rather, the uncapped nature of the TCC rate prong renders it
unreasonable rather than narrowly disruptive.
Balancing the foregoing considerations, the Judges find that
Copyright Owners' disruption-based argument lacks merit.
6. Conclusion Regarding Uncapped TCC Rate Prong
For the foregoing reasons, the Judges decline to adopt the uncapped
TCC rate tier proposed on remand by Copyright Owners.
III. Rejection of Phonorecords II Settlement as a Benchmark
A. D.C. Circuit Ruling
Each of the Streaming Services advanced somewhat different rate
plans, but all four proffered a benchmark that ``broadly sought to
maintain the Phonorecords II rate structure,'' while lowering or
eliminating the mechanical floor.\94\ Johnson, 969 F.3d at 371. With
regard to the Services' proposed benchmark based on the Phonorecords II
rates, rate structure, and terms (hereinafter, PR II-based
benchmark),\95\ the Judges are guided by several rulings in JJohnson.
---------------------------------------------------------------------------
\94\ The ``mechanical floor'' refers to an alternative rate
calculation. ``If the All-In Rate calculation results in a dollar
royalty payment below the stated Mechanical Floor rate, then that
floor rate would bind.'' Determination at 26 n.59.
\95\ See Services' Joint Rate Proposal (in Services' Joint
Written Direct Remand Submission at Tab C) (Apr. 1, 2021). According
to the Services, their rate proposal in this proceeding is meant to
``update the Phonorecords II terms to include terms of the
Determination, as amended during the implementation of the Music
Modernization Act, that were upheld in Johnson . . . including terms
relating to student and family plan products, or that were not
challenged by either the Copyright Owners or the Services.'' Id. at
2. The Services include in their Joint Rate Proposal a chart
summarizing the proposed rates for their offerings. That chart is
attached as an Addendum to this Initial Ruling.
---------------------------------------------------------------------------
In particular, the D.C. Circuit found the Judges' treatment of the
PR II-based benchmark to be ``muddled.'' Johnson, 969 F.3d at 387. The
D.C. Circuit emphasized that the Judges ``failed to
[[Page 54432]]
explain'' their rejection of the PR II-based benchmark. Id. at 367. See
also id. at 376 (Judges ``failed to ``reasonably explain'' rejection).
In the appeal, Copyright Owners attempted to defend the Judges'
reliance on the absence of evidence of the settling parties' subjective
intent in reaching the Phonorecords II terms. Id. at 387. The D.C.
Circuit dismissed Copyright Owners' post hoc attempt, noting that
``nowhere does the [ ] Determination explain why evidence of the
parties' subjective intent in negotiating the Phonorecords II
settlement is a prerequisite to its adoption as a benchmark.'' Id. at
387 (emphasis added).
The D.C. Circuit also criticized the attempt by the Judges'
appellate counsel to ``change tack'' and argue that their rejection of
the PR II-based benchmark was reasonable because: (1) evidence showed
that the prior rates had been set far ``too low'' and (2) it was
``outdated''. The D.C. Circuit found that those arguments also were
``nowhere to be found in the [ ] Determination's discussion'' of the
appropriateness of the Phonorecords II settlement as a potential
benchmark. Id. at 387 (emphasis added).\96\ In the end, the D.C.
Circuit agreed with the Streaming Services that, inter alia, the Judges
failed to reasonably explain their rejection of the benchmark and, for
all of the reasons cited, vacated and remanded the adopted rate
structure and percentages for further proceedings. Id. at 381.
---------------------------------------------------------------------------
\96\ In the present remand ruling, the Judges do not rely on
their appellate counsel's ad hoc arguments that the D.C. Circuit
found to be absent from the Determination. The Judges note though
(as discussed in more detail infra) that in this Initial Ruling they
are increasing the 10.5% royalty rate in the Phonorecords II rates
by 44% to 15.1% (as phased-in by the Determination), thus addressing
appellate counsel's ad hoc assertion that the Phonorecords II rates
were ``too low.'' Similarly, as discussed infra, the Judges address
the notion that the PR II-based benchmark is outdated.
---------------------------------------------------------------------------
B. Remand Procedure Regarding the PR II-Based Benchmark
On December 15, 2020, subsequent to the D.C. Circuit's decision,
the Judges entered an Order Regarding Proceedings on Remand, in which
the Judges stated:
The Judges accept the parties' proposals to resolve the issues
concerning the use of the Phonorecords II settlement as a benchmark.
. . .
. . .
The Services and Copyright Owners also agree that the Judges
should resolve this issue based on the existing record, after
receiving two rounds of additional briefing from the parties.
Remand Order at 1-2.
Based on the ruling in Johnson the Judges reject Copyright Owners'
position that they need not engage in a full analysis of the issue. The
Judges conclude that they must engage in, and fully articulate, a
reasoned analysis that adequately addresses ``the issues concerning the
use of the Phonorecords II settlement as a benchmark.'' Id. (emphasis
added). If the Judges determine that the Majority properly rejected the
Services' proposed use of the PR II-based benchmark, the rejected
portions will play no part in the Judges' remand ruling. On the other
hand, if the Judges find, after engaging in that analysis, that the PR
II-based benchmark was not properly rejected then, as a matter of law
and logic, the Judges must weigh the Services' PR II-based benchmark
for application, in whole or in part.
The Judges reject Copyright Owners' reading of Johnson as holding
that the Judges cannot fully consider the PR II-based benchmark on
remand. Copyright Owners argue that the D.C. Circuit ``did not suggest
the [Judges] substantively erred'' in rejecting that benchmark, or that
they ``needed to reconsider [their] decision,'' but had ``merely
remanded for a `reasoned analysis' . . . as to why it did so.'' CO
Initial Submission at 10; see also Copyright Owners' Reply Remand Brief
at 7-8. Because Johnson ruled that the Majority's reasoning was
muddled, indiscernible, unexplained and lacking in reason, the D.C.
Circuit obviously neither accepted nor rejected the Majority's
disregard for the PR II-based benchmark--thus requiring the CRB Judges
to take a comprehensive look at that benchmark. In this regard, the
Judges agree with the Services that, pursuant to apposite case law, if
the outcome of the remand as to this issue was preordained pending the
further ``reasoned analysis,'' the D.C. Circuit would have expressed a
desire simply to remand without vacating as to this issue. Services'
Joint Remand Reply Brief at 7-8 (citing Allied-Signal, Inc. v. NRC, 988
F.2d 146, 150-51 (D.C. Cir. 1993) (``The decision whether to vacate
depends on the seriousness of the order's deficiencies (and thus the
extent of doubt whether the agency chose correctly) and the disruptive
consequences of an interim change that may itself be changed.'')).\97\
---------------------------------------------------------------------------
\97\ However, the Judges note that section 803(d)(3) may require
the D.C. Circuit to remand rather than reverse when the issue
concerns more than rates alone. Thus, the statute appears to require
a remand in order for the Judges to apply their statutory authority
and expertise in toto.
---------------------------------------------------------------------------
Because Johnson held that the Majority's reasoning was muddled,
indiscernible, unexplained, and lacking in reason, the D.C. Circuit
obviously neither accepted nor rejected the Majority's disregard for
the PR II-based benchmark. Thus, the Judges take a comprehensive look
at that benchmark's rates and rate structure to evaluate its usefulness
in this proceeding.
Relatedly, the Judges also reject Copyright Owners' assertion that
the Judges can only consider on remand the Phonorecords II rates, and
cannot consider on remand the relative strengths and weaknesses of the
structure in which those rates are embedded. See Copyright Owners'
Reply Brief on Remand at 14. This distinction is impractical and
unworkable. If the (non-``headline'' rates \98\) themselves can be
reviewed and found acceptable (as they are infra) into what structure
would they be placed? There are multiple provisions in the Phonorecords
II rate structure providing for different rates, designed to balance
(1) the ability of services to attract consumers with a low
Willingness-to-Pay and/or a low Ability-to-Pay (the price
discriminatory and differentiated features \99\) with (2) the revenue
diminution protections for which Copyright Owners had successfully
negotiated. Moreover, the D.C. Circuit has vacated the Determination,
and in doing so did not make any rulings critical of the rate structure
in the Phonorecords II-based benchmark that would suggest the cramped
review advocated by Copyright Owners. Indeed, the D.C. Circuit
explicitly stated, without distinguishing between rates and structure,
that it ``agree[s] with the Streaming Services that the [Judges] . . .
failed to reasonably explain [their] rejection of the Phonorecords II
settlement as a benchmark . . .'' See Johnson, 969 F.3d at 376; see
also id. at 389 (issues relating to ``rates'' and ``rate structure''
are ``intertwined'').
---------------------------------------------------------------------------
\98\ As explained elsewhere in this Initial Ruling, the Judges
are increasing the ``headline'' rate from 10.5% to 15.1%.
\99\ Specifically, the PR II-based benchmark would incorporate
the price discriminatory features for product differentiation as
between: (1) subscription vs. ad-supported services; (2) portable
and non-portable services; and (3) unbundled vs. bundled services.
See Determination at 10; Dissent at 26. The third category--bundled
vs. unbundled--is discussed infra in the context of the Bundled
Revenue definition.
---------------------------------------------------------------------------
Further, the Judges emphasize that the rate structure of the PR II-
based benchmark provides protection sought by Copyright Owners against
revenue diminution by the Services--protection they would otherwise
lose--because in this Initial Ruling the Judges are not adopting the
vacated uncapped TCC prong for which Copyright Owners are now
advocating, and which they claim
[[Page 54433]]
would have protected them in that regard. Cf. CO Additional Submission
at 4-6 (acknowledging PR II-based benchmark provided some TCC
provisions, allowing for protection against revenue diminution). Thus,
the Judges' remand rulings on the PR II-based benchmark rates and on
the uncapped TCC rate prong are inextricably interlaced. See Johnson,
969 F.3d at 381 (absence of ``reasoned explanation'' for rejecting PR
II-based benchmark was problematic because it occurred ``when'' Judges
adopted an alternative proposal that called for ``setting . . . total
content cost and revenue rates.'') (emphasis added).
The Judges weigh each benchmark's intrinsic strengths and
weaknesses, as well as its comparative advantages and disadvantages
vis-[agrave]-vis other proffered benchmarks. On remand, the
interrelationships of the competing benchmarks are of particular
importance, given Copyright Owners' need for the aforementioned
protections against revenue diminution via price discrimination.\100\
---------------------------------------------------------------------------
\100\ The Judges categorically reject Copyright Owners'
assertion that the PR II-based benchmark cannot be considered
because the parties agreed in the Phonorecords II settlement that
any future statutory mechanical rate determination would made ``de
novo'' vis-[agrave]-vis that settlement determination. In fact, the
industrywide representatives (NMPA and Digital Media Association
(DiMA)) who entered into the settlement conspicuously did not agree
that the existing rate structure or rates could not be considered as
the bases for future rate determinations. By contrast, the
Phonorecords I settlement agreement expressly stated ``[s]uch
royalty rates shall not be cited, relied upon, or proffered as
evidence or otherwise used in the [Phonorecords II] Proceeding.''
Trial Ex.6013, Phonorecords I Agreement at sec. 3. Compare Trial Ex.
6014, Phonorecords II Agreement at sec. 5.5 (omitting clause
precluding reliance on evidentiary value of Phonorecords II royalty
rates and including full-integration clause). This change
objectively demonstrates that the parties to the 2012 settlement
understood the evidentiary value of the Phonorecords II settlement
in the next section 115 proceeding, i.e., this proceeding. See
Dissent at 15-16.
On the other hand, the Judges reject the Services' argument that
the Phonorecords II rates and structure should be retained merely
because the Services relied on their continuation to make
investments in their business models. As Copyright Owners note, the
applicable regulations provide that ``[i]n any future proceedings
the royalty rates payable for a compulsory license shall be
established de novo.'' 37 CFR 385.17; see also 37 CFR 385.26. A
party may feel confident that past is prologue and that the parties
will agree to roll-over the extant rates for another period; a party
could be sanguine as to its ability to make persuasive arguments as
to why the rates should remain unchanged; a party might even
conclude that the mechanical rate is such a small proportion of the
total royalty obligation that its increase would be unlikely to
alter long-term business plans. But for sophisticated commercial
entities to claim that they simply assumed the rates would roll-
over--without the reasonable possibility of significant adjustment
or outright abandonment--strikes the Judges as so irrational and
reckless as to raise serious doubts about the credibility of that
position. (If the Services had made a persuasive argument that
certain fixed cost investments were ``sunk'' and had useful lives
that substantially exceeded the five-year rate term, then such costs
could be considered under Factor C of section 801(b)(1), but they
did not make a persuasive argument in this regard. Cf. SDARS II, 78
FR 23054, 23069 (Apr. 17, 2013) (adjusting rates downward under
Factor C, and distinguishing internet music transmissions, to
reflect that--because Sirius XM needed to make ``unique and
substantial'' investments in the form of ``sunk'' costs paid for
satellites with a useful life of l2-15 years--``it is not
unreasonable for Sirius XM to expect to recoup a certain amount of
those costs over the expected useful life of the [s]atellites,''
which exceeded the five-year rate term.)
---------------------------------------------------------------------------
Through this approach, the Judges ultimately may adopt only one of
the parties' benchmarks or other methodologies, or they may modify the
proposals by combining them, provided such a modification is ``within a
reasonable range of contemplated outcomes . . . piecing together a rate
structure, the economic and policy consequences of which had already
been explored and developed by the parties in the record.'' Johnson,
369 F.3d at 382.
In their consideration of the PR II-based benchmark, the Judges are
not suggesting that this benchmark is the optimal tool to use in order
to identify rates and terms among all approaches that might have been
proffered (but were not). But the Judges are cabined by the evidence
they receive. See 17 U.S.C. 803(a)(1) (``the Judges shall act . . . on
the basis of a written record . . . .''); see also P. Wald, supra,
(noting that parties' economic proposals made in an action ``impose[ ]
a practical constraint'' on judge who will, ``for the most part, be
limited by what the parties serve up to her.''). Based upon the
available record evidence, the Judges find that the Services' PR II-
based benchmark--although not necessarily perfect--is more than
sufficient to satisfy the legal requisites for application, as well as
a practical benchmark, when used in conjunction with the 15.1% headline
revenue rate advocated by Copyright Owners. See generally Nat'l Cable
Television Ass'n v. Copyright Royalty Tribunal, 724 F.2d 176, 182 (D.C.
Cir. 1983) (rate-setting is an intensely practical affair).
C. Parties' Remand Arguments Regarding PR II-Based Benchmark
101
---------------------------------------------------------------------------
\101\ The parties made arguments both in the original hearing
and in this remand proceeding regarding the Services' proffer of the
PR II-based benchmark. Each party's pre-remand and post-remand
arguments overlap to some extent. Examination of the pre-remand
arguments is also necessary because of the findings in Johnson and
because the parties agreed that the evidentiary record on this
remanded issue would not be enlarged.
---------------------------------------------------------------------------
1. Services' Arguments
The Services maintain that their PR II-based benchmark satisfies
the ``reasonable'' rate requirement and is consistent with the four
itemized factors set forth in section 801(b)(1). They make several
arguments in favor of this position.
First, they aver that their PR II-based benchmark possesses all the
characteristics of an ``ideal'' benchmark. Services' Joint Opening
Brief at 19. In this regard, they argue that their proffered benchmark
``involves the same sellers, the same or similar buyers, and the same
rights as at issue in this proceeding,'' and that there has been ``no
material change in the economic circumstances of the marketplace that
would warrant adjusting the rate levels or rate structure in the
benchmark.'' Id. at 20.
Applying the facts to these benchmark characteristics, the Services
assert that the first three elements--same sellers (here, licensors),
same buyers (here, licensees) and same rights (the mechanical license
for interactive streaming) are satisfied. In particular, they note that
the majority of the participants in the present proceeding either
directly participated in the Phonorecords II settlement process or were
active in the market contemporaneous with that settlement. Id. at 20-
21.
Turning to the next benchmark characteristic--the absence of a
``material change in the economic circumstances of the marketplace that
would warrant adjusting the rate levels or rate structure in the
benchmark''--they emphasize that the PR II-based benchmark contains
different rate levels for different product offerings, to account for
(a) consumers' varying willingness-to-pay (WTP) and (b) the zero
marginal physical cost of digital reproductions of sound recordings
containing musical works. Id. at 21-22 (citing multiple experts).
Next, the Services point to the fact that the ``headline'' \102\
royalty rate is based on a percent-of-revenue, so that revenue growth
(or decline) on this rate prong allows for royalty payments to directly
adjust in tandem. Id. Further, the Services assert that the importance
of streaming as ``the future of the music industry'' was known to the
Phonorecords II negotiators, as evidence by the then-recent launch in
the United
[[Page 54434]]
States of the popular Spotify service. Id. at 23.
---------------------------------------------------------------------------
\102\ The Judges and the parties characterize the percent-of-
revenue of revenue rate as the ``headline'' rate. See Johnson, 969
F.3d at 383 n.10.
---------------------------------------------------------------------------
Beyond these benchmark requisites, the Services also emphasize that
the PR II-based benchmark is the product of a settlement whose
negotiated features burnish the value of this benchmark as reflective
of effective competition. Specifically, they note:
The settlement was negotiated in the same statutory
context, concerning the identical rate standard and factors as
applicable to the present proceeding.
Neither side would have accepted a deal materially worse
than what it expected from a section 115 proceeding applying the
section 801(b)(1) considerations.
The statutory alternative diminishes any additional
licensor-side negotiating power arising from ``Must-Have''
complementary oligopoly of the licensors of the musical works
publishers.
Id. at 22. Moving from the negotiating context to market performance
under this standard, the Services aver that this approach has borne
fruit for the industry as a whole. They point to the evidence of the
licensors' consistent profitability and the licensees' ability to
``benefit'' from the Phonorecords II approach. Id. at 23.
The Services also maintain that the Phonorecords II structure
``addresses any concerns with bundling and the potential for revenue
deferment.'' Id. at 24.\103\ They assert that these issues were
specifically addressed by Copyright Owners during the Phonorecords II
negotiation, because ``multiproduct firms such as Yahoo and Microsoft''
that offered streaming services had the capacity to make bundled
offerings to consumers. These concerns were addressed in the
Phonorecords II rate structure, the Services note, through the use of
``multiple rate prongs, minima and floors,'' ensuring that ``the total
musical works royalty for certain types of offerings does not fall
below a specified level,'' thereby ``mitigat[ing] the effect of any
potential revenue deferrals and appropriately address[ing] any concerns
with bundling.'' Id.\104\
---------------------------------------------------------------------------
\103\ The issue of bundling is addressed in this Initial Ruling
infra, in connection with the Judges' definition of Service Revenue
generated through the offering of sound recordings as part of a
bundle containing other goods or services.
\104\ The Services also reiterate their pre-remand argument that
the Phonorecords III settlement of subpart A rates for sales of
physical and digital download phonorecords (now reorganized in
subpart B) confirms the appropriateness of the Phonorecords II-based
benchmark. However, any further reliance by the Services on that
argument is moot, because the D.C. Circuit affirmed the Majority's
analysis of the subpart A rates. Johnson, 969 F.3d at 386 (noting
that the Majority adequately explained treatment of the subpart A
rates as `` `at best' a floor'' below which the mechanical royalty
rates paid by the Services for interactive streaming cold not fall).
---------------------------------------------------------------------------
Finally, the Services maintain that ``[d]irect agreements between
Copyright Owners and Services also support adoption of the PR II-based
benchmark.'' Id. at 34. In particular, they note that many of the
royalty rates (and terms) in these direct agreements apply the
Phonorecords II rates. Moreover, the Services maintain, because these
direct agreements are in the nature of blanket license of a publisher's
entire catalog, they provide an added ``access'' value in the form of
full-repertoire licensing. These direct agreements do not include a
rate above Phonorecords II levels; thus, the Services contend, they
underscore the reasonableness of the Phonorecords II rates. Id. \105\
---------------------------------------------------------------------------
\105\ Under section 115--prior to the effective date of the 2008
Music Modernization Act--an interactive service was required to
serve a ``Notice of Intent'' to use the copyright license (NOI) with
the owner of a copyright for each musical work before streaming the
sound recording embodying that musical work. By contrast, a direct
license with a publisher covers more than an individual musical work
by providing ``access'' value to an entire catalog, without the
transaction cost burden of filing multiple individual NOIs.
---------------------------------------------------------------------------
Finally, the Services aver that the PR II-based benchmark satisfies
the itemized four section 801(b)(1) factors. With regard to Factor A,
they maintain that: (1) the Phonorecords II framework has corresponded
with an increase in the supply of musical works; (2) the PR II-based
benchmark will increase the likelihood that the Services will increase
subscriber counts, generating profitability, which will make streaming
available to more listeners; and (3) the price discriminatory aspects
of this royalty rate structure allows the Services to afford to offer
streamed music to listeners with a low willingness (or ability)-to-pay,
at lower rates or through ad-supported services. Services' Joint
Opening Brief at 25-27.
Regarding Factors B and C (the ``fair return'' and ``relative
contributions'' objectives), the Services emphasize that the PR II-
based benchmark satisfies these statutory elements because it: (1) was
the result of negotiations between industrywide representatives who had
every incentive to obtain a ``fair'' return and to receive recompense
for their ``contributions'' to streaming; and (2) allowed interactive
streaming to become ``a significant means for consumers to listen to
music'' while simultaneously generating growth in annual royalties for
Copyright Owners.'' Id. at 27-29.
Lastly on the subject of the statutory factors--regarding Factor D
(minimizing disruptive impact)--the Services make a succinct argument:
``By renewing the rate levels and structure of Phonorecords II, there
is minimal risk of disruption.'' Id. at 29-30.
The Services also address several further criticisms of the PR II-
based benchmark contained in the Determination. Focusing first on an
issue specifically addressed in Johnson, they assert the irrelevancy of
the ``subjective intent'' of the parties that negotiated the
Phonorecords II settlement--a factor on which the Majority relied in
deciding not to adopt the PR II-based benchmark. In this regard, the
Services are also responding to the D.C. Circuit's concern regarding
this issue. See Johnson, 969 F.3d at 387 (``In rejecting that
settlement as a possible benchmark, the [Judges] faulted the Streaming
Services for failing to explain why the parties to the Phonorecords II
settlement agreed to the rates in that settlement . . . [b]ut nowhere
does the [ ] Determination explain why evidence of the parties'
subjective intent in negotiating the Phonorecords II settlement is a
prerequisite to its adoption as a benchmark.'').
The Services note that no benchmark evidence presented by any party
is proffered with supporting evidence of the subjective intent of the
bargainers who negotiated the benchmark. Moreover, they note that the
Majority in fact acknowledged that ``[r]elying on a benchmark as
objectively useful without [the need for] further inspection'' is
``typical and appropriate for the benchmarking method.'' Id. at 35
(quoting Determination at [55] & n.106 (emphasis added)).
With regard to other criticisms of the Majority's failure to use
the PR II-based benchmark, the Services argue that the Majority
misapplied their previous rulings that they ``cannot and will not set
rates to protect any particular streaming service business model.'' Id.
at 37 (quoting Phonorecords III, 84 FR 1945). The Services find this
principle inapposite, because their point is that the multiple price-
discriminatory aspects of the Phonorecords II approach made it ``a
valuable benchmark . . . because it had allowed for different service
types to emerge and grow, which benefits the entire market.'' Id. at
37. The Services also take issue with the Majority's assertion that the
Phonorecords II rate structure was too complex, deriding it as a
``Rube-Goldberg-esque'' contraption. Id. at 38. Rather, the Services
maintain that the structure was as complex as necessary to effectuate
the parties' needs, particularly the price discriminatory features and
the protections against
[[Page 54435]]
revenue diminution. Id. at 38-39. Further, the Services note that the
record is devoid of any testimony or evidence indicating any actual
confusion caused by the Phonorecords II rate structure. Id. at 39.
Finally in this regard, the Services maintain that the rate structure
adopted by the Majority is essentially as complex as the structure in
Phonorecords II, with the only major change being the replacement of
the capped TCC rates with uncapped TCC rates.\106\ Id.
---------------------------------------------------------------------------
\106\ As discussed infra, the relative complexity or simplicity
of the rate structure is not a statutory factor, nor is it a
decisive element of a reasonable rate structure, when the details of
that structure effectuate price discriminatory configurations that
would increase the availability of music and streaming revenues and
otherwise satisfy the statutory criteria.
---------------------------------------------------------------------------
The Services address another criticism--that the rates in the PR
II-based benchmark are too low. This issue is largely moot, as the D.C.
Circuit's affirmance of the Majority's expert ``line-drawing'' and
``reasoned weighing of the evidence'' confirmed that a rate increase
was necessary. In this Initial Ruling, the Judges have acknowledged
specifically the appropriateness of the 15.1% revenue rate--a 44%
increase over the 10.5% headline rate in the PR II-based
benchmark.\107\
---------------------------------------------------------------------------
\107\ The Judges characterize this issue as largely moot because
the PR II-based benchmark includes on its ``lesser of'' prongs price
discriminatory rates, discussed infra. But those ``lesser of'' rates
are overridden by the ``greater'' 15.1% rate. As also discussed
infra, Mechanical Floors continue to bind at lower mechanical
royalty levels (without reducing the songwriters' ``All-In'' musical
works royalty that includes the performance royalties), because
these floors were retained in the Determination and were not the
subject of appeal.
---------------------------------------------------------------------------
2. Copyright Owners' Arguments
Copyright Owners assert that the record evidence overwhelmingly
supports the Judges' rejection of the PR II-based benchmark. At the
outset, they maintain that the Judges found--and the D.C. Circuit
affirmed--that a rate increase was required in the Phonorecords III
terms. CO Initial Submission at 13. (As noted, an increase in the
headline rate by 44%, to 15.1%, is adopted in this Initial Ruling.)
Next, Copyright Owners maintain that the evidence established that
``market conditions'' were ``radically different'' at the time of the
Phonorecords III proceeding compared with when the parties entered into
their 2012 industrywide agreement in Phonorecords II. Id. at 17. In
particular, Copyright Owners point to testimony describing the
streaming industry as ``nascent'' in 2012, with fewer streams,
subscribers, services, and choices of music; operating in a consumer
environment when download purchases and Pandora's noninteractive
service were the predominant means for consumers to listen digitally to
music. Id. at 18-21. In sum, Copyright Owners maintain, that streaming
was ``economically insignificant'' to the music industry when the PR II
provisions were adopted. Id. at 20.
Copyright Owners particularly emphasize the substantial increase in
streaming revenue during the Phonorecords II period. They point out
that while ``total streaming revenue had ranged from approximately $150
million in 2005 to $212 million in 2010, . . . after 2012[,] annual
[streaming] revenue exploded to reach approximately $1.6 billion by
2015.'' Id. at 23. Further, they note there is no evidence that the
music publishers or anyone else had predicted this substantial rise in
streaming and the revenues it generated, and that in no way could it be
inferred that those rates had ``baked-in'' future growth. In fact,
Copyright Owners assert at the hearing that the PR II rates were merely
``experimental''--consistent with the relatively nascent stage of the
streaming industry. Id. at 25.
Additionally, Copyright Owners maintain that the identities of the
parties involved in the Phonorecords III proceeding are different from
those who established the Phonorecords II framework. Although they
acknowledge the presence of current interactive services Spotify and
Rhapsody in this market prior to the Phonorecords II framework agreed
to by the trade associations for the interactive services and the music
publishers, they point out that ``[n]one of the other participants in
this proceeding even entered the streaming business until after the
Phonorecords II settlement.'' Id. at 21.
Next, Copyright Owners assert that the Services' evidence is
inadequate to support a finding that the rates in their PR II-based
benchmark are suitable for use in setting royalty rates in this
proceeding. First, they echo the Determination, which stated that the
Services (1) did not examine in detail the particular rates within the
existing rate structure; (2) relied on the 2012 rates as objectively
useful without further inspection; and (3) did not call witnesses to
testify regarding the 2012 settlement negotiations. Id. at 27 (citing
Determination, 84 FR 1944 & n.106). Because of the absence of the
foregoing evidence, Copyright Owners assert that the Services were left
with ``no evidence explaining how the particular rates and percentages
in those settlements were calculated or derived, how they were
negotiated, or how they were reasonable in light of the explosive
growth in the streaming marketplace between the time of those
settlements and the Phonorecords III proceeding.'' Id. at 28. The
absence of such evidence, according to Copyright Owners, meant that the
Services had failed to carry their burden of proof under 5 U.S.C.
556(d) with respect to their proposal, a burden Copyright Owners assert
the Services acknowledged they bore. Id. at 29-30.
Additionally, Copyright Owners claim that the D.C. Circuit found
``validity'' in Copyright Owners' assertion that the subjective intent
of the parties to the Phonorecords II settlement is relevant because it
would have revealed whether the agreed-upon rates were based on
economic realities or instead were driven by other considerations. Id.
at 30-31 (citing Johnson, 969 F.3d at 387). However, Copyright Owners
acknowledge that, because this was not a reason given by the Majority,
it carried no weight with the D.C. Circuit on appeal. Id. at 31.
3. Analysis and Decision Regarding PR II-Based Benchmark \108\
---------------------------------------------------------------------------
\108\ The setting of statutory royalty rates involves to a
significant degree the application of economic analysis.
Accordingly, the Judges find it appropriate to set forth certain key
aspects of microeconomics that guide the application of the section
801(b)(1) standard in the present proceeding. That guidance is set
forth more fully in the Dissent at 29-39.
---------------------------------------------------------------------------
a. PR II-Based Benchmark Meets Most of the Requisites for a Useful
Benchmark
The four classic characteristics of an appropriate benchmark are:
(1) the degree of comparability of the negotiating parties to the
parties contending in the rate proceeding,
(2) the comparability of the rights in question,
(3) the similarity of the economic circumstances affecting the
earlier negotiators and the current litigants, and
(4) the degree to which the assertedly analogous market under
examination reflects an adequate degree of competition to justify
reliance on agreements that it has spawned.
In re Pandora Media, 6 F.Supp.3d 317, 354 (S.D.N.Y. 2014, aff'd sub
nom Pandora Media Inc. v. ASCAP, 785 F.3d 73 (2d. Cir. 2015). As
discussed below, the PR II-based benchmark meets criteria (1), (2) and
(4), but requires adjustment to fully satisfy criterion (3).
First, the PR II-based benchmark obviously pertains to the same
rights at issue in this proceeding, as it reflects the licensing
provisions from the immediately preceding mechanical license
proceeding.
Second, the licensors (songwriters and music publishers) and
licensees (interactive streaming services) are
[[Page 54436]]
comparable (albeit not identical). While Copyright Owners emphasize the
different identities and market involvement of the licensees,
particularly the greater market penetration of Amazon, Apple, and
Google, the Services note that even prior to the more significant entry
of these three entities, similar multiproduct firms, such as Yahoo and
Microsoft, were active licensees. The Judges find that the changing
identities of the large multiproduct technology firms does not
demonstrate the absence of comparability between and among such firms
in the Phonorecords II and Phonorecords III rate periods. The shifting
market entries, exits, strategies, successes and setbacks of otherwise
comparable firms are expected occurrences in a dynamic capitalist
market system and are not factors that materially diminish the
necessary comparability of the parties for benchmarking purposes.
Third, important economic fundamentals of the marketplace are
sufficiently similar in crucial respects. First, the heterogeneity of
the willingness-to-pay among subscribers and listeners in the
downstream market continues to support price discrimination and thus
differentiated royalty rates upstream pursuant to the concept of
``derived demand.'' See Determination at 19 (and record citations
therein) (``Weighing all the evidence and based on the reasoning in
this Determination, the Judges conclude that a flexible, revenue-based
rate structure is the most efficient means of facilitating beneficial
price discrimination in the downstream market.''); Dissent at 32, 51,
86, 121, 126 (and record citations therein).\109\ Second, the items
being licensed for transmission--``second copies'' of sound recordings
(with embedded musical works)--have a marginal physical cost of zero, a
critical economic point on which the experts for both parties concur,
and as to which the Majority and the Dissent repeatedly and
significantly rely. See Determination at 18, 21, 36, 59, 80 (and record
citations therein); Dissent at 30-31, 33-34, 37, 47, 49-50, 59, 122,
127-128 (and record citations therein).\110\
---------------------------------------------------------------------------
\109\ The Determination asserts that it includes a price
discriminatory feature because a revenue percentage-based rate is
itself price discriminatory, in that it does not set royalties on a
per-play basis. Determination at 35 n.71. But that ``blunt'' form of
price discrimination does not capture the granular discriminatory
features that the parties had negotiated. There is no sufficient
basis for the Judges to substitute their own blunt conception of the
appropriate form and extent of price discrimination for the
structure generated in negotiations by the market participants. See
Dissent at 37.
\110\ It bears emphasis that the fact ``second copy''
reproductions are physically costless does not even suggest that the
market price should be zero. Rather, in this ``second-best''
economic context, pricing above marginal physical costs is
imperative in order for Copyright Owners to recover their ``first
copy'' costs, avoid ``opportunity costs,'' and earn profits. See
Dissent at 36-38.
---------------------------------------------------------------------------
Copyright Owners are clearly correct, however, in noting a
substantial change in economic circumstances that distinguished the
Phonorecords II negotiations from the current proceeding; viz., the
dramatic growth of interactive streaming revenues.\111\ The economic
impact of this revenue growth is incorporated into the experts' Shapley
Value Models and the Judges' analysis of same. This analysis has
generated the 44% increase in the headline royalty rate, from 10.5% to
15.1% (as phased-in by the Majority and again in this Initial
Ruling).\112\
---------------------------------------------------------------------------
\111\ Copyright Owners also cite data demonstrating the increase
in listeners and the number of streams. The Judges find those data
to be causal for the key point in rate setting in this proceeding--
the significant increase in revenues.
\112\ At first blush it may seem that the increase in
interactive revenues is not an economic fundament that would support
an increase in a percentage-of-revenue based royalty formula.
However, as more fully discussed herein, under the Shapley Value
approach, the increase in revenues has generated an increased
``Shapley Surplus'' (roughly analogous to interactive streaming
industry profits), which the two ``Must Have'' input suppliers
(record companies and Copyright Owners) will essentially split
equally. If this surplus increases faster than the interactive
services' non-content costs (or if those costs remain stable or
fall), the increased revenues would flow disproportionately to
theses input suppliers, thus causing the increase in revenues to
support an increase in the royalty rate, all other things held
constant. And, because the ``Must Have'' input suppliers have
complementary oligopoly power, the Majority relied on a Shapley
model constructed by Spotify's expert, Professor Marx, that adjusted
for this market power.
---------------------------------------------------------------------------
Simply put, three economic principles co-exist. First, the
downstream interactive streaming market remains differentiated among
listeners with different willingnesses and abilities to pay, based on
varied preferences (utility) and disparities in income. Second,
streaming of the ``second copy'' of the sound recordings (with embedded
musical works) remains physically costless (but generates potential
``opportunity costs''). But, third, streaming revenues have grown
substantially. There is no incompatibility or inconsistency in the
simultaneity of these economic principles. Each of them must be taken
into account and they are in this Initial Ruling.
This economic context refutes the arguments made during oral
argument at the D.C. Circuit that the PR II-based benchmark should be
rejected in toto because it was supposedly ``outdated.'' The
heterogeneity of the downstream demand of listeners and the zero
physical cost of ``second copies'' are enduring features that affect
the upstream market via the principle of derived demand. The
substantial growth of streaming revenues, however, necessitated an
increase in the headline rate from 10.5% to 15.1% (as phased-in), for
the reasons discussed in the Judges' analysis in this Initial Ruling of
the interrelationship among: (1) Shapley Value modeling; (2) Nash
Bargaining; (3) complementary oligopoly power; and (4) effective
competition.
Further, the foregoing analysis also undermines the pre-remand
argument made by Copyright Owners that the PR II-based benchmark
reflects a market that was not yet ``mature,'' or was only
``experimental.'' Markets are not ``mature'' as opposed to, say,
``adolescent.'' Indeed, the metaphor is strained because all economic
models are subject to revision if the salient facts have changed,
without rendering the prior models mere ``experiments.'' Markets
simultaneously exhibit enduring characteristics--here, heterogeneous
customers and zero marginal physical costs and dynamic change--here,
significant revenue increases.\113\
---------------------------------------------------------------------------
\113\ If one were to indulge the ``maturity'' metaphor, the
ongoing creative destruction in the streaming industry has only
reinforced the fact that, according to one of Copyright Owners' own
economic expert witnesses, the interactive streaming market (as of
the Phonorecords III hearing) was not yet mature, but rather
remained ``a relatively new enterprise.'' Watt WRT ]] 39-40. Thus,
it is hardly clear from the record that interactive streaming has
``matured'' in a manner that would render anachronistic the enduring
marketplace characteristics.
---------------------------------------------------------------------------
And yet, Copyright Owners seek to deny the idea that these
principles could exist simultaneously. In an attempt to disqualify the
application of the PR II-based benchmark, Copyright Owners complain:
[W]hile streaming activity and revenues grew under the
Phonorecords II royalty rates, the [REDACTED]. For example . . .
[REDACTED].
CO Initial Submission at 15-16 (emphasis added).
But as the Services explained, the economic defect in Copyright
Owners' analysis, is that it ignores the principle of price
discrimination and its beneficial effects:
[A]s [Professor] Hubbard explained, it is ``meaningless'' to
compare growth in streams to growth in royalties in the context of
Prime Music in particular because the record showed that Prime Music
brings ``new people into the market.'' . . . If not for the
flexibility (and beneficial price discrimination) the
[[Page 54437]]
existing Service Provider Revenue definition and rate structure
facilitated, the Copyright Owners ``would have gotten zero'' from
those new listeners. . . . ``So they're better off by that amount''
of royalty growth. . . . The undisputed fact that [REDACTED]--
reflects that the existing rule enables beneficial price
discrimination that expands the total royalty pool and benefits
Copyright Owners.
Services' Reply at 58-59.
This rebuttal by Professor Hubbard is an example of the important
distinction between ``increases in demand'' (when the demand curve
shifts outward) and movements ``down the demand curve'' (when sellers
use price discrimination to generate more revenue without additional
cost to attract buyers with a lower willingness or ability to pay). The
parties' otherwise dueling economists agreed on this point. Compare 4/
3/17 Tr. 4373-74 (Rysman) (Copyright Owners' witness acknowledging that
under the current rate regime overall revenues might be increasing
because of movements ``down the demand curve'' (i.e., changes in
quantity demanded in response to lower prices), rather than because of,
or in addition to, an outward shift of the demand curve (i.e., increase
in demand at every price)) with 3/13/17 Tr. 701 (Katz) (the Services'
witness who likewise noted that the present structure enhances variable
pricing that allows streaming services ``to work[ ][their]way down the
demand curve.'').
Moreover, Copyright Owners baldly cherry-pick the data they
present. [REDACTED] CO Initial Submission at 15-16. So, by their own
data, presented in their own brief, they acknowledge that [REDACTED].
See Services' Reply at 57-58 (Copyright Owners have proven the
``opposite'' of what they intended). This is precisely what beneficial
price discrimination is designed to accomplish.\114\
---------------------------------------------------------------------------
\114\ Further, [REDACTED] because: (1) the marginal physical
cost of ``second-copy'' streams is zero; (2) royalties were
calculated [REDACTED]; and (3) Copyright Owners' original proposed a
per-play (i.e., per-stream) metric, which was rejected by all three
of the Judges.
---------------------------------------------------------------------------
The appropriateness of adopting the price discriminatory rate
provisions of the PR II-based benchmark is further underscored by
Copyright Owners' candid acknowledgement at the hearing that they were
essentially urging the Judges to adopt what is known as the
``Bargaining Room'' approach to rate setting. See Dissent at 24 (and
record citations therein).\115\
---------------------------------------------------------------------------
\115\ The Bargaining Room approach was first proposed for
incorporation into the statutory license standard in 1967 by the
NMPA, to be included in the predecessor section, later reorganized
in section 801(b)(1) that governs this proceeding. See Dissent at
22-24 (and citations therein). Ultimately, Congress punted on the
Bargaining Room approach, and adopted into law the four-factor
language set forth in section 801(b)(1). A subsequent attempt by
NMPA to have the Copyright Royalty Tribunal (CRT) (a predecessor to
the Judges) adopt the Bargaining Room theory was rejected by the
CRT, a rejection that was affirmed on appeal. See Recording Industry
Ass'n. of America v. Copyright Royalty Tribunal, 662 F.2d 1, 37
(D.C. Cir. 1981), aff'g Adjustment of Royalty Payable under
Compulsory License for Making and Distributing Phonorecords, 46 FR
10466, 10478 (1981). See generally, F. Greenman & A. Deutsch, The
Copyright Royalty Tribunal and the Statutory Mechanical Royalty:
History and Prospect, 1 Cardozo Arts & Ent. L.J. 1, 53, 64 (1982).
---------------------------------------------------------------------------
In the present proceeding, the appropriateness, vel non, of the
Bargaining Room approach boils down to the following:
Copyright Owners emphasize the inability of the Judges (or
anyone) to identify present market rates precisely, let alone over
the five-year rate period because the compulsory license set by the
Judges cannot possibly contemplate every single business model that
may develop in the ensuing time. . . . If the statutory rate is set
below market rates, then the parties will never negotiate upward
toward the market rates, because the licensees will always prefer to
invoke the right to use the licensed work at the below-market
statutory rates. However, if the Judges set the statutory rate above
what they find to be market rates, different licensees who each have
a maximum willingness to pay (WTP) below such a statutory rate would
seek to negotiate lower rates with the licensors. In response to
such requests to negotiate, according to this argument, Copyright
Owners would respond by negotiating various lower rates for those
licensees, provided lower rates were also in the self-interest of
Copyright Owners.
Dissent at 24-25 (and record citations therein).
The Judges find no reason to depart from the policy decision in
Phonorecords I that the rate setting policies made explicit in section
801(b)(1) are best discharged if the Judges eschew the Bargaining Room
approach and continue to identify rate structures and rates that
reflect the standards set forth in the statutory provision. To supplant
the statutory factors with a Bargaining Room approach would essentially
be to adopt a purely market-based rate-setting approach that is
inconsistent with section 801(b)(1) and with the Judges' application of
that statute to set rates, rate structures, and terms consonant with
effective competition.
With this background in mind, the Judges turn specifically to the
interrelationship between the price discrimination aspects of the rates
in the PR-II benchmark and the Bargaining Room approach.
Copyright Owners have demonstrated (albeit tacitly) their
understanding that, if the statutory provisions did not contain a price
discriminatory rate structure to reflect the varying WTP, they would
have to invent it. This finding is apparent from their advocacy for the
adoption of a Bargaining Room approach to rate-setting. See, e.g., 4/3/
17 Tr. 4390, 4431 (Rysman) (lauding bargaining room approach as
reflecting ``economical element of price discrimination . . . the
[licensor] is picking its prices carefully.'') (emphasis added); id. at
4431 (explaining that under this approach, when negotiating with
Spotify regarding a rate for ad-supported service, ``Must Have'' music
publishers would ``have the right . . . to set that price.''); 4/4/17
Tr. 483-45 (Eisenach) (acknowledging Copyright Owners' approach was
consistent with Bargaining Room theory because they were seeking rates
so high as to force would-be licensees to negotiate for the ``Must
Have'' mechanical license.).
Thus, the Judges find there to be no real dispute as to whether
there is a market-based need for an upstream discriminatory rate
structure.\116\ Rather, the parties appear to be in disagreement as to
who shall be in control of the setting of rates, the Judges, through
their application of law, or Copyright Owners, through the exercise of
their complementary oligopoly power. The resolution of this choice is
clear; the Judges, not the licensors, are statutorily-charged with
establishing provisions that are reasonable and otherwise properly
reflect the itemized objectives of section 801(b)(1).
---------------------------------------------------------------------------
\116\ The Majority recognized this point as well when--regarding
the ``increase the total revenue that price discrimination enables--
they ask (and answer) rhetorically: ``How could Copyright Owners and
their economic experts argue against a rate structure that inures to
their benefit as well? The answer is: They do not. . . . [T]hey
advocate for a rate set under the bargaining room theory, through
which mutually beneficial rate structures can still be negotiated,
but not subject to the ``reasonable rate'' and itemized factor
analysis required by law.'' Determination at 85 & n.153. The Judges
also note that Copyright Owners' acknowledgement that they too would
set price discriminatory rates and structures is not simply a
feature of this market. Rather, ``discriminatory pricing . . . is
the normal attribute of equilibrium . . . in a broad range of market
types and conditions where consumers can be separated into distinct
groups with different demand elasticities.'' W. Baumol, Regulation
Misled by Misread Theory: Perfect Competition and Competition-
Imposed Price Discrimination at 2 (2002). See also Dissent at 38,
n.74. Given the ubiquity of discriminatory pricing, the Judges also
find that the adoption into the statutory license of such pricing is
not--as Copyright Owners contend--simply the inappropriate favoring
of a particular business model, but rather a necessary reflection of
the fundamental nature of market demand, particularly, the varied
WTP among listeners.
---------------------------------------------------------------------------
Fourth, the PR II-based benchmark reflect a rate structure with an
adequate degree of competition, because there
[[Page 54438]]
was a balance of bargaining power between the two negotiating
industrywide trade associations, offsetting the complementary oligopoly
effects in place when a ``Must Have'' licensor bargains separately with
each licensee. Recently, the Judges discussed in detail how the
presence of countervailing bargaining power generates royalty rates at
effectively competitive levels. See Web V, 86 FR 59452, 59457 (Oct. 27,
2021).
Further with regard to this fourth point, the parties have been
operating over the past ten years under this basic rate structure, with
profits accruing to the licensors and admittedly tolerable losses
befalling the licensees. Moreover, after experience with these rates
and this rate structure in the Phonorecords I period, they renewed and
expanded this structure for use in the Phonorecords II period, when the
alternative of a statutory rate proceeding was available to licensors
and licensee alike. Their mutual willingness to continue in this manner
is important evidence of the workability and reasonableness of this
approach.
b. Evidence of Subjective Intent Not Prerequisite to Partial Adoption
of the PR II-Based Benchmark \117\
---------------------------------------------------------------------------
\117\ At the outset, the Judges reject Copyright Owners'
contention that the D.C. Circuit found ``validity'' in their
assertion that there was merit in Copyright Owners' assertion of the
``subjective intent issue.'' Rather, on this issue, Johnson first
held: ``[N]owhere does the [ ] Determination explain why evidence of
the parties' subjective intent in negotiating the Phonorecords II
settlement is a prerequisite to its adoption as a benchmark.''
Johnson, 969 F.3d at 387. Then, when Copyright Owners' appellate
counsel attempted to cure that failure by making their own
``subjective intent'' argument, the D.C. Circuit responded to that
``subjective intent'' argument with a single word: ``Perhaps.'' Id.
(emphasis added). This does not in any way suggest that Johnson
found ``validity'' in the ``subjective intent'' argument, but rather
was a non-committal response, consistent with the D.C. Circuit's
ruling finding that the Determination had not explained this point.
---------------------------------------------------------------------------
The Judges rely on the PR II-based benchmark as an objective
benchmark. Thus, the absence of testimony regarding what went through
the minds of the negotiators of the Phonorecords II agreement (and the
predecessor Phonorecords I agreement) does not diminish the objective
value of this benchmark. The Judges view the provisions of the PR II-
based benchmark as they would any benchmark, in the context of the
requisite benchmarking elements identified and discussed supra. This
approach allows the factfinder to analyze the benchmark through the
lens of its service in the marketplace as an objective model for the
market at issue, the Phonorecords III market. See, e.g., 3/13/17 Tr.
550-51, 566 (Katz) (knowledge of why parties negotiated specific
provisions is unnecessary, because objective results demonstrate
satisfactory performances of market).
Both Professors Katz and Hubbard noted that the current rate
structure remains useful, not based on consideration of the parties'
subjective understandings at the time of its creation, but because the
market has not since changed in a manner that would create a basis for
departure. Katz WDT ] 80 (``My analysis has identified no changes in
industry conditions since then [2012] that would require changing the
fundamental structure of the percentage-of-revenue prong.''); 4/13/17
Tr. 5977-78 (Hubbard) (changes in market are ``not uncorrelated with
the structure that was in place'' in 2012).\118\
---------------------------------------------------------------------------
\118\ As noted supra, the relevant material change since the
Phonorecords II agreement was reached is the significant growth in
streaming revenues. That change is reflected in the Judges'
application of the Shapley Value analyses, by which the Judges
increased the headline royalty rate by 44%, from 10.5% to 15.1%
(phased-in).
---------------------------------------------------------------------------
In this regard, it bears emphasis that Copyright Owners' own
witness, Dr. Eisenach, relied on several potential approaches that the
Majority characterized as benchmarks for his rate analysis, without
attempting to examine the subjective intent of the parties who
negotiated those agreements. Indeed, the Majority found that the PR II
Rates were properly considered as an objective benchmark, in the same
manner as Dr. Eisenach's proffered benchmarks:
The Services do not examine in detail the particular rates
within the existing rate structure. Rather, they treat the rates
within that structure as benchmarks, i.e., generally indicative of a
sufficiently analogous market that has ``baked-in'' relevant
economic considerations in arriving at an agreement. Dr. Eisenach
did not analyze why he chose the levels for the rates and ratios on
which he relied as benchmarks or consider the subjective
understandings of the parties who negotiated his benchmarks.
Similarly, the Services' economists elected to rely on the 2012
rates as objectively useful without further inspection.
This point is not made to be critical of Dr. Eisenach's
approach, but rather to show that the Services' reliance on the 2012
settlement as a benchmark shares this similar analytical
characteristic, typical and appropriate for the benchmarking method.
(The factual wrinkle here is that, hypothetically, the Services
could have called witnesses and presented testimony regarding the
negotiations that led to the 2012 (and 2008) settlements, but did
not, rendering the 2012 benchmark similar to other benchmarks taken
from other markets.)
Determination at 55 & n.106.\119\
---------------------------------------------------------------------------
\119\ Copyright Owners do not deny that they did not offer
evidence of subjective intent for Dr. Eisenach's benchmarks. Rather,
they assert Dr. Eisenach's reliance on benchmarks without examining
the subjective understandings of the negotiators of the benchmarks
is irrelevant because: (1) Copyright Owners were not seeking the
adoption in toto of the rates contained in any specific benchmark
cited by Dr. Eisenach; (2) Dr. Eisenach analyzed multiple benchmarks
to derive a reasonable range of rates; (3) his benchmarks were not
adopted; and (4) his benchmarks and are not at issue on this remand.
Copyright Owners Reply Brief on Remand at 28 n.19. But Copyright
Owners confuse evidentiary standards with evidentiary application.
Benchmarks are subject to the same evidentiary standards, regardless
of the breadth of purpose for which they are proffered and
regardless of whether they were adopted or rejected. Further, the
fact that Dr. Eisenach's chosen benchmarks are ``not at issue on
this remand'' does not render Copyright Owners' reliance on purely
objective benchmarks uninformative as to their own understanding of
the irrelevancy of the subjective thoughts of benchmark negotiators.
See generally Web IV, 81 FR 26370 (proposed benchmark adjustment
based on alleged ``additional value'' should be supported by
``record evidence . . . to provide a basis for such for such an
adjustment.'').
---------------------------------------------------------------------------
Copyright Owners also aver that they entered into the Phonorecords
II settlement simply to avoid litigation costs. Copyright Owners' Reply
Brief on Remand at 29. At the hearing, this assertion was presented by
David Israelite, NMPA's President. Israelite WRT ] 28; 3/29/17 Tr.
3649-52 (Israelite) (claiming NMPA lacked financial position to fund
rate litigation). The Services countered by noting that there was no
evidence to support Mr. Israelite's testimony in this regard, or how it
may have impacted the NMPA decision to participate. And, the Services
pointed out, notwithstanding his testimony regarding financial
constraints, NMPA had incurred the expense of a year-long negotiation
with the Services to seek higher rates, create new service categories
in subpart C, and change the TCC calculations. Id. at 159, 161-64; 3/
29/17 Tr. 3856 (Israelite).
Further, as a general principle, a party's mere assertion that the
Phonorecords II approach was the product of a settlement that was
predicated on the avoidance of litigation costs savings does not
invalidate its use as a benchmark in proceedings before the Judges,
especially because, by statute, the Judges are authorized to consider
such agreements. See Music Choice v. Copyright Royalty Board, 774 F.3d
1000, 1014-15 (D.C. Cir. 2014) (testimony alleging agreement was
reached to avoid litigation costs does not invalidate evidentiary use
of that agreement for rate setting purposes, absent other evidence
demonstrating settlement was involuntary or otherwise unreasonable.).
Thus, the Judges find that the evidentiary record does not support
Copyright Owners' position that this ``litigation cost avoidance''
assertion constituted a separate, idiosyncratic value that diminishes
the
[[Page 54439]]
Judges' partial reliance on the PR II Rates in this Initial Ruling.
Copyright Owners also mistakenly rely on the fact that the Services
bore the burden of proof regarding the absence of any subjective
idiosyncratic factors that hypothetically could have diminished the
useful value of the PR II-based benchmark. Id. at n.21. The Services
indeed bore the burden of proof (i.e., persuasion) with regard to their
proffered benchmark PR II Rates, and they presented adequate objective
evidence and testimony that this approach has worked in the marketplace
to serve as prima facie proof to support the Judges'(partial) use of
this benchmark in this remand proceeding. And, as explained above, such
subjective intent was not a necessary element of their benchmark
proofs. But, with regard to Copyright Owners' rebuttal to those proofs,
Copyright Owners bore the burden of production, to present sufficient
evidence and/or testimony that the Judges could rely on to reject the
(partial) use of the PR II-based benchmark. This Copyright Owners
failed to do.\120\
---------------------------------------------------------------------------
\120\ As described in this Initial Ruling, the Judges identified
this same distinction between the burden of proof and the burden of
production to find in favor of Copyright Owners' proffered expert
testimony in support of their Nash Bargaining analysis, testimony
which constituted prima facie proof that was not adequately rebutted
by the production of sufficient testimony from the Services' expert
economic witnesses.
---------------------------------------------------------------------------
In fact, given Copyright Owners' reliance on the subjective intent
of the parties to a benchmark, the Judges attempted to identify
potential subjective evidence of how the capped TCC rates in the PR II-
based benchmark \121\ were derived, during the examination of Dr.
Eisenach at the hearing:
---------------------------------------------------------------------------
\121\ The ``capped'' TCC rates are elements of the Phonorecords
II rates.
[JUDGE STRICKLER] Do you discuss, Dr. Eisenach, . . . in your
written direct or written rebuttal testimony how the parties arrived
. . . at the ratios for sound recording to musical works in [witness
interrupts]
[DR. EISENACH] That process is opaque to me, Your Honor.
[JUDGE STRICKLER] Did you [witness interrupts]
[DR. EISENACH] I know--I know there was a 2008 negotiation. I
know there was a 2012 negotiation. I wasn't . . . present, and I'm
not privy to any of the details.
[JUDGE STRICKLER] You were not informed by your client or by any
other source of information as to how they arrived at those
particular ratios?
[DR. EISENACH] When I've asked the question, I've found people
chuckle and--and there doesn't seem to have been too much system--
systematic thought that went into it, but I don't really know that.
I just--when I ask the question, people say: Nobody really knows. .
. . Someone may know, but that's what I've been told.
4/4/17 Tr. 4611 (Eisenach) (emphasis added). The Judges find it
perplexing, to say the least, that Copyright Owners would ``chuckle''
when asked by their expert witness for the very subjective evidence
which they claim to be relevant. But of perhaps greater relevance is
Dr. Eisenach's further testimony, quoted above, that he was also told
by Copyright Owners that ``nobody really knows'' how the parties
arrived at those rate ratios. Copyright Owners' ``chuckle,'' in
response to its expert's critical inquiry as to the derivation of
rates--and that expert's understanding that his client simply did not
know how those rates were derived--undercut Copyright Owners' claim
that subjective understanding of those rates could undermine their
usefulness in the benchmark.\122\
---------------------------------------------------------------------------
\122\ The Judges also find Copyright Owners' assertion that they
did not know how those rates were established is not credible, given
that they and their representatives negotiated those rates.
---------------------------------------------------------------------------
c. Substantial Evidence Demonstrates That PR II Rates, Other Than the
Headline Rate, Are Not ``Too Low''
As noted supra, one reason the D.C. Circuit vacated and remanded
the Determination was because it declined to entertain the argument
made only by appellee's counsel that ``the prior rates had been set far
too low, thus negating the usefulness of the prior settlement as a
benchmark.'' Johnson, 969 F.3d at 387. The Judges have noted throughout
this Initial Ruling their adoption of the Shapley Value modeling
analysis undertaken by the Majority, and raised the headline royalty
rate by 44% from 10.5% to 15.1% (as phased-in), rendering moot
appellate counsel's suggestion regarding the rate level.
Here, the Judges further consider whether other rates within the PR
II-based benchmark are reasonable, not only because they are part and
parcel of the workable structure of that benchmark, but also to
determine if they are supported by record evidence. To put this issue
in context, those rates would apply on the second prong of the
``greater-of'' rate structure in the PR II-based benchmark. The first
prong in the PR II-based benchmark rates is the 10.5% revenue rate--
increased to 15.1% (as phased-in) by this Initial Ruling. The second
prong consists of the ``lesser of'' a TCC rate or a per subscriber
rate.\123\ For certain delivery configurations, these rates also cannot
fall below any applicable Mechanical Floor. See Johnson, 969 F.3d at
370.\124\
---------------------------------------------------------------------------
\123\ This second prong contains only a TCC rate (i.e., an
uncapped rate) for: (1) the ad-supported the service, because there
are no subscribers to such a service; and for (2) bundled
subscription service, for which there is a $0.25 per month floor but
no per-subscriber cap, and Service Revenue for such bundles is
calculated pursuant to 37 CFR 385.11 (``Service Revenue''
definition, ] 5).
\124\ As Johnson explained, the CRB Judges ``retained the
mechanical floor'' because, like so much of the PR II-based
benchmark, it ```appropriately balances the [streaming service
providers'] need for the predictability of an All-In rate with
publishers' and songwriters' need for a failsafe to ensure that
mechanical royalties will not vanish[.]'' Id. at 371-72. It is
noteworthy that Copyright Owners urged the Judges (successfully) to
maintain the Mechanical Floor provisions, which are the product of
the Phonorecords II (and Phonorecords I) negotiations. Thus, it
seems apparent that Copyright Owners as well as the Services
consider provisions from the negotiated rates and rate structure to
be in the nature of benchmarks, although differing as to which
elements such be included or excluded. (The Services unsuccessfully
argued for the elimination of the Mechanical Floors.) This
perspective underscores the correctness of the Judges' decision on
remand to treat the PR II-based benchmark as useful.
---------------------------------------------------------------------------
The Services describe the key feature of these non-headline rates
as the fostering of beneficial price discrimination, i.e., the adoption
of ``different rate levels for different product offering,'' in order
[t]to account for consumers' different willingness to pay [WTP] for
music. Services' Joint Opening Brief (on Remand) at 21. As an example
of how these price discriminatory rates impacted the market, the
Services compare and contrast two Amazon offerings, Amazon Music
Unlimited (for Echo) and Amazon Prime Music.
Amazon Music Unlimited, with more than 30 million available songs
as of the Phonorecords II proceeding period, see Mirchandani WDT ] 41,
[REDACTED].\125\ By contrast, Amazon Prime Music, calculated as a
``bundled subscription'' configuration, makes available only an
abridged repertoire of 2 million songs, see Mirchandani, supra, and
[REDACTED]. See id. at Sec. 385.13(a)(4).
---------------------------------------------------------------------------
\125\ [REDACTED].
---------------------------------------------------------------------------
Thus, Amazon pays [REDACTED] for listening by the more casual
consumers who use the limited catalog Prime Music service at no
additional charge beyond their Prime membership fee, compared to
consumers who want the full repertoire provided by Amazon Music
Unlimited on their Echo devices. See Services' Joint Opening Brief at
71. These royalty obligations demonstrate the combination of price
discrimination, product differentiation and ``derived demand'' in
action; that is, the [REDACTED] are derived from the lower demand of
consumers of the limited Amazon Prime Music service compared with
subscribers to Amazon Music
[[Page 54440]]
Unlimited on their Echo devices, which in turn drive higher revenues.
It is also important to note that these differential rates on the
second prong of the ``greater-of'' structure of the PR II Rates are
overridden by the revenue percentage rate on the first prong if that
first prong rate generates more revenue. For example, [REDACTED], see
Dissent at 29 (Table) and 116; see also [REDACTED]. With the headline
rate now increased on a phased-in basis, the price discriminatory
royalty generated by this [REDACTED].
It is noteworthy that Johnson affirmed the Majority's setting of
other price discriminatory features, e.g., the family and student plan
provisions, based on the Judges' reliance on the Services' expert
testimony regarding the benefits of ``having a way . . . where low
willingness to pay consumers can still access music in a way that still
allows more monetization of that provision of that service.'' Johnson,
969 F.3d at 392-93. In similar fashion, the multi-tiered rates in the
PR II-based benchmark likewise were supported by the same type of
testimony; indeed, from expert testimony proffered by both parties, as
considered below.
First, Professor Katz notes that the existing rate structure
captures two important aspects of the economics of the interactive
streaming market: (1) the variable WTP among listeners; and (2) the
corollary variable demand for streaming services. See 3/13/17 Tr. 586-
87 (Katz); see also Marx WRT ] 239 et seq.; 4/7/17 Tr. 5568 (Marx)
(noting that the present structure serves differentiated products
offered to customer segments with a variety of preferences and WTP). In
more formal economic terms, Professor Katz notes that the present
structure enhances variable pricing that allows streaming services ``to
work [their] way down the demand curve,'' i.e., to engage in price
discrimination that expands the market, providing increased revenue to
the Copyright Owners as well as the Services. 3/13/17Tr. 701 (Katz).
Second, in similar testimony, Professor Hubbard captures the
interrelationship between the economics of this market and the existing
rate structure:
[F]rom an economic perspective, you can think of this market and
this industry as being composed of different customer segments by
tastes and preferences and willingness to pay. And so no rate
structure can really work without understanding that, and no
business model can really work without understanding that.
[I]n terms of rate structures, the Phonorecords II framework
from the previous proceeding does offer a benchmark to start because
it provides for differences in distinct product categories in terms
of music service offerings, pricing possibilities, and so on. And it
has encouraged a very diverse digital music offering set from actual
competitors.
3/21/17 Tr. 2175-76 (Hubbard). Moreover, Professor Hubbard [REDACTED]
4/13/17 Tr. 5978 (Hubbard); see also Hubbard WDT ] 4.7 (the 2012 rate
structure provides the ``necessary flexibility to accommodate the
underlying economics of Amazon's various digital music service
offerings.''). See also 3/15/17 Tr. 1176 (Leonard) (notwithstanding
changes and growth in the streaming marketplace over current rate
period, underlying economic structure of marketplace, which made
percent-of-revenue based royalty appropriate, has not changed).
Third, the Services' experts further assert that the multiple
pricing structures necessary to satisfy the WTP and the differentiated
quality preferences of downstream listeners relate directly to the
upstream rate structure to be established in this proceeding. For
example, Professor Marx opines that the appropriate upstream rate
structure is derived from the characteristics of downstream demand. 3/
20/17 Tr. 1967 (Marx) (agreeing that rate structure upstream should be
derived from need to exploit willingness to pay of various users
downstream via percentage of revenue because downstream listeners have
varying willingness to pay that should be exploited for mutual benefit
of copyright licensees and licensors). Professor Marx further
acknowledged that this upstream:downstream consonance in rate
structures represents an application of the concept of ``derived
demand,'' whereby the demand upstream for inputs is dependent upon the
demand for the final product downstream. Id. Moreover, Dr. Leonard
notes that reliance on the Services to identify segmented demand and
develop price discriminatory approaches is appropriate because ``the
downstream company is going to have a lot more information about . . .
the business, about what makes sense.'' 4/6/17 Tr. 5238 (Leonard).
Regarding a comparison of revenue growth to streaming growth,
Professor Hubbard dismisses as economically ``meaningless'' Copyright
Owners' argument that they have suffered relative economic injury under
the current rate structure simply because the increase in their
revenues from interactive streaming has been proportionately less than
the growth in the number of interactive streams, leading mathematically
to a lower implicit or effective per stream royalty rate. That is, he
notes there is no evidence to rebut this prima facie indication of
beneficial price discrimination, i.e., no contrary evidence indicating
that, if the Services had sought to increase the price of the services
available to these low to zero WTP listeners because of higher
royalties, they would have paid the higher price, rather than declined
to utilize a royalty-bearing interactive streaming service. See 4/13/17
Tr. 5971-73 (Hubbard); see also Dissent at 52.
The Services also link their price discrimination argument to the
fact that the marginal physical cost of streaming is zero to the need
for a flexible rate structure such as now exists. In this regard,
Professor Hubbard notes that, because ``[t]he marginal production cost
at issue here is--is zero. . . . it's not clear why it's not better to
bring new customers into the market on which royalties would be paid
and, of course, zero marginal cost incurred.'' 4/13/17 Tr. 5917-18
(Hubbard). See also Marx WDT ] 97 (``Setting the price of marginal
downstream listening at its marginal cost of zero induces more music
consumption and variety than per-song or per-album pricing.'').
Professor Marx makes the same argument as to the salutary nature of
price discrimination in this context with regard to Spotify's ad-
supported approach. Focusing on the first purpose, Spotify is
attracting ad-supported listeners who have a relatively low WTP,
whether they have low incomes, (a budget constraint) or low interest in
music (low ``utility,'' in the parlance of economists). These
listeners, and the advertising revenue they generate are real and
reflect the WTP of a large swath of all interactive listeners. See Marx
WRT ] 115-16 & Fig. 9 (``While I agree that one aspect of the ad-
supported service is to provide an on-ramp to paid services, it also
has another important aspect, namely to serve low WTP customers. . . .
Copyright Owners' economists err in not calculating the impact of the
Copyright Owners' proposal on ad-supported services. Ad-supported
services currently make up a majority of subscribers and [REDACTED]% of
all streams in the industry.'').
Accordingly, a separate tier for an ad-supported service accounts
for the different nature of the downstream listenership, allowing the
upstream royalty to be based on that characteristic. This
differentiation was essentially acknowledged by Copyright Owners late
(too late, actually) when they proposed in their post-hearing filing
that ``if the Judges intend to include the Spotify ad-supported
[[Page 54441]]
service in the rate structure and rate calculations, that they do so by
establishing separate rates and terms for the ad-supported service. See
COPCOL (Corrected) ] 228 & n.34. But the PR II-benchmark already
incorporates separate rates for free/ad-supported services!\126\
---------------------------------------------------------------------------
\126\ Copyright Owners also belatedly proposed that the Judges
establish specific functionality limits on a separate ad-supported
prong to avoid cannibalization of subscriber-based streaming with
fuller functionality. Id. [REDACTED].
---------------------------------------------------------------------------
Another important evidentiary factor buttressing the need for price
discriminatory rates and structures was the testimony of the Services'
survey expert, Mr. Robert Klein, Chair and co-founder of Applied
Marketing Systems, Inc. Mr. Klein surveyed 2,101 people (the Klein
Survey) who were listeners to streamed music and found, inter alia,
that: (1) the majority of listeners would not pay for a monthly
streaming subscription; and (2) for those who do subscribe, their
demand was elastic, with increases in subscription prices causing
overall greater percentage reductions in quantity demanded, moving
customers to free, ad-supported and non-streaming alternatives. See
Klein WRT ]] 60-67. By contrast, Copyright Owners did not present any
survey testimony. The Determination fully credited the Klein Survey,
finding as follows:
It is important to note that Copyright Owners' attacks on the
Klein Survey are not levelled by any witnesses, nor contradicted by
their own survey expert, because Copyright Owners elected not to
proffer such an expert in their direct (or rebuttal) cases. Rather,
Copyright Owners elected to make a descriptive argument regarding
the elasticity of demand among different segments of the market, as
opposed to a survey-based or econometric study of price elasticity.
[Although] Copyright Owners attack the Klein Survey on several
fronts[,] [t]he arguments made by Copyright Owners are insufficient
. . . to seriously weaken the probative value of the Klein Survey.
In the end, the Judges are not persuaded by the Copyright Owners'
revenue bundling arguments not to adopt a flexible, revenue-based
royalty rate.
Determination at 22-23 & n.53; see also Dissent at 64-67 (including
point-by-point rejection of Copyright Owners' non-expert criticisms of
Klein Survey).
The Services also note that the existing rate structure has
produced generally positive practical consequences in the marketplace.
Their joint accounting expert, Professor Mark Zmijewski, testified that
the [REDACTED] from the sale of product under (former) Subpart A since
2014 has been [REDACTED] over the same period. Expert Report of Mark E.
Zmijewski February 15, 2017 ]] 38, 40 (Zmijewski WRT); 4/12/17 Tr. 5783
(Zmijewski); see also 4/13/17 Tr. 5897 (Hubbard) (``the evidence that I
reviewed suggests that the copyright holders have actually benefitted
from this structure. . . .'').
More particularly, Professor Zmijewski testified that:
Total revenues reported by the NMPA for NMPA members from
all royalty sources [REDACTED]. Zmijewski WRT ] 41.
This [REDACTED]. Id.
The [REDACTED]. Id.
Mechanical royalty revenue for the sale of downloads and
physical phonorecords [REDACTED]. Id. ] 38.\127\
---------------------------------------------------------------------------
\127\ By contrast, Copyright Owners assert that the appropriate
approach would only consider interactive service payment of
mechanical royalties, and exclude performance royalties. On that
basis, revenue, for the sale of digital downloads and physical
phonorecords mechanical royalty revenue [REDACTED] from [REDACTED]
in 2014 to [REDACTED] (as noted in (4) above, whereas mechanical
royalty from streaming [REDACTED] from [REDACTED] in 2014 to
[REDACTED] in 2015. Thus, the [REDACTED] in mechanical royalty
revenue from streaming [REDACTED] in mechanical royalty revenue from
the sale of digital and physical phonorecords. The Judges do not
agree with Copyright Owners. Performance royalty and mechanical
royalty payments made by the Services are for perfect complements--
neither license has any value to the Services unless they acquire
both. Indeed, that is a critical reason why the mechanical rate is
calculated on an ``All-In'' basis. Thus, it makes sense to make the
comparison in the manner undertaken by Professor Zmijewski.
---------------------------------------------------------------------------
In sum, the foregoing analysis demonstrates the economic
reasonableness and appropriateness of the price discriminatory
Phonorecords II rate structure and its negotiated safeguards to address
the real possibility of revenue diminution. As discussed below, the
record evidence also supports royalty rates within the PR II-based
benchmark.\128\
---------------------------------------------------------------------------
\128\ Again, to be clear, the Judges are substituting the 15.1%
revenue rate for the 10.5% revenue rate as the headline rate in the
``greater-of'' structure of the Phonorecords II benchmark. Thus, the
price discriminatory royalty rates discussed below would apply only
if they generated a ``greater'' level of revenue than the headline
15.1% revenue rate. And, although the Mechanical Floor rate is not
tied directly as an alternative to the ``greater-of'' revenue rate
(now 15.1% as phased-in), it is not a floor that ignores the effect
of that ``greater of'' rate. For example, assume the popular
standalone portable subscription streaming service that people
access on their mobile phones would pay an ``All-In'' musical works
royalty of 15.1% based on the application of the two ``greater-of''
prongs. However, assume also the ``Performance Royalty'' that must
be subtracted is 12%. That would leave 3.1% of service revenue
attributable to the mechanical right. However, if that revenue rate
of 3.1% yielded mechanical royalty revenue that was less than the
royalty revenue generated by the applicable monthly mechanical floor
of $0.50 per subscriber, then the mechanical floor would control.
This application, like any other application of the mechanical
floor, does not diminish the value of the 15.1% right, but rather
limits its reduction under the ``All-In'' calculation. Recall also
that the Determination, Dissent and Johnson do not disturb the All-
In and Mechanical Floor features of the Phonorecords II benchmark.)
And finally, with regard to the actual per subscriber monetary
values in the mechanical floors, no party suggested changes from
rate levels in the PR II-based benchmark, including in the
mechanical floor rates. The Judges recognize, as did Dr. Katz,
Pandora's economic expert witness, that alternate values might have
been preferable for rates contained in the PR II-based benchmark,
but none were in the record. See 4/15/17 Tr. 5056-58 (Katz).
---------------------------------------------------------------------------
The PR II-based benchmark contain several alternate rates
explicitly calculated as a percentage of payments made by interactive
streaming services to the record companies for sound recording rights.
See Addendum to this Initial Ruling. In the Subpart relating to
streaming, the (former) subpart B category, the TCC is 22% for ad-
supported services and 21% for portable subscriptions. Id.; see also 37
CFR 385.13(b)(2) and (c)(2). These percentage figures correspond to
sound recording: musical works royalty ratios of 4.55:1 and 4.76:1,
respectively.
With regard to these ratios, Copyright Owners' economic expert
witness, Dr. Eisenach, stated: ``In my opinion, the evidence . . .
indicates that the relative valuation ratios implied by the current
Section 115 compulsory license . . . represent an upper bound on the
relative market valuations of the sound recording and musical works
rights.'' Id. ] 92 (emphasis added). (As an ``upper bound,'' these
ratios would represent the lower bound on the relative market
valuations of the reciprocal percentage of the value musical works
rights relative to sound recording rights, again, 22% and 21%.\129\)
Thus, there appears to be consensus between Copyright Owners' witness
and the Services (who advocate for applying these rates on the price
discriminatory tier of their benchmark) that these rates constitute
``relative market valuations'' (even if they are not Dr. Eisenach's
preferred market valuations within the bounded zone of such values).
---------------------------------------------------------------------------
\129\ 1 / 4.55 = .219, or 22% (rounded); 1 / 4.76 = .210 (21%).
---------------------------------------------------------------------------
Dr. Eisenach's testimony regarding the ``bounds'' of useful market
valuations is noteworthy because his acknowledgement is consonant with
judicial precedent. The Judges' setting of reasonable rates often
requires them to identify a ``zone of reasonableness,'' within which
they identify appropriate statutory rates. See, e.g., Intercollegiate
Broadcasting System, Inc. v. Copyright Royalty Board, 684 F.3d 1332,
1340 (D.C. Cir. 2012) (The CRB Judges' rate setting can necessitate the
finding of a ``zone of reasonableness [because] ``[s]tatutory
reasonableness is an
[[Page 54442]]
abstract quality represented by an area rather than a pinpoint.'').
The 21% and 22% TCC rates within section 115 identified by Dr.
Eisenach as generating the ``lower bound on the relative market
valuations'' imply certain approximate percent-of-revenue rates, i.e.,
percent of total service revenue (not percent of sound recording
revenue). See Dissent at 91, n.133 (sound recording rates clustered
between [REDACTED]% and [REDACTED]% of revenue). For example, if the
sound recording royalty rate for interactive streaming is [REDACTED]%
of revenue, then the musical works rate would be calculated as 0.21 x
[REDACTED], which equals [REDACTED]%, (or as .22 x [REDACTED] which
equals [REDACTED]%). At the low end of the range, if the sound
recording royalty rate is [REDACTED]%, then, applying these TCC
figures, the implied musical work royalty rate would be calculated as
[REDACTED]% (.21 x [REDACTED]) or [REDACTED]% (.22 x [REDACTED]).\130\
---------------------------------------------------------------------------
\130\ Dr. Eisenach's identification of the 21%-22% TCC as within
the bounds of market valuations may appear surprising at first in
light of the higher 26.2% uncapped TCC rate pursued (unsuccessfully)
on remand by Copyright Owners. But in the context of his testimony,
Dr. Eisenach's opinion is understandable. The former headline rate
of 10.5%, when sound recording rates ranged from approximately
[REDACTED]% to [REDACTED]% of streaming revenues, yielded TCC rates
between [REDACTED]% and [REDACTED]%. Thus, Dr. Eisenach was
identifying a market valuation [REDACTED] (at his lower bound)
between [REDACTED]% (the difference between 21% and [REDACTED]%) and
[REDACTED]% (the difference between 22% and [REDACTED]%). Again, for
context, this Initial Ruling raises the percentage rate by 44% when
fully phased-in (based on the experts' Shapley analyses,
significantly above the TCC rates advocated by Dr. Eisenach, even
assuming the [REDACTED]%-[REDACTED]% sound recording rates on which
he relied.
---------------------------------------------------------------------------
It is important to emphasize and detail the context of these price
discriminatory rates. These capped TCC rates are on the ``greater of
prong'' that is compared with the headline 15.1% revenue rate (phased-
in) that the Judges are also adopting in this Initial Ruling. As phased
in, the headline rate is greater than all the capped TCC-based rates
identified in Dr. Eisenach's testimony, supra, [REDACTED]. For 2019,
the phased-in headline percentage rate, 12.3%, is [REDACTED] the
[REDACTED]% and [REDACTED]% revenue rates derived if the sound
recording rates was [REDACTED]%. For 2018, the phased-in headline
percentage rate, 11.4%, is [REDACTED] all the rates derived from the
capped TCC rates Dr. Eisenach identified as ``market valuations''
(albeit the lower bound in his opinion). But that is of no negative
consequence for Copyright Owners, because they would get paid on the
``greater-of'' metric (capped TCC or headline rate) under the
Phonorecords II-based rate structure the Judges are adopting (For the
portable subscriptions, even though the 80 cents/subscriber ``lesser-
of'' portion of the non-headline prong would apply on that prong if it
was lower than the capped TCC rate, the actual rate could not be lower
than the phased-in headline rate.)
Dr. Eisenach also examined direct agreements between record
companies and interactive streaming services that contain rates for
sound recordings and mechanical royalties, respectively. See, e.g., id.
]] at 84-91. In such cases, the ratio of sound-recording to musical-
works royalties ranged tightly between [REDACTED] and [REDACTED],
closely tracking the regulatory ratios implicit in the section 115 TCC.
Id. ] 92. (The [REDACTED] ratio equates to a TCC rate of [REDACTED]%,
and the [REDACTED] ratio equates to a mechanical rate of [REDACTED]%.).
He concluded, as he did with regard to the actual section 115 license
rates: ``In my opinion, the evidence presented . . . indicates that the
relative valuation ratios implied by the . . . negotiations under [the
statutory] shadow--ranging from [REDACTED] [[REDACTED]%] to [REDACTED]
[[REDACTED]%]--represent an upper bound on the relative market
valuations of the sound recording and musical works rights.'' Eisenach
WDT ] 92 (emphasis added).
Dr. Eisenach also identified several additional useful benchmarks.
First, he identified what was coined the ``Pandora Opt-Out Agreement''
benchmark,\131\ which reflected a ratio of [REDACTED] of sound-
recordings to musical-works in a comparable benchmark setting. This
ratio translates to a TCC percent of [REDACTED]%. With sound recording
royalty rates of approximately [REDACTED]% to [REDACTED]%, this TCC
reflects an effective percentage of total revenue equal to [REDACTED]%
to [REDACTED]%.
---------------------------------------------------------------------------
\131\ Pandora was only a noninteractive service at that time,
and thus only paid the performance right royalty, not the mechanical
right royalty, for the right to use musical works. Because the
parties agree that the performance right and the mechanical right
are perfect complements, Pandora's payments for the performance
right are thus relevant and probative, as they reflect the full
value of the musical works royalty to a noninteractive service.
These factors became relevant because major music publishers had
negotiated direct licensing agreements with Pandora for its
noninteractive service covering the period from 2012 through 2018.
Eisenach WDT ] 103. They negotiated these direct agreements after
certain publishers had decided to ``opt-out,'' i.e., to withdraw
their digital music performance rights from PROs, and asserted the
right to negotiate directly with a digital streaming service.
Pandora thus negotiated several such ``Opt-Out'' Agreements with an
understanding that the rates contained in those direct agreements
might not be subject to rate court review and thus could reflect
market-based rates. Given this unique circumstance, and given that
the markets and parties involved in the Pandora Opt-Out agreements
are somewhat comparable to the markets and parties at issue in this
proceeding, Dr. Eisenach concluded that these agreements provided
``significant insight into the relative value of the sound recording
and musical works rights in this proceeding.'' Id. (emphasis added).
(The Judges did not adopt Dr. Eisenach's speculation that this
performance royalty would continue to grow after 2018. See
Determination at 51; Dissent at 102-103.)
---------------------------------------------------------------------------
Second, Dr. Eisenach identified YouTube agreements with music
publishers that relate to the combination of a commercial sound
recording and a ``static image.'' The YouTube agreements contain an
explicit royalty of [REDACTED].\132\ That [REDACTED]% royalty is a
denominator in the ratio concept utilized by Dr. Eisenach, and the
numerator is the [REDACTED] sound recording royalty paid to the record
companies. YouTube had agreed to pay [REDACTED]% of its revenues, and
had agreed to pay [REDACTED] and other record companies [REDACTED]% of
revenues. The [REDACTED] ratio reduces to [REDACTED], implying a TCC
([REDACTED]) of [REDACTED]%. The [REDACTED] ratio reduces to
[REDACTED], implying a TCC ([REDACTED]) of [REDACTED]%. See Dissent at
101-102.
---------------------------------------------------------------------------
\132\ Dr. Eisenach preferred to use YouTube agreements that
included [REDACTED], but the Judges relied on [REDACTED] as more
comparative. Determination at 50; Dissent at 102.
---------------------------------------------------------------------------
These additional rates identified in Dr. Eisenach's testimony
further confirm the reasonableness of the non-headline rates within the
PR II-based benchmark.
Finally, the Judges look at the effective rates paid by Spotify,
the largest interactive streaming service in terms of in terms of the
number of subscriber-months and the number of plays. See Marx WRT ]]
37-38 & Figs. 8 & 9. Under the PR II based benchmark, Spotify paid on
its subscription service an effective ``All-In'' royalty rate of
[REDACTED]% of its total revenues. See Dissent at 80, 115, 149 (and
record citations therein). Spotify paid this effective percent-of-
revenue rate [REDACTED]. See id. at 29 (Table).
Turning to Spotify's free/ad-supported offering (and as noted
supra), Spotify paid royalties under the PR II Rates at an effective
``All-In'' royalty rate of [REDACTED]%. Spotify paid this effective
percent-of-revenue rate [REDACTED]. See id. When Spotify's two tiers
are blended and averaged, the effective percent-of-revenue rate is
[REDACTED]% of revenue. See id. at 116. The average rate has salience
in this proceeding because Spotify's two
[[Page 54443]]
tiers are interrelated, in that free/ad-supported listeners constitute
a pool of potential converts to the subscription tier under this
``freemium'' model, even as this offering generates royalties under the
PR II-based benchmark.
d. Copyright Owners' Concern Regarding Revenue Diminution Is
Insufficient To Reject the PR II-Based Benchmark
Copyright Owners argue that what the Services tout as beneficial
price discrimination generates an ``incredible'' level of revenue
diminution, including displacement, resulting in a ``major problem''
that reduces reportable revenues and thus the royalty base. See, e.g.,
3/7/22 Tr. 193 (Copyright Owners' counsel). This argument is based upon
documents and evidence that demonstrated the following:
[REDACTED];
[REDACTED].
[REDACTED];
[REDACTED];
[REDACTED];
[REDACTED]
[REDACTED];
[REDACTED];
[REDACTED]; and
Copyright Owners' expert, Dr. Rysman, testified that
interactive services often elect to forgo current profit maximization,
e.g., by charging lower prices, in order to build a customer base and
greater long-run profitability or value, from selling music and non-
music products or services to its customers.
CO Initial Submission at 40-42 (and record citations herein).
The Services' economic experts do not ignore the fact that there
can be revenue attribution problems when interactive streaming is
combined with other products or services. They acknowledge that, even
absent any wrongful intent with regard to the identification and
measurement of revenue, attribution of revenue across product/service
lines of various services can be difficult and imprecise. See, e.g., 4/
5/17 Tr. 5000 (Katz) (problem of measuring revenue ``certainly a factor
that goes into thinking about reasonableness.'').
However, Professor Katz testified that the existing rate structure
agreed to by the parties accommodates these bundling, deferral, and
displacement issues via the use of an alternative rate prong that would
be triggered if the royalty revenue resulting from the headline rate of
10.5% of streaming revenue fell below the royalty revenue generated by
that second prong. Katz WDT ]] 82-83; 3/13/17 Tr. 670 (Katz). Moreover,
Professor Katz concluded that, because the marketplace appears to be
functioning (in the sense that publishers are earning profits and new
and existing interactive streaming services continue to operate despite
accounting losses), these revenue-measurement issues are being
adequately handled by the alternative rate prong, even if an altered
second prong might work better. Id. at 738-39. More generally,
Professor Katz further noted that, the existing rates within the PR II-
based benchmark were performing well, and even if alternative minima
might be preferable, no such alternative rates were in the record. See
4/15/17 Tr. 5056-58 (Katz) (under the PR II-based benchmark ``the
industry . . . was performing well,'' but ``if someone had a proposal
[with] a specific reason why we should adjust this minimum that's
something I would have examined,''). But Copyright Owners did not
propose alternative rates or minima within the PR II-based benchmark,
but instead urged the Judges to disregard the benchmark writ large.
Accordingly, there were no alternative rates or minima in the record.
Professor Katz further noted that the PR II-based benchmark rates
were established when ``ecosystem'' entities such as Yahoo--akin to
Amazon, Apple, and Google--were in the marketplace. 4/5/17 Tr. 5055-57
(Katz); see also Determination at 31 (and record citations therein)
(noting the presence of Microsoft as well as Yahoo as licensees in the
interactive market during the Phonorecords II negotiations).
More broadly, the Services' position regarding the use of the two
prongs and their alternate rates to ameliorate the revenue-measurement
problems is summed up by Professor Katz as follows:
[T]he primary reason [for the two rate prongs] . . . is because
of the measurement issues that can come up when having royalties
based on a . . . percentage of revenues because there can be issues
about how to appropriately assign revenues to a service. And so I
think the minim[a] can play an important role when those--you know,
when those measurement problems are severe, you can turn to the
minimum instead. . . . [W]hat I have in mind, right, is that what
would happen if you could imagine an entrepreneur coming along and
saying we want to have a service and have some incredibly low price
and not a very good monetization model, where a copyright owner
would say--in an effectively competitive market, would say, wait a
minute, I don't want to license to you on those terms. It's--I just
think the possibility of getting a return is so low, I'm not going
to do it, even though you, as an entrepreneur, are willing to try
this. I as the copyright owner want some sort of, you know, return
on it. And that's what the minimum also helps to do.
3/13/17 Tr. 599 (Katz.); see also 3/20/17 Tr. 1900-01 (Marx) (minima
protect against revenue measurement problems); 4/7/17 Tr. 5584 (Marx)
(statutory minima play ``two roles''--protecting the Copyright Owners
from ``revenue mismeasurement'' by creating the ``greater of'' prong,''
but incorporating per subscriber rate prong in ``lesser of'' component
to protect services from the record companies' use of their market
power to engage in ``manipulation of the sound recording royalties'' on
which the TCC prong is calculated).
After considering the record, the Judges determine that the
Majority had not found--as Copyright Owners claim--that the activities
and strategies by the Services were ``incredible'' or a ``major
problem. Rather, the Majority's characterization was measured, stating
repeatedly that the Services engaged ``to some extent'' in revenue
diminution because they ``might focus on long-term profit maximization
to promote their long-term growth strategy, which occurs ``even absent
wrongful intent.'' Determination at 20-21, 36, 90; accord, Dissent at
59. In fact, the Majority specifically stated: ``The Judges agree that
there is no support for any sweeping inference that cross-selling has
diminished the revenue base.'' Id. at 21 (emphasis added). The Majority
(and the Dissent) thus acknowledged the reasonableness of both sides of
this issue, recognizing both the Services' use of price discriminatory
approaches that can lower per user or per-stream revenues but grow
royalties, market share and revenue, as well as Copyright Owners'
concomitant desire to protect themselves from reductions in the royalty
revenue base, however limited in extent, that would only serve to
diminish royalties.
One way the input supplier can avoid this impact is to refuse to
accept a percent of revenue form of payment and move to a fixed per-
unit input price. This is what Copyright Owners originally and
unsuccessfully sought in this proceeding, subject to a bargaining room
approach by which they could switch back to the old approach (or any
other approach) through purely market-based negotiations, unbounded by
the statutory and regulatory standards of ``fairness'' and ``effective
competition.'' See Dissent at 60.
The Judges must reconcile the parties' competing considerations. A
way by which they are both accommodated is through a pricing structure
with
[[Page 54444]]
alternate rate prongs and floors, below which the royalty revenue
cannot fall. This is precisely the bargain struck between Copyright
Owners and services in 2008 and 2012, and that has been the rate
structure through 2017. And, because the Majority and the Dissent found
that revenue diminution occurred only ``to an extent,'' rather than in
the pervasive (sweeping'') manner averred by Copyright Owners, there is
no sufficient reason in the record to depart from the bargained-for
multi-tiered rate structure in Phonorecords II that allows for price
discrimination but tempers its impact on royalties through the use of
minima and floors.
e. Copyright Owners' Claim of ``Inherent'' Economic Value Is Belied by
the Record, Including Their Own Arguments
Pre-remand, Copyright Owners approached this rate setting process
with an overarching premise: A musical work has an ``inherent value''
that must be reflected in the royalty rates. As the NMPA's president,
Mr. Israelite testified, when asked how ``inherent'' value is defined:
[W]hoever owns an individual copyright is the one to define it.
I think that would be the most appropriate definition of it. What
someone is willing to license it for would be that inherent value to
that owner . . . That would be market value.
3/29/17 Tr. 3707 (Israelite).
If the market for musical works was as atomistic as the above quote
assumes, the songwriter of an individual musical work could indeed set
his or her own royalty rate, and refuse to license to any streaming
service or other distributor who refused to pay that royalty. But that
is not how the licensing market works.\133\ Songwriters typically
assign their licensing rights to music publishers (to avoid ruinous
transaction costs). These music publishers control huge ``Must Have''
repertoires that are offered under blanket licenses to streaming
services. (The musical works market of course is subject to a
compulsory license, but this is precisely how the unregulated market
works for the licensing of sound recordings by labels to interactive
streaming services.) It is acknowledged even by Copyright Owners' own
expert witness, Professor Watt that the creation of these large
collectives generates market power that necessitates rate regulation.
See R. Watt, Copyright and Economic Theory: Friends or Foes at 163, 190
(2000) (quoted in Dissent at 35).
---------------------------------------------------------------------------
\133\ The record does not include evidence of self-marketing by
songwriters through social media or via negotiation of individual
royalty contracts by the exercise of overwhelming star power,
whether through traditional payment mechanisms or new methods, such
as the murky mechanism of non-fungible tokens (NFTs). The absence of
incidents of such self-marketing from the record evidence in this
proceeding suggests that they likely constitute but a small segment
of the songwriter/publisher market. Accordingly, such self-marketing
and individual negotiations do not impact the Judges' setting of
statutory rates in this proceeding.
---------------------------------------------------------------------------
Further, this ``inherent'' market value notion is antiquated as a
matter of economics. Although an individual Copyright Owner can
announce his or her ``asking'' royalty, that is not sufficient to
generate a ``market'' royalty, unless and until a licensee agrees to
pay it. In market-based economics. that is to say, the economic
consensus that has governed economics since the ``marginal revolution''
in the mid to late 19th century, value is ascertained through the
intersection of supply and demand, with the price established at the
margin representing the market value of the good or service bought and
sold.\134\ If there is no demand for a product, be it a musical work or
anything else, it has no economic value. Even though costs have been
incurred to produce the product, those costs cannot be recovered (or
profit earned) absent a sufficient WTP in the market. And, as noted
supra, the product being offered and at issue here is comprised of
``second copies'' of sound recordings (with embedded musical works),
which are costless to reproduce for streaming purposes. Of course,
these ``second copies'' do have actual value when they are in demand,
and the royalties that their licensing generates must cover: (1) the
first copy (creative) costs; (2) the ``opportunity cost'' (measured by
the next best alternative for royalty earnings if the ``second copies''
could have been supplied through another distribution channel that paid
higher royalties to attract the end-user/consumer at issue); and (3)
profits to induce the creation of musical works.
---------------------------------------------------------------------------
\134\ As one scholar has summarized the 19th century transition
from classical to neoclassical economics: ``By the early 1870s,
economics reached a tipping point, and it ushered in a revolution in
thought, signaling the beginning of the ``modern,'' or
``neoclassical'' era. Marginalists flipped classical economics on
its head. Instead of focusing on the production side of economics,
they turned to consumption. It is the satisfaction of the wants of
consumers that matters for value, not the labor required for
production. What established the overall value of a good is the
value fetched by the final unit of that item on the market. As more
units of a good are produced, the marginal value of the last unit
tends to decrease. . . . According to marginal utility, the
consumer, not the producer, therefore drives the valuation
process.'' J. Wasserman, The Marginal Revolutionaries at 28 (2019).
This transformation reflected the abandonment of the ``labor theory
of value''--the cornerstone of Marxian economics. See E.R.
Canterbury, A Brief History of Economics at 111 (2001) (``Marx's
devotion to a labor theory of value was complete.''). It initially
appears as irony that Copyright Owners espouse a Marxian approach to
value while preaching the virtues of unregulated markets. The
initial whiff of irony dissipates when one appreciates that a
collective licensor with the market power of control over a ``Must
Have'' input has every incentive to urge a pricing or valuation
method that takes the focus away from the force of consumer demand
in an effectively competitive market, which is a hallmark of
neoclassical economics.
---------------------------------------------------------------------------
Second, the fact that Copyright Owners originally proposed a per-
subscriber alternative rate to their per-play rate itself belies their
conviction that some ``inherent'' economic value exists. When the
metric of value switches from ``per-play'' to ``per-subscriber,'' the
focus of value likewise shifts from an emphasis on producer value to
consumer value. That is, if there is truly an ``inherent'' value for a
product or service, that singular value cannot divide into two distinct
values with the ``greater-of'' the two controlling. Such an argument
gives away the game, so to speak, demonstrating, perhaps
unsurprisingly, that economic arguments (not unlike legal advocacy) are
often situational--designed to support maximalist positions and the
exercise of market power, however acquired. See also Determination at
28 n.64 (rejecting the ``inherent value'' argument).
f. PR II-Based Benchmark Not ``Too Complex''
Copyright Owners and the Majority complained that the PR II-based
benchmark is too complex. See Copyright Owners' PFF ] 12 (criticizing
complexity of PR II Rates as lacking ``transparency''); Determination
at 36 (characterizing parties' negotiated, renewed, and expanded rate
structure as Rube-Goldberg-esque in complexity and impenetrability.'')
After considering this issue on remand, the Judges disagree. If
some songwriters or lyricists have been confused by their royalty
statements, their confusion of course should be resolved. However, one
of the benefits of a collective is that it possesses the expertise and
resources to identify and explain how royalties are computed and
distributed. Moreover, this claim of complexity cannot serve as a basis
to override the multi-part negotiated benchmark that the parties,
through their respective trade associations, negotiated and
implemented. As the Dissent stated: ``There is no good reason why the
rate structure that is consonant with the parties' ten-year history and
with the relevant economic model should be sacrificed on the slender
[[Page 54445]]
argument that ``simpler is better than complicated.'' Dissent at
88.\135\
---------------------------------------------------------------------------
\135\ Copyright Owners' concern for transparency has apparently
evaporated in connection with its eagerness to adopt the proffered
uncapped TCC rates. Under that approach, the definition of revenue,
the handling of bundled products and the exclusion of certain
consideration from the royalty base will remain opaque to
songwriters--and to the Judges.
---------------------------------------------------------------------------
Further, section 801(b)(1) does not identify ``simplicity'' as a
statutory goal for the setting of rates, rate structure, and terms.
Although there is certainly no need for gratuitous complexity, the
price discriminatory structure and the associated levels of rates in
the PR II-based benchmark that were eliminated by the Majority (while
maintaining all the remaining complexity) were most certainly not
gratuitous, but rather designed, after negotiations, to establish a
structure that would expand the revenues and royalties to the benefit
of Copyright Owners and Services alike, while also protecting Copyright
Owners from potential revenue diminution by the Services. Moreover,
when the market itself is complex--in that the WTP across consumer
groups is heterogeneous and the offerings reflect that fact--it is
unsurprising that the regulatory provisions would resemble the complex
terms in a commercial agreement negotiated in such a setting. For the
Judges to demand simplicity in this context would be to sacrifice the
specificity that an effectively competitive market requires. See
Dissent at 88 (rejecting the simplicity argument by invoking the advice
attributed to Albert Einstein that ``[e]verything should be made as
simple as possible, but no simpler.''
g. So-Called Statutory ``Shadow'' Does Not Diminish Value of the PR II-
Based Benchmark Rates
Copyright Owners maintain that the rates in the PR II-based
benchmark are infirm because, like any benchmark for which a statutory
rate is the default, they are not actual market rates. That is, such a
rate is said to exist in the so-called ``shadow'' of the statutory
rate. See Dissent at 70 (and citations therein).
The Judges reject this argument for several reasons. First, the
argument is undercut by the explicit language of section 115 of the
Copyright Act, which states: ``In addition to the objectives set forth
in section 801(b)(1), in establishing such rates and terms, the
Copyright Royalty Judges may consider rates and terms under voluntary
license agreements described in subparagraphs (B) and (C).'' 17 U.S.C.
115(c)(3)(D). Subparagraphs (B) and (C), respectively, refer to
agreements on ``the terms and rates of royalty payments under this
section'' by ``persons entitled to obtain a compulsory license under
[17 U.S.C. 115(a)(1)]; and ``licenses'' covering ``digital phonorecord
deliveries.'' Id. Thus, it is beyond dispute that Congress has
authorized the Judges, in their discretion, to consider such agreements
as evidence, irrespective of--or perhaps because of--the shadow cast by
the compulsory license. Thus, the appropriate question is how much
weight the Judges, in their discretion, should afford such benchmarks
in any particular proceeding.
There is no basis to find, as Copyright Owners suggest, that
statutorily-based or influenced benchmarks, including specifically the
PR II-based benchmark in this proceeding, are per se inferior to other
benchmarks or alternative economic evidence (e.g., from models, surveys
or experiments) that may be unaffected by the shadow. Those other
benchmarks or forms of evidence will also be subject to their own
imperfections and incompatibilities with the target market and must be
identified and weighed accordingly.\136\ Thus, the Judges must not only
consider (i) the importance, vel non, of any potential so-called
``shadow-based'' distortionary effects from a benchmark derived from a
regulated statutory benchmark market, but also (ii) how any such
purported ``shadow'' effects compare to any distortions generated by
other proffered benchmarks and competing alternative economic evidence,
e.g., distortions based on complementary oligopoly power, bargaining
constraints and product differentiation in other benchmarks, models,
surveys or experiments.\137\
---------------------------------------------------------------------------
\136\ It has been famously and wisely said that ``all models are
wrong, but some are useful.'' G. Box & N. Draper, Empirical Model-
Building at 424 (1987). Benchmarks, Shapley, and Nash models,
surveys and experiments are all models, in that ``[a] model is a
representation of something beyond itself . . . being used as a
representative of that something, and in prompting questions of
resemblance between the model and that something . . . substitute
systems . . . directly examined . . . to indirectly acquire
information about their target systems.''). U. Maki, Models are
Experiments, Experiments are Models, 12 J. Econ. Meth. 303 (2005).
\137\ It is also important to note that the reasonable rate and
rate structure identified under the section 801(b)(1) standard
(before considering the four itemized statutory factors) need not be
a market-based rate, as discussed infra.
---------------------------------------------------------------------------
The Services' experts discount the foregoing shadow-based
criticism. Moreover, the Services laud a statutorily-influenced
benchmark in general, and the specific PR II-based benchmark in
particular, because the latter reflects more equal bargaining power
between licensors and licensees. In this regard, one of the Services'
economic expert witnesses, Professor Katz, points out that rates set
voluntarily by the parties in a settlement under the ``shadow'' provide
two important benefits. First, with a statutory rate-setting proceeding
as a backstop, large licensors cannot credibly threaten to ``hold out''
and ``walk away'' from the negotiations without an agreement, thereby
negating their ability to use their ``must have'' status to obtain
rates above effectively competitive levels. Second, when, as here, such
negotiations are conducted with all the parties at the figurative
table--including here, trade associations--no single party has
disproportionate market power in the negotiations. See 3/13/17 Tr. 661
(Katz).
The Judges agree that settlement agreements reached in the
statutory shadow are useful. Although imperfect when considered in
isolation, in that the statutory proceeding is the default backstop, in
context they negate the power of any entity simply to refuse to strike
a deal. The negation of that power blunts the complementary oligopoly
power of licensors of ``Must Have'' repertoires (whether musical works
or sound recordings), making a benchmark agreement reached in the so-
called ``shadow'' advantageous in establishing an effectively
competitive rate. See Web IV, supra, 26,316, 26,330-31 (May 2, 2016)
(noting counterbalancing effect of statutory license in establishing
effectively competitive rates). Further, when such settlement
agreements are industrywide, they tend to eliminate disproportionate
market power, See Dissent at 72; Web III, 79 FR 23102, 23111 (Apr. 25,
2014), aff'd Intercollegiate Broad. Sys., Inc. v. Copyright Royalty
Bd., Case No. 14-1098 (D.C. Cir. Aug. 11, 2015) (relying on two
settlement agreements).
Nonetheless, Copyright Owners are correct to note that,
hypothetically, some licenses might have otherwise been negotiated at
rates higher than the settlement rate that was affected by the so-
called shadow. But that is simply the tradeoff that the statutory
scheme makes in its identification of settlement rates as evidentiary
benchmarks. Such a theoretical problem cannot serve to override the
salutary aspects of benchmark settlement agreements. See Web IV, supra
at 26,630 (rejecting same argument as speculative and ``too untethered
from the facts to be predictive or useful in adjusting for the supposed
shadow of the existing statutory rate.'').
Lastly, with regard to a benchmark affected by the so-called
``shadow,'' the Judges find that, with regard to the application of the
itemized factors in
[[Page 54446]]
section 801(b)(1), they have the same duty to independently weigh those
factors as they do for all otherwise reasonable rates. Thus, the Judges
reject the idea that rates and terms reached through a settlement must
be understood to supersede--or can be assumed to embody--the Judges'
current thinking as to the application of the statutory elements set
forth in section 801(b)(1). The Judges are obliged to conduct the four-
factor analysis anew when considering a previously adopted settlement
in a subsequent proceeding--and they do so infra. Of course, if on such
further analysis, the Judges find that the provisions in an otherwise
useful benchmark agreement (including those in a benchmark influenced
by the so-called ``shadow'') do appropriately reflect the four itemized
statutory factors in section 801(b)(1), then the Judges may adopt the
provisions of that settlement without a factor-based adjustment.
h. Conclusion Regarding PR II-Based Benchmark
Accordingly, the Judges find the PR II Rates to be a useful
benchmark. However, this benchmark is modified by the Judges'
substitution of the 15.1% headline percentage rate for the 10.5%
headline percentage rate in the benchmark.
D. Precedent Permits Judges To Apply Elements of PR II Rates, Rate
Structure and Terms Even if Those Are Not Proffered as Benchmarks
The D.C. Circuit has previously held that the Judges have the
authority to adopt elements from the existing rate provisions, if they
find that those prevailing provisions better satisfy the statutory
requisites than any other proposed structures and rates discernible
from the record evidence. Music Choice v. Copyright Royalty Bd., 774
F.3d 1000, 1009 (D.C. Cir. 2014). This authority exists even when no
party has proffered those provisions in the form of a benchmark.
In Music Choice (concerning the setting of satellite radio royalty
rates under the same section 801(b)(1) standard), the CRB Judges
rejected the parties' proffered benchmarks and instead relied on a
percent-of-revenue rate (13%) that was neither a benchmark nor even the
prior statutory rate, but merely ``a component of a prior
determination.'' Id. at 1009. The licensor-party, SoundExchange,
argued, like Copyright Owners here, that this component of a prior rate
was ``stale,'' ``outdated,'' or ``obsolete.'' Rejecting this argument
as ``erroneous,'' the D.C. Circuit stated that ``the Judges did not
consider the 13% rate as a current benchmark,'' but rather used it to
``bridge the gap'' caused by the inadequacies of the parties' rejected
benchmarks. Id. In so doing, the D.C. Circuit held that the Judges
properly resolved ``serious problems'' with the licensor's proposal,
even as it had ``partially credited it'' and also ``used permissible
indicia of reasonableness to help fix the rate.'' Id.
Music Choice is highly instructive. Here, on remand, the Judges
adopt a modified version of the prior rate structure and rates in
Phonorecords II. The fact that it was also proffered as a benchmark, in
another modified form by the Services, does not render Music Choice
inapposite. Rather, because the Phonorecords II provisions were
proffered as benchmark evidence, these provisions were placed squarely
into the record, allowing the parties and the Judges to address the
relative merits. A fortiori, Music Choice underscores the propriety of
the Judges approach in this proceeding. That is, even if the Services
had not proffered this approach as a benchmark, Music Choice allows the
Phonorecords II approach to serve as a guidepost for establishing the
rates and rate structure in this proceeding.
Further, here the Judges are adopting actual elements from the
prior rate provisions, rather than, as in Music Choice, a mere
``component'' used to generate the prior rate. A fortiori yet again,
Music Choice allows the Judges to prudently utilize the prior rate and
rate structure regulations to synthesize a determination in this
proceeding. The analogous nature of Music Choice is also seen in the
Judges' use in the present case of the ``headline'' 15.1% revenue rate
proposed by Copyright Owners on remand combined with elements from the
PR-II regulatory provisions, including its price discriminatory rates.
In Music Choice, the Judges likewise ``partially credited'' the
licensor's proposal, which, as noted supra, the D.C. Circuit affirmed.
Finally, the Judges take note that Music Choice also addressed the
Judges' findings regarding the setting of another statutory license,
for Preexisting Subscription Services (PSS), by using a rate in a
settlement from a prior period. This context is also analogous here,
because Copyright Owners object to the use of the Phonorecords II rate
structure and rates as the product of a settlement. It is instructive
to consider how the arguments of the licensor (SoundExchange) in Music
Choice mirror those of Copyright Owners in this proceeding:
SoundExchange notes that this rate ``is the product of
settlement negotiations that occurred in SDARS I between Music Choice
and SoundExchange.''
SoundExchange argues that the Judges arbitrarily rejected
. . . more recent data points in favor of the ``outdated'' settlement
rate.
SoundExchange maintains that the Judges conceded that the
prevailing rate had limited value, as the settlement rate ``was
negotiated in the shadow of the statutory licensing system and cannot
properly be said to be a market benchmark rate.''
SoundExchange also argues that simply reciting that
``nothing in the record persuades the Judges'' that the prevailing rate
is unreasonable . . . does not show that [it] is reasonable, or that it
is supported by the written record.
[G]iven the lack of creditable benchmarks in the record,
the Judges did not err when they used the prevailing rate as the
starting point of their Section 801(b) analysis.
The Copyright Act contemplates that the Judges would . . .
consider ``prior determinations'' and rates established ``under
voluntary license agreements.''
[T]he Judges did not err when relying on the settlement
rate. The Judges conceded that the settlement rate does not represent a
market rate. . . . But . . . the relevant portion of the Copyright Act
does not use the term ``market rates,'' nor does it require that the
term ```reasonable rates'' be defined as market rates. . . . The Act
authorizes the Judges to consider rates set ``under voluntary license
agreements.''
Music Choice complains that it agreed to a higher rate to
avoid litigation costs, but has not introduced evidence that the
settlement was involuntary or otherwise unreasonable. It was not
arbitrary, then, for the Judges to consider the voluntary settlement
rate.
Music Choice, 774 F.3d at 1012-15. These aspects of Music Choice are
highly instructive, considering the Judges' parallel findings regarding
the same and similar arguments as discussed supra regarding prior
settlement agreements and the so-called ``shadow'' of the statutory
rates.
In sum, Music Choice provides ample support for the conclusion
that, even if the Services had not proffered their PR II-based
benchmark, the Judges would have acted well within their authority to
give the same weight to the PR II rates and structure as they have in
this Initial Ruling.\138\
---------------------------------------------------------------------------
\138\ This ruling is in no way conflicts with the Judges' duty
to set rates, rate structures, and terms de novo in each rate
proceeding, as discussed supra. The de novo process requires the
Judges to weigh new evidence regarding potential new rates, rate
structures, and terms, but that is not inconsistent with the Judges'
ability, as explicated by the D.C. Circuit in Music Choice, to adopt
prior rates, rate structures, and terms in whole or in part if, in
their discretion, the new evidence is deficient. See Music Choice,
supra, at 1012 (``The Judges were under no obligation to salvage
benchmarks they found to have fundamental problems.'').
---------------------------------------------------------------------------
[[Page 54447]]
E. Four Itemized Factors in Section 801(b)(1)
The Judges have considered the application of the four itemized
statutory factors A through D, in connection with their application of
the 15.1% revenue rate and their partial use of the PR II-based
benchmark.
1. Factor A
The Judges have explained supra that price discrimination is a
``win-win'' for Copyright Owners and the Services. By serving low WTP
listeners, it brings in new listeners and subscribers who increase
royalty payments as well as revenues. Any licensor would prefer to
increase its royalties, rather than ``leave money on the table,'' and a
rate structure that effects such an increase (through the concept of
``derived demand'') is appropriate. Moreover, for purposes of applying
Factor A, a rate structure that increases royalties, ceteris paribus,
would induce more production of musical works, a result that Copyright
Owners should desire.
This point appears to raise a question: How could Copyright Owners
and their economic experts object to a rate structure that inures to
their benefit as well? The answer is: They do not object. They are not
economic naifs. As stated supra, they advocate for a rate set under the
bargaining room theory, through which rate structures can still be
negotiated, but not subject to the ``reasonable rate'' and itemized
factor analysis required by law. In those negotiations, as Dr. Eisenach
candidly acknowledged, Copyright Owners would have a different threat
point to use in order to obtain better rates and terms. 4/4/17 Tr.
4845-46 (Eisenach).
Second, given a heterogeneous downstream WTP, it would not be more
profitable simply to equate ``availability'' with a higher rate. As
noted supra, any product that is priced beyond the WTP of a significant
portion of the public is unavailable to that segment.\139\ Royalties
that are aligned with the varying WTP of different classes of listeners
will make downstream price discrimination more affordable to the
services, driving new revenue and royalties--precisely as the PR II-
based benchmark allows.\140\ In this regard, Copyright Owners have
taken a cramped and unrealistic view of such incentives. In particular,
the Judges disagree with Copyright Owners' expert economic witness,
Professor Rysman, who startlingly asserted in response to a
hypothetical from the bench that even a $10,000 per month subscription
price would increase ``availability.'' 4/3/17 Tr. 4397 (Rysman).
---------------------------------------------------------------------------
\139\ The concept of willingness-to-pay (WTP) as used by
economists is an antiseptic phrase, because it includes not merely
people who do not value a music streaming subscription highly, but
also individuals and families who are ``income constrained'' (yet
another antiseptic phrase, read ``low income'' people and families)
who lack the ``ability-to-pay'' for an interactive subscription.
That segment of the population likely reflects a significant portion
of the nation, because ``40% of Americans would struggle to come up
with even $400 to pay for an unexpected bill,'' let alone pay for a
music streaming service. See https://www.minneapolisfed.org/article/2021/what-a-400-dollar-emergency-expense-tells-us-bout-the-economy.
When the royalty rates paid by interactive services enable streaming
services to satisfy the demand of these low-income consumers
(through the principle of ``derived demand'') that segment of
American society can enjoy the benefits of listening to interactive
streamed music, even if the offerings they can afford lack the large
catalogs and ``bells and whistles'' of a pricier service.
\140\ To be sure, royalties will not increase in equal
proportions with increases in the number of streams or listeners,
but that is a feature of price discrimination, not a bug. The goal
is to generate revenues from low WTP listeners who otherwise would
be lost as sources of revenues and royalties to both the interactive
services and Copyright Owners.
---------------------------------------------------------------------------
The Judges find Professor Rysman misapprehends the nature of a
price signal. If the price is so high as to eliminate or reduce total
revenue to creators, in no way will higher rates simply induce the
supply of creative works over time. Indeed, even monopolists do not
seek the highest price possible, but rather seek to maximize profits.
See E. Mansfield & G. Yohe, Microeconomics at 362-63 (11th ed. 2004)
(``Monopolies maximize profits by producing where marginal cost equals
marginal revenue.''). Thus, even monopolists, who have the most market
power, are constrained in their pricing by the demand curve and the
marginal revenue it creates. Simply put, although a higher royalty rate
might have an immediate superficial appeal, if the consequence will be
lower revenues, the high per-play rate would reveal itself as a form of
fool's gold.
In sum, the Judges find that the Factor A objective of ``maximizing
the availability of creative works'' is furthered by an upstream rate
structure that contains multiple royalty rates reflective of and
derived from downstream variable WTP, because it will facilitate
beneficial price discrimination. Such price discrimination allows for
access to be afforded ``down the demand curve,'' making musical works
available to more members of the public. However, there is no evidence
to suggest that the price discriminatory rates should be changed, in
order to address the connection between price discrimination and the
objective of Factor (A). Accordingly, the Judges find no basis to
adjust either the rate structure or the rates based on Factor (A).
2. Factors B and C
The concepts of ``fair income,'' ``fair return'' and recompense for
costs and other contributions was considered in connection with the
setting of the 15.1% revenue rate. In that context, the Judges analyzed
the Shapley Value modeling that was designed to generate ``fair'' rates
that allowed the parties to recover their costs and to share the
surplus (over and above costs) in a manner that: (1) prevented the
``Must Have'' Input Suppliers (the record companies and Copyright
Owners) from using the essential aspect of their inputs to engage in
hold-up by threatening to withhold their respective repertoires; and
(2) allocated surplus shares according to each party's contribution to
the surplus (as calculated though the ``arrival orderings'' in the
Shapley model).\141\
---------------------------------------------------------------------------
\141\ As noted elsewhere in this Initial Ruling, Professor Marx,
Spotify's economic expert witness, reduced the relative market power
of the input suppliers in her model which she claimed would be
consonant with the ``fairness'' objectives in Factor B. On behalf of
Copyright Owners, Professor Watt disagreed, arguing that the Shapley
approach takes the existing market power as reflective of the
parties' market contributions, and thus needs no adjustment. The
Majority utilized Professor Marx's Shapley-based calculation of a
total royalty payment of [REDACTED]% of service revenue in setting a
15.1% revenue rate (phased-in), which the Judges are adopting in
this Initial Ruling. The Majority also used Professor Marx's
calculation to find that Factors B and C were satisfied without
further adjustment. See Determination at 68 & n.120, 75, 86-87. But
this issue is not relevant to the present discussion of Factors B
and C with regard to the application of the PR II-based benchmark.
---------------------------------------------------------------------------
The PR II-based benchmark was the product of an industrywide
negotiation, with the music publishers represented by the NMPA and the
interactive streaming services represented by DiMA, their respective
trade associations. As explained in the Dissent, supra, at pp. 137-39,
when an industrywide settlement is reached, particularly when the
default procedure is a contested rate proceeding before the Judges, it
contains the same benefits with regard to the avoidance of the ``hold-
out'' effect and the equalizing of bargaining power as produced by
Professor Marx's Shapley value modeling. See 3/13/17 Tr. 577 (Katz)
(``I think of the shadow as balancing the bargaining power between the
two
[[Page 54448]]
parties.''); Katz CWRT 136, n.236 (``there are market forces that
promote the achievement of the statutory objectives in private
agreements, such as the 2012 Settlement, when the parties are equally
matched (it was an industry-wide negotiation) and the negotiations are
conducted in the shadow of a pending rate-setting proceeding that can
be expected to set reasonable rates in the event that the private
parties do not reach agreement.'').
Accordingly, this benchmark already incorporates the dynamics of a
negotiation between parties with mutually countervailing power
(although those dynamics required updating of the headline rate to
15.1% to account for the higher revenues, as undertaken by the
Majority's Shapley analysis). See Web V, 86 FR 59452, 59456 (Oct. 27,
2021) (``the licensor-side complementary oligopoly power could be
ameliorated by the ``countervailing power'' of a licensee'').
Therefore, the Judges do not make any adjustment in their
application of the PR II-based benchmark pursuant to Factors B and C.
3. Factor D
As noted supra, the Judges understand that a Factor D adjustment is
warranted if the rate the Judges would otherwise establish
directly produces an adverse impact that is substantial, immediate
and irreversible in the short-run because there is insufficient time
for either [party] to adequately adapt to the changed circumstance
produced by the rate change and, as a consequence, such adverse
impacts threaten the viability of the music delivery service
currently offered to consumers under this license.
Determination at 87.
There is no record evidence to suggest that the Services' PR II-
based benchmark, as utilized by the Judges in this Initial Ruling,
would create the requisite ``adverse impact'' to trigger Factor D. The
Services certainly do not assert that their own proffered benchmark
would be disruptive. With regard to Copyright Owners, the Judges cannot
identify any aspect of the PR II-based benchmark that would cause the
type of disruption that can serve as an adjustment under the statutory
language of Factor D or the Judges' application of same, as quoted
above. The Judges understand Copyright Owners' complaint to be
principally that [REDACTED] during the Phonorecords II period,
[REDACTED] the number of musical works streamed via sound recordings
performed on interactive services. However, that is most certainly not
any sort of disruption, let alone a disruption cognizable under section
801(b)(1) and under the Judges' application of that provision.
F. Subpart C Offerings Covered by Foregoing Analysis
The Phonorecords II parties also negotiated several new service
types--paid locker services, purchased content locker services, mixed
service bundles, music bundles and limited offerings. These service
configurations were described in subpart C of 37 CFR 385 under the
Phonorecords II regulatory provisions.\142\ Parness WDT ] 13; Levine
WDT ]] 38-39; Israelite WDT ]] 28-30. These negotiations spanned more
than a year. See 3/29/17 Tr. 3652-55 (Israelite) (involved protracted
bargaining, in which NMPA rejected some categories, while others were
accepted and became part of subpart C). Id. at 3654-56. The parties
ultimately agreed on a structure for subpart C that resembled the
subpart B structure, including a headline percentage of revenue royalty
rate and per-subscriber and TCC minima. Parness WDT ] 14; see also 37
CFR 385.22. As with the bundling negotiations relating to subpart B,
the parties negotiated and created a bundled service category under
subpart C (with certain adjustments to the definition of ``revenue.'')
3/8/17 Tr. 161-64 (Levine); 37 CFR 385.21.
---------------------------------------------------------------------------
\142\ The interactive steaming (and limited download) provisions
that are the principal subject of this proceeding were contained in
subpart B of the Phonorecords II (and Phonorecords I) regulations.
(These subparts were reorganized pursuant to the now vacated
Determination.)
---------------------------------------------------------------------------
Copyright Owners urge the elimination of the subpart C provisions
as essentially obsolete because locker services for ``purchased
content'' (new download purchases) and for ``paid'' downloads (already
owned) have largely disappeared, as listeners transitioned away from
ownership models to access models. See 3/8/17 Tr. 159-160 (Levine); 3/
16/17 Tr. 1458-1461 (Mirchandani); Mirchandani WDT ] 33; 3/22/17 Tr.
2523 (Dorn). Copyright Owners also re-assert the same arguments with
respect to subpart C as they have for interactive streaming in subpart
B. See CORPFF-JS at p.2.
The Services argue that Copyright Owners do not point to any
evidence to show that locker services have completely disappeared,
emphasizing that Apple and Amazon continue to offer locker service.
Joyce WDT ] 5; Mirchandani WDT ]] 16-17; 3/22/17 Tr. 2523-25 (Dorn);
Ramaprasad WDT, Table 3. More generally, the Services urge the Judges
to use the subpart C rate structure as the benchmark for rates in the
forthcoming period for the same reasons as they urge the use of the
subpart B rates as an appropriate benchmark. See Mirchandani WDT ]] 58-
62.
The Judges find no reason on remand to treat the subpart C
offerings differently than the manner in which they are treating the
subpart B interactive streaming offerings, for the reasons set forth in
the Dissent at 118-119. That means, however, that the various
``headline'' rates for these subpart C offerings must also adjust to
15.1%,\143\ whereas the alternative rates (identified in subpart C as
``minima'' and ``subminima)'' rates shall remain unchanged.
---------------------------------------------------------------------------
\143\ Accordingly, in the PR II-based benchmark, the subpart C
``headline'' rates that shall adjust to 15.1% are: 11.35% for Mixed
Service Bundles; 11.35% for Music Bundles; 10.5% for Limited
Offerings; 12% for Paid Locker Services; and 12% for Purchased
Content Locker Services. See 37 CFR 385.22(a)(1) (Step 1);
385.23(a)(1) through (5).
---------------------------------------------------------------------------
IV. Change in Definition of Service Revenue for Bundles 144
---------------------------------------------------------------------------
\144\ Judge Strickler disagrees with the procedural analysis of
a different majority by which they readopt the Bundled Revenue
definition from the Initial Determination, and he dissents on that
specific issue. However, Judge Strickler concurs and joins with the
Majority regarding the substantive re-adoption of that definition
from the Initial Determination. Judge Strickler has drafted a
separate opinion on this Bundled Revenue issue.
---------------------------------------------------------------------------
The Judges analyze the definition of ``Service Revenue'' for
bundled offerings in the context of the partial adoption of the PR II-
based benchmark. As discussed supra, the Judges have found that the PR
II-based benchmark is a useful benchmark, particularly because of its
features that incentivize beneficial downstream price discrimination
that generates more listeners, revenues, and royalties.
A. Background
In their Initial Determination, the Judges adopted a definition of
``Service Revenue'' (i.e., a royalty base) for a ``Bundle'' \145\ that
provided, in pertinent part:
---------------------------------------------------------------------------
\145\ For interactive streaming, the Judges' Initial
Determination defined a ``bundle'' (in pertinent part) as an
offering which combined the delivery of streamed music: ``together
with one or more non-music services . . . or non-music products . .
. as part of one transaction without pricing for the music services
or music products separate from the whole offering. . . .'' Initial
Determination, Attachment A at 2 (Sec. 385.2 therein).
Service Revenue shall be the revenue recognized from End Users for
the Bundle less the standalone published price for End Users for
---------------------------------------------------------------------------
each of the other component(s) of the Bundle . . .
Initial Determination, Attachment A at 7 (Sec. 382.2 therein).\146\
---------------------------------------------------------------------------
\146\ The definition added: ``[I]f there is no standalone
published price for a component of the Bundle, then the Service
shall use the average standalone published price for End Users for
the most closely comparable product or service in the U.S. or, if
more than one comparable exists, the average of standalone prices
for comparables.'' Id. at 7-8.
---------------------------------------------------------------------------
[[Page 54449]]
After the Judges issued their Initial Determination, Copyright
Owners submitted a Motion for Clarification or Correction of
Typographical Errors and Certain Regulatory Terms which disclaimed any
intent to seek rehearing, but sought ``clarification or correction'' of
certain regulatory terms to conform them to what Copyright Owners
claimed to be the apparent intent of the Initial Determination. (Motion
for Clarification).\147\ Copyright Owners purported to bring their
motion under the Judges' general regulations governing motions. See 37
CFR 303.3 and 303.4 (formerly codified at 37 CFR 350.3 and 305.4).
---------------------------------------------------------------------------
\147\ Streaming Services submitted a motion for rehearing that
was limited to fixing clerical errors and clarifying existing
ambiguities in the proposed regulatory terms appended to the Initial
Determination.
---------------------------------------------------------------------------
The Motion for Clarification argued, among other things, that the
definition of Service Revenue as applied to bundled offerings should be
reworked. Copyright Owners argued that defining the revenue as the
total price of the bundle, minus the standalone published prices for
the non-streaming offerings in the bundle, undervalued the revenue
created by the streaming offerings. They proposed that ``Service
Revenue'' for bundled offerings be defined as the standalone price of
the offering (or comparable offerings).
The Services objected to Copyright Owners' styling of their motion
as something other than a motion for rehearing. The Services also
objected that Copyright Owners had not previously proposed a definition
of ``Service Revenue'' for bundled offerings, and that their ``late-
proposed'' definition was unsupported by the record.
On October 29, 2018, the Judges issued an Order concluding neither
party had met the exceptional standard for granting rehearing
motions,\148\ stating that the parties had failed to present ``even a
prima facie case for rehearing under the applicable standard''. Amended
Order Granting in Part and Denying in Part Motions for Rehearing (Order
on Rehearing) (Jan. 4, 2019).\149\
---------------------------------------------------------------------------
\148\ The standard is set forth in the Order on Rehearing at 2
n.3. The Judges discuss and apply this standard infra, pursuant to
Johnson, and in the context of this remand proceeding.
\149\ Judge Strickler, who had dissented from the Initial
Determination and the Determinations, did not join in this Order on
Rehearing.
---------------------------------------------------------------------------
The Judges explained that they nevertheless found it appropriate to
resolve the issues that the parties had raised. Order on Rehearing at
2. The Judges added that, to the extent such resolution could be
considered a rehearing under 17 U.S.C. 803(c)(2), the Judges resolved
the motions on the papers without oral argument. Id.
Regarding the definition of ``Service Revenue'' for bundled
offerings, the Judges summarized the parties' competing arguments:
Copyright Owners presented evidence that the existing approach
led, in some cases, to an inappropriately low revenue base--but did
so in service to their argument that the Judges should reject
revenue-based royalty structures. They did not present evidence to
support a different measure of bundled revenue because their rate
proposal was not revenue-based. The Services rely on the fact that
the approach to bundled revenue in the extant regulations is derived
from the 2012 Settlement. The Judges have, however, declined to rely
on the 2012 Settlement as a benchmark, as the basis for the rate
structure, or, therefore, as regulatory guidance.
The Services have observed correctly that the evidentiary
records in Web IV and SDARS III differ from the record in this
proceeding.\150\
---------------------------------------------------------------------------
\150\ In Web IV and SDARS III, unlike under the Phonorecords II-
based benchmark, there were no minima or floors to provide licensors
with royalties in the event bundled offerings would otherwise fail
to generate royalties.
Order on Rehearing at 17 (emphasis added).
Despite these arguments, the Judges found that neither party
presented evidence adequate to support the approach advocated in post-
determination filings, because ``the `economic indeterminacy' problem
inherent in bundling'' remained unresolved.'' Id.\151\ The Judges
stated that the Services were the party in possession of the relevant
information, and concluded that the Services bore the burden of
providing evidence that might mitigate the ``indeterminacy problem''
inherent in bundling. Because the Judges concluded that the Services
had not met that burden, they ruled that they must adopt an approach to
valuing bundled revenue that is in line with what the Copyright Owners
proposed. As a result, the Judges discarded the formula in the Initial
Determination and ruled, instead, that streaming service providers will
use their own standalone price (or comparable) for the music component
(not to exceed the value of the entire bundle) when allocating bundled
revenue. Id. at 16-18.
---------------------------------------------------------------------------
\151\ The ``economic indeterminacy'' problem was described in
SDARS III: ``Such bundling [for full quotation, see eCRB no. 27063
n.140].'' SDARS III, 83 FR 65264. As discussed in this Initial
Ruling, this indeterminacy problem was addressed by the Phonorecords
II-based benchmark through negotiated alternative royalty provisions
for bundled offerings.
---------------------------------------------------------------------------
Consistent with the Judges' Order on Rehearing, the Judges'
replaced the definition of ``Service Revenue'' for a ``Bundle'' that
they had included in the Initial Determination with a new definition in
the Determination. The final definition provided, in pertinent part:
Service Revenue shall be the lesser of the revenue recognized
from End Users for the bundle and the aggregate standalone published
prices for End Users for each of the component(s) of the bundle that
are Licensed Activities . . . [or] if there is no [such] standalone
price, then the average standalone . . . price . . . for the most
closely comparable product or service . . . or . . . the average of
standalone prices for comparables.
Determination, Attachment A at 8.
The Services, Copyright Owners and George Johnson appealed the
Judges' Determination to the D.C. Circuit. See Johnson, 969 F.3d 363.
The Services challenged both the Judges' legal authority and the
substantive soundness of the decision to reformulate the definition of
``Service Revenue'' for bundled offerings, after the Judges had issued
the Initial Determination.
The D.C. Circuit examined several authorities under which the
Judges may revisit and amend a determination. It addressed the three
ways identified in the statute: ``(i) order rehearing `in exceptional
cases' in response to a party's motion, 17 U.S.C. 803(c)(2)(A); (ii)
correct `technical or clerical errors,' id. Sec. 803(c)(4); and (iii)
`modify the terms, but not the rates' of a royalty payment, `in
response to unforeseen circumstances that would frustrate the proper
implementation of [the] determination.' '' Johnson, 969 F.3d at 390.
The D.C. Circuit found that the Judges' reformulation of the definition
of ``Service Revenue'' fit none of those categories.
The D.C. Circuit noted that the Judges were explicit that they did
not treat the Motion for Clarification as a motion for rehearing under
17 U.S.C. 803(c)(2). Id. Furthermore, the D.C. Circuit noted the
Judges' own findings that the Motion for Clarification did not meet the
exceptional standard for granting rehearing motions under section
803(c)(2) and that the Copyright Owners failed to make even a prima
facie case under the rehearing standard.
In Johnson, the D.C. Circuit found that the change to the
definition of Service Revenue for bundled offerings was not an exercise
of the Judges' authority under section 803(c)(4) to ``correct any
technical or clerical errors in the determination[.]'' 17 U.S.C.
803(c)(4).
[[Page 54450]]
The D.C. Circuit observed the substantive nature of the change to the
definition and determined that there was nothing technical or clerical
about the amendment. The D.C. Circuit found that the Judges did not
even purport to modify the terms in response to unforeseen
circumstances that would frustrate the proper implementation of the
Initial Determination. The D.C. Circuit observed that the Judges never
mentioned section 803(c)(4) or unforeseen circumstances as the basis
for revamping the Service Revenue definition.
Beyond the explicit statutory authorities for amendments to
determinations, the D.C. Circuit addressed arguments for inherent
authority to make sua sponte any appropriate substantive or fundamental
changes after the Initial Determination. The D.C. Circuit foreclosed
reliance on inherent authority, finding that Congress's decision to
limit rehearing to exceptional cases, and to confine other post hoc
amendments to cases involving technical or clerical errors, would be a
nullity if the Judges also had plenary authority to revise their
determinations whenever they thought appropriate. The D.C. Circuit
noted that the Judges' decision to amend the definition said nothing of
the sort, and prior decisions are silent on that topic.
In sum, the D.C. Circuit found that the Judges failed to explain
the legal authority for reformulating the definition of ``Service
Revenue.'' In relevant part, the D.C. Circuit ruled
we must vacate the [ ] Determination's bundled offering Service
Revenue definition and remand for the [Judges] . . . either to
provide `a fuller explanation of the agency's reasoning at the time
of the agency action[,]' or to take `new agency action' accompanied
by the appropriate procedures.
Id. at 392 (citing Regents, 140 S.Ct. at 1908).
Because the D.C. Circuit determined that the Judges failed to
identify any legal authority for adopting the new Service Revenue
definition, it found no occasion to address the Streaming Services'
separate argument that the definition was arbitrary, capricious, or
unsupported by substantial evidence. Id.
The Services and Copyright Owners agreed that the Judges should
resolve the definitional issue based on the existing record, after
receiving two rounds of additional briefing from the parties.\152\ See
Services' Proposal for Remand Proceedings (Dec. 10, 2020) (Services'
Proposal) at 5-6, 9-10; Proposal of the Copyright Owners for Conduct
and Resolution of the Remand (Public) (Dec. 10, 2020) (Copyright
Owners' Proposal) at 4-6. The Judges issued an Order Regarding
Proceedings on Remand, which, in part, opened briefing on the issue of
the adoption of a revised definition of ``service revenue'' for bundled
offerings between issuing the Initial Determination and the
Determination. Order Regarding Proceedings on Remand (Dec. 15, 2020).
The Judges received the following relevant briefing.
---------------------------------------------------------------------------
\152\ As indicated below, during the remand proceedings, the
Judges solicited two rounds of additional briefing addressing
specific issues.
CO Initial Submission
Services' Initial Submission
CO Reply
Services' Reply
On December 9, 2021, the Judges requested additional briefing. Dec.
9 Order. The Dec. 9 Order sought additional briefing setting forth the
parties' views on whether this proceeding constitutes the type of new
agency action addressed by the D.C. Circuit, which would allow adoption
of a Service Revenue definition without limitation to the definition
expressed in the Initial Determination. Additionally, the Judges
requested additional evidence that the parties might offer to support
adoption of the Service Revenue definitions expressed in either the
Initial Determination or the Determination. In response to the Dec. 9
Order, the Judges received the following relevant briefing.
CO Additional Submission
Services' Additional Submission
On February 9, 2022, the Judges solicited further briefing on
``Whether the D.C. Circuit's Johnson decision permitting the Judges to
engage in new agency action in this remand proceeding allows the Judges
to engage in new agency action through a reconsideration of Copyright
Owners' February 12, 2018 Motion for Clarification as a Motion for
`rehearing' pursuant to 17 U.S.C. 803(c)(2)(A) and 37 CFR 353.1.'' Sua
Sponte Order Regarding Additional Briefing (Feb. 9 Order). In response
to the Feb. 9 Order, the Judges received the following relevant
briefing.
Copyright Owners' Brief Responding to Judges' February 9, 2022
Sua Sponte Order Regarding Additional Briefing on New Agency Action
Question, and Replying to Services' New Agency Action Arguments in
their Joint Supplemental Brief Addressing the Judges' Working Proposal
(in Additional Materials Rebuttal Submission of Copyright Owners at Tab
B) (Feb. 24, 2022) (``CO Further Briefing'')
Services' Joint Response to the Judges' February 9, 2022 Sua
Sponte Order Regarding Additional Briefing and Rebuttal Regarding ``New
Agency Action'' (Feb. 24, 2022) (``Services' Further Briefing'')
B. Authority for Modification to the Initial Determination
1. Copyright Owners' Position
Copyright Owners assert that this remand proceeding offers a
straightforward path to take new agency action and that the law makes
clear that new agency action can consist of issuing a new determination
on remand. CO Initial Submission at 71. Copyright Owners maintain that:
[T]the new agency action here is a determination after remand
proceedings, the Board is largely free to chart its own procedural
course, and the Board has done so in its December 15 Order. The
Board is not required to undertake any of the procedural steps set
forth in 17 U.S.C. 803(b) in order to take such ``new agency
action.'' See 17 U.S.C. 803(d)(3) (requiring only that on remand
further proceedings be taken ``in accordance with subsection (a)'');
37 CFR 351.15; Intercollegiate Broad. Sys., Inc., 796 F.3d at 125
(``[N]either the Copyright Act nor the Board's regulations prescribe
any particular procedures on remand.'') The Circuit's instruction
that the action be ``accompanied by the appropriate procedures[,]''
Johnson, 969 F.3d at 392, does not dictate what those ``appropriate
procedures'' must be but instead plainly refers to these flexible
rules. See also Oceana, Inc., 321 F. Supp. 3d at 136 (explaining
that when remanding to an agency, a court generally ``may not
dictate to the agency the methods, procedures, or time dimension,
for its reconsideration'').
CO Initial Submission at 71, FN 33.
Copyright Owners acknowledge the Services' position that the
asserted procedural error is an ``absence of authority'' that can never
be cured. Id. at 74 (citing Services' Proposal for Remand Proceedings
at 10). They note that the D.C. Circuit did not say the Judges lacked
the authority to revisit the service revenue definition for bundles on
remand. Nor, they observe, did it say the Judges have no authority to
review the record evidence and the parties' arguments and reach the
same conclusion or a different conclusion on remand. Copyright Owners
opine that if the only possible outcome were for the Judges to
reinstate a definition that lacked any explanation or evidentiary
support solely because it was present in the Initial Determination,
then the D.C. Circuit would not have remanded the issue but would have
simply reversed and reinstated the Initial Determination definition.
But instead, they note, the D.C. Circuit remanded and said the
[[Page 54451]]
Judges could take new agency action precisely to cure the asserted
procedural defect. Copyright Owners assert that the remand allowed the
parties to present the record evidence and their arguments so that the
Judges can address the definition ``afresh'' in the remand
determination. Id. at 74.
Copyright Owners argue that 17 U.S.C. 803(d)(3) states only that
proceedings on remand must be in accordance with 17 U.S.C. 803(a). They
contend that remand proceedings need not be confined to procedures the
Services claim are too late in the game for the Judges to follow. The
Copyright Owners point to the D.C. Circuit's ruling in Intercollegiate
Broad. Sys., Inc. v. Copyright Royalty Bd., that ``neither the
Copyright Act nor the Board's regulations prescribe any particular
procedures on remand.'' 796 F.3d 111, 125 (D.C. Cir. 2015) (citing 17
U.S.C. 803(a), (d)(3)). Accordingly, they argue, the Judges can
reaffirm the adopted bundled service revenue definition following their
review of the parties' submissions without regard to section 803(c)(2)
or 803(c)(4). CO Reply at 65-66.\153\
---------------------------------------------------------------------------
\153\ Copyright Owners reiterate this argument in the CO
Additional Submission. Copyright Owners added that the parties in
this remand were afforded the opportunity for further briefing and,
if they wished, to submit additional evidence on this issue, thus
providing broader opportunity for submission than in Fisher v.
Pension Benefit Guaranty Corp., 994 F.3d 664, 670 (D.C. Cir. 2021),
in which the D.C. Circuit upheld new agency action after remand even
though the agency did not provide appellant the opportunity to
submit new briefing or exhibits. CO Additional Submission at 35-36;
38.
---------------------------------------------------------------------------
Copyright Owners further argue that the Judges may properly justify
the changed definition under section 803(c) as a fuller explanation of
the agency's reasoning at the time it was made. They urge that the
Judges could explain that, especially in light of the evidence of how
the Services misused the prior definition to make service revenue
completely disappear, carrying over the prior bundle service revenue
from Phonorecords II into the Initial Determination was unintended and
inadvertent.\154\ CO Reply at 69. Copyright Owners also assert that the
Judges could explain that Copyright Owners had, in their Motion for
Clarification, identified an ``exceptional case'' under section
803(c)(2) because the prior definition failed to comport with Judges'
precedent and economic principles, and was unsupported by
evidence.\155\ In addition,
---------------------------------------------------------------------------
\154\ Copyright Owners assert that the definition in the Initial
Determination conflicted with, the Board's findings in the Initial
Determination, including its findings that the adopted rates and
terms would afford copyright owners a fair return for their creative
works, thereby satisfying factor B of the 801(b) standard and thus
needed to be revised so as to not ``frustrate the proper
implementation of'' the Final Determination. CO Reply at 69 (citing
17 U.S.C. 801(b) and 803(c)(4)).
\155\ In response to an Order by the Judges, Copyright Owners
provided additional briefing regarding reconsideration of the motion
for clarification as a motion for ``rehearing'' which is addressed
separately infra.
---------------------------------------------------------------------------
Copyright Owners note that the Judges reheard the evidence and
legal arguments as presented in the parties' briefs on the issue and,
as a result, may choose to adopt the revised definition. Copyright
Owners maintain that for the Judges to do so would not be impermissible
post-hoc reasoning, because the D.C. Circuit remanded precisely because
the Judges did not provide any reason in the Determination for revising
the bundle revenue definition. CO Reply at 69-71.
2. Services' Position
The Services assert that the D.C. Circuit found only ``three ways
in which the Board can revise Initial Determinations'' and that the
Judges had failed to establish that the change to the service revenue
definition fit any of those three categories. Services' Initial
Submission at 64-65 (citing Johnson at 390).
According to the Services the first way the Judges may revise an
Initial Determination is to ``order rehearing `in exceptional cases' in
response to a party's motion, 17 U.S.C. 803(c)(2)(A).'' Services'
Initial Submission at 65 (citing Johnson at 390).\156\ The Services
argue that the D.C. Circuit held in Johnson that the Judges' ``material
revision of the `Service Revenue' definition for bundled offerings does
not fall within the Board's rehearing authority under section
803(c)(2)(A)'' because ``the Board itself . . . was explicit that it
`did not treat the [Copyright Owners'] motion[ ]' . . . `as [a] motion[
] for rehearing under 17 U.S.C. 803(c)(2).' '' The D.C. Circuit also
noted that ``as the Board found, the Copyright Owners' motion did `not
meet [the] exceptional standard for granting rehearing motions' under
section 803(c)(2).'' Id. (citing Johnson at 390). The Services assert
that the Judges were not able to make ``a volte-face'' and justify on
appeal their revision to the definition as an exercise of rehearing
authority. As the D.C. Circuit held, agency action must be justified by
``reasons invoked by the agency at the time it took the challenged
action,'' and post-hoc rationalizations are insufficient. Id. (citing
Johnson at 390).
---------------------------------------------------------------------------
\156\ In response to an Order by the Judges, the Services
provided additional briefing regarding reconsideration of the motion
for clarification as a motion for ``rehearing'' which is addressed
separately infra.
---------------------------------------------------------------------------
The Services add their view that the Judges cannot revisit the
decision to deny rehearing without engaging in impermissible post-hoc
reasoning. They note that the Supreme Court has explained that, while
an agency may ``elaborate later'' on its ``initial explanation'' of the
reason (or reasons) for its action, it ``may not provide new ones.''
Services' Initial Submission at 66 (citing e.g., Regents, 140 S. Ct. at
1908). The Services offer that the Judges, having stated that they did
not consider the Copyright Owners' motion to revise the definition to
be a motion for rehearing, cannot now conclude that the motion
qualified as one for rehearing and that the Judges in fact engaged in
rehearing. Id.
The Services add that under section 803(c)(2)(A), the Judges can
only use their rehearing authority `` `in exceptional cases' in
response to a party's motion.'' Id. (citing Johnson at 390). The
Services argue that the Motion for Clarification cannot be found to
have satisfied that standard. The Copyright Owners did not argue that
their motion satisfied the ``exceptional cases'' standard before the
Judges or the D.C. Circuit, and have therefore waived that argument.
Id.
According to the Services, the second way the Judges may revise an
Initial Determination, viz. action to correct a technical or clerical
error under section 803(c)(4), cannot be used now to justify any
modification of the Service Revenue definition in the Initial
Determination. The Services note that the D.C. Circuit held
specifically that the Judges' change to the Service Revenue definition
could not be construed as correcting a technical or clerical error
because it involved a substantive rewrite of the Service revenue
definition. Id. at 67 (citing Johnson at 391).
The Services aver that the third way the Judges may revise the
terms in an Initial Determination is in response to unforeseen
circumstances that would frustrate the proper implementation of the
determination. Id. at 67. The Services note that the D.C. Circuit held
in Johnson that this authority did not justify the Judges' change to
the Service Revenue definition because the Judges did not invoke this
authority and ``the need to ground the original definition in the
record'' could not credibly be described as ``an unforeseen
circumstance.'' Id. (citing Johnson at 391).
The Services also note that the D.C. Circuit rejected the argument
that the Judges have ``inherent authority'' to make changes to the
Initial
[[Page 54452]]
Determination. The D.C. Circuit explained that the specific
restrictions Congress placed on the Judges' authority in section 803
``would be a nullity if the Board also had plenary authority to revise
its determinations whenever it thought appropriate.'' Id. (citing
Johnson at 391-92). The Services add that even if the Judges offered a
new source of authority capable of justifying substantive changes to
the Service Revenue definition now, the Judges would be unable to rely
on this ``uninvoked authority'' without engaging in impermissible post-
hoc reasoning. Id.
The Services counter Copyright Owners' position that the Judges
need not respond to the error the D.C. Circuit identified with this
aspect of the Determination and that the Judges' ``new agency action''
may consist of issuing a new determination on remand. The Services
argue that failure to address the legal and factual issues on which the
D.C. Circuit remanded would violate the D.C. Circuit's order and would
result in a second remand. The Services surmise that the issue of
authority to make the changes to the Initial Determination are
particularly important in this context, where the D.C. Circuit
recognized that the Copyright Act places limits on the Judges'
authority to alter an initial determination by defining conditions for
rehearing and the types of changes that are permitted absent a
rehearing. In this regard, the Services maintain that the Judges cannot
do on remand what they lacked authority to do in the first instance.
The Services assert that the Judges must resolve the legal question
whether there is authority to alter the revenue definition in the
Initial Determination. They urge that the remanded issue is not what
the substance of the service revenue definition should be as a matter
of first impression, but instead is whether the Judges have properly
exercised authority to alter the Initial Determination's definition.
Services Reply at 52-54.\157\
---------------------------------------------------------------------------
\157\ The Services agree that this remand proceeding qualifies
as ``new agency action'' but again urge that failure to address the
legal and factual issues on which the court remanded would
nonetheless violate the D.C. Circuit's order. Services' Additional
Submission at 38-42.
---------------------------------------------------------------------------
The Services assert that the Judges have two paths available to
them: (1) to provide a ``fuller explanation'' of the prior conclusion
that the Judges had legal authority to revise the Service Revenue
definition in the Initial Determination or (2) answer that threshold
question through new agency action. The Services maintain that, if they
pursue the ``fuller explanation'' path, the Judges are limited to
elaborating on what they said previously, and that they cannot add new
reasons they did not initially provide. With regard to what may
constitute new agency action, the Services assert that path gives the
Judges freedom to consider new reasons that the Copyright Act provided
the Judges with the authority to make this change to the Initial
Determination. The Services argue, however, that undertaking a new
agency action does not, as Copyright Owners claim, obviate the need for
the Judges to identify proper legal authority before substantively
changing the Initial Determination, such authorities being limited to
the authority of section 803(c)(4) or the rehearing authority of
section 803(c)(2). Id. at 54-55.
The Services address Copyright Owners' position that if the only
possible outcome were for the Judges to reinstate a definition that
lacked any explanation or evidentiary support solely because it was
present in the Initial Determination, then the D.C. Circuit would not
have remanded the issue but would have simply reversed and reinstated
the Initial Determination definition. The Services urge that the D.C.
Circuit could not reverse because the Department of Justice raised for
the first time on appeal new justifications for the Judges' decision to
change the Initial Determination. Instead, the Services maintain, the
D.C. Circuit had to remand and give the Judges the opportunity to
address the Department of Justice's new justifications in the first
instance, as the D.C. Circuit could not rule them out given the posture
of the appeal. Id. at 56.
In the Services' Additional Submission, they concede that this
remand proceeding is new agency action and that the Judges have
provided the parties with sufficient procedural opportunities to
present any new evidence and raise any additional arguments regarding
the question the D.C. Circuit remanded. Services' Additional Submission
at 38. But the Services still insist that the Judges may not alter the
Service Revenue definition without first identifying legal authority in
the Copyright Act for modifying the Initial Determination. In the
Services' view the new agency action avenue provided by the D.C.
Circuit merely offers a singular path beyond the Judges' ability to
offer a ``fuller explanation'' of their previous reasoning for
revisiting the definition in the Rehearing Order. According to the
Services' argument, the new agency action provided for in this remand
only offers the additional opportunity to offer new reasons supporting
any legal authority for altering the Initial Determination's Service
Revenue definition, beyond those that were raised in the appeal.
Services' Additional Submission at 38-42
C. Reconsideration of Motion for Clarification as Motion for
``Rehearing'' 158
---------------------------------------------------------------------------
\158\ The Judges consider the briefs filed in response to the
Feb. 9, 2022 Order only to the extent that they are responsive to
the Feb. 9, 2022 Order, which requested briefing on the specific
matter of whether the D.C. Circuit's Johnson decision permitting the
Judges to engage in new agency action in this remand proceeding
allows the Judges to engage in new agency action through a
reconsideration of Copyright Owners' February 12, 2018 Motion for
Clarification as a Motion for ``rehearing,'' pursuant to 17 U.S.C.
803(c)(2)(A) and 37 CFR 353.1.
---------------------------------------------------------------------------
1. Copyright Owners' Position
Copyright Owners argue that the Judges have the authority to engage
in new agency action in this remand proceeding through a
reconsideration of the Motion for Clarification as a motion for
rehearing, pursuant to 17 U.S.C. 803(c)(2)(A) and 37 CFR 353.1.
Copyright Owners urge, however, that proceeding in that fashion would
add an entirely unnecessary and complicating step. They again suggest
that there is no need to reconsider or recharacterize the Motion for
Clarification as a motion for rehearing because the remand itself
affords the opportunity for the Judges to take new agency action,
which, as in a rehearing, permits them to reconsider evidence and
arguments, but, unlike a rehearing, is not limited by the constraints
of section 803(c)(2). CO Further Briefing, Tab B at 7-8.
Copyright Owners posit that if the Judges engage in new agency
action to reconsider the Motion for Clarification as a motion for
rehearing under 803(c), and to decide that motion based on all of the
evidence in the record supporting the adopted bundle revenue definition
and showing the prior bundle revenue definition to be unsupported and
unreasonable, they may properly do so. They assert that the while they
did not make a request for rehearing on the face of the Motion for
Clarification, that is not the same as a finding that the standard
could not have been met. The Judges may consider whether, based on the
evidence in the record, the rehearing standard has been satisfied on
this remand. In Copyright Owners' view, the Judges could conclude,
revisiting on remand the question of whether the rehearing standard has
now been met, that Copyright Owners have satisfied the ``exceptional
case'' standard for granting rehearing motions under
[[Page 54453]]
section 803(c)(2). Copyright Owners note that if the Judges do engage
in new agency action that reconsiders the Motion for Clarification as a
motion for rehearing, the Judges should fully explain their reasoning.
Id. Tab B at 8-10.\159\
---------------------------------------------------------------------------
\159\ With regard to the obligation to fully explain their
reasoning for any reconsideration, the Copyright Owners point to
United Food & Com. Workers Union, Loc. No. 663 v. U.S. Department of
Agriculture., 532 F. Supp. 3d 741, 769 (D. Minn. 2021) (``When an
agency takes a new course of action, it must `display awareness that
it is changing position' and `show that there are good reasons for
the new policy.' ''), quoting FCC v. Fox Television Stations, Inc.,
556 U.S. 502, 515 (2009) (emphasis in original).
---------------------------------------------------------------------------
2. Services' Position
The Services assert that the Judges cannot invoke their rehearing
authority by construing the Motion for Clarification as a rehearing
motion. They maintain that the D.C. Circuit expressly found that the
revision of the Service Revenue definition for bundled offerings does
not fall within the Judges' rehearing authority under section
803(c)(2)(A). The Services assert that Copyright Owners did not satisfy
either prong of section 803(c)(2)(A), which authorizes rehearing only
``upon motion of a participant'' and ``in exceptional cases.'' They
note that the D.C. Circuit agreed with the Judges' decision not to
treat Copyright Owners' motion as one for rehearing and that the D.C.
Circuit also agreed with the Judges' further finding that ``Copyright
Owners' motion did not meet the exceptional standard for granting
rehearing motions.'' Services' Further Briefing at 7 (citing Johnson at
390).
The Services add their view that the Judges are bound by the D.C.
Circuit's conclusions on this issue. They maintain that because the
Judges' section 803(c)(2)(A) rehearing authority is among the grounds
that Johnson addressed and determined, the Judges cannot rely on that
authority on remand. Id. at 8-9. The Services urge that the Judges
already correctly concluded that the Motion for Clarification was not a
motion for rehearing, and note that Copyright Owners never presented
their motion as one for rehearing. The Services add that because
Copyright Owners did not challenge that decision on appeal, it is too
late for them to do so now.\160\ Id. at 9-10.
---------------------------------------------------------------------------
\160\ In fact, the issue of whether to recharacterize the Motion
for Clarification as a motion for rehearing is not one raised by
Copyright Owners, but by the Judges sua sponte. The Services'
estoppel argument as to the Copyright Owners cannot apply to the
Judges' action.
---------------------------------------------------------------------------
The Services argue that Copyright Owners' Motion did not make any
attempt to satisfy the exceptional cases standard set out in 17 U.S.C.
803(c)(2)(A). They argue that Copyright Owners did not purport to
identify any new evidence, new legal authority, or even a substantive
error in the Judges' reasoning in the Initial Determination, but
instead the motion asserted that the Judges' inclusion of the
definition of service revenue in the Initial Determination was
supposedly inadvertent. The Services add that Copyright Owners did not
identify any specific evidence in the Phonorecords III record or any
aspect of the Initial Determination that suggested the inclusion of
this definition was a mistake. Id. at 10.
The Services point out that Copyright Owners' motion did not comply
with the procedural requirements for a motion for rehearing. They then
urge that the Judges cannot invoke their section 803(c)(2)(A) authority
by rewriting a participant's motion to say it is seeking rehearing when
that participant specifically and unambiguously disclaimed any intent
to seek rehearing. Id. at 11.
The Services note that the Judges previous conclusion that even if
the Motion for Clarification had requested rehearing, that motion would
not and does not meet that exceptional standard for granting rehearing
and failed to make even a prima facie case for rehearing. The Services
observe that the Judges apply a strict standard to rehearing motions to
prevent parties from using the rehearing process to seek a second bite
at the apple by advancing theories and arguments that could have been
advanced earlier during the proceeding. Id. at 12. The Services
reiterate their view that Copyright Owners' motion did not point to any
evidence in the Phonorecords III record at all, and, that the only
evidence in the Phonorecords III record concerning bundles supports the
longstanding definition of Service Revenue which has been effective in
encouraging the Services to offer bundles that benefit Copyright Owners
by growing the market for music streaming services. Id. at 14.
The Services finally assert that this is not an extraordinary case
where a party has identified an error that, if left uncorrected, would
result in manifest injustice. Id. at 15-16. The Services conclude by
urging that given this procedural history and the unchanged state of
the record since the initial hearing, any claim that Copyright Owners
have somehow now satisfied the exceptional case standard would be clear
error. Id. at 17.
D. Record Evidence Regarding Definition of Service Revenue
1. Copyright Owners' Position
Copyright Owners assert that the prior bundle revenue definition
(published in the Initial Determination) failed to address the ``
`economic indeterminacy' problem inherent in bundling'' appropriately
and in a way consistent with Judges' precedent. CO Initial Submission
at 75 (citing Order on Rehearing at 16-18). Copyright Owners proceeded
to cite several portions of testimony from the Services' economic
experts who acknowledged this problem. Id. They then point to hearing
testimony in which Copyright Owners repeatedly raised the ``economic
indeterminacy'' problem and demonstrated what they characterized as the
absurd results to which the prior definition had led. Id. at 76. They
point out that under the prior definition, service revenue for bundled
subscriptions started with revenues recognized from the bundle (i.e.,
the price paid by the subscriber) and subtracted ``the standalone
published price'' for all non-music components of the bundle.
[REDACTED]. Id.
Copyright Owners point out that the Judges already found with
respect to other licenses that such an approach is not only
fundamentally unfair, but ``absurd.'' Id. (citing 81 FR 26316, 26382
(May 2, 2016) (webcaster licenses)); see also 83 FR 65210, 65264 (Dec.
19, 2018) (SDARS licenses) (rejecting proposed deductions by service
for bundle revenues because of the ``acknowledged `economic
indeterminacy' problem inherent in bundling''). The Copyright Owners
concur with the Judges' correct conclusion that the same reasoning
applies to Phonorecords III. Id. at 76-77 (citing Order on Rehearing at
18) (``the `economic indeterminacy' problem inherent in bundling is
common to all three proceedings.''). The Copyright Owners offer that
Spotify conceded to this flaw in the definition in the Initial
Determination, but offered an alternative that contained the same
loophole. Id. at 77-78.
Copyright Owners point out that the proponent of a term bears the
burden of proof as to adoption. The Judges made clear that the licensee
who wishes to offer bundles must bear the burden of providing evidence
that might mitigate the acknowledged economic indeterminacy problem
inherent in bundling, because any such evidence would be in its
possession, not in the possession of the licensors. Id. at 79 (citing
SDARS III Determination, 83 FR 65210, 65264) (``bundling [is]
undertaken to increase [the Services']
[[Page 54454]]
revenues and it would be reasonable to assume that [the Services have]
information relevant to the economic allocation of the bundled
revenue.''). The Copyright Owners contend they presented unrebutted
evidence showing the unreasonableness of the Services' proposed
definition while the Services offered no evidence to support their
definition. Id. at 78, 79 (citing Order on Rehearing at 18). Copyright
Owners maintain that no Service offered evidence concerning the
separate values of the constituent parts of the bundles, or any other
evidence concerning the economic allocation of bundled revenue, let
alone the reasonableness of the definition in the Initial
Determination. Id. at 80. Copyright Owners assert that in the absence
of evidence to support the proposed definition, the Judges may adopt or
fashion a definition of service revenue for bundled offerings that
comports with the record evidence, which is precisely what the Judges
did and can, through new agency action, do again. Id. at 81.
Copyright Owners dispute the Services' assertion that there is
support for the Phonorecords II approach to bundles in the record of
this proceeding. Instead, Copyright Owners argue, the Services'
purported evidence at most supports the benefits of the practice or
strategy of bundling. They maintain that the strategy of bundling
covered music services with other products or services has nothing to
do with whether the Services should be free to reduce the revenue
allocable to music to zero. They offer that the definition in the
Initial Determination has nothing to do with such benefits, and that
those benefits may be equally served by a definition that ensures value
is apportioned to the music component in the bundle. CO Reply at 73-76.
2. Services' Position
The Services argue that the evidence in the existing written record
addressing bundles shows both that this definition is supported by the
Phonorecords II benchmark and that it has proven, industry-wide
benefits. Services' Initial Submission at 68. They offer that the
Copyright Owners did not propose an alternative definition of service
revenue until after the Judges issued the Initial Determination and
that any definition they propose now would fail the basic requirement
that the Judges must adopt rules ``on the basis of a written record.''
Id. (citing 17 U.S.C. 803(a)(1) and 803(c)(3)).
Addressing the merits of the definition contained in the Initial
Determination, the Services argue that it best serves the goals of the
Copyright Act; that as a bright-line, easily administered rule, it
continues the broad industry agreement from Phonorecords II. The
Services contend the prior definition increases output and incentivizes
beneficial price discrimination to reach listeners who would otherwise
not pay for music. They argue that the record evidence confirms that
the prior treatment of bundles enabled experimentation and variation in
the distribution of music with long-term benefits for all parties. They
state that Copyright Owners' argument that Services [REDACTED] also
demonstrates the broad benefits of the definition of Service Revenue in
Phonorecords II because the record showed that arrangement enabled
funneling of many of listeners into full-priced, full-catalog
services--such treatment of bundles enabled the flexibility and price
discrimination that yielded beneficial growth of the royalty pool.\161\
The Services allege that Copyright Owners also ignore the extensive
royalties that were generated. They add that with the per-subscriber
minimum guarantees that the Copyright Owners will still be paid a fair
royalty. The Services then cite several portions of testimony from
various Services' economic experts who point out the realization of an
expanded royalty pool, which the Services offer as proving a
functioning marketplace. Id. at 68-74.\162\
---------------------------------------------------------------------------
\161\ Notably, the Services do not deny that the former
definition did, in fact, [REDACTED].
\162\ The Services' Reply reiterates this point and offers that
the testimony cited by the Copyright Owners also shows why the
Initial Determination's Service Revenue definition works for bundles
and grows royalties. Services Reply at 57-58.
---------------------------------------------------------------------------
The Services then assert that no other definition of service
revenue for bundles that has been before the Judges combines both the
administrative simplicity of the Initial Determination's definition and
the broad price discrimination benefits of promoting discounted
bundles. They maintain that while neither the Services nor Copyright
Owners submitted evidence specifically addressing the way that
customers, Services, or Copyright Owners might value the component
parts of bundles, such subjective valuations are unnecessary for the
Judges to find ample support for the Phonorecords II approach to
bundles in the record. Id. at 75-76.
The Services also argue that while the Judges' decision in SDARS
III did involve valuation of the music and non-music components of a
bundle, the resolution in SDARS III is inapposite because, here, the
rate structure has a way of ensuring that Copyright Owners are fairly
compensated for bundles: the statutory minimum payment. Services Reply
at 62.
E. Analysis and Conclusions Regarding Definition
1. Remand Proceeding as New Agency Action
Having considered the entirety of the record of this proceeding, a
majority of the Judges (Definition Majority) conclude that this remand
constitutes ``new agency action'' and meets all of the criteria to
qualify as new agency action. The Judges thus have the opportunity to
consider the issue afresh consistent with their procedural rules
regarding remands.
The Definition Majority finds that it is unnecessary to attempt to
distinguish new ``agency action'' from ``new agency action.'' Neither
approach is endorsed clearly by the varied judicial interpretations of
a new agency action. See R.J. Krotoszynski, Jr., Administrative Law
Discussion Forum: ``History Belongs to the Winners'': the Bazelon-
Leventhal Debate and the Continuing Relevance of the Process/Substance
Dichotomy in Judicial Review of Agency Action, 58 Admin. L. Rev. 995
(Fall 2006). As noted by Judge Bazelon, the D.C. Circuit ``believed in
process-based review, [but] he argued that it was improper for judges
to prescribe specific procedures.'' Id. at 1001. Judge Bazelon's remand
orders focused on providing ``genuine opportunities to participate in a
meaningful way'' and ``genuine dialogue'' with interested parties,
while leaving the agency ``free to decide which specific procedures to
undertake.'' Id.
Several reported cases point to new action as an alternative to a
fuller explanation. But few define ``new agency action'' other than to
say, as did the Johnson court, that the agency must take it
``accompanied by the [unspecified] appropriate procedures.'' Johnson,
969 F.3d at 392. Parties to the original action, already familiar with
the issue and the factual and legal background, recognized that the
D.C. Circuit identified the adoption of a modified definition in the
Determination as one of three issues on remand. In repeated rounds of
remand submissions, both the Services and the Copyright Owners included
the definition issue. The Judges were not satisfied with the parties'
lack of focus on the issue, however, and ordered expressly further
briefing on the new agency action issue and sub-issues relating to the
adoption of a definition of Service Revenue as it relates to
[[Page 54455]]
bundled service offerings. See (Dec. 9 Order) at 4; Sua Sponte Order
Regarding Additional Briefing (Feb. 9, 2022).
New agency action is not synonymous with justification, or
confirmation, of the prior action. New agency action is a procedural
mechanism for reconsideration of the record, reopening the record for
additional evidence and argument, and adoption of a conclusion based on
the expanded record. In this instance, the presentations, written and
oral, of participants on remand, together with a re-examination of the
original record, support reversion to the definition originally
announced in the Initial Determination. Ultimately, given repeated
opportunities for legal analysis on the issue, both sides agreed that
the remand proceeding itself, with ample notice and multiple
opportunities for input was sufficient to constitute new agency action.
See CO Further Briefing at 3, 7.
The Services argued, however, that notwithstanding this appropriate
new agency action, the Judges remained without authority to adopt the
revised definition as a term governing the royalty rates determined in
this proceeding. Their arguments regarding procedures undertaken in the
Determination are superseded by the Judges' conduct of extensive remand
proceedings.\163\ The gravamen of the Administrative Procedure Act is
transparency in agency \164\ rulemaking. Agencies must publish notice
of their intentions, provide opportunities for interested parties to
comment and object, and finalize regulations only after reconciling
objections with the policies and purposes of proposed regulations. The
adjudication of this remand proceeding was conducted openly. Interested
parties had ample opportunity to object, to comment, and to brief legal
and factual issues relating to the Judges' approach to promulgating an
appropriate definition of bundled service revenue.
---------------------------------------------------------------------------
\163\ Furthermore, the issue of the Judges' authority to take an
action in issuing the Determination is moot. The Judges, after new
agency action, have chosen not to defend the definition in the
Determination but rather to conclude, following that new agency
action, that the definition in the Initial Determination is more
appropriate in these circumstances. Whether the Judges had the
authority in the first instance is not at issue, as they are not
repeating the former action.
\164\ The proceedings of the Copyright Royalty Board (CRB) are
subject to the standards of the Administrative Procedure Act. See 17
U.S.C. 803(a)(1).
---------------------------------------------------------------------------
The present analytic approach merely takes the position that the
Judges engaged in new agency action by conducting a fully open and
broadly explored remand proceeding. Unlike a rehearing or exercise of
continuing jurisdiction, this remand proceeding is not limited by the
constraints of sections 803(c)(2) or 803(c)(4). Contrary to the
Services' assertion, the Judges address the issue on which the D.C.
Circuit remanded, the need to exercise authority within the lines drawn
by the authorizing statute. This remand proceeding does not, therefore,
violate the D.C. Circuit's order.
The Johnson opinion clearly states the two paths by which the
Judges may address the issues presented to them on remand; they may
either (1) provide ``a fuller explanation of the agency's reasoning at
the time of the agency action[,]'' or (2) to take ``new agency action''
accompanied by the appropriate procedures. Johnson, 369 F.3d at 392.
The Judges chose to pursue the second option: this new agency action.
The Judges reiterate: the Services concede that, through this
proceeding the Judges have provided the participants with adequate
procedural opportunities to present any new evidence on the proper
Service Revenue definition for bundles. The Judges also acknowledge,
but disagree with, the Services' position that that they must return to
the issues as they were presented after issuance of the Initial
Determination, regardless of the admittedly complete and valid remand
procedure, which constitutes new agency action.
The Judges (the majority on this issue) determine that any
confining action on remand to the provisions of sections 803(c)(2)(A)
or 803(c)(4) would misconstrue the clear expression of the ``new agency
action'' alternative presented by the D.C. Circuit,\165\ as well as
chapter 8 of title 17. As the Copyright Owners correctly observed, in a
remand proceeding, the Judges are not required to undertake any of the
procedural steps set forth in section 803(b) nor are the Judges
compelled to consider or be limited by sections 803(c)(2)(A) or
803(c)(4). The statute only requires that the Judges' remand
proceedings are in accordance with section 803(a).\166\
---------------------------------------------------------------------------
\165\ The case that the D.C. Circuit points to for the new
agency action path clarifies that ``An agency taking this [new
agency action] route is not limited to its prior reasons but must
comply with the procedural requirements for new agency action.''
Regents, 140 S. Ct. at 1908).
\166\ ``The court [United States Court of Appeals for the
District of Columbia Circuit] may also vacate the determination of
the Copyright Royalty Judges and remand the case to the Copyright
Royalty Judges for further proceedings in accordance with subsection
(a).'' 17 U.S.C. 803(d)(3).
---------------------------------------------------------------------------
The D.C. Circuit observed that the Judges have ``considerable
freedom to determine [their] own procedures.'' SoundExchange v. CRB,
904 F.3d 41 at 61. The D.C. Circuit also cautions that such flexibility
must be exercised within the lines drawn by the authorizing statute.
Here, the Judges operate within the lines drawn with respect to remand
proceedings set forth in chapter 8 of title 17.
2. ``Fuller Explanation'' of Modification to Initial Determination
Case law regarding development of a ``fuller explanation'' of an
agency's action emphasizes that the agency cannot adopt post hoc
reasoning on the same record. See, e.g., SEC v. Chenery Corp., 332 U.S.
194, 201 (1947) (after remand, agency bound to ``deal with the problem
afresh . . . .''). Certainly, adopting a post hoc argument of appellate
counsel, just because it offers a rationale for the agency's original
action is impermissible.\167\ On the other hand, if the record in the
initial proceeding is sufficiently robust to support a reinterpretation
or additional reasoning, the agency may justify its initial action with
that ``fuller explanation'' without considering any new evidence. See,
Fisher v. Pension Benefit Guar. Corp., 468 F.Supp.3d 7, 20 (D.C.D.C.
2020), aff'd Fisher v. Pension Benefit Guar. Corp., 994 R.3d 664 (D.C.
Cir. 2021), rehearing en banc denied, Fisher v. Pension Ben. Guar.
Corp., 2021 U.S. App. LEXIS 18793 (D.C. Cir., June 23, 2021)
(requirement of new evidence a ``novel proposition of law'' without
precedent). On remand, an agency may elaborate on its prior reasoning,
but it may not provide new reasons for the original decision. Fisher,
994 F.3d at 669. If the Judges had chosen in this remand to rest on
their Determination regarding the service revenue definition, they
might have done so only if they could elaborate on the existing
record.\168\ In the alternative, the Judges issue a new decision after
new agency action. Id.
---------------------------------------------------------------------------
\167\ A rationalization is not post hoc simply because it is
iterated by counsel. Denomination of a rationalization as post hoc
is a matter of timing, not of the offeror.
\168\ In this instance, had the Judges decided to keep the
definition in the Determination, they probably could have given a
fuller explanation based on the record in the underlying proceeding.
Because the Judges have opted to rely on the fresh-look approach in
the ``new agency action'' alternative and because the prior
definition is appropriate given adoption of the PR II rate
structure, development of that fuller explanation based on the
record is unnecessary.
---------------------------------------------------------------------------
The Judges, having engaged in new agency action to settle on the
definition of service revenue for bundled offerings, do not find a need
to address the statutory avenues or the confines that are provided for
rehearing or continuing jurisdiction, nor do the Judges pursue the
propriety of reconsideration of the
[[Page 54456]]
Motion for Clarification as a motion for rehearing.\169\
---------------------------------------------------------------------------
\169\ The Judges also find no need to consider any inherent
authority that may remain for consideration.
---------------------------------------------------------------------------
3. Substantive Analysis of Dueling Definitions of Bundled Revenue
The fundamental difference between the impact of the two
alternative definitions is simply stated:
Under the Initial Determination: downstream bundling and its price
discriminatory effect would be incentivized by a royalty structure that
reflects the lower WTP of consumers who subscribe by paying for a
Bundle;
Under the Determination: downstream bundling and its price
discriminatory effect would not be incentivized by a royalty structure
that reflects the lower WTP of consumers who subscribe by paying for a
Bundle.
To explain this difference, the Judges find it helpful to describe
(as in the Determination and Dissent) how bundling facilitates price
discrimination and how lower royalties for bundled streaming services
incentivize such bundling.
Price discrimination occurs when a seller offers different units of
output at different prices. See, e.g., H. Varian, Intermediate
Economics at 462 (8th ed. 2010). The benefit to the seller arises from
attempting to ``charge each customer the maximum price that the
customer is willing to pay for each unit bought.'' R. Pindyck & D.
Rubinfeld, Microeconomics at 401 (8th ed. 2013). For all goods, and
intellectual property goods such as copyrights in particular,\170\ the
social benefit is that price discrimination more closely matches the
quantity sold with the competitive quantity as the seller or licensor
better aligns the price with the WTP of different categories of buyers
or licensees. See W. Fisher, Reconstructing the Fair Use Doctrine, 101
Harv. L. Rev. 1659, 1701 (1988).
---------------------------------------------------------------------------
\170\ Streamed copies of intellectual property, such as musical
works and sound recordings, have a marginal production cost of
essentially zero, making price discrimination particularly
beneficial, because charging any positive price, even to a buyer
with the lowest WTP, still exceeds the zero marginal production
costs. See Dissent at passim.
---------------------------------------------------------------------------
A seller can engage in price discrimination in several ways. One
form is known as ``second-degree price discrimination,'' by which
buyers self-sort the packages and quantities they purchase.\171\ See W.
Adams & J. Yellen, Commodity Bundling and the Burden of Monopoly, 90 Q.
J. Econ. 470, 476 (1976) (the profitability of bundling ``stem[s] from
its ability to sort customers into groups with different reservation
price [WTP] characteristics.''). Bundling, i.e., the ``practice of
selling two or more products as a package,'' Pindyck & Rubinfeld, supra
at 419, is thus a type of second-degree price discrimination. See A.
Boik & H. Takahashi, Fighting Bundles: The Effects of Competition on
Second Degree Price Competition, 12 a.m. Econ. J. 156, 157 (2020).
---------------------------------------------------------------------------
\171\ ``First-degree'' price discrimination is a hypothetical
construct by which a seller can identify the WTP of every buyer.
``Third-degree'' price discrimination occurs when the seller offers
different prices to buyers based on their different characteristics
(e.g., a senior citizen discount). See Pindyck & Rubinfeld, supra,
at 402, 404-05.
---------------------------------------------------------------------------
The applicability of these basic economic principles was understood
and explained by the parties' experts at the hearing. See, e.g., 3/15/
17 Tr. 1224-25 (Leonard) (Google's economic expert testifying that
price discrimination through bundling is ``very, very common . . . even
by pretty competitively positioned firms . . . to sort out customers
into willingness-to-pay groups.''); 3/30/17 Tr. 3983 (Gans) (Copyright
Owners' economic expert acknowledging that bundling is a form of price
discrimination); see also Dissent at 69 (same).
How does this downstream (retail level) benefit of price
discrimination impact the setting of upstream royalty rates? As the
Majority explained (in summarizing the Services' expert testimony) the
linkage is explained by the economic concept of ``derived demand'':
[M]ultiple pricing structures necessary to satisfy the WTP and
the differentiated quality preferences of downstream listeners
relate directly to the upstream rate structure to be established in
this proceeding. Professor Marx opines that the appropriate upstream
rate structure is derived from the characteristics of downstream
demand. 3/20/17 Tr. 1967 (Marx) (rate structure upstream should be
derived from need to exploit WTP of users downstream via a
percentage of revenue). This upstream to downstream consonance in
rate structures represents an application of the concept of
``derived demand,'' whereby the demand upstream for inputs is
dependent upon the demand for the final product downstream. Id.; see
P. Krugman & R. Wells, Microeconomics at 511 (2d ed. 2009)
(``[D]emand in a factor market is . . . derived demand . . . [t]hat
is, demand for the factor is derived from the [downstream] firm's
output choice'').
Determination at 19; accord Dissent at 32 (noting that ``the upstream
demand of the interactive streaming services for musical works (and the
sound recordings in which they are embodied)--known as ``factors'' of
production or ``inputs''--is derived from the downstream demand of
listeners to and users of the interactive streaming services . . . This
interdependency causes upstream demand to be characterized as ``derived
demand.'').
In the present proceeding, the PR II-based benchmark embodies the
parties' negotiated definition of Bundled Revenue for purposes of
calculating royalties on bundled interactive offerings. This is
definition in the Initial Determination. Copyright Owners' preferred
definition for Bundled Revenue--the Determination's definition--would
not only ignore this agreed-upon definition, but would also de-link the
royalty rate from the WTP of purchasers of bundles.\172\ The Judges
recognize that Copyright Owners have expressed concern the Services
could use such bundling in order to diminish revenue otherwise payable
on a higher royalty tier. However, the Majority noted that the evidence
indicated such diminishment only occurred ``in some cases.''
Clarification Order at 17. Thus, the Judges find that eliminating the
incentive for price discrimination via bundling would be a
disproportionate response and inconsistent with the broad price
discriminatory PR II-based benchmark they find useful in this
proceeding.
---------------------------------------------------------------------------
\172\ To see the incentivizing effect of the link between the
royalty level and variable WTP, consider the following example.
Assume a hypothetical bundle consists of a subscription to the
``Acme'' interactive music streaming service and the sports service
NFL Sunday Ticket. Assume also that Acme and NFL Sunday Ticket have
standalone monthly subscription prices of $9.99/month and $149.99/
month respectively, so that purchasing both separately would cost
$159.98/month. But assume the bundle price is only $140/month.
Acme's purpose in bundling its interactive music streaming service
subscription offering with NFL Sunday Ticket would be to attract
customers who had a WTP for the standalone Acme service below $9.99/
month, but a WTP at or above the $140/month for the bundle.
Under the definition in the Determination, royalties would be
paid on the standalone $9.99/month Acme price. But the purpose of
the bundling was to attract subscribers who would not pay the
standalone $9.99/month price, so no such would-be subscribers would
sign-up, and no royalties would be generated by them.
By contrast, under the Initial Determination, the standalone
price of NFL Sunday Ticket, $159.98/month, would be subtracted from
the $140/month bundle price. Although that would preclude a payment
of royalties on a revenue prong, royalties still would be paid,
under a different tier or on the mechanical floor.
---------------------------------------------------------------------------
Expert testimony in this regard is ``substantial evidence'' on
which the Judges can rely. For example, the D.C. Circuit also relied in
Johnson on the testimony of the same witness, Spotify's economic expert
witness, Professor Marx, who explained how a downstream ``lower
willingness (or ability) to pay'' among some cohorts of consumers
supports definitional terms, for student and family subscribers, that
lower royalty rates in order to further ``economic efficiency'' in a
manner that
[[Page 54457]]
``still allows more monetization of that provision of that service.''
Johnson at 392-93. Broadening her lens, Professor Marx also explained
that this price discriminatory approach is appropriate ``across all
types of services and subscribers,'' as in ``[t]he current law [and in
the PR II-based benchmark]'' which ``accommodates . . . ad-supported
services . . . and `bundled services' through different rate
provisions.'' Marx WRT ] 41 (emphasis added). See also 3/21/17 2182-83
(Hubbard) (Amazon's expert witness testifying that ``Prime Music, which
is bundled with an Amazon Prime service . . . sort[s] out customers'
willingness to pay, with an idea of trying to maximize the number of
customers,'' and agreeing that this approach constitutes ``sorting by
way of bundling.'') (emphasis added). Further, Professor Hubbard opined
that, given the revenue attribution ``measurement problem'' associated
with bundled products, the ``Phonorecords II'' approach ``with the
different categories and the minima . . . address this sort of problem
[in] a very good way.'' 3/15/17 Tr. 1221 (Hubbard).
As in the case of family and student price discrimination, the
beneficial effect of such differential pricing was supported by
industry witnesses as well as expert witnesses. See, e.g., Mirchandani
WDT ] 71 (Amazon executive citing the Phonorecords II-based benchmark
provisions regarding bundling that ``allowed Amazon to bundle Prime
Music with Amazon Prime, enabling Amazon to bring a limited catalog of
music [REDACTED]''). In sum, the same type of witness testimony that
the D.C. Circuit found sufficient to support price discriminatory
student and family plans also supports the use of the price
discriminatory bundled definition contained in the Initial
Determination.
Given the overall benefits from price discrimination, at first
blush it is curious that Copyright Owners would risk ``leaving money on
the table'' by removing the royalty-based incentive for price
discrimination via bundling. The Judges have identified this problem
earlier in this Initial Ruling, in connection with the broader issue of
the overall beneficial price discriminatory structure of the PR II-
based benchmark. As the Judges noted in that general price
discrimination context, Copyright Owners' own expert economic witnesses
acknowledged that they would not irrationally ``leave money on the
table.'' In fact, Copyright owners' aim, according to that testimony,
is to create an unregulated space--per the Bargaining Room theory--and
to use their complementary oligopoly power to negotiate price
discriminatory rates (in bundles or otherwise), which would free them
from the section 801(b)(1) requirements of reasonableness and fairness.
The Judges further find that their prior ruling on this issue in
SDARS III is distinguishable. There, a proffered bundled revenue
definition eliminated the payment of any royalty at all. Copyright
Owners quite correctly describe that result as ``absurd,'' but that is
not the result here. Rather, in the present case, the parties'
negotiated an approach that the Judges adopted in the Initial
Determination requiring royalties to be paid on interactive services
bundled with other products or services.
Even more distinguishable is Copyright Owners' assertion that Web
IV provides support for their preferred definition of service revenue.
The argument is immediately suspect, because Web IV involved per-play
royalty rates--not percent-of-revenue rates, making the definition of
revenue wholly inapposite. Further, the discussion of the price of an
``ice cream cone'' in Web IV--on which Copyright Owners rely--had
nothing to do with bundling or isolating the WTP for different products
or services. Rather, there the Judges criticized a bizarre argument
made by a licensee (who had a quantity discount for plays steered in
its direction), that was tantamount to arguing that if a vendor sells
one ice cream cone for $1.06 but a buyer could buy two for $1.06, that
the market price of an ice cream cone is thus only $.06. This argument
was indeed fallacious, because the price of an ice cream cone would be
the average of the total cost for the two cones, i.e., $.53/cone. Here,
the issue is how to address the WTP of different classes of buyers with
heterogeneous WTP, not the pricing of a discount for all purchasers
buying the same quantity. The parties utilized the Bundled Revenue
definition from the PR II-based benchmark (and in the Initial
Determination) to address the indeterminacy inherent in the variable
WTP among purchasers of the bundles, by setting floors and minima,
rather than attempt to sort out the WTP of individual (or individual
blocs) of subscribers.\173\
---------------------------------------------------------------------------
\173\ Accordingly, Copyright Owners' assertion that the Services
did not satisfy their burden of proof with regard to the Bundled
Revenue definition misses the point. The Services' burden was to
show the reasonableness of utilizing the Bundled Revenue definition
in the PR II-based benchmark, not to show that their proffered
approach measured the WTP of individual subscribers (or blocs of
subscribers). Such an alternative approach might have had merit but
no alternative approach was presented to the Judges.
To be clear, the Judges are not declaring that an alternative
Bundled Revenue definition and/or alternative rates and structures
for bundle, might not have been preferable. See 4/15/17 Tr. 5056-58
(Katz) (``[I]f someone had a proposal [with] a specific reason why
we should adjust this minimum that's something I would have
examined''); see also 3/15/17 Tr. 1227-28 (Leonard) (Google's
economic expert testifying that ``if somebody had . . . suggest[ed]
. . . a different sort of bucket that should be created . . . that's
a good idea.''). But Copyright Owners did not propose such
alternatives at the hearing, and the alternative in their Motion for
Clarification simply eviscerated the ``derived demand''-based link
between royalties and bundled offerings. As the Judges have noted
supra, in the words of Judge Patricia Wald, all judges are cabined
by the record evidence introduced by the parties. Therefore (in the
absence of a way in which to synthesize the parties' proposals in a
manner that does not ``blindside'' the parties) the Judges must
choose between the proposals that are in the record, not potentially
superior proposals that are not in the record. Here, the Judges
favor the Bundled Revenue definition in the Initial Determination
that was negotiated by the parties, incentivizes price
discrimination and pays royalties on the bundled music, over the
substituted definition in the Determination pursued by Copyright
Owners that would eliminate price discrimination, except under the
terms Copyright Owners could impose via their complementary
oligopoly power, and without regard to the statutory requirements of
a ``reasonable rate'' and a ``fair income'' for the Services.
---------------------------------------------------------------------------
For the foregoing reasons, the Judges find that the definition in
the Initial Determination (unlike the definition in the Determination)
is consistent with the Judges' other substantive rulings herein. That
is, just as the Majority abandoned its Bundled Revenue definition in
its Initial Determination because it refused to credit the PR II-based
benchmark (even as ``guidance''), the Judges here do partially rely on
the PR II-based benchmark, and thus find that it supports the Bundled
Revenue definition contained in the Initial Determination.
4. Application of Four Itemized Statutory Factors
As the forgoing analysis explains, bundling is a form of price
discrimination. Accordingly, the Judges' explanation of how price
discriminatory rates in the PR II-based benchmark interrelate with the
Factor (A) through (D) objectives in section 801(b)(1) are equally
applicable here. Accordingly, the Judges adopt by reference their
discussion of those four factors set forth supra in connection with the
PR II-based benchmark, and find that there is no basis pursuant to
those four factors to adjust the PR II-based benchmark definition of
Bundled Revenue.
V. Conclusion
On the basis of the foregoing analyses, and in consideration of the
entirety of the record, the Judges make the following determination
relating to the issues on remand from the D.C. Circuit.
[[Page 54458]]
As noted at the outset, the headline rate for all offerings
throughout the Phonorecords III period shall be as follows:
2018-2022 All-In Headline Royalty Rates
----------------------------------------------------------------------------------------------------------------
2018 2019 2020 2021 2022
----------------------------------------------------------------------------------------------------------------
Percent of Revenue........................ 11.4% 12.3% 13.3% 14.2% 15.1%
----------------------------------------------------------------------------------------------------------------
In all other respects, the rates and rate structure of the PR II-
based benchmark shall be effective as the rates and structure
throughout the Phonorecords III period.
The definition of Service Revenue for bundled offerings throughout
the Phonorecords III period shall be the definition contained in the
Initial Determination.
VI. Order
In light of the foregoing analyses and conclusions, the Judges
hereby order that the participants in this remand proceeding prepare
and submit regulatory provisions consistent with this ruling.\174\ The
participants shall file agreed regulatory language within ten days of
the date of this ruling.
---------------------------------------------------------------------------
\174\ The Judges adopt this process in order to avoid a dispute
regarding the regulatory provisions issued in connection with their
ruling. Because this is a remanded proceeding, the Judges are not
restricted to the procedures that would control in an original
proceeding, and are exercising their authority to ``make any
necessary procedural . . . rulings in any proceeding under this
chapter.'' 17 U.S.C. 801(c).
---------------------------------------------------------------------------
The Judges further order that if the participants cannot agree on a
joint submission, the Judges will accept separate submissions
respectively from (1) Copyright Owners and (2) Services, jointly. In
absence of an agreed submission, the participants shall file separate
submissions not later than 15 days after the date of this ruling.\175\
---------------------------------------------------------------------------
\175\ In their agreed upon or separate submissions, the parties
shall address the issue identified in note 135 infra, regarding
Copyright Owners' assertion that the Services omitted from their
proposed subpart C rates a portion of the Phonorecords II rates.
---------------------------------------------------------------------------
The Judges further order that parties shall not file, and the
Judges shall not consider, briefing or legal argument beyond necessary
explanatory notes to the proposed language, section by section, not to
exceed 250 words per proposed section.\176\ The Judges specifically
admonish the parties that they shall not use these submissions as a
basis to object to this Initial Ruling, either explicitly or implicitly
by proposing regulatory provisions inconsistent with this Initial
Ruling.
---------------------------------------------------------------------------
\176\ A section of the regulations is designated by a number
following the decimal after the part number, for example, Sec.
385.5. The regulations relevant to this proceeding are found in part
385.
---------------------------------------------------------------------------
The Judges further order that, within 30 days of the date of this
Initial Ruling and the attendant dissenting documents, the parties
shall file an agreed redacted version of this Initial Ruling, and the
dissents, for public viewing.
After the Judges have reviewed the parties' regulatory submissions,
the Judges shall adopt and format the necessary regulatory language
format terms relevant to this ruling and issue a restricted Initial
Determination after Remand, which shall embody their determination of
rates and terms. The parties will have an opportunity to suggest
redactions from the Initial Determination after Remand before it is
issued as a public version.
The parties shall not file any motions seeking rehearing or
reconsideration of this Initial Ruling. Subsequent to the Judges'
issuance of their Initial Determination after Remand as identified in
the immediately preceding paragraph, any party may file a Motion for
Rehearing within 15 days of the issuance of said Initial Determination
after Remand.
After ruling on any and all Motions for Rehearing as identified in
the immediately preceding paragraph, the Judges shall issue a Final
Determination after Remand.
So ordered.
Issue Date: July 1, 2022.
Stephen S. Ruwe,
Copyright Royalty Judge.
David R. Strickler,
Copyright Royalty Judge.
Suzanne M. Barnett,
Chief Copyright Royalty Judge.
Addendum to Final Ruling and Order
----------------------------------------------------------------------------------------------------------------
% of Service
provider ``Mechanical-only''
Offering revenue TCC % or TCC amount royalty floor
(percent)
----------------------------------------------------------------------------------------------------------------
Standalone Non-Portable Subscription 10.5 The lesser of 22% of TCC for 15 cents per subscriber
Offering--Streaming Only. the Accounting Period or 50 per month.
cents per subscriber per
month.
Standalone Non-Portable Subscription 10.5 The lesser of 21% of TCC for 30 cents per subscriber
Offering--Mixed. the Accounting Period or 50 per month.
cents per subscriber per
month.
Standalone Portable Subscription 10.5 The lesser of 21% of TCC for 50 cents per subscriber
Offering. the Accounting Period or 80 per month.
cents per subscriber per
month.
Bundled Subscription Offering......... 10.5 21% of TCC for the Accounting 25 cents per month for
Period. each Active Subscriber
during that month.
Mixed Service Bundle.................. 11.35 21% of TCC for the Accounting n/a.
Period.
Limited Offering...................... 10.5 21% of TCC for the Accounting n/a.
Period.
Paid Locker Service................... 12 20.65% of TCC for the n/a.
Accounting Period.
Purchased Content Locker Service...... 12 22% of TCC for the Accounting n/a.
Period.
Free nonsubscription/ad-supported 10.5 22% of TCC for the Accounting n/a.
services free of any charge to the Period.
End User.
----------------------------------------------------------------------------------------------------------------
[[Page 54459]]
B. Order 43 on Phonorecords III Regulatory Provisions (Public Version
With Federal Register Naming and Formatting Conventions)
Introduction
The present Order concerns a single issue in dispute among the
parties \177\ regarding regulatory language implementing the Judges'
Initial Ruling and Order after Remand (``Initial Ruling'') entered in
this proceeding.\178\ Subsequent to filing dueling submissions (see
footnote 2 infra), the parties filed a Joint Submission, informing the
Judges that they had ``agree[d] on all of the regulatory language''
except for certain rate percentages contained in Table 2 of the
proposed Sec. 385.21. Joint Submission . . . Regarding Regulatory
Provisions Following Initial Ruling and Order (after Remand) at 1 (Nov.
30, 2022) (``Joint Submission'') (eCRB no. 27337).
---------------------------------------------------------------------------
\177\ The parties who have joined on this dispute (through
filings after the issuance of the Initial Ruling) are the National
Music Publishers' Association and Nashville Songwriters Association
International (collectively, ``Copyright Owners'') and Amazon.com
Services LLC, Google LLC, Pandora Media, LLC, and Spotify USA Inc.
(collectively, the ``Services''). (Copyright Owners have informed
the Judges that another party, George Johnson, joins in Copyright
Owners' position with respect to the issue considered in this
Order.)
\178\ The Judges instructed the parties to ``prepare and submit
regulatory provisions consistent with this ruling.'' Initial Ruling
and Order after Remand at 114 (July 1, 2022) (eCRB nos. 26938,
27063). The Judges further instructed that, ``if the participants
cannot agree on a joint submission, the Judges will accept separate
submissions respectively from (1) Copyright Owners and (2) Services,
jointly.'' Id. The parties did not initially file an agreed-upon
joint submission as to regulatory provisions, but rather filed the
permitted separate submissions.
---------------------------------------------------------------------------
The Regulatory Language in Dispute
The dispute between the parties is whether the Judges should adopt
in the Phonorecords III regulations: (1) the several ``Total Content
Cost'' (``TCC'') rates \179\ set forth in the Phonorecords II-based
benchmark; or (2) the single 26.2% TCC rate discussed in the Initial
Ruling. This dispute relates to nine offerings made by interactive
streaming services, as detailed below:
---------------------------------------------------------------------------
\179\ TCC is defined in the Initial Ruling as ``a shorthand
reference to the extant regulatory language describing generally the
amount paid by a service to a record company for the section 114
right to perform digitally a sound recording.'' Initial Ruling at 4
n.8 (citations omitted).
------------------------------------------------------------------------
Copyright
owners'
Offering proposal Services' proposal
(percent)
------------------------------------------------------------------------
Standalone Non-Portable 26.2 The lesser of 22% of
Subscription Offering-- TCC for the Accounting
Streaming Only. Period or 50 cents per
subscriber per month.
Standalone Non-Portable 26.2 The lesser of 21% of
Subscription Offering--Mixed. TCC for the Accounting
Period or 50 cents per
subscriber per month.
Standalone Portable 26.2 The lesser of 21% of
Subscription Offering. TCC for the Accounting
Period or 80 cents per
subscriber per month.
Bundled Subscription Offering.. 26.2 21% of TCC for the
Accounting Period.
Free nonsubscription/ad- 26.2 22% of TCC for the
supported services free of any Accounting Period.
charge to the End User.
Mixed Service Bundle........... 26.2 21% of TCC for the
Accounting Period.
Purchased Content Locker 26.2 22% of TCC for the
Service. Accounting Period.
Limited Offering............... 26.2 21% of TCC for the
Accounting Period.
Paid Locker Service............ 26.2 20.65% of TCC for the
Accounting Period.
------------------------------------------------------------------------
Sources: Offering column text from Exhibit A to Joint Submission . . .
Regarding Regulatory Provisions Following Initial Ruling and Order
(after Remand) at 17 (Nov. 30, 2022) (eCRB no. 27338); Services'
Proposal column text from Services' Joint Submission of Regulatory
Provisions Ex. A at 11 (July 18, 2022) (eCRB no. 27005).
The Issue
At a high level, the remaining regulatory issue is the following:
Whether a 26.2% TCC rate identified in the hearing record, and
discussed both on appeal and on remand by the D.C. Circuit, should
substitute for TCC rates in the Phonorecords III period, or whether
these uncapped TCC rates should be set at the specific levels
ranging between 20.65% and 22% set forth in the Phonorecords II-
based benchmark adopted by the Judges in the Initial Ruling.
To frame, address, and rule on this issue, in this Order the Judges
place the parties' dispute in the context of the prior rulings by the
D.C. Circuit and the Judges in connection with this proceeding.
Background
On January 5, 2016, the Judges initiated proceedings to determine
the appropriate mechanical license royalty rates and terms for the
January 1, 2018 to December 31, 2022 period. See Notice Announcing
Commencement of Proceedings in Phonorecords III, 81 FR 255 (Jan. 5,
2016). After the parties filed their written and rebuttal testimonies
and engaged in discovery, they participated in a five-week evidentiary
hearing presided over by the Judges. See Determination of Royalty Rates
and Terms for Making and Distributing Phonorecords, 84 FR 1918, 1920,
1923-1925 (Feb. 5, 2019).\180\
---------------------------------------------------------------------------
\180\ The Determination was not unanimous. Judge David Strickler
dissented from the Majority's setting of the TCC rate, and he
proposed that the appropriate rates should essentially be those
proposed in the Phonorecords II-based benchmark proposed by several
of the Services. Thus, for clarity, this Order refers to the
``Majority Opinion'' and the ``Dissenting Opinion,'' rather than the
``Final Determination,'' when discussing the respective opinions.
---------------------------------------------------------------------------
In the Majority Opinion, the Judges adopted a ``greater-of''
royalty rate structure for the mechanical license, which contained a
TCC rate applicable to all categories of offerings.\181\ See 84 FR
1963; see also Johnson v. Copyright Royalty Board, 969 F.3d 363, 372
(D.C. Cir. 2020) (summarizing the Majority Opinion). More particularly,
the Majority adopted the following rates and rate structure:
---------------------------------------------------------------------------
\181\ The other prong in the ``greater-of'' rate structure is
the percent-of-revenue generated by the interactive streaming
service, i.e., ``service revenue.''
[[Page 54460]]
2018-2022 All-In Royalty Rates: the Greater of:
----------------------------------------------------------------------------------------------------------------
2018 2019 2020 2021 2022
----------------------------------------------------------------------------------------------------------------
Percent of Revenue.............. 11.4% 12.3% 13.3% 14.2% 15.1%
Percent of TCC.................. 22.0% 23.1% 24.1% 25.2% 26.2%
----------------------------------------------------------------------------------------------------------------
Majority Opinion at 1918, 1960.
The Services appealed.\182\ Among their arguments were the
assertions--pertinent to this Order--that the Majority: (i) violated
the Services' procedural right to fair notice by choosing a structure
that was not advanced by any party; (ii) acted arbitrarily and
capriciously by simultaneously combining a TCC prong (phased-in to
26.2% of TCC) with an increase in the percentages on the revenue prong
(phased-in to 15.1%); and (iii) failed to reasonably explain its
rejection of the Phonorecords II settlement as a benchmark. Johnson,
supra, at 376, 380-81.\183\
---------------------------------------------------------------------------
\182\ The Copyright Owners and George Johnson also appealed; all
three parties' appeals were consolidated by the D.C. Circuit.
Johnson at 375.
\183\ The annual phased-in rates are set forth in the Table
supra.
---------------------------------------------------------------------------
Copyright Owners argued in opposition that: (i) the Services'
procedural rights were not violated because ``every component'' of the
Majority's approach was contained in the hearing record; (ii) the
Majority's rate and rate structure rulings were well-reasoned,
factually supported and, therefore, not arbitrary and capricious; and
(iii) sufficient reasons existed in the record to support the
Majority's rejection of the Phonorecords II-based benchmark. Johnson,
supra, at 382-383; 387.
The D.C. Circuit vacated and remanded. More particularly, Johnson
holds as follows:
1. The Majority Determination ``failed to provide adequate
notice of the drastically modified rate structure [they] ultimately
adopted,'' which was beyond ``a reasonable range of contemplated
outcomes'' in ``the parties' pre-hearing proposals, the arguments
made at the evidentiary hearing, and the preexisting rate
structures.'' Johnson at 381-82. Accordingly, as to this issue,
``[i]f the [Judges] wish[ ] to pursue [their] novel rate structure,
[they] will need to reopen the evidentiary record.'' Id. at 383.
2. The appellate issue of whether the Majority's adoption of the
(phased-in) 26.2% TCC royalty rate was ``arbitrary and capricious''
could not be addressed--given the absence of ``adequate notice''
cited in point (1) above. Id.
3. The Majority's derivation, calculation and application of the
royalty rate of 15.1% on the revenue prong was proper.\184\ The D.C.
Circuit explained that, as to this issue, the Majority had engaged
in the ``type of line-drawing and reasoned weighing of the evidence
[that] falls squarely within the [Judges'] wheelhouse as an expert
administrative agency.'' Johnson at 386. More particularly, the D.C.
Circuit approved of the Judges' reliance on ``substantial evidence''
in the form of expert testimony to set the 15.1% service revenue
rate. Johnson, at 384-85 (emphasis added). See also id. at 388
(finding ``substantial evidence'' for the Judges' finding that an
increase in the mechanical royalty rate was necessary to address a
``marked decline in mechanical royalty income. . . .'').
---------------------------------------------------------------------------
\184\ The italicization of the word ``application'' serves to
foreshadow a critical point discussed infra: The D.C. Circuit did
not affirm any application of the 26.2% TCC rate, except for the use
of that 26.2% rate as an input derived from a specific dataset, to
set the 15.1% service revenue-based royalty rate. Johnson, supra, at
385-86; see also at 386 n.11.
---------------------------------------------------------------------------
4. The Majority's rejection of the Phonorecords II-based
benchmark is remanded because the D.C. Circuit ``cannot discern the
basis on which the [Judges] rejected the Phonorecords II rates as a
benchmark in [their] analysis, that issue is remanded to the
[Judges] for a reasoned analysis.'' Johnson at 387.
On remand, the Judges adopted procedures that mainly followed the
parties' requests. More particularly, the Judges followed the D.C.
Circuit's directive and reopened the evidentiary record to receive
evidence and testimony relating to the TCC issues. See Order Regarding
Proceedings on Remand at 2 (Dec. 15, 2020). The post-remand
supplementary record added: (1) rate evidence for the 33-months from
January 2018 through September 2020, when the parties operated under
the Majority's new (but subsequently vacated) regulations including the
TCC rates; and (2) new testimony from economic expert witnesses on
behalf of Copyright Owners and the Services. See Initial Ruling,
passim. However, none of the post-remand evidence submitted and relied
upon by the parties specifically addressed as a separate issue the
rates for the nine offerings that are the subject of the present Order.
On July 1, 2022, the Judges issued their Initial Ruling \185\--
applying Johnson and considering the entire record developed pre-remand
and post-remand. In their Initial Ruling, the Judges made several
findings that bear upon the issue at hand, viz., whether to adopt in
the Phonorecords III regulations the 26.2% TCC rate or the TCC rates
(ranging from 20.65% to 22%) from the Phonorecords II-based benchmark.
In particular, in the Initial Ruling, the Judges stated the following:
---------------------------------------------------------------------------
\185\ The findings and conclusions in the Initial Ruling were
adopted by a majority of the Judges, but two Judges filed separate
opinions. See Initial Ruling at 2 n.5. One Judge, former Chief Judge
Suzanne Barnett, dissented from the Majority's adoption in the
Initial Ruling regarding the Phonorecords II rate structure (section
II of the Initial Ruling), though not from the exception to that
benchmark with regard to the headline rate of 15.1% and the
imposition of a cap on the TCC rate prong. See Chief Judge Barnett's
``Dissent re Benchmark'' (July 1, 2022) (eCRB no. 26943). The other
opinion was issued by Judge Strickler, who dissented from the
reasoning relating to the adoption of the definition of Service
Revenue (section V), but concurred in the adoption of that
definition. See Judge Strickler's ``Dissent in Part as to Section IV
of the Initial Ruling and Order after Remand'' (July 1, 2022) (eCRB
no. 26965).
1. The Phonorecords II-based benchmark incorporates price
discriminatory features for product differentiation as between: (a)
subscription and ad-supported services; (b) portable and non-
portable services; and (c) unbundled and bundled services. See
Initial Ruling at 67-68 (noting the salutary price discriminatory
nature of the Phonorecords II-based benchmark).
2. The Phonorecords II-based benchmark ``reflect[s] a rate
structure with an adequate degree of competition, because there was
a balance of bargaining power [``countervailing power''] between the
two negotiating industrywide trade associations, offsetting the
complementary oligopoly effects in place when a ``Must Have''
licensor bargains separately with each licensee.'' Initial Ruling at
69.
3. Based upon the available record evidence, the Judges find . .
. the Services' Phonorecords II-based benchmark . . . ``more than
sufficient to satisfy the legal requisites for application, as well
as a practical benchmark, when used in conjunction with the 15.1%
headline revenue rate advocated by Copyright Owners.'' Initial
Ruling at 59.
4. ``Substantial evidence demonstrates that the Phonorecords II-
based benchmark rates, other than the headline rate, are not `too
low.' '' Initial Ruling at 73.
5. A Copyright Owner expert witness opined that ``the evidence .
. . indicates that the relative valuation ratios implied by the
current Section 115 compulsory license [i.e., the Phonorecords II-
based benchmark] implies a ``lower bound on the relative market
valuations of the reciprocal percentage of the value musical works
rights relative to sound recording rights [i.e., TCC rates] [of] 22%
and 21%.'' Initial Ruling at 78 (emphasis therein).
6. The royalty rates and terms within subpart C of the
Phonorecords II-based benchmark--which include the rates and term
for the offerings at issue in this Order--
[[Page 54461]]
are expressly ``covered by [the] foregoing analysis.'' Initial
Ruling at 93. In rejecting all of Copyright Owners' arguments for
different treatment of Phonorecords II-based benchmark rates in
Subpart C therein, the Judges declined to adopt Copyright Owners'
``re-assert[ion] [of] the same arguments with respect to subpart C''
that Copyright Owners advanced in opposing the Phonorecords II-based
benchmark ``for interactive streaming in subpart B.'' See Initial
Ruling at 93-94 (``The Judges find no reason on remand to treat the
subpart C offerings differently than the manner in which they are
treating the subpart B interactive streaming offerings . . . . That
means, however, that the various ``headline'' rates for these
subpart C offerings must also adjust to 15.1%, 131 whereas the
alternative rates (identified in subpart C as ``minima'' and
``subminima'') rates shall remain unchanged.'') (emphasis added).
7. The D.C. Circuit had affirmed that: (a) the ``headline''
percentage royalty rate (not a TCC rate) of 10.5% was too low; and
(b) that the Majority had not improperly exercised its authority
when it increased that revenue royalty rate to 15.1% (as phased-in
over the five-year rate term). Accordingly, on remand, the Judges
maintained the 15.1% (phased-in) percentage royalty rate. See, e.g.,
Initial Ruling at 4, 17.
8. The D.C. Circuit affirmed the Majority's derivation and
calculation of the 26.2% TCC rate for use as an input in calculating
the 15.1% (phased-in) service revenue percentage royalty rate.
However, Johnson vacated and remanded the Majority's application and
inclusion of the 26.2% TCC rate. Initial Ruling at 19-20.
For these reasons, the Judges decided in the Interim Ruling that:
(1) the overall Phonorecords II rates comprise a ``useful benchmark,''
when the 15.1% headline percentage rate replaces the 10.5% headline
percentage rate for the offerings in Subparts B and C of the
Phonorecords II-based benchmark; and (2) ``[t]he (phased-in) 26.2% rate
[is] unreasonable.'' Initial Ruling at 50 n.77; 88; and 93-94.
Procedures Following the Post-Remand Initial Ruling
In the Initial Ruling, the Judges directed the parties to attempt
to submit jointly agreed-upon regulatory provisions implementing the
Initial Ruling, for the Judges to consider. The Judges further ruled
that, if the parties could not agree on all the regulatory language,
they should make separate submissions regarding regulatory provisions
in dispute. See Initial Ruling at 114.
The parties agreed to many regulatory provisions but disagreed as
to several such provisions. Accordingly, they filed separate
submissions and respective replies, regarding the regulatory
provisions. Services' Joint Submission of Regulatory Provisions (July
18, 2022); Copyright Owners' Submission of Regulatory Provisions to
Implement the Initial Ruling (July 18, 2022); Services' Joint Response
to Copyright Owners' Submission of Regulatory Provisions (Aug. 5,
2022); Copyright Owners' Response to Judges' July 27, 2022 Order
Soliciting Responses Regarding Regulatory Provisions (Aug. 5, 2022).
The Judges considered those submissions and entered an order
addressing the disputed regulatory provisions. See Corrected Order
regarding Regulatory Provisions following Initial Ruling and Order
(After Remand) (Nov. 10, 2022) (``November 10th Order'').\186\
---------------------------------------------------------------------------
\186\ The November 10th Order corrected an otherwise
substantively identical order issued two days earlier, on November
8, 2023, which had inadvertently included a small amount of text.
See November 10th Order at 1.
---------------------------------------------------------------------------
In the November 10th Order, the Judges directed the parties once
more to file a joint submission ``of regulatory provisions that embody
the rulings set forth in Johnson, the Initial Ruling and this [November
10th] Order, and any aspects of the [Majority] Determination (pre-
remand) that the parties understand to remain effective after the
foregoing rulings.'' November 10th Order at 31.
On November 30, 2022, the parties made the Joint Submission (as
also identified at the outset of the present Order), in which they
provided joint regulatory language no longer in dispute that applied
the binding rulings of the Judges and the D.C. Circuit. However, as
also noted above, the parties identified the single issue in dispute
that relates to the nine service offerings described supra.\187\
---------------------------------------------------------------------------
\187\ On January 10, 2023, Spotify USA Inc., Amazon.com Services
LLC, Google LLC, Pandora Media, LLC, National Music Publishers'
Association, Inc. and the Nashville Songwriters Association
International filed a joint Motion (eCRB no. 27418) requesting
modification of the previously proposed language for 37 CFR 385.3,
which governs fees owed for late payment. There was no opposition to
the January 10, 2023 joint Motion. The Judges find good cause to
adopt the modified language, which provides that ``where payment is
due to the mechanical licensing collective under 17 U.S.C.
115(d)(4)(A)(i), late fees shall accrue from the due date until the
mechanical licensing collective receives payment.''
---------------------------------------------------------------------------
The Parties' Respective Arguments in Their November 30th Joint
Submission
Copyright Owners' Arguments
According to Copyright Owners, the Initial Ruling ``appears to
plainly acknowledge that, in light of Johnson, the derivation and
calculation of the (phased-in) 26.2% TCC rate percentage cannot be
changed.'' Joint Submission at 6. More particularly, Copyright Owners
aver that, according to the Judges' Initial Ruling, ``the D.C. Circuit
affirmed the Majority's derivation and calculation of the 26.[2]% . . .
TCC rate'' and further that ``both rate prongs''--the service revenue
rate and the TCC rate--were ``derived from the same analyses.'' Initial
Ruling at 19; Joint Submission at 6-7 (quoting Initial Ruling at 19
(emphasis removed)). Further to this point, Copyright Owners rely on
the Judges' additional language in the Initial Ruling that the pre-
remand Final Determination's ``derivation and calculation of the TCC
rate [i.e., the 26.2% rate] . . . is not subject to further
consideration on remand by the Judges.'' Joint Submission at 7 (quoting
Initial Ruling at 20 (emphasis in Initial Ruling)).\188\
---------------------------------------------------------------------------
\188\ However, Copyright Owners disregard the Initial Ruling's
observation that Johnson vacated and remanded the Majority's
application and inclusion of the 26.2% TCC rate. Initial Ruling at
19.
---------------------------------------------------------------------------
According to Copyright Owners, the foregoing points are consistent
with the limited scope of the remand, which ``was not opened for new
evidence concerning TCC rate percentages.'' Joint Submission at 7
(citations omitted). Accordingly, Copyright Owners emphasize that
``there is no evidence in the record after remand to support changing
the (phased-in) 26.2% TCC rate percentage.'' Joint Submission at 7.
Copyright Owners--characterizing the former Phonorecords II TCC rates
now at issue as newly derived and calculated--maintain that these
``new'' TCC rate percentages therefore are ``foreclosed'' by the
Initial Ruling and post-remand orders cited above. Joint Submission at
7-8.
Copyright Owners also assert that the TCC rate at issue here--``was
not appealed by the Services or challenged during the remand, nor
called into question by the Circuit in Johnson.'' Joint Submission at 8
(emphasis removed). The absence of an appeal as to this issue,
according to Copyright Owners, means that the only TCC rate supported
by Johnson is the 26.2% TCC rate. Joint Submission at 8.
The Services' Arguments
According to the Services, the Judges should adopt in the
regulations the TCC percentage rates--ranging from 20.65% to 22%--
because those rates are contained in the Phonorecords II-based
benchmark adopted by the Judges and thus essentially have been
``expressly set out by the Judges'' in two prior decisions. Joint
Submission at 2 (citing Initial Ruling at 2; November 10th Order at 6
n.13). In light of these prior Orders, the Services characterize
Copyright Owners' position as the new argument, improperly seeking
regulatory provisions that ``reflect the 26.2% rate
[[Page 54462]]
previously imposed by the [M]ajority in the now-vacated pre-remand
Final Determination.'' Id.
More pointedly, the Services argue that the Judges' Initial Ruling
already expressly considered and rejected application of the 26.2% TCC
rate. Id. (citations omitted). Further, the Services maintain that it
is because the Judges rejected the 26.2% TCC rate in the Initial Ruling
that the Judges had no need to ``substantively address the topic of TCC
rates'' in their November 10th Order. Id. at 4.
The Services further maintain that ``Johnson does not compel the
Judges to simply reinstate their original pre-remand TCC rates.'' Id.
To this point, the Services rely on the Judges' post-remand finding
that, although the error made by the Majority in adopting the 26.2% TCC
rate in the pre-appeal Phonorecords III Determination was procedural,
the ``consequence . . . was substantive.'' Id. (emphasis herein).
For the above reasons, the Services maintain that the Judges could
not possibly be required on remand to adopt an express 26.2% in any
portion of the Phonorecords III regulations.
Turning from their argument that the 26.2% TCC rate was rejected by
the Judges, the Services focus on the Judges' finding in the post-
remand Initial Ruling that the ``Phonorecords II benchmark . . . is the
`better of the benchmarks proposed by the parties . . . one that
satisfies the requirements of 17 U.S.C. 801(b)(1) in all respects,' ''
Joint Submission at 5 (quoting Initial Ruling at 2). Because the
Phonorecords II benchmark includes the TCC rates now at issue--ranging
from 20.65% to 22%--the Services maintain that those rates should
properly be included in the Phonorecords III regulations. Id.\189\
---------------------------------------------------------------------------
\189\ The Services also argue that Copyright Owners' assertion
at this time that the 26.2% TCC rate should substitute for the
Phonorecords II-based benchmark rates is procedurally untimely and
improper. The Judges only partially agree with Services' argument in
this regard. If Copyright Owners had wanted to timely make this
argument, they should have done so during the post-remand period
before the Judges entered their Initial Ruling (or, of course,
during the initial proceeding pre-appeal). In that sense, Copyright
Owners failed to avail themselves procedurally of the right to make
this substantive challenge. However, the Judges have afforded the
parties the procedural right to propose regulatory language that
they claim would implement the Initial Ruling; a procedural right
exercised by both parties, as evidenced by, for example, their
arguments in the Joint Submission. In that narrow sense, Copyright
Owners' present argument is not procedurally improper. As a matter
of substance though, as explained in ``The Judges Analysis and
Ruling'' infra, the Judges have considered herein Copyright Owners'
present arguments and found them inconsistent with the Initial
Ruling.
Finally, with regard to subsequent substantive challenges to the
Initial Ruling, the parties correctly understand that such
challenges can be made after the Judges issue their post-remand
``Initial Determination'' (a statutorily-mandated ruling). See Joint
Submission at 9 (Services agreeing with Copyright Owners'
understanding that they continue to properly ``reserve all rights
with respect to the Initial Ruling, any implementing regulations and
any Initial and Final Determination, including the right to
challenge any of the foregoing.'').
---------------------------------------------------------------------------
The Judges' Analysis and Ruling
Having considered the parties' submissions, the Initial Ruling and
all other pertinent material, the Judges rule that the 26.2% TCC rate
cannot and shall not be applied in the regulatory provisions now at
issue. Rather, the Judges rule that the TCC rates set forth in the
Phonorecords II-based benchmark shall be applied in the nine regulatory
provisions now at issue, because they are consistent with and give
effect to the Judges' Initial Ruling. The more particular bases for
this ruling are set forth below.
Most fundamentally, the Judges note at the outset that in the
Initial Ruling they expressly did not apply the 26.2% TCC rate in any
manner other than as an input--using that TCC rate only as the D.C.
Circuit directed--to calculate the 15.1% of service-revenue royalty
rate. See, e.g., Initial Ruling at 41 (``[A] careful reading of the
remand testimony by Copyright Owners' economists, Professors Watt and
Spulber, reveals that neither of them actually testifies that there is
sufficient theoretical and empirical evidence to support the . . .
26.2% TCC rate . . . .'') (emphasis in original). See also id. at 40-41
n.69 (contrasting the improper application of the 26.2% TCC as a
separate statutory rate from the use of the 26.2% TCC rate as input
from a ``bargaining model'' solely to increase the service revenue rate
to 15.1%.).\190\
---------------------------------------------------------------------------
\190\ The Services claim that this distinction constitutes a
semantic twisting of words. See Joint Submission at 7. The Judges
reject that characterization. Rather, their ruling is substantive,
not semantic, because they have relied upon the testimony of several
economic expert witnesses, including one of Copyright Owners' own
economic experts, who identified five reasons that the Judges found
to preclude adoption of the 26.2% TCC rate as a separate statutory
rate. See, e.g., Initial Ruling at 41. Moreover, not a single
economist who testified at the hearing proposed that the Judges
adopt the 26.2% TCC rate as a statutory rate, see Initial Ruling at
38, further supporting the Judges' adoption in the Initial Ruling of
the consensual negotiated TCC rates contained in the Phonorecords
II-based benchmark for the nine offerings at issue.
---------------------------------------------------------------------------
In this regard, the Initial Ruling has relied upon the clear
distinction made in Johnson between the 15.1% service revenue rate and
the 26.2% TCC rate. Compare Johnson, supra, at 385 (affirming the
Majority's application of the ``revenue rate of 15.1%'' as ``the type
of line-drawing and reasoned weighing of the evidence falls squarely
within the[ir] wheelhouse as an expert administrative agency'') with
id. at 382-83 (vacating the Majority's decision for ``significantly
hiking the TCC rate to 26.2% from approximately 17% to 22%'' without
allowing the Services an opportunity to address the issue--an error
that was even ``worse'' than the elimination of caps on certain other
TCC offerings.).
Further, the offerings now at issue were contained in the
Phonorecords II-based benchmark, and the Judges' application of that
benchmark in the Initial Ruling is unambiguous: Other than the new and
increased headline rate of 15.1%, ``the rates and rate structure of the
Phonorecords II-based benchmark proposed by the Services . . .) shall
constitute the rates and rate structure for the Phonorecords III
period.'' Initial Ruling at 2. Accordingly, with regard to the single
remaining issue, pertaining to the nine offerings listed supra, the
regulatory provisions proposed by the Services in the Joint Submission
are fully consistent with the Initial Ruling.
By contrast, Copyright Owners' proposed language introduces a
change in the Phonorecords II-based benchmark rates that was never the
subject of an evidentiary proceeding pre-or post-remand, whether
through live or written testimony. But perhaps more importantly, as a
matter of substance, Copyright Owners' proposed regulatory provisions
are inconsistent with the language and a key purpose of the Initial
Ruling, which is to adopt the Phonorecords II-based benchmark rates,
the basis of which were generated consensually by the parties, through
negotiations between industrywide trade associations, which prevented
unwarranted and disproportionate complementary oligopoly market power
from affecting the royalty rates. See Initial Ruling at 69-70.\191\
---------------------------------------------------------------------------
\191\ The Judges also note that their adoption of these 20.65%
through 22% TCC rates in the Phonorecords II-based benchmark--
because they are lower than the 26.2% rate proposed by Copyright
Owners--is consistent with their rationale for adopting that
benchmark. As the Judges explained repeatedly and throughout the
Initial Ruling, their adoption of the Phonorecords II-based
benchmark purposefully incorporates into the Phonorecords III
regulations the beneficial price discriminatory features that are
hallmarks of that benchmark. See, e.g., Initial Ruling at 65 n.98
(``[T]the granular discriminatory features that the parties had
negotiated . . . reflect an ``appropriate form and extent of price
discrimination . . . .'' The Judges emphasized this point
repeatedly. See generally Initial Ruling, passim.
Further, as the Services note, Copyright Owners themselves--even
when advocating for an otherwise across-the-board 26.2% TCC prong--
had continued to propose the 20.65% to 22% TCC rates for the nine
offerings at issue now. See Copyright Owners' Submission of
Regulatory Provisions to Implement the Initial Ruling at 15-16)
(July 18, 2022); see also Joint Submission at 6.
---------------------------------------------------------------------------
[[Page 54463]]
The Judges also reject Copyright Owners' argument that by
maintaining the 20.65% through 22% TCC rates in the Phonorecords II-
based benchmark they would be violating their prior rulings regarding
the scope of the remand. Citing to the Judges' Order Regarding
Proceedings on Remand at 1 (eCRB no. 23390) (``Remand Order''),
Copyright Owners state in their Joint Submission that that the remand
``was not opened for new evidence concerning TCC rate percentages.''
Joint Submission at 7. But the decision to re-open the existing, and
robust, evidentiary record only as to rate structure, did not limit the
scope of the remand itself, nor consideration of evidence from the
underlying proceeding.
Moreover, the Judges find no language in either the Remand Order or
the Remand Scheduling Order, and no other basis, that would support
Copyright Owners' characterization of the 20.65% through 22% TCC rates
in the Phonorecords III-based benchmark as new evidence, given that
they were expressly included in that benchmark which had been proffered
at the hearing prior to the remand.
Further, the present issue of whether the regulatory provisions
implementing the Initial Ruling should apply the Phonorecords II-based
benchmark TCC rates or the 26.2% TCC rate is not a dispute regarding
the derivation or calculation of a new TCC rate. The Phonorecords II-
based benchmark rates are self-evidently not new rates, because they
existed in that prior benchmark. Moreover, the present dispute relates
to whether the language and reasoning in the Initial Ruling are
consistent with maintaining the rates contained in the Phonorecords II-
based benchmark for the nine offerings at issue, or whether the Initial
Ruling calls for abandoning those benchmark rates and replacing them
with the 26.2% TCC rate proffered by Copyright Owners. As explained
supra, the 26.2% TCC rate was properly utilized by the Majority as an
input (combined with other evidence) in order to calculate the 15.1%
service revenue royalty rate. The record reflects no other context in
which the 26.2% TCC rate can be utilized, let alone must be utilized.
Indeed, as explained supra, the record reflects the Judges' rejection
of the 26.2% TCC rate as a stand-alone statutory royalty rate.
The Judges also reject Copyright Owners' argument that the Services
somehow waived their argument for maintaining the 20.65% through 22%
TCC Phonorecords II-based benchmark rates. More particularly, Copyright
Owners incorrectly assert that these rates were ``not appealed by the
Services. . . .'' Joint Submission at 8. Rather, the D.C. Circuit
stated unambiguously: ``[T]he Streaming Services object to the
[Judges'] . . . rejection of the Phonorecords II . . . settlement[ ] as
[a] rate benchmark[ ].'' Johnson, 969 F.3d at 384; see also id. at 386
(``The Streaming Services argue . . . that the [Judges] arbitrarily
rejected . . . [a] potential rate benchmark[ ] . . . the Phonorecords
II settlement--without adequate explanation.'').
Moreover, the D.C. Circuit repeatedly noted that it was vacating
and remanding the Majority's Determination with regard to, inter alia,
the Majority's improper decision to reject the Phonorecords II-based
benchmark writ large, i.e., without qualification by the appellate
panel that some parts of that proffered benchmark might have been
correctly rejected. See Johnson, 969 F.3d at 367, 376, 381, 387.
Obviously, virtually all the elements of the Phonorecords II-based
benchmark--including the offerings now at issue--were appealed, and not
waived, foregone or forfeited by the Services.
Likewise, Copyright Owners are wrong in their claim that the
Services had never ``challenged'' these rate issues ``during the
remand.'' Joint Submission at 8. Rather, the Services argued on remand
for the Phonorecords II-based benchmark to be applied comprehensively,
without itemizing every element of that proffered benchmark. See
Services' Joint Opening Brief (post-remand) at 19-44 (Apr. 1, 2021)
(detailing why ``the Services' proposal based on the Phonorecords II
settlement is reasonable . . . .''); see also Services' . . .
Submission of Regulatory Provisions at 2 (July 18, 2022) (``Services'
July 18th Submission'') (``[T]he Services have faithfully implemented
the task at hand--to use the rates and rate structure of the
``Phonorecords II-based benchmark'' proposed by the Services during the
remand proceeding . . . .'').\192\
---------------------------------------------------------------------------
\192\ The decision in Johnson could be construed as rejecting
one element of the Phonorecords II-based benchmark, viz., the 10.5%
headline rate, because the appellate panel affirmed the higher
Majority's adoption of the (phased-in) 15.1% headline royalty
revenue rate. The Initial Ruling is consistent with that ruling, and
this rate is not now in dispute. See Services' July 18th Submission
at 2 (the Services acknowledge that in their proposed regulatory
provisions they ``replac[ed] the headline rate'' of 10.5% with the
headline royalty rate ``set by the Judges [15.1%] in the Initial
Ruling.'').
---------------------------------------------------------------------------
Finally, the Judges find and conclude that their ruling in this
Order sets forth reasonable rates satisfying the four objectives in the
then-applicable (but now superseded) statutory rate standard contained
in 17 U.S.C. 801(b)(1).\193\ First, with regard to Factor (A),\194\ the
Judges recognize and follow the D.C. Circuit's ruling that the
Majority's decision to increase in the ``headline'' service revenue
royalty rate by 44% from 10.5% to 15.1% was supported by substantial
evidence. Johnson at 387-88.
---------------------------------------------------------------------------
\193\ The D.C. Circuit expressly declined to adopt most of the
Majority's application of the explicit statutory objectives. As to
Factor (A), regarding the objective of ``maximiz[ing] the
availability of creative works to the public,'' the D.C. Circuit
held that the Majority's finding that ``an increase in the royalty
rates for mechanical licenses was necessary to ensure the continued
viability of songwriting as a profession'' was ``supported by
substantial evidence.'' Johnson at 387-388. However, with regard to
the remaining statutory factors, Johnson instead vacated and
remanded consideration of those matters to the Judges. See Johnson
at 389. The Initial Ruling after remand considered these statutory
objectives in detail. See Initial Ruling at 90-93. (The parties made
no express argument regarding the application of these statutory
objectives in their Joint Submission.).
\194\ Factor (A) provides that rates shall be calculated to
achieve the objective of ``maximize[ing] the availability of
creative works to the public.'' 17 U.S.C. 801(b)(1)(A).
---------------------------------------------------------------------------
Further with regard to Factor (A), the Judges understand their
analysis and reasoning in the Initial Ruling--applying the Phonorecords
II-based benchmark and thus rejecting the 26.2% TCC rate--to be
applicable to the present dispute regarding the adoption of regulations
to implement the Initial Ruling. Accordingly, the Judges adopt by
reference herein their analysis and reasoning set forth at pages 90-91
of the Initial Ruling. For those reasons, the Judges decide, as they
did in the Initial Ruling, that there is no basis for yet a further
increase in the royalty rate based on Factor (A), finding ``no evidence
to suggest that the price discriminatory rates should be changed, in
order to address the connection between price discrimination and the
objective of Factor (A).'' Id. at 91.
Next, in considering Factors (B) and (C),\195\ the Judges' Initial
Ruling adopts the Majority's reasoning that the 15.1% service revenue
royalty rate provided a ``fair allocation of revenue between copyright
owners and services'' and it would be ``substantively unwarranted to
engage in any new consideration on remand of the impact, if any, of
Factors
[[Page 54464]]
(B) and (C) on the otherwise reasonable 15.1% revenue rate.'' Id. at
15-16.
---------------------------------------------------------------------------
\195\ The Factor (B) objectives (providing a ``fair return'' and
a ``fair income'' to the licensors and licensees respectively) and
Factor (C) objectives reflecting their relative roles in making the
streamed music available to the public) are typically considered
jointly, because of their overlapping concerns. See Initial Ruling
at 15 n.31 (citing Johnson, 969 at 388). In this Order, the Judges
likewise jointly address Factors (B) and (C).
---------------------------------------------------------------------------
In their Joint Submission, the parties have presented no arguments
specifically addressing how Factors (B) or (C) might support their
proposed TCC rates now at issue. Examining the record, the Judges find
and conclude that maintaining the Phonorecords II-based rates ranging
from 20.65% to 22% embodies the fairness associated with rates
negotiated between industrywide trade associations wielding relatively
comparable bargaining power, as discussed supra and in the Initial
Ruling.\196\ This notion of fairness is embodied in the determination
of the reasonable rate and, as can be the case, when one of the four
itemized statutory objectives of section 801(b)(1) is bound-up and
appropriately addressed within the broader context of setting a
reasonable rate, no further adjustment is necessary through an
invocation of an itemized statutory factor. See Determination of
Royalty Rates and Terms for Making and Distributing Phonorecords
(Phonorecords III) 84 FR 1918, 1955, 2015 (Feb. 5, 2019) (Majority and
Dissenting Opinions agreeing that ``to the extent market factors may
implicitly address any (or all) of the four itemized factors, the
reasonable, market-based rates may remain unadjusted.'').
---------------------------------------------------------------------------
\196\ In this regard, the Judges agree with the Services'
argument. See Initial Ruling at 61 (summarizing the Services'
position as to Factors (B) and (C)).
---------------------------------------------------------------------------
Finally, the Judges see no reason to alter their adoption of the
Phonorecords II-based benchmark rates for the nine offerings at issue
in this Order based upon the final listed statutory objective, Factor
(D).\197\ In the Joint Submission, Copyright Owners did not make an
express argument relating to this factor (nor did the Services).
Independently considering the potential application of Factor (D), the
Judges find no evidence that the continuation of the Phonorecords II-
based benchmark rates for the offerings at issue in this Order would
cause any disruption that Factor (D) is intended to address. Further,
as noted supra, the Judges have phased-in an increase in the headline
service revenue royalty rate from 10.5% to 15.1%--a 44% increase--
rendering unreasonable any argument that the present decision to
maintain the Phonorecords II-based TCC rates is ``disruptive'' to
Copyright Owners under the statutory Factor (D) standard.
---------------------------------------------------------------------------
\197\ ``Factor (D) . . . instructs the Judges to consider the
`competing priority' of `minimiz[ing] any disruptive impact on the
structure of the industries involved and on generally prevailing
industry practices.''' Initial Ruling at 16. More particularly,
``disruption'' potentially remediable under Factor (D) requires that
the contemplated rate ``directly produce[ ] an adverse impact that
is substantial, immediate and in the short-run because there is
insufficient time for either [party] to adequately adapt to the
changed circumstance produced by the rate change . . . .'' Initial
Ruling at 53-54.
---------------------------------------------------------------------------
Moreover, the Judges reassert their point in the Initial Ruling
that there is no need to independently consider any potential
disruption under the Factor (D) standard because the Judges have
already found an application of that rate to be unreasonable. See
Initial Ruling at 50 n.77. Further, the D.C. Circuit was aware of the
existence of the 20.65% to 22% TCC rates in the Phonorecords II-based
benchmark for these nine offerings now at issue, and not only declined
to affirm the Majority's increase in those rates to 26.2%--a
significant increase of 19% to 27% \198\--but also condemned that
increase. See Johnson at 383 (``Worse still . . .the [Judges] also
raised the total content cost [TCC] rate to 26.2%. . . .That rate
previously fell between approximately 17% and 22%''). Nothing in the
record suggest that the Judges can or should utilize the narrow
statutory ``disruption'' standard in Factor (D) of section 801(b)(1) as
a basis to override the position of the D.C. Circuit or the Judges'
analysis in the Initial Ruling as to the inapplicability of the
proffered 26.2% royalty rate.
---------------------------------------------------------------------------
\198\ An increase from 20.65% to 26.2% is a 5.55 percentage
point increase, which is an increase of 27% (rounded). An increase
from 22% to 26.2% is a 4.2 percentage point increase, which is an
increase of 19% (rounded).
---------------------------------------------------------------------------
Order
For the foregoing reasons, the Judges shall adopt in the regulatory
provisions \199\ the several ``Total Content Cost'' (``TCC'') rates set
forth in the Phonorecords II-based benchmark as proposed by the
Services.\200\
---------------------------------------------------------------------------
\199\ As addressed herein, the Judges find good cause to adopt
the joint proposal for modified language regarding late fees, in 37
CFR 385.3.
\200\ The Initial Determination shall issue forthwith.
---------------------------------------------------------------------------
Within two days of the date of issuance of this Restricted Order,
the parties shall file an agreed proposed redacted version for public
viewing.
Issue Date: April 26, 2023.
David P. Shaw
Chief Copyright Royalty Judge
C. Dissent in Part as to Section IV of the Initial Ruling and Order
After Remand by Judge David R. Strickler 201 (Redacted
Version With Federal Register Naming and Formatting Conventions)
---------------------------------------------------------------------------
\201\ I am concurring in the Majority's substantive re-adoption
of the Bundled Service Revenue definition from the Initial
Determination. As explained herein, I disagree with the Majority
regarding the procedural manner in which the Judges may reach this
result. Thus, it would be more accurate to describe this ``Dissent''
as a ``Concurring Opinion'', or an ``Opinion Concurring in Part and
Dissenting in Part.'' However, the Copyright Act does not expressly
authorize Judges to issue a ``concurring opinion,'' but rather
references the issuance of a ``dissenting opinion.'' See 17 U.S.C.
803(a)(3). Accordingly, I identify this opinion as a ``Dissent in
Part as to Section IV of the Initial Ruling and Order after
Remand.''
---------------------------------------------------------------------------
I. The Contours of This Partial Dissent
I respectfully Dissent from Section IV of the Initial Ruling and
Order after Remand (Initial Ruling). As explained herein, I conclude
that the D.C. Circuit's rulings in Johnson preclude the Judges from
engaging in ``new `agency action.' '' \202\ See Johnson v. Copyright
Royalty Board, 969 F.3d 363, 386 (D.C. Cir. 2020). Accordingly, I
cannot join with the present Majority in its determination that this
remand proceeding constitutes ``new `agency action' '' consistent with
Johnson. That argument is circular and renders useless the D.C.
Circuit's careful analysis of the procedures that are and are not
available to the Judges after they have issued their Initial
Determination.
---------------------------------------------------------------------------
\202\ I place the phrase agency action within quotation marks
inside the broader phrase new agency action to avoid potential
ambiguity and inconsistency with the directives in Johnson. There,
the D.C. Circuit held that the Judges cannot assert ``plenary
authority to revise [their] determinations whenever [they] thought
appropriate,'' because such a power grab would render ``a nullity .
. . the lines drawn by the authorizing statute . . . to confine . .
. post hoc amendments'' to statutorily identified circumstances.''
Johnson at 392. So, ``new'' means the new application of an existing
statutorily available ``agency action'' that had not previously been
invoked--not ``new'' in the sense of a form of action conjured up to
meet the moment. (When this phrase is used in a quotation I do not
use the double quotation marks.) This distinction is important
because the Majority and Copyright Owners advance new forms of
(extra-statutory) agency action, not merely new applications of
statutorily-authorized agency actions.
---------------------------------------------------------------------------
As further explained herein, the argument is circular because it
begins with the D.C. Circuit's ruling that the Determination \203\ was
improper because it invented a new procedure to change
[[Page 54465]]
the Bundled Revenue definition that was in the Initial
Determination,\204\ only to circle back to where it started by
creating--through the D.C. Circuit's own remand no less--a further and
extra-statutory ``new `agency action'''.
---------------------------------------------------------------------------
\203\ Determination of Royalty Rates and Terms for Making and
Distributing Phonorecords (Phonorecords III), 84 FR 1918 (Copyright
Royalty Board Feb. 5, 2019) (final rule and order)
(``Determination''); See also Final Determination, 16-CRB-0003-PR
(2018-2022) (Nov. 5, 2018) (citations to the Determination and to
the Dissent in this Dissent in Part are found in this document). The
Dissent is appended to and part of the same document as the
Determination.
\204\ Initial Determination, 16-CRB-0003-PR (2018-2022) (Jan.
27, 2018).
---------------------------------------------------------------------------
The Majority also renders Johnson useless, by adopting a process by
which--after the D.C. Circuit has remanded an issue because the Judges
lacked procedural authority to rule--the procedural error is
essentially honored in the breach, because the remand neuters the
effect of the D.C. Circuit's ruling.\205\
---------------------------------------------------------------------------
\205\ The Initial Ruling suggests that the Judges could have
utilized a ``further explanation'' for the switched Bundled Revenue
definition, as opposed to using ``new `agency action.''' I do not
dissent from that general point. However, even though the Majority
did not utilize this alternative approach on remand, I dissent to
the extent that section could be read to allow a fuller explanation
that would conflict with Johnson.
---------------------------------------------------------------------------
I join with the Majority though on its substantive decision to re-
adopt the definition of Bundled Revenue set forth in the Initial
Determination. As explained infra, I too find that it is clearly
preferable to the definition that was swapped into the (Final)
Determination. But as explained herein, I reconcile the procedural and
substantive points differently. I apply what I believe to be the proper
understanding of the D.C. Circuit's ruling--finding, contrary to the
Majority, no avenue for ``new `agency action''' post-remand. Rather,
the Judges must revert to the original--and substantively appropriate--
definition of Bundled Revenue in the Initial Determination.
To explicate the bases of this Dissent, my opinion as to this issue
is set forth below.
II. Introduction
The Majority and I analyze the definition of ``Service Revenue''
from ``Bundled Offerings'' (henceforth ``Bundled Revenue'' definition)
in the context of our partial adoption of the PR II-based benchmark. As
discussed supra, the Remand Majority found that the PR II-based
benchmark is a useful benchmark, particularly because of its features
that incentivize beneficial downstream price discrimination and
generate more listeners, revenues, and royalties. As explained below,
the Bundled Revenue definition--itself an element within the PR II-
based benchmark--also embodies such price discriminatory incentives.
Thus, the Judges' analysis of the PR II-based benchmark and the Bundled
Revenue definition are connected.
In the Determination, the earlier Majority likewise found the
issues relating to the PR II-based benchmark to be bound-up with the
question of the appropriate Bundled Revenue definition. But because
that earlier Majority rejected the PR II-based benchmark, it likewise
rejected the Bundled Revenue definition contained in the Initial
Determination. The definition in the Determination thus eliminated the
royalty-based incentive to engage in price discrimination via bundling.
In the interregnum between the Initial Determination and the
(Final) Determination, the Judges considered Copyright Owners' post-
hearing motion which sought, inter alia, to strike the Bundled Revenue
definition in the Initial Determination. The Majority agreed with
Copyright Owners that the definition in the Initial Determination
should be replaced. An important rationale--highly relevant in the
present context--was as follows: ``The Judges have . . . declined to
rely on the 2012 . . . benchmark . . . as the basis for the rate
structure, or, therefore, as regulatory guidance.'' Amended Order
Granting in Part and Denying in Part Motions for Rehearing at 17 (Jan.
4, 2019) (Clarification Order).\206\
---------------------------------------------------------------------------
\206\ This January 4, 2019 Order was issued in response to two
motions; the Services' ``Joint Motion for Rehearing to Clarify the
Regulations'' and Copyright Owners' ``Motion for Clarification or
Correction of Typographical Errors and Certain Regulatory Terms.''
As explained infra, Copyright Owners did not style their motion as a
``rehearing'' motion and expressly declined to argue that their
motion met the statutory and regulatory requisites for rehearing.
This remand issue pertains only to the post-hearing switch in the
Bundled Revenue definition sought and obtained by Copyright Owners
via their motion. Accordingly, it is clearer to refer herein to the
Judges' January 4, 2019 Order as the ``Clarification Order,'' rather
than as a ``Rehearing Order,'' because the semantic distinction
carries substantive overtones. (I had dissented from the Initial
Determination and the Determination, and thus did not join in the
Clarification Order.)
---------------------------------------------------------------------------
Unlike in the Determination, in this Initial Remand Ruling the
Judges do rely on the PR II-based benchmark in part because of its
price discriminatory aspects. More particularly, because the bundling
of interactive services also constitutes a form of price
discrimination, the Judges find the PR II-based benchmark definition of
Bundled Revenue set forth in the Initial Determination to be
substantively reasonable and otherwise consistent with the four
itemized factors in section 801(b)(1).
As a procedural matter though, I can neither: (1) offer any further
or fuller explanation for why the Majority made this change in the
Bundled Revenue definition nor (2) identify any ``new `agency action'''
that would permit this definitional switch. And contrary to present
Majority on remand, I also cannot identify a ``new `agency action'''
that the Judges can now take to return to the definition in the Initial
Determination. But, as explained infra, the Judges need not identify
such action, because the absence of a justification for the
definitional switch requires the Judges to revert back to the
definition in the Initial Determination.
As a substantive matter though, the Judges unanimously agree to
replace the post-hearing definition of Bundled Revenue in the
Determination and reinstate the definition set forth in the Initial
Determination.
III. Background
In this remand proceeding, the parties propose two starkly
different definitions of Bundled Revenue. Each has a dramatically
different impact on the use of the royalty structure and levels to
incentivize price discrimination in the downstream market.
The Services argue in favor of the language contained in the
Initial Determination, i.e., in their PR II-based benchmark, which
defines Bundled Revenue, in pertinent part, as
the revenue recognized from End Users [i.e., consumers] for the
Bundle less the standalone published price for End users for each of
the other component(s) of the Bundle . . . .
Initial Determination, Attachment A at 7 (Sec. 382.2 therein).
By contrast, Copyright Owners support the Majority's substituted
language contained in the Determination, which defines Bundled Revenue,
in pertinent part, as
the lesser of the revenue recognized from End Users [i.e.,
consumers] for the bundle and the aggregate standalone published
prices for End Users for each of the component(s) of the bundle that
are License Activities . . . .
Determination, Attachment A at 8 (Sec. 382.2 therein).
In Johnson, the D.C. Circuit succinctly summarized these
conflicting definitions as follows:
In its Initial Determination, the [Judges] directed that the
revenue from streaming services that are included in bundled
offerings would generally be measured by the value remaining after
subtracting the prices attributable to the other products in the
bundle.
When the Copyright Owners objected to the substance of that
definition in their motion for ``clarification,'' the Board adopted
an entirely new definition of Service Revenue for bundled offerings.
. . . This new definition generally measured the value of the
streaming component of a bundle as the standalone price of the
streaming component.
[[Page 54466]]
Johnson at 389.\207\
---------------------------------------------------------------------------
\207\ As explained infra (including by way of an example), the
Bundled Revenue definition in the Initial Determination aligns with
and incentivizes price discrimination in the downstream market, but
the definition in the Determination does not.
---------------------------------------------------------------------------
In the Clarification Order, the Judges succinctly summarized the
parties' respective positions. Id. at 17. They noted that Copyright
Owners had presented evidence that the PR II-based benchmark definition
contained in the Initial Determination ``led in some cases to an
inappropriately low revenue base,'' although the Judges ``agree that
there is no support for any sweeping inference that cross-selling has
diminished the revenue base.'' Id. at 17, 21 (emphasis added). The
Judges further noted the Services' assertion that the Bundled Revenue
definition in the Initial Determination is consistent with the Judges'
``endorsement of the classic price discrimination enabled by bundling
strategies.'' Id.\208\
---------------------------------------------------------------------------
\208\ The parties' substantive arguments are discussed in more
detail infra.
---------------------------------------------------------------------------
The Majority resolved this issue in the Clarification Order in
favor of Copyright Owners. Specifically, the Majority found that,
because of the ``indeterminacy problem'' \209\ inherent in bundling,
``the Services--not the Copyright Owners--. . . are in a position to
provide evidence of how they price bundles and value the component
parts thereof.'' Id. at 17-18. However, according to the Majority,
although the Services ``bore the burden of providing evidence
concerning the proper economic allocation of bundled revenue,'' they
``failed to do so,'' and ``[b]y default . . . the Judges must adopt an
approach to valuing bundled revenue that is in line with what the
Copyright Owners have proposed.'' Id. at 18.
---------------------------------------------------------------------------
\209\ The ``economic indeterminacy arises when ``the input
suppler . . . is paid as a percent of retail revenue, and the
bundled revenue consists of some revenue attributable to the royalty
base and other revenue excluded from the royalty base, the economic
indeterminacy of the revenue attributable to each bucket creates a
measurement problem, absent further information regarding the WTP
[Willingness-to-Pay] of buyers/subscribers to the bundle.'' SDARS
III, 83 FR 65264. As explained infra, the PR II-based benchmark
addresses this informational uncertainty with the parties'
negotiated alternative rate prongs and floors that guarantee
royalties are paid, whereas the definition in the Determination
eliminated the alignment of royalties to price discriminatory
bundles designed to increase downstream access to musical works.
---------------------------------------------------------------------------
IV. The Rulings in Johnson Regarding the Bundled Revenue Definition
The Services appealed the Majority's abandonment of the Bundled
Revenue definition in their Initial Determination. Their appeal
``challenge[s] both the legal authority and the substantive soundness''
of this switch.
First, the Services argued that the Majority failed to identify and
explain the procedural basis for making the switch after the hearing
had concluded. Second, the Services argued that, substantively, the
replacement definition in the Determination ``was arbitrary,
capricious, or unsupported by substantial evidence.'' Johnson at 389,
392.
The D.C. Circuit agreed with the Services regarding the procedural
issue and therefore vacated and remanded that aspect of the Bundled
Revenue definitional switch. In light of its procedural ruling, the
D.C. Circuit explicitly declined to rule on the Services' substantive
argument relating to the definitional switch. Id. at 392. (``Because
the Board failed to explain the legal authority for its late-breaking
rewrite, we vacate and remand that aspect of the decision [and] we have
no occasion to address the Streaming Services' separate argument that
the definition was arbitrary, capricious, or unsupported by substantial
evidence.'').
The D.C. Circuit's rulings in Johnson pertaining to this Bundled
Revenue Definition were clearly articulated. The D.C. Circuit found
that the Majority ``failed to explain under what authority'' it made a
material change to the definition ``so late in the game.'' Johnson at
389, 392. The D.C. Circuit noted that the Judges expressly declined to
treat the Clarification Motion as a motion for rehearing; consequently,
the motion did not request and the Judges did not reconsider either
evidence or legal argument. Id. at 390. Although appellate counsel
offered rationales, the D.C. Circuit rejected counsel's post hoc
reasoning. Id. and 391-92. Ultimately, the D.C. Circuit remanded the
adopted regulation ``either to provide `a fuller explanation of the
[Judges'] reasoning at the time of the agency action[,]' or to take
`new agency action' accompanied by the appropriate procedures.'' Id. at
392, citing Dep't of Homeland Sec. v. Regents of Univ. of Cal. 140
S.Ct. 1891, 1908 (2020).
To be precise, I take note of the following specific rulings in
JJohnson:
1. ``The problem is that the [Majority] has completely failed to
explain under what authority it was able to materially rework that
definition so late in the game.'' Id. at 389.
2. ``The [Majority] did not treat Copyright Owners' motion to have
the definition changed as a motion for rehearing . . . [because]
Copyright Owners' motion did not request a literal rehearing of
evidence or legal argument.'' Id. at 390 (cleaned up).
3. ``The [Majority] nowhere in its order or the [ ] Determination
explains the source of its power to make `fundamental' changes under
the authorizing statute . . . .'' Id. at 392. [same as #1]
4. ``[I]t should go without saying that we may not sustain the
Board's action based on its attorney's theorizing at oral argument . .
. vacillating gestures to uninvoked authority will not do.'' Id. at
391-92 (the D.C. Circuit alluding to its rejection of arguments also
made only by appellate counsel in support of the Majority's rejection
of the PR II-based benchmark earlier in the decision).\210\
---------------------------------------------------------------------------
\210\ Going beyond the Majority's actual rulings, the CRB
Judges' appellate counsel argued that the Majority's authority for
this definitional switch fell under either or both of the
``inherent'' statutory powers of the Judges or their ``rehearing
power.'' Id. at 392. (The D.C. Circuit rejecting appellate counsel's
argument that it was unnecessary ``for this Court to address which
one it is because . . . it could properly be understood as both.'').
---------------------------------------------------------------------------
``We must vacate the [ ] Determination's bundled offering Service
Revenue definition and remand for the [CRB Judges] either to provide
`fuller explanation of the agency's reasoning at the time of the agency
action[,]' or to take `new agency action' accompanied by the
appropriate procedures.'' Id. at 392.
V. Remand Procedure Regarding Bundled Revenue Definition
Post-remand, the Judges stated their understanding, as well as the
parties' understanding, of the issue on remand with respect to the
Bundled Revenue definition:
The Services and Copyright Owners agree that the proceedings on
remand should be limited to three issues: * * * [3] the adoption of
a revised definition of ``service revenue'' for bundled offerings
between issuing their Initial Determination and [their]
Determination.
Order Regarding Proceedings on Remand at 1 (Dec. 15, 2020) (Remand
Order).
The parties proposed, and the Judges agreed, that the record would
not be re-opened with regard to the Bundled Revenue definitional issue.
Rather, the Remand Order permitted the parties only to provide further
briefing on this matter. Id. Specifically, the Judges subsequently
permitted each party to file simultaneous Initial Remand Submissions
and simultaneous Reply Remand Submissions. See Order Adopting Schedule
for Proceedings on Remand (Dec. 20, 2020). Thereafter, seeking further
analysis on the question of ``new agency action,'' the Judges
solicited, and received, further briefing on this issue. See Notice and
Sua Sponte Order Directing the Parties to Provide Additional Materials
(Dec. 9,
[[Page 54467]]
2021) (Feb. 9, 2021); Sua Sponte Order Regarding Additional Briefing
(Feb. 9, 2021).
VI. The Parties' Submissions Regarding Bundled Revenue Definition
In their respective briefing, Copyright Owners and the Services
made arguments relating to: (1) the procedural issue, i.e., the Judges'
authority, vel non, to switch to a new Bundled Revenue definition in
the Determination; and (2) the substantive issue, i.e., the relative
merits of the two conflicting Bundled Revenue definitions. See Initial
Remand Submission of Copyright Owners at 7-10 (Apr. 1, 2021) (CO
Initial Submission); Services' Joint Opening Brief (in Services' Joint
Written Direct Remand Submission at Tab D) at 64-76 (Apr. 1, 2021)
(Services' Initial Submission); Copyright Owners' Reply Brief on Remand
(in Reply Remand Submission of Copyright Owners, Vol. 1) at 64-88 (CO
Reply); Services' Joint Reply Brief at 52-63 (Services' Reply).
A. The Procedural Issue
1. Copyright Owners' Arguments
Copyright Owners assert first that the Judges can preserve their
post-hearing switch of the Bundled Revenue definition by sidestepping
the D.C. Circuit's holding and rationale in Johnson. That is, Copyright
Owners maintain that this remand proceeding itself constitutes the
necessary form of ``new `agency action''' that Johnson invites, while
also liberating the Judges from the consequences of the procedural
infirmities identified by the D.C. Circuit. More particularly,
Copyright Owners argue:
[T]he new agency action here is a determination after remand
proceedings[.] [T]he [Judges are] largely free to chart [their] own
procedural course, and [they] ha[ve] done so in [their] [Remand]
Order. The [Judges are] not required to undertake any of the
procedural steps set forth in 17 U.S.C. 803(b) in order to take such
``new agency action.'' See 17 U.S.C. 803(d)(3) (requiring only that
on remand further proceedings be taken ``in accordance with
subsection (a)''); 37 CFR 351.15; Intercollegiate Broad. Sys., Inc.,
796 F.3d at 125 (``[N]either the Copyright Act nor the [Judge's]
regulations prescribe any particular procedures on remand.'') The
Circuit's instruction that the action be ``accompanied by the
appropriate procedures[,]'' Johnson, 969 F.3d at 392, does not
dictate what those ``appropriate procedures'' must be but instead
plainly refers to these flexible rules. See also Oceana, Inc., 321
F. Supp. 3d at 136 (explaining that when remanding to an agency, a
court generally ``may not dictate to the agency the methods,
procedures, or time dimension, for its reconsideration'').
CO Initial Submission at 71 n.33.
Copyright Owners reject the Services' position that the asserted
procedural error is an ``absence of authority'' that can never be
cured. Id. at 74 (citing Services' Proposal for Remand Proceedings at
10). They note that the D.C. Circuit did not say the Judges lacked the
authority to revisit the service revenue definition from bundles on
remand. Nor, they observe, did it say the Judges have no authority to
review the record evidence and the parties' arguments and reach the
same conclusion or a different conclusion on remand.
Copyright Owners further opine that if the only possible outcome
were for the Judges to reinstate a definition that lacked any
explanation or evidentiary support solely because it was present in the
Initial Determination, then the D.C. Circuit would not have remanded
the issue but would have simply reversed and reinstated the Initial
Determination definition. But instead, they note, the D.C. Circuit
remanded and said the Judges could take ``new agency action'' precisely
to cure the asserted procedural defect. Copyright Owners assert that
the remand allowed the parties to present the record evidence and their
arguments so that the Judges can address the definition ``afresh'' in
the remand determination. Id. at 74.
Further, Copyright Owners argue that 17 U.S.C. 803(d)(3) states
only that proceedings on remand must be in accordance with 17 U.S.C.
803(a). They contend that remand proceedings need not be confined to
procedures the Services claim are too late in the game for the Judges
to follow, again relying on the holding in Intercollegiate Broad. Sys.,
supra, that ``neither the Copyright Act nor the Board's regulations
prescribe any particular procedures on remand.'' Id. at 125.
Accordingly, they argue, the Judges can reaffirm the adopted bundled
service revenue definition following their review of the parties'
submissions without invoking section 803(c)(2) or 803(c)(4) that were
ruled inapplicable in Johnson. CO Reply at 65-66.
Also, Copyright Owners argue that the Judges may properly justify
the changed definition under section 803(c) as a fuller explanation of
the agency's reasoning at the time it was made. They urge that the
Judges could explain that, especially in light of the evidence of how
(in Copyright Owners' characterization) the Services misused the prior
definition to make service revenue completely disappear, the Judges
carry-over of the prior Bundled Revenue definition from Phonorecords II
into the Initial Determination was unintended and inadvertent.\211\ CO
Reply at 69.
---------------------------------------------------------------------------
\211\ Copyright Owners assert that the definition in the Initial
Determination conflicted with the CRB Judges' finding in the Initial
Determination that the adopted rates and terms would afford
Copyright Owners a fair return for their creative works, thereby
satisfying Factor B of the 801(b) standard. Thus, they maintain that
the definitional switch was necessary so as to not ``frustrate the
proper implementation of'' the Determination. CO Reply at 69 (citing
17 U.S.C. 801(b) and 803(c)(4)).
---------------------------------------------------------------------------
Copyright Owners also assert that, on remand, the Judges could
explain that Copyright Owners had, in their Motion for Clarification,
identified an ``exceptional case'' under section 803(c)(2) because the
prior definition failed to comport with Judges' precedent and economic
principles, and was unsupported by evidence. In addition, the Judges
reheard the evidence and legal arguments as presented in the parties'
briefs on the issue and, as a result, chose to adopt the revised
definition. Copyright Owners maintain that for the Judges to do so
would not be impermissible post-hoc reasoning. They note that the D.C.
Circuit remanded precisely because the Judges did not provide any
reason in the Determination for revising the Bundled Revenue
definition. Copyright Owners note that it was the Services, not
Copyright Owners, who appealed the Judges' modification of the bundled
service revenue definition; thus, Copyright Owners cannot be penalized
for not making every possible argument for affirmance. CO Reply at 70.
Further, and again notwithstanding the holding in Johnson,
Copyright Owners argue that the Judges have the authority to engage in
new agency action in this remand proceeding through a recasting of the
Motion for Clarification as a motion for rehearing, pursuant to 17
U.S.C. 803(c)(2)(A) and 37 CFR 353.1. In this regard, Copyright Owners
dismiss the point, raised by the D.C. Circuit, that their Motion for
Clarification could not be recast as a motion for rehearing because
Copyright Owners had explicitly disavowed that their motion sought
rehearing under the statute, and that the Judges agreed. Rather,
Copyright Owners maintain that the foregoing is not the same as a
finding that the standard could not have been met. In Copyright Owners'
view, the Judges could revisit on remand the question of whether the
rehearing standard has now been met, and find that Copyright Owners
have satisfied the ``exceptional case'' standard for granting rehearing
motions under section 803(c)(2).\212\ Copyright Owners
[[Page 54468]]
add that if the Judges do engage in new agency action that reconsiders
the Motion for Clarification as a motion for rehearing, the Judges
should fully explain their reasoning. Id. at 8-10.
---------------------------------------------------------------------------
\212\ The Majority set forth the rehearing standard in the
Clarification Order: ``According to the Copyright Act, the Judges
may grant a motion for rehearing in exceptional circumstances,
provided the moving party shows that an aspect of the determination
is ``erroneous.'' See 17 U.S.C. 803(c)(2); 37 CFR 353.1. The moving
participant must identify the aspects of the determination that it
asserts are ``without evidentiary support in the record or contrary
to legal requirements.'' 37 CFR 353.2. In general, the Judges grant
rehearing only ``when (1) there has been an intervening change in
controlling law; (2) new evidence is available; or (3) there is a
need to correct a clear error or prevent manifest injustice.'' See,
e.g., Order Denying Motion for Reh'g at 1, Docket No. 2006-1 CRB
DSTRA (Jan. 8, 2008) (SDARS I Rehearing Order) (applying federal
district court standard under Fed. R. Civ. P. 59(e)).''
Clarification Order at 2, n.3.
---------------------------------------------------------------------------
However, Copyright Owners urge that proceeding in that fashion
would add an entirely unnecessary and complicating step. They again
suggest that there is no need to reconsider or recharacterize the
Motion for Clarification as a motion for rehearing because the remand
itself affords the opportunity for the Judges to take new agency
action, which, as in a rehearing, permits them to reconsider evidence
and arguments, but, unlike a rehearing, is not limited by the
constraints of section 803(c)(2). See Copyright Owners' . . .
Additional Briefing on New Agency Action . . . Question, etc., Tab B at
7-8 (Feb. 24, 2021).
2. The Services' Arguments
The Services' arguments are based on the reasoning of the D.C.
Circuit in Johnson. Specifically, they assert that the D.C. Circuit
found only ``three ways in which the [Judges] can revise Initial
Determinations'' via ``new agency action,'' and the Judges failed to
establish that the change to the service revenue definition fit any of
those three categories. Services' Initial Submission at 64-65 (citing
Johnson at 390).\213\
---------------------------------------------------------------------------
\213\ The Services acknowledge that the Judges could
alternatively have attempted to provide on remand a fuller
explanation of their prior reasoning (in lieu of engaging in ``new
`agency action'''). That issue is considered infra.
---------------------------------------------------------------------------
According to the Services, the first statutory way the Judges may
revise an Initial Determination is to ``order rehearing `in exceptional
cases' in response to a party's motion, 17 U.S.C. 803(c)(2)(A).''
Services' Initial Submission at 65 (citing Johnson at 390). The
Services argue that the D.C. Circuit held in Johnson that the Judges'
``material revision of the `[Bundled] Revenue' definition . . . does
not fall within the [Judges'] rehearing authority under section
803(c)(2)(A)'' because ``the [Judges] [themselves] . . . w[ere]
explicit that [they] `did not treat the [Copyright Owners'] motion[ ]'
. . . `as [a] motion[ ] for rehearing under 17 U.S.C. 803(c)(2).''' Id.
The D.C. Circuit also noted that ``as the [Judges] found, . . .
Copyright Owners' motion did `not meet [the] exceptional standard for
granting rehearing motions' under section 803(c)(2).'' Id. (citing
Johnson at 390). The Services assert, quoting Johnson once more, that
the Judges were not able to make ``a volte-face'' and justify on appeal
their revision to the definition as an exercise of rehearing authority.
As the D.C. Circuit held, agency action must be justified by ``reasons
invoked by the agency at the time it took the challenged action,'' and
post-hoc rationalizations are insufficient. Id. (citing Johnson at
390).
The Services add their view that the Judges cannot revisit the
decision to deny rehearing without engaging in impermissible post-hoc
reasoning. They note the Supreme Court has explained that, while an
agency may ``elaborate later'' on its ``initial explanation'' of the
reason (or reasons) for its action, it ``may not provide new ones.''
Services' Initial Submission at 66, citing e.g., Regents at 1908. The
Services offer that the Judges, having stated that they did not
consider the Copyright Owners' motion to revise the definition to be a
motion for rehearing, cannot now conclude that the motion qualified as
one for rehearing and that the Judges in fact engaged in rehearing.
Id.\214\
---------------------------------------------------------------------------
\214\ In fact, the issue of whether to characterize Copyright
Owners' Motion for Clarification as a motion for rehearing is not
one raised by Copyright Owners, but rather by the Judges sua sponte.
---------------------------------------------------------------------------
The Services next argue, relatedly, that the Judges cannot simply
recast the Services Motion for Clarification as a rehearing motion in
an attempt to satisfy the rehearing standard. In this regard, they
maintain that Copyright Owners did not argue before the Judges or the
D.C. Circuit that their Motion for Clarification satisfied the
``exceptional cases'' standard, and have therefore waived that
argument. Id.
The Services assert that the second statutory way the Judges may
revise an Initial Determination, viz. taking ``new agency action'' to
correct a technical or clerical error under section 803(c)(4), cannot
be used to justify the modification of the Bundled Revenue definition
in the Initial Determination. The Services note that the D.C. Circuit
held specifically that the Judges' change in the Bundled Revenue
definition could not be construed as correcting a technical or clerical
error because it involved a substantive rewrite of the Service revenue
definition. Id. at 67 (citing Johnson at 391).
The Services argue that the third and final statutory justification
for the Judges to engage in ``new agency action'' is to revise the
terms in an Initial Determination is in response to ``unforeseen
circumstances'' that would frustrate the proper implementation of the
determination. Id. at 67. The Services note that the D.C. Circuit held
in Johnson that this authority did not justify the Judges' change to
the Bundled Revenue definition because the Judges did not invoke this
authority and ``the need to ground the original definition in the
record'' could not credibly be described as ``an unforeseen
circumstance.'' Id. (citing Johnson at 391).
The Services also note that the D.C. Circuit rejected the argument
that the Judges have an ``inherent authority''--unmentioned in the
statute--to make changes to the Initial Determination. The D.C. Circuit
explained that the specific restrictions Congress placed on the
[Judges'] authority in section 803 ``would be a nullity if [they] also
had plenary authority to revise [their] determinations whenever [they]
thought appropriate.'' Id. (citing Johnson at 391-92). The Services add
that even if the Judges offered a new source of authority capable of
justifying substantive changes to the [Bundled] Revenue definition now,
the Judges would be unable to rely on this ``uninvoked authority''
without engaging in impermissible post-hoc reasoning. Id.
The Services also reject Copyright Owners' position that the Judges
may sidestep the D.C. Circuit's ruling by issuing a new determination
on remand and simply arguing that any ruling after remand qualifies as
new agency action pursuant to Johnson. The Services argue that failure
to address the legal and factual issues on which the court remanded
would violate the D.C. Circuit's decision and would result in yet
another remand. The Services emphasize that the issue of authority to
make the changes to the Initial Determination are especially important
in this context, because the D.C. Circuit recognized that the Copyright
Act places limits on the Judges' authority to alter an initial
determination by defining conditions for rehearing and the types of
changes that are permitted absent a rehearing. In this regard, the
Services maintain that the Judges cannot do on remand what they lacked
authority to do in the first instance. The Services assert that the
Judges must resolve the legal question of whether authority exists to
alter the revenue definition in
[[Page 54469]]
the Initial Determination. Services' Reply at 52-54.\215\
---------------------------------------------------------------------------
\215\ In The Services agree that this remand proceeding
qualifies as a ``new agency action'' but do not maintain that a
ruling on remand that is inconsistent with Johnson would be the type
of ``new `agency action''' that Johnson permits. See Services
Additional Submission at 38-42.
---------------------------------------------------------------------------
The Services also take note of the alternative path available to
the Judges: to provide a ``fuller explanation'' of the prior conclusion
that the Judges had legal authority to revise the Service Revenue
definition. The Services maintain that if the Judges pursue the
``fuller explanation'' path, the Judges are limited to elaborating on
what they said previously, and that they cannot add new reasons they
did not initially provide. Id. at 54-55; see also Services' Joint
Rebuttal Brief Addressing the Judges' Working Proposal at 38-42 (Feb.
24, 2022) (``Services' Additional Submission'').
The Services address Copyright Owners' position that if the only
possible outcome were for the Judges to reinstate a definition that
lacked any explanation or evidentiary support solely because it was
present in the Initial Determination, then the D.C. Circuit would not
have remanded the issue but would have simply reversed and reinstated
the Initial Determination definition. The Services urge that the D.C.
Circuit could not reverse because the CRB's appellate counsel had
raised--for the first time on appeal--new justifications for the
Judges' decision to change the Initial Determination. Instead, the
Services maintain, the D.C. Circuit had to remand and give the Judges
the opportunity to address appellate counsel's new justifications in
the first instance, as the D.C. Circuit could not rule them out given
the posture of the appeal. Services' Reply at 56.
VII. Analysis and Decision
A. The Procedural Issue: Is There ``New Agency Action'' Available to
the Judges?
Having considered the parties' arguments, I conclude that the
rulings in Johnson, which clearly rejected all of the Majority's
procedural arguments seeking to justify their switch in the Bundled
Revenue definition, foreclose any avenue for procedurally justifying
this definitional switch. More particularly, I conclude that none of
the procedural avenues proffered by Copyright Owners would constitute
``new `agency action''' consonant with the holdings in Johnson.
Further, I cannot identify any other procedural device (i.e., an extra-
statutory form of agency action) that would permit the switched
definition in a manner consistent with Johnson.\216\ In addition, I
cannot identify any further or fuller explanation that might support
the Majority's procedural reasoning for swapping out the Bundled
Revenue definition in the Initial Determination and substituting the
definition in the Determination.
---------------------------------------------------------------------------
\216\ In this section, Copyright Owners' arguments regarding
recasting their Motion for Clarification as a request for rehearing,
a correction for technical or clerical errors, or for unforeseen
circumstances would constitute a new application of an existing
``form of agency action'' that the D.C. Circuit had rejected. But
Copyright Owners' argument in favor of the Judges' supposed
``inherent authority'' to enlarge their post-hearing jurisdiction is
an argument creating a new form of agency action, not an argument in
favor of new application of an existing form of authority. Likewise,
the next approach proffered by Copyright Owners, i.e. construing the
remand itself as generating the requisite agency action, which is
also the Majority's approach, is an example of an agency action that
is not statutorily specified and, as explained infra, is
inconsistent with section 803(a).
---------------------------------------------------------------------------
In reaching this conclusion, I take note of the following specific
language in Johnson:
Section 803 identifies three ways in which the Board can revise
Initial Determinations. It can (i) order rehearing ``in exceptional
cases'' in response to a party's motion; (ii) correct ``technical or
clerical errors,''; and (iii) ``modify the terms, but not the
rates'' of a royalty payment, ``in response to unforeseen
circumstances that would frustrate the proper implementation of
[the] determination.''
Johnson at 390 (citations omitted). After identifying these three
alternatives, the D.C. Circuit concluded that the CRB Judges ``rollout
of an entirely new manner for calculating the streaming service revenue
from bundled offerings fit none of those categories.'' Id.
First, I consider whether in the present case they can engage in
``new `agency action''' pursuant to 17 U.S.C. 803(c)(2)(A) by recasting
Copyright Owners' Motion for Clarification as a Motion for Rehearing. I
conclude that this avenue has been unambiguously cut-off by Johnson
and, indeed (as noted in Johnson), by the Judges' own prior ruling:
The [CRB Judges'] material revision of the Bundled Revenue
definition . . . does not fall within [their] rehearing authority
under Section 803(c)(2)(A). We have that on no less an authority
than the [CRB Judges themselves], [who were] explicit that [they]
``did not treat the Copyright Owners' motion'' to have the
definition changed ``as a motion] for rehearing under 17 U.S.C.
803(c)(2).'' That is because the Copyright Owners' motion did not
``request[ ] a literal rehearing of evidence or legal argument.''
Nor could they have because, as the [CRB Judges] found, the
Copyright Owners' motion did ``not meet [the] exceptional standard
for granting rehearing motions'' under Section 803(c)(2). . . . [The
CRB Judges] explain[ed] that . . . Copyright Owners ``failed to make
even a prima facie case for rehearing under the [rehearing]
standard''.
Johnson, 369 F.3d at 390.
Further cutting off this ``rehearing'' approach, Johnson also
expressly holds that it is a ``forceful'' principle that the D.C.
Circuit ``cannot sustain action on grounds that the agency itself
specifically disavowed. Id. Moreover, in this Initial Remand Ruling I
echo the Majority's ruling in the Clarification Order that Copyright
Owners had failed to present ``even a prima facie case for rehearing
under the applicable standard''. Clarification Order at 2.\217\
---------------------------------------------------------------------------
\217\ The first two bases for rehearing under the statute, viz.,
change in the controlling law and the availability of new evidence,
clearly do not apply. The third basis, i.e., to correct a clear
error or prevent manifest injustice, also does not apply. As
explained herein, the substantive difference between the conflicting
Bundled Revenue definitions should be resolved consistent with the
Judges' adoption of the PR II-based benchmark and the parties'
negotiated compromise of the ``price discrimination vs. revenue
diminution'' dilemma. This resolution does not constitute an
``error,'' let alone a ``clear error,'' and maintaining the parties'
rate architecture from the Initial Determination does not generate
any ``injustice,'' ``manifest'' or otherwise.
---------------------------------------------------------------------------
Next, I consider whether the Judges can engage in ``new `agency
action''' by recharacterizing their switch of the Bundled Revenue
definition as an attempted correction of ``technical or clerical
errors,'' pursuant to their ``continuing jurisdiction'' under section
803(c)(4). Once again, they cannot, and the D.C. Circuit has
effectively explained why this is so:
The [Judges] do[ ] not even try to squeeze [their] substantive
rewrite of the Service Revenue definition into that [Sec.
803(c)(4)] category. Quite the opposite, the [Judges] admit[ ] that
the new definition ``represent[s] a departure'' from the definition
in the Initial Determination, and was a substantive swap designed to
``mitigate'' the alleged ``problem'' of the original definition
leaving the interactive streaming service providers free to
``obscure royalty-based streaming revenue by offering product
bundles that include music service offerings with other goods and
services[.]'' . . . To that same point, the order itself labels the
initial and new definitions ``diametrically-opposed approaches to
valuing bundled revenues.'' . . . . Nothing technical or clerical
about that.
Johnson at 391.
On remand, I am unable to ascertain any basis for describing or
justifying the changed Bundled Revenue definition as a technical or
clerical correction. Thus, I conclude that the Judges cannot engage in
``new `agency action''' pursuant to this section.
Next, I consider whether the Judges can engage in ``new `agency
action'''--
[[Page 54470]]
by trying to squeeze the square peg of their definitional swap into the
round hole that is the ``unforeseen circumstances'' clause in section
803(c)(4). That provision permits the Judges to exercise ``continuing
jurisdiction'' if necessary to modify a regulatory term in a
determination in response to ``unforeseen circumstances,'' if the
absence of modification would frustrate the proper implementation of
the determination. Once again, Johnson shuts the door:
Come oral argument, the [Judges] attempted to explain that ``the
unforeseen circumstances would be that [they] initially adopted a
definition that was not supported by the record, and that was in
fact substantively unreasonable and would frustrate the proper
implementation of their determination.'' . . . It is hard to see how
the need to ground the original definition in the record was an
unforeseen circumstance. That is Administrative Law 101. See also 17
U.S.C. 803(c)(3) (``A determination of the [Judges] shall be
supported by the written record.'').
Johnson at 391 (cleaned up). I agree. The present panel of Judges is
bound by the D.C. Circuit's ruling that the overlooking of the need to
ground in the factual record the Bundled Revenue definition in the
Initial Determination cannot constitute an ``unforeseen circumstance.''
Accordingly, I am unable to ascertain any basis for describing or
justifying the changed Bundled Revenue definition as an ``unforeseen
circumstance'' that would justify their invocation of ``continuing
jurisdiction.''
I further consider the argument (made by the Judges' appellate
counsel and by Copyright Owners) that the Judges have the ``inherent
authority sua sponte to make any `appropriate' substantive . . . or
`fundamental' changes after the Initial Determination . . . that [they]
believe[ ] serve `the interests of enhancing the clarity and
administrability of the regulatory terms accompanying the [ ]
Determination.''' Johnson at 391. The D.C. Circuit made short work of
this argument as well, stating that, although the CRB Judges have
``considerable freedom'' with regard to determining their own
procedures
that flexibility must be exercised within the lines drawn by the
authorizing statute. Congress's decision to limit rehearing to
``exceptional cases,'' and to confine other post hoc amendments to
cases involving ``technical or clerical errors,'' would be a nullity
if the [Judges] also had plenary authority to revise [their]
determinations whenever [they] thought appropriate. The [Judges]
nowhere in [their] order or the [ ] Determination explain[ ] the
source of [their] power to make ``fundamental'' changes under the
authorizing statute . . . any time [they] deem such changes
``appropriate'' . . . even after the Initial Determination.
Johnson at 392.\218\
---------------------------------------------------------------------------
\218\ By the same reasoning, Johnson also rejected the Judges'
explanation in the Determination that they were permitted to treat
Copyright Owners' request as a general motion under Sec. 350.4) of
their regulations. Id.
---------------------------------------------------------------------------
As with regard to the proffered rationales discussed supra, I
cannot identify any authority that would allow the Judges to declare
for themselves in the present factual and legal context an ``inherent''
authority to override the Copyright Act and declare their right to
engage in ``new `agency action.'''
Finally, I consider Copyright Owners' suggestion that the remand
itself by the D.C. Circuit permits the Judges, pursuant to the
Copyright Act, to engage in any procedure necessary to support their
switch in the Bundled Revenue definition. The present Majority
essentially adopts this procedural approach. However, I reject that
argument as meritless.
The argument begins with a correct premise but seriously veers off
course. Copyright Owners correctly note (and the Services do not
disagree) that this remand proceeding constitutes ``new `agency
action.''' Copyright Owners then maintain that, because the Copyright
Act does not provide for procedures that govern remand proceedings, the
Judges are statutorily unconstrained with regard to the procedures they
may adopt. This premise, although perhaps correct in other contexts, is
most definitely incorrect in this specific context, given the clear
holding in Johnson.
Here, the D.C. Circuit has been unequivocal in identifying the
statutory limitations that precluded the Judges from switching out the
Bundled Revenue definition in their Initial Determination and replacing
it with a different definition in the Determination that was, to use
the Majority's phrase, ``diametrically opposed'' to the prior
definition, in that it would eliminate the royalty-based incentive to
price discriminate via bundling.\219\ But Copyright Owners assert that
the remand itself clothes the Judges with the procedural authority to
make the very switch that Johnson forbids! I do not understand the D.C.
Circuit to have admonished the Majority for its failure to respect the
boundaries of its jurisdiction, only to provide them, via remand, with
a back-door through which they may circle-back and exceed those very
boundaries.
---------------------------------------------------------------------------
\219\ This substantive impact of the definitional switch is
discussed infra.
---------------------------------------------------------------------------
A reading of section 803(a), upon which Copyright Owners rely,
provides a further demonstration of the error in this argument. This
subsection lists the authorities whose pronouncements the Judges must
``act in accordance with,'' including, quite unsurprisingly, ``the
decisions of the court of appeals under this chapter.'' 17 U.S.C.
803(a). In the instant case, the D.C. Circuit has unambiguously held
that the Judges lacked the statutory authority to make the definitional
switch at issue. For the Judges to construe that clear ruling as an
implicit invitation to create new extra-statutory remand procedures
that contradict the D.C. Circuit's rationale for the remand would be
inexplicable and would render useless the procedural ruling in
Johnson.\220\
---------------------------------------------------------------------------
\220\ In fact, this argument is dangerous. The CRB Judges or any
administrative agency, could willfully engage in extra-statutory
procedures to obtain a particular substantive result. If there is no
appeal, the extra-statutory procedure would be successful. But if
the extra-statutory procedure was the subject of a successful appeal
resulting in a remand, the CRB Judges (or any agency) could declare
the remand as license to engage once more in extra-statutory
procedures in order to obtain the same substantive result. This is a
``heads-I-win, tails-you-lose'' strategy.
---------------------------------------------------------------------------
In sum, I cannot and do not understand that the D.C. Circuit
intended in Johnson simply to write a meaningless procedural opinion
that the Judges could not merely ignore, but use to cleanse the very
procedural error the D.C. Circuit had condemned.\221\
---------------------------------------------------------------------------
\221\ Copyright Owners also argue that if the D.C. Circuit had
intended in Johnson to prohibit the Judges from engaging in ``new
`agency action''' on remand, they would have reversed and reinstated
the Initial Determination, rather than vacated and remanded that
aspect of the Determination. But that argument confuses prudence
with uncertainty. The D.C. Circuit prudently allowed the Judges, who
are presumed to have particular knowledge of their duties, to
consider whether there exist further explanations of their reasoning
or ``new `agency actions''' they could invoke to support their
definitional switch. That prudence hardly suggests that the D.C.
Circuit was sanguine about the existence of further explanations or
additional actions that might support the switch.
Also, 17 U.S.C. 803(d)(3) explicitly allows the D.C. Circuit to
``vacate [a] determination of the . . . Judges and remand the case
to the . . . Judges for further proceedings,'' but only expressly
allows the court to ``enter its own determination'''' in connection
with ``the amount or distribution of royalty fees and costs, and
order the repayment of any excess fees, the payment of any underpaid
fees and the payment of interest pertaining respectively thereto . .
. .'' Id. Thus, it is hardly clear that the D.C. Circuit understood
it had any choice upon vacating, save to remand for further
proceedings.
---------------------------------------------------------------------------
Accordingly, the Bundled Revenue definition in the Initial
Determination should be reinstated. As explained in the portion of the
Initial Remand Ruling in which I join, this reinstatement is harmonious
with the entirety of the Judges' findings and conclusions regarding the
other remanded issues.
[[Page 54471]]
B. The Substantive Issue: The Dueling Definitions of Bundled Revenue
1. Introduction: The Issue as Framed in the Clarification Order
Regarding the definition of ``Service Revenue'' from bundled
offerings, the Judges summarized the parties' competing arguments:
Copyright Owners presented evidence that the existing approach
led, in some cases, to an inappropriately low revenue base--but did
so in service to their argument that the Judges should reject
revenue-based royalty structures. They did not present evidence to
support a different measure of bundled revenue because their rate
proposal was not revenue-based.
The Services rely on the fact that the approach to bundled
revenue in the extant regulations is derived from the 2012
Settlement. The Judges have, however, declined to rely on the 2012
Settlement as a benchmark, as the basis for the rate structure, or,
therefore, as regulatory guidance. The Services have observed
correctly that the evidentiary records in Web IV and SDARS III
differ from the record in this proceeding.\222\
---------------------------------------------------------------------------
\222\ In Web IV and SDARS III, unlike under the Phonorecords II-
based benchmark, there were no minima or floors to provide licensors
with royalties in the event bundled offerings would otherwise fail
to generate royalties.
Clarification Order at 17 (emphasis added).
Despite these arguments, the Judges found that neither party
presented evidence adequate to support the approach advocated in post-
determination filings, because the ``economic indeterminacy problem
inherent in bundling'' remained unresolved. Id. The Judges stated that
the Services were the party in possession of the relevant information,
and concluded that the Services bore the burden of providing evidence
that might mitigate the ``indeterminacy problem'' inherent in bundling.
Because the Judges concluded that the Services had not met that burden,
they ruled that they must adopt an approach to valuing bundled revenue
that is in line with what the Copyright Owners proposed. As a result,
the Judges discarded the formula in the Initial Determination and
ruled, instead, that streaming service providers will use their own
standalone price (or comparable) for the music component (not to exceed
the value of the entire bundle) when allocating bundled revenue. Id. at
16-18.
On remand, the parties have made the following arguments regarding
the substance of the Bundled Revenue definition:
2. Copyright Owners
According to Copyright Owners, the prior Bundled Revenue definition
in the Initial Determination failed to address the `` `economic
indeterminacy' problem inherent in bundling'' appropriately and in a
way consistent with Judges' precedent. CO Initial Submission at 75
(citing Clarification Order at 16-18). Copyright Owners proceeded to
cite several portions of testimony from the Services' economic experts
who acknowledged this problem. Id. They then point to hearing testimony
in which Copyright Owners repeatedly raised the ``economic
indeterminacy'' problem and demonstrated what they characterized as the
absurd results to which the prior definition had led. Id. at 76. They
pointed out that under the Initial Determination, the first step in
computing Bundled Revenue was to identify revenues recognized from the
entire bundle (i.e., the price paid by the subscriber). The second step
was to subtract ``the standalone published price'' for all non-music
components of the bundle. According to Copyright Owners, [REDACTED].
Id. at 76, 83.
Copyright Owners point out that the Judges already found with
respect to other licenses that such an approach is not only
fundamentally unfair, but ``absurd.'' Id. (citing Web IV, 81 FR 26316,
26382 (May 2, 2016) (webcaster licenses); see also SDARS III, 83 FR
65210, 65264 (Dec. 19, 2018) (SDARS licenses) (rejecting proposed
deductions by service from bundle revenues because of the
``acknowledged `economic indeterminacy' problem inherent in
bundling''). Copyright Owners concur with the Judges' conclusion that
the same reasoning applies to Phonorecords III. Id. at 76-77 (citing
Clarification Order at 18 (``the `economic indeterminacy' problem
inherent in bundling is common to all three proceedings.'')). Copyright
Owners offer that Spotify conceded to this flaw in the definition in
the Initial Determination, but offered an alternative that contained
the same loophole. Id. at 77-78.
Copyright Owners also point out that the proponent of a term bears
the burden of proof as to adoption. The Judges made clear that the
licensee who wishes to offer bundles must bear the burden of providing
evidence that might mitigate the acknowledged economic indeterminacy
problem inherent in bundling, because any such evidence would be in its
possession, not in the possession of the licensors. Id. at 79 (citing
SDARS III, 83 FR 65264 (``bundling [is] undertaken to increase [the
Services'] revenues and it would be reasonable to assume that [the
Services have] information relevant to the economic allocation of the
bundled revenue.'')). Copyright Owners contend they presented
unrebutted evidence showing the unreasonableness of the Services'
proposed definition while the Services offered no evidence to support
their definition. Id. at 78, 79 (citing Clarification Order at 18).
Copyright Owners maintain that no Service offered evidence concerning
the separate values of the constituent parts of the bundles, or any
other evidence concerning the economic allocation of bundled revenue,
let alone the reasonableness of the definition in the Initial
Determination. Id. at 80. Copyright Owners assert that in the absence
of evidence to support the proposed definition, the Judges may adopt or
fashion a definition of service revenue for bundled offerings that
comports with the record evidence, which is precisely what the Judges
did and, through new agency action, do again. Id. at 81.
They further argue that the hearing record and the Judges'
precedent and reasoning further explain the unreasonableness of the
prior definition and support the adopted bundle revenue definition. Id.
at 82. Copyright Owners offer that in contrast to the Services'
evidentiary failure, they have provided sufficient evidence showing the
unreasonableness of the Services' proposed definition. They maintain
that the definition adopted by the Judges in the Determination was
consistent with the statutory factors and the evidence in the
proceeding showing how the prior definition had been manipulated and
``led, in some cases, to an inappropriately low revenue base.'' Id. at
83 (citing Clarification Order at 17-18).
Copyright Owners dispute the Services' assertion that there is
support for the Phonorecords II approach to bundles in the record of
this proceeding. Instead, Copyright Owners argue, the Services'
purported evidence at most supports the benefits of the practice or
strategy of bundling. They maintain that the strategy of bundling
covered music services with other products or services has nothing to
do with [REDACTED]. They offer that the definition in the Initial
Determination has nothing to do with such benefits, and that those
benefits may be equally served by a definition that ensures value is
apportioned to the music component in the bundle. CO Reply at 73-76.
3. The Services
The Services argue that the evidence in the existing written record
addressing bundles shows both that this definition is supported by the
Phonorecords II benchmark and that it has proven industry-wide
benefits. Services' Initial Submission at 68. They emphasize that
[[Page 54472]]
Copyright Owners did not propose an alternative definition of service
revenue until after the Judges issued the Initial Determination and
that any definition they propose now would fail the basic requirement
that the Judges must adopt rules ``on the basis of a written record.''
Id. (citing 17 U.S.C. 803(a)(1) and 803(c)(3)).
Addressing the merits of the definition contained in the Initial
Determination, the Services argue that it best serves the goals of the
Copyright Act; that as a bright-line, easily administered rule, it
continues the broad industry agreement from Phonorecords II, which
``was negotiated voluntarily between the Services and . . . Copyright
Owners--strong evidence that its terms are mutually beneficial.''
Services' Initial Submission at 69.
The Services contend the prior negotiated definition increases
output and incentivizes beneficial price discrimination to reach casual
and passive listeners who would otherwise not pay for music and thus
would not generate revenue from which royalties could be paid. With
regard to [REDACTED]. Id. at 71 (and record citations therein).
They further state that the definition of Bundled Revenue in
Phonorecords II also enabled funneling of many of listeners into full-
priced, full-catalog services. The Services allege that Copyright
Owners also ignore the extensive royalties that were generated. They
add that, for casual/passive listeners and those who may be funneled to
subscription services, the per-subscriber minimum guarantees that the
Copyright Owners will still be paid a fair royalty. The Services then
cite several portions of testimony from various Services' economic
experts who point out the realization of an expanded royalty pool,
which the Services offer as proving a functioning marketplace. Id. at
68-74.\223\
---------------------------------------------------------------------------
\223\ The Services' Reply reiterates this point and offers that
the testimony cited by the Copyright Owners also shows why the
Initial Determination's Service Revenue definition works for bundles
and grows royalties. Services' Reply at 57-58.
---------------------------------------------------------------------------
The Services maintain that while neither the Services nor Copyright
Owners submitted evidence specifically addressing the way that
customers, Services, or Copyright Owners might value the component
parts of bundles, such subjective valuations are unnecessary--given
that the parties' negotiated handling of the bundling issues provides
the Judges with ample support for the PR II-based benchmark definition
in the Initial Determination. See id. at 75-76.
The Services also argue that while the Judges' decision in SDARS
III did involve valuation of the music and non-music components of a
bundle, the resolution in SDARS III is inapposite because, here, the
rate structure has a way of ensuring that Copyright Owners are fairly
compensated from bundles: the statutory minimum payment. Services'
Reply at 62.
C. Analysis and Decision
The fundamental difference between the impact of the two
alternative definitions is simply stated:
Under the Initial Determination: downstream bundling and its price
discriminatory effect would be incentivized by a royalty structure that
reflects the lower WTP of consumers who subscribe by paying for a
Bundle;
Under the (Final)Determination: downstream bundling and its price
discriminatory effect would not be incentivized by a royalty structure
that reflects the lower WTP of consumers who subscribe by paying for a
Bundle.
To explain this difference, the Judges find it helpful to describe
(as in the Determination and Dissent) how bundling facilitates price
discrimination and how lower royalties for bundled streaming services
incentivize such bundling.
Price discrimination occurs when a seller offers different units of
output at different prices. See, e.g., H. Varian, Intermediate
Economics at 462 (8th ed. 2010). The benefit to the seller arises from
attempting to ``charge each customer the maximum price that the
customer is willing to pay for each unit bought.'' R. Pindyck & D.
Rubinfeld, Microeconomics at 401 (8th ed. 2013). For all goods, and
intellectual property goods such as copyrights in particular,\224\ the
social benefit is that price discrimination more closely matches the
quantity sold with the competitive quantity as the seller or licensor
better aligns the price with the WTP of different categories of buyers
or licensees. See W. Fisher, Reconstructing the Fair Use Doctrine, 101
Harv. L. Rev. 1659, 1701 (1988).
---------------------------------------------------------------------------
\224\ Streamed copies of intellectual property, such as musical
works and sound recordings, have a marginal production cost of
essentially zero, making price discrimination particularly
beneficial, because charging any positive price, even to a buyer
with the lowest WTP, still exceeds the zero marginal production
costs. See Dissent, passim.
---------------------------------------------------------------------------
A seller can engage in price discrimination in several ways. One
form is known as ``second-degree price discrimination,'' by which
buyers self-sort the packages and quantities they purchase.\225\ See W.
Adams & J. Yellen, Commodity Bundling and the Burden of Monopoly, 90 Q.
J. Econ. 470, 476 (1976) (the profitability of bundling ``stem[s] from
its ability to sort customers into groups with different reservation
price [WTP] characteristics.''). Bundling, i.e., the ``practice of
selling two or more products as a package,'' Pindyck & Rubinfeld, supra
at 419, is thus a type of second-degree price discrimination. See A.
Boik & H. Takahashi, Fighting Bundles: The Effects of Competition on
Second Degree Price Competition, 12 a.m. Econ. J. 156, 157 (2020).
---------------------------------------------------------------------------
\225\ ``First-degree'' price discrimination is a hypothetical
construct by which a seller can identify the WTP of every buyer.
``Third-degree'' price discrimination occurs when the seller offers
different prices to buyers based on their different characteristics
(e.g., a senior citizen discount). See Pindyck & Rubinfeld, supra,
at 402, 404-05.
---------------------------------------------------------------------------
The applicability of these basic economic principles was understood
and explained by the parties' experts at the hearing. See, e.g., 3/15/
17 Tr. 1224-25 (Leonard) (Google's economic expert testifying that
price discrimination through bundling is ``very, very common . . . even
by pretty competitively positioned firms . . . to sort out customers
into willingness-to-pay groups.''); 3/30/17 Tr. 3983 (Gans) (Copyright
Owners' economic expert acknowledging that bundling is a form of price
discrimination); see also Dissent at 69 (same).
How does this downstream (retail level) benefit of price
discrimination impact the setting of upstream royalty rates? As the
Majority explained (in summarizing the Services' expert testimony) the
linkage is explained by the economic concept of ``derived demand'':
[M]ultiple pricing structures necessary to satisfy the WTP and
the differentiated quality preferences of downstream listeners
relate directly to the upstream rate structure to be established in
this proceeding. Professor Marx opines that the appropriate upstream
rate structure is derived from the characteristics of downstream
demand. 3/20/17 Tr. 1967 (Marx) (rate structure upstream should be
derived from need to exploit WTP of users downstream via a
percentage of revenue). This upstream to downstream consonance in
rate structures represents an application of the concept of
``derived demand,'' whereby the demand upstream for inputs is
dependent upon the demand for the final product downstream. Id.; see
P. Krugman & R. Wells, Microeconomics at 511 (2d ed. 2009)
(``[D]emand in a factor market is . . . derived demand . . . [t]hat
is, demand for the factor is derived from the [downstream] firm's
output choice'').
Determination at 19; accord Dissent at 32 (noting that ``the upstream
demand of the interactive streaming services for musical works (and the
sound recordings in which they are embodied)--known as `factors' of
[[Page 54473]]
production or `inputs'--is derived from the downstream demand of
listeners to and users of the interactive streaming services . . . This
interdependency causes upstream demand to be characterized as ``derived
demand.'').
In the present proceeding, the PR II-based benchmark embodies the
parties' negotiated definition of Bundled Revenue for purposes of
calculating royalties on bundled interactive offerings. This is the
definition in the Initial Determination. Copyright Owners' preferred
definition for Bundled Revenue--the Determination's definition--would
not only ignore this agreed-upon definition, but would also de-link the
royalty rate from the WTP of purchasers of bundles.\226\ The Judges
recognize that Copyright Owners have expressed concern the Services
could use such bundling in order to diminish revenue otherwise payable
on a higher royalty tier. However, the Majority noted that the evidence
indicated such diminishment only occurred ``in some cases'' and that
such practices were not ``sweeping.'' Clarification Order at 17, 21.
Thus, the Judges find that eliminating the incentive for price
discrimination via bundling would be a disproportionate response and
inconsistent with the broad price discriminatory PR II-based benchmark
they find useful in this proceeding.
---------------------------------------------------------------------------
\226\ To see the incentivizing effect of the link between the
royalty level and variable WTP, consider the following example.
Assume a hypothetical bundle consists of a subscription to the
``Acme'' interactive music streaming service and the sports service
NFL Sunday Ticket. Assume also that Acme and NFL Sunday Ticket have
standalone monthly subscription prices of $9.99/month and $149.99/
month respectively, so that purchasing both separately would cost
$159.98/month. But assume the bundle price is only $140/month.
Acme's purpose in bundling its interactive music streaming service
subscription offering with NFL Sunday Ticket would be to attract
customers who had a WTP for the standalone Acme service below $9.99/
month, but a WTP at or above the $140/month for the bundle.
Under the definition in the Determination, royalties would be
paid on the standalone $9.99/month Acme price. But the purpose of
the bundling was to attract subscribers who would not pay the
standalone $9.99/month price, so no such would-be subscribers would
sign-up, and no royalties would be generated by them.
By contrast, under the Initial Determination, the standalone
price of NFL Sunday Ticket, $159.98/month, would be subtracted from
the $140/month bundle price. Although that would preclude a payment
of royalties on a revenue prong, royalties still would be paid,
under a different tier or on the mechanical floor.
---------------------------------------------------------------------------
Expert testimony in this regard is ``substantial evidence'' on
which the Judges can rely. For example, the D.C. Circuit also relied in
Johnson on the testimony of the same witness, Spotify's economic expert
witness, Professor Marx, to affirm the inclusion of the price
discriminatory structure for student and family plans. Johnson, 969
F.3d at 392-94. Professor Marx explained how a downstream ``lower
willingness (or ability) to pay'' among some cohorts of consumers
supports definitional terms, for student and family subscribers, that
lower royalty rates in order to further ``economic efficiency'' in a
manner that ``still allows more monetization of that provision of that
service.'' Johnson at 392-93. Broadening her lens, Professor Marx also
explained that this price discriminatory approach is appropriate
``across all types of services and subscribers,'' as in ``[t]he current
law [and in the PR II-based benchmark]'' which ``accommodates . . . ad-
supported services . . . and `bundled services' through different rate
provisions.'' Marx WRT ] 41 (emphasis added). See also 3/21/17 2182-83
(Hubbard) (Amazon's expert witness testifying that ``Prime Music, which
is bundled with an Amazon Prime service . . . sort[s] out customers'
willingness to pay, with an idea of trying to maximize the number of
customers,'' and agreeing that this approach constitutes ``sorting by
way of bundling.'') (emphasis added). Further, Professor Hubbard opined
that, given the revenue attribution ``measurement problem'' associated
with bundled products, the ``Phonorecords II'' approach ``with the
different categories and the minima . . . address this sort of problem
[in] a very good way.'' 3/15/17 Tr. 1221 (Hubbard).
As in the case of family and student price discrimination, the
beneficial effect of such differential pricing was supported by
industry witnesses as well as expert witnesses. See, e.g., Mirchandani
WDT ] 71 (Amazon executive citing the Phonorecords II-based benchmark
provisions regarding bundling that ``allowed Amazon to bundle Prime
Music with Amazon Prime, enabling Amazon to bring a limited catalog of
music [REDACTED]''). In sum, the same type of witness testimony that
the D.C. Circuit found sufficient to support price discriminatory
student and family plans also supports the use of the price
discriminatory bundled definition contained in the Initial
Determination.
Given the overall benefits from price discrimination, at first
blush it is curious that Copyright Owners would risk ``leaving money on
the table'' by seeking to remove the royalty-based incentive for price
discrimination via bundling. The Judges have identified this problem
earlier in this Initial Remand Ruling, in connection with the broader
issue of the overall beneficial price discriminatory structure of the
PR-based benchmark. As the Judges noted in that general price
discrimination context, Copyright Owners' own expert economic witnesses
acknowledged that they would not irrationally leave money on the table.
In fact, Copyright owners' aim, according to that testimony, is to
create an unregulated space--per the Bargaining Room theory--and to use
their complementary oligopoly power to negotiate price discriminatory
rates (in bundles or otherwise), which would free them from the section
801(b)(1) requirements of reasonableness and fairness.
The Judges further find that their prior ruling on this issue in
SDARS III is distinguishable. There, a proffered bundled revenue
definition eliminated the payment of any royalty at all. Copyright
Owners quite correctly describe that result as ``absurd,'' but that is
not the result here. Rather, in the present case, the parties'
negotiated an approach that the Judges adopted in the Initial
Determination requiring royalties to be paid on interactive services
bundled with other products or services.
Even more distinguishable is Copyright Owners' assertion that Web
IV provides support for their preferred definition of service revenue.
The argument is immediately suspect, because Web IV involved per-play
royalty rates--not percent-of-revenue rates, making the definition of
revenue wholly inapposite. Further, the discussion of the price of an
``ice cream cone'' in Web IV--on which Copyright Owners rely--had
nothing to do with bundling or isolating the WTP for different products
or services. Rather, there the Judges criticized a bizarre argument
made by a licensee (who had a quantity discount for plays steered in
its direction), that was tantamount to arguing that if a vendor sells
one ice cream cone for $1.06 but a buyer could buy two for $1.06, that
the market price of an ice cream cone is thus only $.06. This argument
was indeed fallacious, because the price of an ice cream cone would be
reasonably identified as the average of the total cost for the two
cones, i.e., $.53/cone, and never as $.06 per cone.
Here, the issue, is how to address the WTP of different classes of
buyers with heterogeneous WTP, not the pricing of a quantity discount.
The parties addressed this issue by utilizing the Bundled Revenue
definition contained in the PR II-based benchmark (and in the Initial
Determination) to address the indeterminacy inherent in the variable
WTP among purchasers of the bundles, by setting floors and minima,
rather than attempt to sort out the WTP of individual (or individual
blocs) of
[[Page 54474]]
subscribers. The ``ice cream cone'' issue in Web IV is wholly
unrelated, and the SDARS III situation, as explained supra, is also
distinguishable.\227\
---------------------------------------------------------------------------
\227\ The foregoing analysis also explains why Copyright Owners'
assertion that the Services did not satisfy their burden of proof
with regard to the Bundled Revenue definition misses the point. The
Services' burden was to show the reasonableness of utilizing the
Bundled Revenue definition in the PR II-based benchmark, not to show
that their proffered approach measured the WTP of individual
subscribers (or blocs of subscribers). Such an alternative approach
might have had merit but no alternative approach was presented to
the Judges.
To be clear, the Judges are not declaring that an alternative
Bundled Revenue definition and/or alternative rates and structures
for bundle, might not have been preferable. See 4/15/17 Tr. 5056-58
(Katz) (``[I]f someone had a proposal [with] a specific reason why
we should adjust this minimum that's something I would have
examined,''). See also 3/15/17 Tr. 1227-28 (Leonard) (Google's
economic expert testifying that ``if somebody had . . . suggest[ed]
. . . a different sort of bucket that should be created . . . that's
a good idea.''). But Copyright Owners did not propose such
alternatives at the hearing, and the alternative in their Motion for
Clarification simply eviscerated the ``derived demand''-based link
between royalties and bundled offerings. As the Judges have noted
supra, in the words of Judge Patricia Wald, all judges are cabined
by the record evidence introduced by the parties. Therefore (in the
absence of a way in which to synthesize the parties' proposals in a
manner that does not ``blindside'' the parties) the Judges must
choose between the proposals that are in the record, not potentially
superior proposals that are not in the record. Here, the Judges
favor the Bundled Revenue definition in the Initial Determination
that was negotiated by the parties, incentivizes price
discrimination and pays royalties on the bundled music, over the
substituted definition in the Determination pursued by Copyright
Owners that would eliminate price discrimination, except under the
terms Copyright Owners could impose via their complementary
oligopoly power, and without regard to the statutory requirements of
a ``reasonable rate'' and a ``fair income'' for the Services.
---------------------------------------------------------------------------
For the foregoing reasons, I find that--even if the Judges had a
procedural mechanism by which to support the switch in the Bundled
Revenue definition--I would decline to utilize it in this Initial
Remand Ruling, because the definition in the Initial Determination
(unlike the definition in the Determination) is consistent with the
Judges' other substantive rulings herein. That is, just as the Majority
abandoned its Bundled Revenue definition in its Initial Determination
because it refused to credit the PR II-based benchmark (even as
``guidance''), the Judges here do partially rely on the PR II-based
benchmark, and thus find that it supports the Bundled Revenue
definition contained in the Initial Determination.
VIII. Application of the Four Itemized Statutory Factors
As the forgoing analysis explains, bundling is a form of price
discrimination. Accordingly, the Judges' explanation of how price
discriminatory rates in the PR II-based benchmark interrelate with the
Factor A through D objectives in section 801(b)(1) are equally
applicable here. Accordingly, the Judges incorporate by reference here
their discussion of those four factors set forth supra in connection
with the PR II-based benchmark, and find that there is no basis
pursuant to those four factors to adjust the PR II-based benchmark
definition of Bundled Revenue.
IX. Conclusion
This Dissent in part is issued as a RESTRICTED document. Within 30
days of the date of issuance, the participants shall file a version of
this Dissent with agreed redactions to permit viewing by the public.
Issue Date: July 2, 2022.\228\
---------------------------------------------------------------------------
\228\ Technical difficulties on July 1 caused the delay in
filing of this Dissent until July 2.
---------------------------------------------------------------------------
David R. Strickler,
Copyright Royalty Judge
D. Dissent in Part Re Benchmark (Redacted Version With Federal Register
Naming and Formatting Conventions)
The Copyright Royalty Judges (Judges) sit as a panel in all
determination proceedings. See 17 U.S.C. 803(a)(2). A majority of two
Judges is sufficient to issue a determination. See 17 U.S.C. 803(a)(3).
If any Judge dissents from the majority determination, that dissenting
Judge may issue a dissenting opinion and file it with the majority's
determination. Id. The Judges accept this same standard with regard to
their issuance of the present Initial Ruling and Order after Remand
(Initial Ruling).
The undersigned Judge, author of this dissent in part (Benchmark
Dissent) respectfully dissents \229\ from the Initial Ruling of the
majority (Remand Majority) on the issue of adopting as a benchmark for
current rates and terms the rates and terms adopted after a settlement
by the parties to the preceding phonorecords proceeding.\230\ It should
be noted that the Remand Majority adopts the rate structure from
Phonorecords II, but retains the headline percent-of-revenue rate
adopted in the Determination.\231\
---------------------------------------------------------------------------
\229\ The dissenting Judge does not fault the economic analysis
of the Remand Majority on this issue. The dissenting Judge is not
the Judge selected for ``a significant knowledge of economics.'' See
17 U.S.C. 802(a)(1). This Benchmark Dissent is based on a broader
reading of the requirements of section 801 of the Copyright Act,
viz. ``to make determinations of reasonable terms and rates. . .''
consistent, of course, with the record evidence and sound legal and
economic analysis. The role of the Judge is to weigh evidence; two
Judges might rightfully and respectfully disagree on where that
scale balances. The Remand Majority's analysis led those Judges to
conclude that they were bound to re-introduce the rate structure
devised in the Phonorecords II proceeding. The Benchmark Dissent
concludes that the economic analysis outlined in the Initial Ruling
supports, but does not dictate, that result, but that the goal of
reasonableness can be met with different structure(s). The Benchmark
Dissent does not construct or propose a detailed, different
structure. To do so would be an inefficient application of judicial
resources at this late stage of this proceeding. The Benchmark
Dissent finds, however, that both licensor and licensee participants
agreed in this proceeding that a less complex rate structure is
warranted.
\230\ The preceding proceeding, referred to as Phonorecords II,
consisted of a final rule adopting the participants' settlement
agreement as regulatory terms and rates. See Final Rule, Adjustment
of Determination of Compulsory License Rates for Mechanical and
Digital Phonorecords, Docket No. 2011-3 CRB Phonorecords II, 78 FR
67938 (Nov. 13, 2013), Technical Amendment at 78 FR 76987 (Dec. 20,
2013). In this partial dissent, references to Phonorecords II, PR
II, and PR II-based benchmark are references to this final rule.
\231\ Determination of Royalty Rates and Terms for Making and
Distributing Phonorecords (Phonorecords III), 84 FR 1918 (Copyright
Royalty Board Feb. 5, 2019) (final rule and order) (Determination);
See also Final Determination, 16-CRB-0003-PR (2018-2022) (Nov. 5,
2018).
---------------------------------------------------------------------------
I. Areas of Concurrence
A. Background Statements
The Benchmark Dissent adopts the statements regarding the
background and procedural posture of this remand proceeding. See
Initial Ruling at 1-2.
B. Percent of Revenue Rate
The Benchmark Dissent agrees with the Remand Majority's retention
of the headline percent-of-revenue rate and its phase-in over the
period at issue.
C. Definition of Service Revenue for Bundled Offerings
For the reasons articulated in the Initial Ruling and the reasoning
of the judge dissenting from that portion of the Initial Ruling, the
definition of Service Revenue for bundled offerings contained in the
Initial Determination must be adopted. See Initial Determination (Jan.
27, 2018). Adoption of the Phonorecords II (PR II) rate structure
requires that the original definition pertain.
II. Area of Dissent
The first function of the Judges is ``to make determinations . . .
of reasonable terms and rates of royalty payments. . . .'' 17 U.S.C.
801(b)(1). Under the statute in effect during the captioned proceeding,
the rates shall be calculated to achieve four statutory objectives. Id.
The terms of payment of the rates, however, are not subject to any
particular statutory restrictions or guidelines. See, e.g., Live365 v.
Copyright Royalty Bd., 698 F. Supp. 2d 25, 29-30 (D.D.C. 2010) (``In
performing their duties, the [Judges have] broad discretion to . . .
impose regulations
[[Page 54475]]
governing the rates and terms of copyright royalties. . . .'').\232\
---------------------------------------------------------------------------
\232\ The Judges' regulations are, of course, subject to
approval by the Librarian of Congress. 17 U.S.C. 802(f)(A)(i); see
Live365 v. Copyright Royalty Bd., 698 F. Supp. 2d 25, 29-30 (D.D.C.
2010).
---------------------------------------------------------------------------
In general, in promulgating regulations the Judges aim to effect
efficient and effective payment of royalty license fees. Regulations
relating to license royalty rates describe the rates the Judges
determine to be reasonable, whether presented by agreement of the
affected parties or after adjudication. The regulations include, where
necessary, methods of calculation of the payable royalties. The
regulations also include such provisions as recordkeeping requirements,
late fee assessments, and audit authority. As the Remand Majority
points out, simplicity and clarity were not among the statutory factors
applicable to determining royalty rates in the captioned underlying
proceeding. Simplicity and clarity should, however, be paramount among
the Judges' considerations in governing rate payment procedures.
In recent proceedings, the Judges have emphasized that the statute
requires that they set both rates and terms. At the end of a different
royalty rate proceeding, having been confronted with competing proposed
regulations, or even with largely agreed regulatory terms, upon which
the parties had proffered no evidence, the Judges cautioned counsel in
this proceeding:
Please be reminded that the Judges have an obligation to set
both rates and terms. . . . In any proceeding, just because a
regulation is in the current Code of Federal Regulations does not
mean that the Judges are adopting that term. . . . The Judges cannot
determine rates or terms without an evidentiary record. . . . The
Judges cannot adopt any terms of royalty administration unless the
parties present evidence to support their proposed terms.
Tr. 03/08/2017 (Barnett, J.) While chapter 8 of the Copyright Act
encourages settlement, the Judges are not mandated to adopt parties'
settlements if they find they face opposition that discounts
reasonableness or if the proposed regulations are contrary to law. See,
e.g., Determination of Royalty Rates and Terms . . . (Phonorecords IV),
87 FR 18342, 18347, 18349 (Mar. 30, 2022).
In the proceeding underlying the Determination, the parties
proffered a variety of proposed regulations.\233\ Copyright Owners
contended that the extant rate structure ``should be modified and
simplified.'' Copyright Owners' Amended Proposed Rates and Terms (5/17/
2017) at 2. Copyright Owners argued that the ten different rate
categories should be ``no longer applicable'' as Copyright Owners
proposed application of the same rates and rate structure to ``all
interactive streams and/or limited downloads [except bundles],
regardless of the business model employed.'' Id. at 3. Copyright
Owners' rate proposal hinged on a per-unit calculation across the
board: the greater of a per-play amount or a per end user amount.
---------------------------------------------------------------------------
\233\ Spotify, as the only pure-play service, offered simplified
regulations, but only because it did not propose any rates or terms
for bundled or locker services. Spotify advocated elimination of the
per-subscriber stop-gap alternative in the greater-of percent-of-
revenue/percent-of-TCC calculation.
---------------------------------------------------------------------------
Amazon proposed retaining the PR II rate structure. See Proposed
Findings . . . of Amazon (May 13, 2017) ] AM-F-25. Amazon argued that
the PR II rate structure ``enabled Amazon to develop a varied
assortment of services. . . .'' Id. Amazon contended that the different
royalty rates permit price discrimination by the Services. Id. ]] AM-F-
47, 49. Amazon conflates price discrimination with provision of
heterogenous musical tastes and preferences. Id. ] AM-F-48. Amazon's
proposal mimicked the regulations adopted by agreement in the
immediately prior proceeding.
Apple proposed a per-play rate calculation, which would render the
PR II rates and rate structure obsolete. Notwithstanding the different
structure, however, Apple offered valid criticisms of the PR II rate
structure. Apple termed the PR II rate structure ``problematic.'' See
Apple Inc.'s Findings . . . and Conclusions . . . (May 11, 2017) at 30.
Apple argued that the PR II rate structure was ``overly complex,
economically unsound, and unpredictable.'' Id. ] APL-F65. Apple
acknowledged that these shortcomings resulted in ``a loss of trust and
overall dissatisfaction with interactive streaming among songwriters. .
. .'' Id.
Apple noted that across the ten rate categories in the PR II rates,
``there are roughly 79 different calculations that can be made.'' Id. ]
APL-F67. Apple argued that the PR II rate structure was ``not
transparent or easy to understand'' for copyright owners and created
``uncertainty for services, who may find it difficult to predict which
prong . . . will kick in in any given month.'' Id. ]] 68-69. Apple
opined that, rather than encouraging new business models, the PR II
rate structure ``tends to stifle innovation around new pricing or
distribution models, as services are incentivized to create businesses
that fit into the ten pre-defined `boxes.' '' Id. ] 70. Apple further
argued that the PR II rates were economically unsound because they are
based on revenue, which is unrelated to demand for a given copyright
owner's song. Id. ] 71.
Google's proposal, from which the Majority derived the uncapped TCC
rate prong of the Determination, contended that the ``fragmented
service categories are unnecessary under [its] proposal. . . .''
Google, Inc.'s Proposed Findings . . . and Conclusions. . . (May 11,
2017) ] GPFF58. Google acknowledged questions regarding the complexity
of the PR II rate structure. Google, therefore proposed a rate
structure that would both streamline the regulations and protect
Copyright Owners' concerns regarding Services' revenue deferment and
displacement. Id. ] GPFF57.
In the captioned underlying proceeding, the Judges heard little
evidence offered in resounding support or vehement objection to the
regulations the parties proffered. No party argued or supported the
proposition that the PR II rate structure was the only way, or even the
best way, to achieve license fee payment.\234\
---------------------------------------------------------------------------
\234\ The Benchmark Dissent does not argue that the PR II rate
structure did not achieve its purpose. Indeed, the all-in, greater-
of, lesser-of scheme with payment minima and mechanical floors
achieved the goals of (1) supporting increased absolute revenue
through downstream price discrimination and (2) protecting creators
from potential loss resulting from licensees' revenue deferral or
displacement. The Judges have never denied the value of price
discrimination in these or other rate setting proceedings.
---------------------------------------------------------------------------
In this remand proceeding, no party argued against the all-in
approach to rate calculation. The parties disagreed regarding retention
of ``mechanical floors'' for configurations for which the Services must
pay mechanical royalties both to Copyright Owners in this proceeding
under section 115 and to Performing Rights Organizations (PROs)
according to the determinations of the ``Rate Court.'' \235\ The
parties disagreed over imposition of a cap on the TCC prong \236\ in
the greater-of percent-of-revenue calculation. They also disagreed over
retention or elimination of the per subscriber sub-minima that were
featured in the PR II rates.
---------------------------------------------------------------------------
\235\ The District Court of the Southern District of New York
determines performing rights royalties. Parties to those rate
proceedings refer to that court, when engaged in the rate-setting
cases, as the ``Rate Court.''
\236\ ``TCC'' refers to a streaming services' costs of content,
referring in this proceeding to the cost of sound recording
royalties the streaming services pay to record companies.
---------------------------------------------------------------------------
The Remand Majority cites with approval the remand parties'
criticism of the simplified rate structure in the Determination, viz.,
that it is ``virtually as complex as'' the PR II rate structure. See
Services' Joint Opening Brief (Apr. 1, 2021) at 39. This
characterization is
[[Page 54476]]
a bit of hyperbole. The rate structure in the Determination is an all-
in rate with ``mechanical floors'' where those are warranted. Except
for the fundamentally different configurations included in subpart B,
it does not set out separate calculations for different delivery
configurations. On remand, the Remand Majority chooses to reinstate the
PR II rate structure in its entirety, with all of its 79 permutations,
changing only the headline percent-of-revenue rate and adding a cap on
the TCC rate prong (which is an element of the structure itself). The
Benchmark Dissent does not dispute the necessity and propriety of the
increased headline percent-of-revenue rate or the cap on the TCC rate
prong. Indeed, as noted in the Remand Majority, the D.C. Circuit
endorsed the rate increase as well-reasoned and determined well within
the Judges' discretion. The D.C. Circuit also found fault with
``yoking'' the TCC rate alternative to sound recording royalty rates,
not subject to the Judges' control, without reins. The basis of this
Benchmark Dissent is simply that the regulatory scheme is not
efficient, transparent, or mandated by credible evidence; nor is the
structure necessary to achieve the purposes of reasonableness and
equity.\237\
---------------------------------------------------------------------------
\237\ As part of the Judges' discretion to promulgate
regulations to effect license rate collection, the Majority
reorganized the regulations in part 385. This reorganization was
completed to further the goal of clarity and conciseness. No party
objected to or sought to overturn that reorganization of the
regulations. Apparently, the perceived sanctity of the PR II rate
structure is not unassailable. Reorganization can perhaps be seen as
a first step to toward clarity, transparency, and simplicity for
licensors and licensees.
---------------------------------------------------------------------------
A. Acceptance of Phonorecords II Settlement as a Proper Benchmark
This is a Dissent in Part. The undersigned Judge does not disagree
with the headline rate being retained at 15.1% or with the imposition
of a TCC cap, for the reasons elucidated by the Remand Majority.
Nonetheless, the Benchmark Dissent continues to disagree with adoption
of the entirety of the rate structure adopted by Phonorecords II. As
noted above, the Judges solicited evidence to support adoption of
regulatory language to effect payment of the rates they established.
Copyright Owners, Google, and Apple submitted rate proposals that
greatly simplified the rate structure. Their rate structure regulation
proposals were crafted to support their varying approaches to rate
calculations not adopted by the Judges. Their criticisms of the PR II
rate structure are valid, nonetheless, and support the Benchmark
Dissent's analysis.
In the underlying proceeding, the Majority declined to label the
rate structure and resulting rates incorporated in the regulations
promulgated after the Phonorecords II proceeding (rates and rate
structure) as a benchmark, or starting point, for determination of new
rates and terms in that proceeding. In the Determination in the extant
proceeding, the Majority alluded to reasons they found the PR II rates
to be inadequate to serve current circumstances.\238\ The D.C. Circuit
noted that appellate counsel offered further explanation on appeal for
the rejection of the PR II rates and rate structure as a benchmark. See
Johnson v. Copyright Royalty Board, 969 F.3d 363, 387 (D.C. Cir. 2020).
Nevertheless, the D.C. Circuit faulted the Majority for not providing
adequate explanation of their rejection of a PR II-based benchmark in
the first instance. See id. Indeed, the D.C. Circuit found the
Majority's reasoning on the issue in the Determination to be
``muddled.'' Id. at 386-87.
---------------------------------------------------------------------------
\238\ The D.C. Circuit found that the Majority articulated a
reasoned and reasonable rejection of the negotiated rates applicable
to the categories of phonorecords included in ``Subpart A'' of the
regulations as a benchmark in this proceeding. The issue on remand
is articulation of a reason for not using the other subparts of 37
CFR 385 as a benchmark in this proceeding. See Johnson v. Copyright
Royalty Board, 969 F.3d 363, 386 (D.C. Cir. 2020).
---------------------------------------------------------------------------
Copyright Owners argue that the D.C. Circuit's remand for further
explanation did not equate to finding error in the Judges' rejection of
the PR II-based benchmark. See Initial Remand Submission of Copyright
Owners (Apr. 1, 2021) 1, 10 (CO Initial Submission). Notably, the
Services did not address the question of a finding of error, but
proposed on remand a rate structure substantially similar to that in PR
II and offered a benchmark analysis therefor. See Services' Joint
Opening Brief (in Services' Joint Written Direct Remand Submission at
Tab D) (Apr. 1, 2021) at 19 (Services' Initial Submission).
While the Copyright Owners' parsing of Johnson might be technically
correct, the Benchmark Dissent nonetheless accepts the wisdom of
revisiting the analysis of the PR II rates and rate structure, focusing
on the intricacies of the structure that ultimately come into play in
determining the amount of royalty payable. The Benchmark Dissent
disagrees that the record in this case demands adoption of the PR II
rate structure as a suitable benchmark. The Benchmark Dissent hereby
provides a full analysis of this issue, which includes a fuller
explanation of the conclusions in the Determination and supports and
justifies rejection of the Phonorecords II rate structure.
B. Attributes of a Useful Benchmark
As repeated by the parties in the initial proceeding and in their
remand submissions, for an exemplar to serve as a useful benchmark, it
must be compared to the target market. The hallmarks of a useful
benchmark are: (1) unity of products, (2) unity of sellers, and (3)
unity of buyers. In addition, (4) economic circumstances and market
conditions can influence the value of a benchmark. See Services'
Initial Submission at 20 (citing Determination of Royalt[ies] for
Transmission of Sound Recordings. . ., 83 FR 65210, 65214 (Dec. 19,
2018) (SDARS III).
In the Remand Majority opinion, the Judges argue that the PR II
rate structure meets ``most of the requisites for a useful benchmark.''
See Initial Ruling, section III. C. 3. Assuredly, in the real world one
is unlikely to find a perfect benchmark; consequently, the Judges in
these proceedings look to the best available benchmark(s) and make
adjustments to compensate for their shortcomings when compared to the
attributes and circumstances of the target rates. The Benchmark Dissent
is not so sanguine about one's ability to reconcile the PR II rate
structure with current market circumstances pertaining to music
streaming (including participants and volumes of sales) almost a decade
after the parties agreed to that structure. Because of the recognized
gulf in market conditions between Phonorecords II and this Phonorecords
III proceeding, the Benchmark Dissent rejects attempts to fit that
square peg into the current round hole.
1. Unity of Products--the Same Rights
The PR II rates regulated ``sales'' of the same licensing rights as
those at issue in the current underlying proceeding, viz., the
statutory license to utilize musical works embodied in the sound
recordings that are the lifeblood of the music streaming services. This
factor was not and is not in controversy. In this respect, the Judges
could look to the PR II rates as a benchmark.
2. Unity of Sellers--Rightsholders
The songwriter or songwriters own the copyright for musical works,
that is, the musical notes and lyrics. In general, songwriters sell or
license their works to publishers who fix the works to a physical
medium, for example, piano rolls or sheet music. Music publishers also
market the musical works licenses to record companies for their sound
recordings. In today's market,
[[Page 54477]]
publishers and songwriters exist in a symbiotic relationship. Without
new works, the publishers have no new product to market.\239\ To ensure
a flow of new product, publishers often subsidize songwriters by
providing working space or monetary advances on future sales of
licensed work, or publishers might purchase outright the songwriters'
copyrights. Whether the rightsholder is a writer, composer, or
publisher, the rights are the same, those derived from 17 U.S.C. 106
and limited by 17 U.S.C. 115. See 17 U.S.C. 106(1), (3) (exclusive
rights); sec. 115 (compulsory licensing). The sellers' interests are
aligned.
---------------------------------------------------------------------------
\239\ Publishers may retain rights to songs no longer considered
``new'' or ``popular'' that might nonetheless still be subject to
the section 115 license. The Services' revenue is driven, however,
by streaming new music. They understand that reselling older music,
even in new packaging (covers) would lower their desirability and
decrease the sources of revenue, their end users.
---------------------------------------------------------------------------
3. Unity of Buyers--Streaming Services
The Services argue unity of rights and sellers between the time of
the PR II rates and the current proceeding. With respect to buyers, the
Services allege that the current buyers are ``the same or similar. . .
.'' Services' Initial Submission at 20. The Services argue that the PR
II rates involved ``either the same type of buyers or the very same
buyers as this proceeding.'' Id. The license delimits the users it
binds. It is axiomatic that current licensees are ``of the same type''
as licensees in 2012. Describing participants as ``similar to those
currently in the market'' or ``of the same type'' as current
participants is sufficiently imprecise to call into question the unity
of buyers required to give great weight to a potential benchmark.
The Services allege that ``[m]ost of the participants in
Phonorecords III were either directly involved in the Phonorecords II
settlement or operated in the market at the time of the settlement.''
Id. ``Most of the participants'' does not reveal which participants
were active in Phonorecords II or the reasons for their participation.
Amazon began an MP3 digital music service in 2004; it launched steaming
in mid-2014. See Written Direct Testimony of Jeffrey Eisenach (Nov. 3,
2016) (Eisenach WDT) ] 51. Tab. 2. Apple launched its streaming service
in 2019. During the Phonorecords II negotiations, Apple's primary
interest was digital downloads from the iTunes store. According to one
of its witnesses, Google was, at the time of the Phonorecords II
negotiations, ``planning to launch a store, a locker, and a
subscription service.'' Google's participation in the Phonorecords II
negotiations was ``primarily designed to make sure that our interests
were met in--for our forthcoming music service.'' 3/8/17 Tr. 157:2-
158:2 (Zahavah Levine).
Although the Services argue that the buyers in the current market
are the same as, or similar to, buyers at the time of adoption of the
PR II rates, the Services then and now advocate differing rate
calculations for each music delivery configuration. Indeed, between
2008 and 2012, the delivery configurations multiplied and the parties
negotiated different rate structures for those multiple configurations.
Acknowledging participation by a service with one configuration--or a
plan to launch one configuration--is insufficient to establish a unity
of buyers for purposes of rate setting. Almost a decade after the
effectuation of the 2012 rates, with new businesses tacking music
streaming onto their digital ecosystems, the development of new and
different delivery configurations continues to evolve.\240\
Nonetheless, the Services would have the Judges adopt a rate structure
that specifies current delivery configurations but excludes some
current innovations and cannot encompass the next innovations, whatever
form they might take.
---------------------------------------------------------------------------
\240\ Some services offer different levels of access to
consumers using their proprietary devices, e.g., Amazon Echo. Some
(non-satellite) music streaming services are now available directly
via a button on a vehicle dashboard.
---------------------------------------------------------------------------
The Benchmark Dissent acknowledges that buyers of the musical works
for which licenses are at issue in this proceeding are of the ``same
type'' as the Phonorecords II buyers. In some instances, they are the
same participants. In the current landscape, however, the interests of
those buyers are vastly different. The extent to which Apple, Amazon,
and Google, were involved in Phonorecords II negotiations bears no
resemblance to the interests of those services and their current
service configurations. Without greater unity of buyers, the Benchmark
Dissent must discount the viability of the PR II rates or rate
structure as a useful benchmark in this proceeding.
4. Economic and Market Conditions
The Services argue that the music streaming industry in 2018 was
essentially unchanged from 2008 or 2012.\241\ See Services' Initial
Submission at 20-21. The evidence in this proceeding compels a contrary
conclusion. In 2008, musical works distribution consisted primarily of
sound recordings reproduced in physical formats (vinyl and CDs) and
digital downloads. See Eisenacht WRT ] 33 (Feb. 13, 2017). The record
reflects that in 2008, of record labels' revenues 96% were derived from
sales of physical and digitally downloaded sound recordings; 2.5% from
interactive streaming.\242\ By 2012, at the inception of the rates that
were re-adopted as the PR II rates, musical works sales were beginning
to shift from physical media to digital forms. In 2012, 8.1% of record
label revenues were attributable to interactive streaming. Id. By 2015,
evidence available in this proceeding showed that record labels'
revenues from digital downloads approximately equaled revenues from
streaming and digital sales were more than double the sales of physical
configurations, such as vinyl and CDs. Id. ]] 44-45 and accompanying
tables.
---------------------------------------------------------------------------
\241\ The PR II rates and rate structure were the product of a
negotiated settlement that began and ended with reference to the
negotiated rates adopted in 2008. Some additional categories of
service were added to the 2008 structure, e.g. locker services. Of
those categories added in 2012, few remain a significant part of the
current streaming industry.
\242\ The difference is attributable to sound recording revenues
from non-interactive streaming.
---------------------------------------------------------------------------
Spotify, the dominant pure play streaming service in the U.S., did
not enter the U.S. market until mid-2011. See CO Initial Submission at
20-21 (Apr. 1, 2021) and evidence cited therein. Spotify did not
participate in the negotiations leading up to the adoption of the 2012
musical works royalty rates. See Eisenacht WRT ] 35, n.38. In fact, the
record contains evidence that music streaming was not a major factor in
setting mechanical license rates in 2008 or 2012.\243\ See CO Initial
Submission at 19-21, and evidence cited therein. As more and larger
streaming services entered the market, music consumption changed in
character. Music consumption in the 2018 market had changed character
completely from an ownership model to an access model. See
Determination at 6.
---------------------------------------------------------------------------
\243\ The Services argue that only Mr. Israelite testified that
the 2008 and 2012 rates were ``experimental'' and that the market is
significantly changed since 2012. The Majority found, based upon the
totality of the evidence, that Mr. Israelite's testimony was
credible and accorded it due weight.
---------------------------------------------------------------------------
Further, three of the Services participating in the current
proceeding are not pure play streaming services but are
multidimensional marketing firms for whom music streaming is only one
small facet of the business. From the perspective of those current
licensees, the music streaming license is relatively insignificant to
their overall financial
[[Page 54478]]
health. The Judges must, therefore, value the license objectively to
assure the conglomerate licensees do not manipulate their revenues so
as to reduce music streaming rights below what is fair and reasonable
to the rightsholders.
The Services further advocate use of the PR II rates and rate
structure as a benchmark because they assert that the multifaceted rate
structure is reflective of the Services' own price discriminatory
services. The Majority noted the Services' price discrimination as a
way to optimally monetize segments of the market with a lesser
willingness to pay.\244\ Greater accommodation of users less willing to
pay results in more streaming and more revenue for the Services at
minimal to no marginal cost. A rate determined as a percentage of a
service's revenue allows that price discrimination to continue,
resulting in additional royalties. The Benchmark Dissent contends,
however, that the Judges need not adopt a rate structure with ten
different service categories to allow the Services to continue their
price discriminatory downstream sales. The payable royalties are a
percent of revenue. If the Services receive relatively less revenue by
marketing a family plan, for instance, that reduced revenue is the
basis for the royalty calculation. Nothing in a simplified rate
structure would inhibit price discriminatory service plans. The PR II
rates' multi-category structure might encompass the price
discrimination the Services employ, but that does not make it a
mandatory benchmark for current rates, especially if the target rate
structure permits the same flexibility.
---------------------------------------------------------------------------
\244\ The adopted Phonorecords III rate regulations acknowledged
price discrimination by, inter alia, permitting Services to account
for discounted subscriptions in different ways. See Determination at
34.
---------------------------------------------------------------------------
C. Adoption of PR II Rates and Rate Structure in Direct Licenses
The Services assert that the PR II rates and rate structure have
been adopted in negotiated direct licenses they have signed with
rightsholders rendering those rates and that rate structure a valuable
benchmark. The Services' witnesses analyzed direct licenses and
concluded that the rates closely matched the rates in the PR II
regulations. [REDACTED].\245\ Analysis of direct licenses executed
belie the Services' assertion that the PR II rates structure is
embraced by rightsholders.\246\
---------------------------------------------------------------------------
\245\ The [REDACTED] direct licenses reportedly adopt the rates
in part 385, which open-ended adoption could indicate acceptance of
both rates and rate structure.
\246\ [REDACTED] See AWDT Leonard ]] 63-64.
[REDACTED]. See Leonard AWDT ] 70-71.
[REDACTED]. See AWDT Leonard ] 54. (calculation is ``effectively
simplified'').
[REDACTED].
[REDACTED].
---------------------------------------------------------------------------
D. Additional Shortcomings of PR II Rates as a Benchmark
The D.C. Circuit dismissed the Majority's argument on appeal that
(1) the PR II rates were too low and (2) the PR II rates were outdated.
The D.C. Circuit noted that these two reasons might support the
Majority's conclusions, but they could not be asserted in the first
instance on appeal. See Johnson at 386.
1. Rates Too Low
The D.C. Circuit found that the Judges' finding that the PR II
rates were too low was not fully articulated until the matter was on
appeal. As a result, the D.C. Circuit could not evaluate that reason as
support for the final rates. Indirectly, however, the D.C. Circuit
nonetheless accepted that underlying reason for the rate changes when
it approved the higher rates themselves. See Johnson at 384-86. The
adopted rates were soundly grounded in the record evidence. See id. By
implication, acceptance of increased rates means the PR II rates were
too low to be continued. With or without the ``too low'' rationale, the
final adopted rates prove the point.\247\
---------------------------------------------------------------------------
\247\ The Services argue that an agreed continuation of the
Subpart A (now Subpart B) rates for, inter alia, physical
phonorecords and permanent downloads, proves that the Phonorecords
II rates are appropriate. See Services' Initial Submission at 30.
This argument asserts a false equivalency. Physical Phonorecords and
permanent downloads are fundamentally different in character from
streamed music. Further, the evidence indicates that the prominence
of streaming access over ownership of recordings is waning. The
parties' agreement to maintain the Phonorecords II rates for this
declining segment of the market does not equate to a mandate to
adopt the entirety of the PR II rate structure.
---------------------------------------------------------------------------
2. Rate Structure Outdated
In the Determination, the Majority cited several factors that
implied the inadequacy of the PR II rates and rate structure as a
compelling benchmark for Phonorecords III. As discussed above, the
music streaming industry in 2018 was completely transformed from 2008
or 2012. Both the buyers and the economic market conditions were
markedly changed. Referring to the PR II rates as ``outdated''
encompasses both a temporal element and a structural component.
a. Significance of the Passage of Time
Music streaming in the earlier rate setting periods was in its
infancy. Listeners had not yet fully embraced the subscribed access
model for music consumption. By 2018, listeners could choose from ``a
diverse array of streaming offerings.'' See WDT of Rishi Mirchandani ]
63. Such industry shifts alone could render the PR II rates
``outdated.''
b. Clarity and Simplicity
Another salient factor the Majority addressed is the rate structure
itself. To understand the PR II rate structure, one needed ten separate
full-page flow chart diagrams, each featuring three formulae for
calculating greater-of and lesser-of rate components. See Trial Ex.
846. The rates for some consumption configurations included a per-
subscriber ``mechanical floor'' as a failsafe against overreaching by
PROs, should the Rate Court increase their rates to an extent that all
of the section 115 all-in percent of revenue royalty be consumed by the
PROs. See, e.g., [FORMER] 37 CFR 385.13(a)(1) (Standalone non-portable
subscription--streaming only [$.15 per subscriber]); [FORMER]
385.13(a)(2) (Standalone portable subscription--mixed use [$.50 per
subscriber]) (2018).\248\ Other consumption configurations included
``minima;'' that is a lesser-of calculation comparing a percent of
sound recording license costs (TCC) and a per subscriber amount. See,
e.g., [FORMER] 37 CFR 385.13(b) (2018). Further, rate calculations
differed depending upon, for example, whether the listener streamed on
a portable device or a non-portable device; or whether the listener
purchased access to the music alone from a pure-play streaming service
or as part of a bundled offering, such as ``free'' streaming for a
limited period included in the purchase price of the streaming
device.\249\
---------------------------------------------------------------------------
\248\ The Majority reintroduced these ``mechanical floor''
safeguards, notwithstanding a lack of evidence to explain, let alone
justify, the difference between $0.15 and $0.50 per subscriber (the
latter being 300% greater than the former) simply because one
consumer listened to a song on a standalone non-portable device and
another consumer listened to a song on a standalone portable device.
\249\ The Services have not offered convincing, substantive
evidence or argument to support the fractured structure of the PR II
rates. Tellingly, the user's choice of consumption device is not a
factor in license rates for other services. See, e.g., 17 CFR 380.10
(Webcasters rates differentiate between commercial and non-
commercial licensees, not based on users' reception devices);
Sec. Sec. 382.3, 382.12 (rates for satellite radio and pre-existing
subscription services do not differentiate based on users' reception
devices).
---------------------------------------------------------------------------
The rationale for these convoluted rate calculation differences is
[[Page 54479]]
unknown.\250\ They were the product of confidential negotiations among
the parties involved in the music streaming business in the first
decade of the 21st century. One side of the negotiating table sought
reconsideration of those rates. The current licensees are not the same
as those who negotiated the 2012 rollover of the 2008 rate scheme.
Music streaming business models have witnessed significant growth and
change. Meanwhile, the business models employed by songwriters and
publishers remain largely unchanged--and not realizing a proportionate
capture of the stream of dollars realized by the Services' monetization
of ever-more consumption configurations. The marginal cost to the
Services of additional streams, regardless of the business
configuration or the user's reception device, is zero. The Services,
therefore, are in a position to capture increased revenue without an
increase in cost of goods sold.
---------------------------------------------------------------------------
\250\ Prof. Katz asserted that ``economic analysis'' indicates
that varying rates based on the characteristics of the service
``facilitates continuing innovation, experimentation, and
differentiation in means of making music accessible to consumers.''
Katz WDT ] 85. Prof. Katz did not identify that economic analysis.
He asserted that the fractured rates allow services to benefit
despite different consumers' willingness to pay. Nothing in the PR
III rate structure at issue in any way inhibited services adapting
to meet consumers' willingness to pay. The rates are, in the main,
revenue based--even if the services choose to market the service at
a lower rate to a particular segment of the market.
---------------------------------------------------------------------------
In the end, a sound recording embodying a licensed musical work is
being delivered to an end ``user'': one song; one listener. The
calculation of what royalty the songwriter is entitled to should not
rest on the medium of transmission or the location of the listening.
See WDS Steve Bogard ] 34 (``Streaming music anytime, any place, on any
device is the way today's music fans want to enjoy their music.
Notwithstanding that the inherent value of a song is the same whether
the consumer chooses to buy an album, permanently download an album or
a single, or stream music on demand. . . .''). The incremental
difference in value to the listener of hearing a song in the car as
opposed to through earbuds during a workout is not likely measurable.
Certainly, no participant in this proceeding presented any evidence of
the relative value of a song to a listener depending on the delivery
configuration.\251\
---------------------------------------------------------------------------
\251\ The Remand Majority dubs analysis of value based on the
cost of production rather than willingness to purchase as old-
fashioned economic analysis. So it may be. In the modern economist's
widget market, if buyers are unwilling to pay enough to cover the
cost of widget components, then widget production ceases. But in the
old-fashioned creativity market, the goods are not fungible. The
inputs to a hit song are ephemeral; sometimes plentiful, sometimes
elusive; they either coalesce or they do not. Songwriters will
persevere because they cannot do otherwise. The demand for music
continues to grow with each new innovation in delivery methods. The
United States Constitution provides for protection of art and the
creators of art. U.S. Const. art. I, sec. 8. Congress has specified
how to protect, inter alia, the copyrights of songwriters. The
Judges' small part in that effort is to continue to assure that
royalty rates are reasonable--for both creators and exploiters. In
the music streaming industry, the evidence supports devoting a
greater share of licensees' increased wealth to the ``widget
makers.'' The Dissent contends that the increase in the percent-of-
revenue headline rate is a good step forward, but only the first
step to assuring equity in the market. Streamlining, simplifying,
and generally ``cleaning up'' payment calculations would go a long
way in the right direction by removing twists and turns and
confusing signals along the path of the royalty dollar from end user
to creator.
---------------------------------------------------------------------------
In the interest of making government more transparent and
accessible to interested citizens, less is more. Opaque systems and
formulae are or should be, in a word, outdated. The fact of settlement
does not cure or even address the unnecessary complication of paying a
royalty for the use of a statutory license under the PR II rates
structure. More importantly, owners of the copyrights being licensed
should be able to comprehend, calculate, and verify the sources and
amounts of their royalty payments.
3. Not Business Model Neutral
The Services contend that the PR II rate structure is preferable as
it is business model neutral. Nothing in the record supports that
assertion. In fact, Apple argued that the PR II rate structure stifled
innovation as streaming services sought to fit any new business into a
business model already defined as one of the ten identified models in
the Phonorecords II regulations. The statute does not require that rate
structures be business model neutral. The reasonableness requirement
demands, however, that the Judges find and adopt reasons for
differentiation in rates based on business models.
4. No Evidence of Settling Parties' Subjective Intent
Copyright Owners participating in the current proceeding argued
that the Judges should consider the subjective intent of the parties in
agreeing to ``roll over'' the 2008 rates and rate structure into the PR
II regulations. The Services countered that subjective intent is
irrelevant, as the product of those negotiations serves as objective
evidence of the parties' intents. On this question, the Services are
correct. The negotiated rates show, objectively, that the negotiating
parties agreed to a certain rate structure. The D.C. Circuit criticized
the Majority for not including in the Final Determination an
explanation of why the subjective intent of the parties to the
settlement was a ``prerequisite'' to adoption of that settlement as a
benchmark. See Johnson at 387. The Judges need not, however, accept
that objective evidence uncritically.
Negotiating parties' subjective state of mind can serve as
convincing evidence of the economic circumstances and the state of the
market at the time of the negotiations. While ascertaining the parties'
subjective intent in reaching the settlement is not a ``prerequisite''
to examination of the terms as a benchmark, the Benchmark Dissent finds
subjective intent informative and useful as one factor in weighing the
value of the settlement as a benchmark.
E. Statutory Factors
The Services argued to the D.C. Circuit that the Majority's
rejection of the PR II rates and rate structure was erroneous because
the Majority failed to evaluate that structure and those rates under
the statutory factors delineated in 17 U.S.C. 801(b)(1). Evaluation
under section 801(b)(1) is required by the statute applicable to this
proceeding.\252\ Nothing in section 801(b)(1) compels the Judges to
evaluate compliance with the statutory factors of every proposed
potential rate or rate structure. Neither are the Judges required to
evaluate every potential benchmark or past rate structure under section
801(b)(1). The Judges are obliged to evaluate any rate structure they
intend to adopt against the requirements of section 801(b)(1). If the
Judges' promulgated rate structure meets the section 801(b)(1)
standard, then the promulgated rate structure can be adopted. Whether
other possible proposals might also meet the section 801(b)(1) standard
is not at issue in a proceeding.
---------------------------------------------------------------------------
\252\ With the passage of the Orrin G. Hatch--Bob Goodlatte
Music Modernization Act, Congress eliminated the four statutory
factors for evaluating license royalty rates. See Public Law 115-
264, 132 Stat. 3676 (2018) (codified in scattered sections of title
17, U.S.C.
---------------------------------------------------------------------------
1. Maximize the Availability of Creative Works to the Public
The Services argue that the PR II rates and rate structure support
and contribute to the maximization of musical works. As evidence, they
cite the growth of music streaming overall, the profitability of all
segments of the music industry.\253\ It is beyond question that music
consumption has grown exponentially since the co-incident
[[Page 54480]]
introduction of portable devices and streaming services. Growth
continues as those devices and services become increasingly easy to
actuate in vehicles.
---------------------------------------------------------------------------
\253\ According to the Services, all segments of the music
industry are thriving [REDACTED].
---------------------------------------------------------------------------
No participant alleged, however, that music industry success is
caused by or even correlated to the PR II rate structure. Coincidence
is not probative evidence.
2. Assure Fair Return to Copyright Owner and Fair Income to the
Licensee
The Services argued they were receiving a fair income and copyright
owners were receiving a fair return under the PR II regulations.
Although the Services argued that overall music royalties absorbed an
inordinate portion of their revenues, none expressly laid that lack of
available revenue at the door of mechanical royalties. Amazon's
witness, Dr. Glenn Hubbard described a growing increase in streaming
industry revenues and forecasts of continuing growth. See WRT of Glenn
Hubbard (Feb. 15, 2017) ] 2.23-24 (Hubbard WRT). Dr. Hubbard
deconstructed Amazon's increased revenues and concluded that the growth
in streaming services' revenue resulted in increased royalty payments
to music publishers and other rights holders. Id. ] 3.10. When royalty
rates are calculated on a percent-of-revenue, the royalty payments
increase when revenues increase.
The difficulty with this tautological argument is that revenue
growth as between services and rightsholders has not been proportional.
And, as Copyright Owners have argued, the rate at which the services
share with mechanical rightsholders is the issue in this proceeding.
The Judges are not called upon to set annual royalty payment dollar
amounts; rather they are mandated to set the rates that drive those
dollar amounts. And to adopt regulations that most closely effectuate
actual payment to rightsholders, minimizing revenue deferral and other
such loopholes. For all of the reasons provided in the Determination
and in this Benchmark Dissent, the PR II-based rates and the
controlling rate structure do not balance the section 115 fair income-
fair return scale appropriately and reasonably.
3. Weigh Relative Roles of Licensors and Licensees in Making the Works
Available to the Public
No participant presented evidence to elucidate specifically the
relative roles of the parties relating to musical works. Economic
evidence assumed that the marginal cost of streaming more music is
minimal. This does not discount the services' sunk costs, such as the
original technological or capital investments. With respect to the
contributions of the copyright owners, the contribution is clear. It
all begins with a song. Without new music, the Services could continue
by streaming unregulated works, new arrangements or covers of existing
works, and non-music content. Whether they would continue to enjoy the
growth they have enjoyed over the last decade is unknown. The PR II
rates might be a contributing factor to both stability and growth of
the industry, but based on the totality of the evidence, the Dissent
concludes that with regard to musical works, the relative role of the
creator of the musical works, and to a lesser extent, the music
publisher, is undervalued.
4. Minimize Disruption
The language for the fourth statutory factor requires the Judges to
establish a rate structure in such a way as ``[t]o minimize any
disruptive impact on the structure of the industries involved and on
generally prevailing industry practices.'' [FORMER] 17 U.S.C.
801(b)(1)(D). The Services argue that the change in rate structure
determined by the Majority in this proceeding is massively, and
potentially fatally, disruptive to music streaming services.
Ironically, the music industry has been in a constant state of
disruption since the introduction of digital music. From peer-to-peer
sharing, to purchased permanent downloads, to interactive and non-
interactive streaming, the history of modern music consumption has been
a model of disruption. Entry into the streaming market by multifaceted
digital ecosystem providers is just the latest significant change in
music delivery to consumers. Innovation in music delivery is constant.
Allegedly to minimize disruption, the Services advocated retention
of the PR II rates and rate structure.\254\ While every aspect of the
music industry is experiencing explosive growth, maintenance of the
inadequate rates for mechanical licenses is unfathomable. Some change,
phased in over time, might be uncomfortable for the licensees, but
failure to change rates to acknowledge the music delivery revolution is
not an option. With such a dynamic history and uncertain future, a
change in mechanical license rates is not just inevitable, but
mandatory.
---------------------------------------------------------------------------
\254\ Tellingly, on remand, the Services did not pursue any
argument that the changes in the rates or rate structure in the
Determination were disruptive.
---------------------------------------------------------------------------
Indeed, the Benchmark Dissent's approach in this proceeding
advances the notion that streamed music is streamed music. This is
certainly true from the viewpoint of the songwriters and publishers,
and of music consumers. Rather than introduce separate rate structures
for each new delivery technology or streaming business model, the
Judges need to establish a rate that will fairly compensate Copyright
Owners for the use of their works and permit a fair return to
licensees, regardless of what next technological disruption they might
choose to introduce to the industry. In the captioned proceeding, the
Majority declined to label the rate structure and resulting rates
incorporated in the regulations promulgated after the Phonorecords II
proceeding as a benchmark, or starting point, for determination of new
rates and terms in this proceeding.
In the Determination, the Majority alluded to reasons they found
the PR II rates to be inadequate to serve current circumstances.\255\
Nevertheless, the D.C. Circuit faulted the Majority for not providing
adequate explanation of their rejection of the PR II benchmark in the
first instance. See Johnson at 386-87. Indeed, the D.C. Circuit found
the Majority's reasoning on the issue in the Determination to be
``muddled.'' Id.
---------------------------------------------------------------------------
\255\ The D.C. Circuit found that the Majority articulated a
reasoned and reasonable rejection of the negotiated rates applicable
to the categories of phonorecords included in [FORMER] subpart A of
the regulations as a benchmark in this proceeding. The issue on
remand is articulation of a reason for not using the other subparts
of 37 CFR part 385 as a benchmark in this proceeding. See Johnson at
386.
---------------------------------------------------------------------------
F. Rate Structure
For all of the reasons outlined above, the Remand Majority's
acceptance and adoption of the Phonorecords II rate structure results
in a rate structure in this proceeding that suffers from the same
deficits the Benchmark Dissent believes to be inherent in that rate
structure. Changing the headline rate and capping the TCC rate prong do
not cure the ills of the rate structure itself. True, the PR II-based
rates permit price discrimination, which increases revenue, and
therefore royalties, in absolute terms. Reinstatement of minima in the
TCC prong introduces a failsafe to runaway TCC-based rates. The
mechanical floors adopted in the Determination continue, protecting
mechanical license rightsholders from runaway performance royalties.
The Benchmark Dissent maintains that all these goals could be met
equally well with a streamlined, transparent, fair, and reasonable rate
structure, as several of the participants in this proceeding advocated.
[[Page 54481]]
III. Conclusion
This Dissent in part is issued as a RESTRICTED document. Within 30
days of the date of issuance, the participants shall file a version of
this Dissent with agreed redactions to permit viewing by the public.
Issue Date: July 1, 2022.
Suzanne M. Barnett
Chief Copyright Royalty Judge
List of Subjects in 37 CFR Part 385
Copyright, Phonorecords, Recordings.
For the reasons set forth in the preamble, the Copyright Royalty
Judges amend 37 CFR part 385 as follows.
PART 385--RATES AND TERMS FOR USE OF NONDRAMATIC MUSICAL WORKS IN
THE MAKING AND DISTRIBUTING OF PHYSICAL AND DIGITAL PHONORECORDS
0
1. The authority citation for part 385 continues to read as follows:
Authority: 17 U.S.C. 115, 801(b)(1), 804(b)(4).
0
2. Add appendix A to read as follows:
Appendix A to Part 385--Part 385 Applicable to the Period January 1,
2018, through December 31, 2022, as clarified on August 10, 2023
Note: Cross-references to part 385 in this appendix are to those
provisions as contained within this appendix.
PART 385--RATES AND TERMS FOR USE OF MUSICAL WORKS UNDER COMPULSORY
LICENSE FOR MAKING AND DISTRIBUTING PHYSICAL AND DIGITAL
PHONORECORDS
Subpart A--Regulations of General Application
385.1 General.
385.2 Definitions.
385.3 Late payments.
385.4 Recordkeeping for promotional or free trial non-royalty-
bearing uses.
Subpart B--Physical Phonorecord Deliveries, Permanent Downloads,
Ringtones, and Music Bundles
385.10 Scope.
385.11 Royalty rates.
Subpart C--Eligible Interactive Streaming, Eligible Limited Downloads,
Limited Offerings, Mixed Service Bundles, Bundled Subscription
Offerings, Locker Services, and Other Delivery Configurations
385.20 Scope.
385.21 Royalty rates and calculations.
385.22 Royalty floors for specific types of Offerings.
Subpart D--Promotional Offerings, Free Trial Offerings and Certain
Purchased Content Locker Services
385.30 Scope.
385.31 Royalty rates.
Subpart A--Regulations of General Application
Sec. 385.1 General.
(a) Scope. This part establishes rates and terms of royalty
payments for the use of nondramatic musical works in making and
distributing of physical and digital phonorecords in accordance with
the provisions of 17 U.S.C. 115. This subpart contains regulations
of general application to the making and distributing of
phonorecords subject to the license under 17 U.S.C. 115 (section 115
license).
(b) Legal compliance. Licensees relying on the compulsory
license detailed in 17 U.S.C. 115 shall comply with the requirements
of that section, the rates and terms of this part, and any other
applicable regulations. This part describes rates and terms for the
compulsory license only.
(c) Interpretation. This part is intended only to set rates and
terms for situations in which the exclusive rights of a Copyright
Owner are implicated and a compulsory license pursuant to 17 U.S.C.
115 is obtained. Neither this part nor the act of obtaining a
license under 17 U.S.C. 115 is intended to express or imply any
conclusion as to the circumstances in which a user must obtain a
compulsory license pursuant to 17 U.S.C. 115.
(d) Relationship to voluntary agreements. The rates and terms of
any license agreements entered into by Copyright Owners and
Licensees relating to use of musical works within the scope of those
license agreements shall apply in lieu of the rates and terms of
this part.
Sec. 385.2 Definitions.
For the purposes of this part, the following definitions apply:
Accounting Period means the monthly period specified in 17
U.S.C. 115(c)(2)(I) and (d)(4)(A)(i), and any related regulations in
this chapter, as applicable.
Active Subscriber means an End User of a Bundled Subscription
Offering who has made at least one Play during the Accounting
Period.
Affiliate means an entity controlling, controlled by, or under
common control with another entity, except that an affiliate of a
Sound Recording Company shall not include a Copyright Owner to the
extent it is engaging in business as to musical works.
Bundled Subscription Offering means a Subscription Offering
providing Licensed Activity consisting of Eligible Interactive
Streams or Eligible Limited Downloads that is made available to End
Users with one or more other products or services (including
products or services subject to other subparts) as part of a single
transaction without pricing for the subscription service providing
Licensed Activity separate from the product(s) or service(s) with
which it is made available (e.g., a case in which a user can buy a
portable device and one-year access to a subscription service
providing Licensed Activity for a single price).
Copyright Owner(s) are nondramatic musical works copyright
owners who are entitled to royalty payments made under this part
pursuant to the compulsory license under 17 U.S.C. 115.
Digital Phonorecord Delivery has the same meaning as in 17
U.S.C. 115(e)(10).
Eligible Interactive Stream means a Stream in which the
performance of the sound recording is not exempt from the sound
recording performance royalty under 17 U.S.C. 114(d)(1) and does not
in itself, or as a result of a program in which it is included,
qualify for statutory licensing under 17 U.S.C. 114(d)(2).
Eligible Limited Download means a Limited Download as defined in
17 U.S.C. 115(e)(16) that is only accessible for listening for--
(1) An amount of time not to exceed one month from the time of
the transmission (unless the Licensee, in lieu of retransmitting the
same sound recording as another Eligible Limited Download,
separately, and upon specific request of the End User made through a
live network connection, reauthorizes use for another time period
not to exceed one month), or in the case of a subscription plan, a
period of time following the end of the applicable subscription no
longer than a subscription renewal period or three months, whichever
is shorter; or
(2) A number of times not to exceed 12 (unless the Licensee, in
lieu of retransmitting the same sound recording as another Eligible
Limited Download, separately, and upon specific request of the End
User made through a live network connection, reauthorizes use of
another series of 12 or fewer plays), or in the case of a
subscription transmission, 12 times after the end of the applicable
subscription.
End User means each unique person that:
(1) Pays a subscription fee for an Offering during the relevant
Accounting Period; or
(2) Makes at least one Play during the relevant Accounting
Period.
Family Plan means a discounted Subscription Offering to be
shared by two or more family members for a single subscription
price.
Free Trial Offering means a subscription to a Service Provider's
transmissions of sound recordings embodying musical works when:
(1) Neither the Service Provider, the Sound Recording Company,
the Copyright Owner, nor any person or entity acting on behalf of or
in lieu of any of them receives any monetary consideration for the
Offering;
(2) The free usage does not exceed 30 consecutive days per
subscriber per two-year period;
(3) In connection with the Offering, the Service Provider is
operating with appropriate musical license authority and complies
with the recordkeeping requirements in Sec. 385.4;
(4) Upon receipt by the Service Provider of written notice from
the Copyright Owner or its agent stating in good faith that the
Service Provider is in a material manner operating without
appropriate license authority from the Copyright Owner under 17
U.S.C. 115, the Service Provider shall within 5 business days cease
transmission of the sound recording embodying that musical work and
withdraw it from the repertoire available as part of a Free Trial
Offering;
[[Page 54482]]
(5) The Free Trial Offering is made available to the End User
free of any charge; and
(6) The Service Provider offers the End User periodically during
the free usage an opportunity to subscribe to a non-Free Trial
Offering of the Service Provider.
GAAP means U.S. Generally Accepted Accounting Principles in
effect at the relevant time, except that if the U.S. Securities and
Exchange Commission permits or requires entities with securities
that are publicly traded in the U.S. to employ International
Financial Reporting Standards in lieu of Generally Accepted
Accounting Principles, then that entity may employ International
Financial Reporting Standards as ``GAAP'' for purposes of this
subpart.
Licensee means any entity availing itself of the compulsory
license under 17 U.S.C. 115 to use copyrighted musical works in the
making or distributing of physical or digital phonorecords.
Licensed Activity, as the term is used in subpart B of this
part, means delivery of musical works, under voluntary or statutory
license, via physical phonorecords and Digital Phonorecord
Deliveries in connection with Permanent Downloads, Ringtones, and
Music Bundles; and, as the term is used in subparts C and D of this
part, means delivery of musical works, under voluntary or statutory
license, via Digital Phonorecord Deliveries in connection with
Eligible Interactive Streams, Eligible Limited Downloads, Limited
Offerings, mixed Bundles, and Locker Services.
Limited Offering means a Subscription Offering providing
Eligible Interactive Streams or Eligible Limited Downloads for
which--
(1) An End User cannot choose to listen to a particular sound
recording (i.e., the Service Provider does not provide Eligible
Interactive Streams of individual recordings that are on-demand, and
Eligible Limited Downloads are rendered only as part of programs
rather than as individual recordings that are on-demand); or
(2) The particular sound recordings available to the End User
over a period of time are substantially limited relative to Service
Providers in the marketplace providing access to a comprehensive
catalog of recordings (e.g., a product limited to a particular genre
or permitting Eligible Interactive Streams only from a monthly
playlist consisting of a limited set of recordings).
Locker Service means an Offering providing digital access to
sound recordings of musical works in the form of Eligible
Interactive Streams, Permanent Downloads, Restricted Downloads or
Ringtones where the Service Provider has reasonably determined that
the End User has purchased or is otherwise in possession of the
subject phonorecords of the applicable sound recording prior to the
End User's first request to use the sound recording via the Locker
Service. The term Locker Service does not mean any part of a Service
Provider's products otherwise meeting this definition, but as to
which the Service Provider has not obtained a section 115 license.
Mixed Service Bundle means one or more of Permanent Downloads,
Ringtones, Locker Services, or Limited Offerings a Service Provider
delivers to End Users together with one or more non-music services
(e.g., internet access service, mobile phone service) or non-music
products (e.g., a telephone device) of more than token value and
provided to users as part of one transaction without pricing for the
music services or music products separate from the whole Offering.
Music Bundle means two or more of physical phonorecords,
Permanent Downloads, or Ringtones delivered as part of one
transaction (e.g., download plus ringtone, CD plus downloads). In
the case of Music Bundles containing one or more physical
phonorecords, the Service Provider must sell the physical
phonorecord component of the Music Bundle under a single catalog
number, and the musical works embodied in the Digital Phonorecord
Delivery configurations in the Music Bundle must be the same as, or
a subset of, the musical works embodied in the physical
phonorecords; provided that when the Music Bundle contains a set of
Digital Phonorecord Deliveries sold by the same Sound Recording
Company under substantially the same title as the physical
phonorecord (e.g., a corresponding digital album), the Service
Provider may include in the same bundle up to 5 sound recordings of
musical works that are included in the stand-alone version of the
set of digital phonorecord deliveries but not included on the
physical phonorecord. In addition, the Service Provider must
permanently part with possession of the physical phonorecord or
phonorecords it sells as part of the Music Bundle. In the case of
Music Bundles composed solely of digital phonorecord deliveries, the
number of digital phonorecord deliveries in either configuration
cannot exceed 20, and the musical works embodied in each
configuration in the Music Bundle must be the same as, or a subset
of, the musical works embodied in the configuration containing the
most musical works.
Offering means a Service Provider's engagement in Licensed
Activity covered by subparts C and D of this part.
Paid Locker Service means a Locker Service for which the End
User pays a fee to the Service Provider.
Performance Royalty means the license fee payable for the right
to perform publicly musical works in any of the forms covered by
subparts C and D this part.
Permanent Download has the same meaning as in 17 U.S.C.
115(e)(24).
Play means an Eligible Interactive Stream, or a play of an
Eligible Limited Download, lasting 30 seconds or more and, if a
track lasts in its entirety under 30 seconds, an Eligible
Interactive Stream or a play of an Eligible Limited Download of the
entire duration of the track. A Play excludes an Eligible
Interactive Stream or a play of an Eligible Limited Download that
has not been initiated or requested by a human user. If a single End
User plays the same track more than 50 straight times, all plays
after play 50 shall be deemed not to have been initiated or
requested by a human user.
Promotional Offering means a digital transmission of a sound
recording, in the form of an Eligible Interactive Stream or an
Eligible Limited Download, embodying a musical work, the primary
purpose of which is to promote the sale or other paid use of that
sound recording or to promote the artist performing on that sound
recording and not to promote or suggest promotion or endorsement of
any other good or service and:
(1) A Sound Recording Company is lawfully distributing the sound
recording through established retail channels or, if the sound
recording is not yet released, the Sound Recording Company has a
good faith intention to lawfully distribute the sound recording or a
different version of the sound recording embodying the same musical
work;
(2) For Eligible Interactive Streams or Eligible Limited
Downloads, the Sound Recording Company requires a writing signed by
an authorized representative of the Service Provider representing
that the Service Provider is operating with appropriate musical
works license authority and that the Service Provider is in
compliance with the recordkeeping requirements of Sec. 385.4;
(3) For Eligible Interactive Streams of segments of sound
recordings not exceeding 90 seconds, the Sound Recording Company
delivers or authorizes delivery of the segments for promotional
purposes and neither the Service Provider nor the Sound Recording
Company creates or uses a segment of a sound recording in violation
of 17 U.S.C. 106(2) or 115(a)(2);
(4) The Promotional Offering is made available to an End User
free of any charge; and
(5) The Service Provider provides to the End User at the same
time as the Promotional Offering Stream an opportunity to purchase
the sound recording or the Service Provider periodically offers End
Users the opportunity to subscribe to a paid Offering of the Service
Provider.
Purchased Content Locker Service means a Locker Service made
available to End User purchasers of Permanent Downloads, Ringtones,
or physical phonorecords at no incremental charge above the
otherwise applicable purchase price of the Permanent Downloads,
Ringtones, or physical phonorecords acquired from a qualifying
seller. With a Purchased Content Locker Service, an End User may
receive one or more additional phonorecords of the purchased sound
recordings of musical works in the form of Permanent Downloads or
Ringtones at the time of purchase, or subsequently have digital
access to the purchased sound recordings of musical works in the
form of Eligible Interactive Streams, additional Permanent
Downloads, Restricted Downloads, or Ringtones.
(1) A qualifying seller for purposes of this definition is the
entity operating the Service Provider, including Affiliates,
predecessors, or successors in interest, or--
(i) In the case of Permanent Downloads or Ringtones, a seller
having a legitimate connection to the locker service provider
pursuant to one or more written agreements (including that the
Purchased Content Locker Service and Permanent Downloads or
Ringtones are offered through the same third party); or
[[Page 54483]]
(ii) In the case of physical phonorecords:
(A) The seller of the physical phonorecord has an agreement with
the Purchased Content Locker Service provider establishing an
integrated offer that creates a consumer experience commensurate
with having the same Service Provider both sell the physical
phonorecord and offer the integrated locker service; or
(B) The Service Provider has an agreement with the entity
offering the Purchased Content Locker Service establishing an
integrated offer that creates a consumer experience commensurate
with having the same Service Provider both sell the physical
phonorecord and offer the integrated locker service.
(2) [Reserved]
Relevant Page means an electronic display (for example, a web
page or screen) from which a Service Provider's Offering consisting
of Eligible Interactive Streams or Eligible Limited Downloads is
directly available to End Users, but only when the Offering and
content directly relating to the Offering (e.g., an image of the
artist, information about the artist or album, reviews, credits, and
music player controls) comprises 75% or more of the space on that
display, excluding any space occupied by advertising. An Offering is
directly available to End Users from a page if End Users can receive
sound recordings of musical works (in most cases this will be the
page on which the Eligible Limited Download or Eligible Interactive
Stream takes place).
Restricted Download means a Digital Phonorecord Delivery in a
form that cannot be retained and replayed on a permanent basis. The
term Restricted Download includes an Eligible Limited Download.
Ringtone means a phonorecord of a part of a musical work
distributed as a Digital Phonorecord Delivery in a format to be made
resident on a telecommunications device for use to announce the
reception of an incoming telephone call or other communication or
message or to alert the receiver to the fact that there is a
communication or message.
Service Provider means that entity governed by subparts C and D
of this part, which might or might not be the Licensee, that with
respect to the section 115 license:
(1) Contracts with or has a direct relationship with End Users
or otherwise controls the content made available to End Users;
(2) Is able to report fully on Service Provider Revenue from the
provision of musical works embodied in phonorecords to the public,
and to the extent applicable, verify Service Provider Revenue
through an audit; and
(3) Is able to report fully on its usage of musical works, or
procure such reporting and, to the extent applicable, verify usage
through an audit.
Service Provider Revenue, as used in this part:
(1) Subject to paragraphs (2) through (5) of this definition and
subject to GAAP, Service Provider Revenue shall mean:
(i) All revenue from End Users recognized by a Service Provider
for the provision of any Offering;
(ii) All revenue recognized by a Service Provider by way of
sponsorship and commissions as a result of the inclusion of third-
party ``in-stream'' or ``in-download'' advertising as part of any
Offering, i.e., advertising placed immediately at the start or end
of, or during the actual delivery of, a musical work, by way of
Eligible Interactive Streaming or Eligible Limited Downloads; and
(iii) All revenue recognized by the Service Provider, including
by way of sponsorship and commissions, as a result of the placement
of third-party advertising on a Relevant Page of the Service
Provider or on any page that directly follows a Relevant Page
leading up to and including the Eligible Limited Download or
Eligible Interactive Stream of a musical work; provided that, in
case more than one Offering is available to End Users from a
Relevant Page, any advertising revenue shall be allocated between or
among the Service Providers on the basis of the relative amounts of
the page they occupy.
(2) Service Provider Revenue shall:
(i) Include revenue recognized by the Service Provider, or by
any associate, Affiliate, agent, or representative of the Service
Provider in lieu of its being recognized by the Service Provider;
and
(ii) Include the value of any barter or other nonmonetary
consideration; and
(iii) Except as expressly detailed in this part, not be subject
to any other deduction or set-off other than refunds to End Users
for Offerings that the End Users were unable to use because of
technical faults in the Offering or other bona fide refunds or
credits issued to End Users in the ordinary course of business.
(3) Service Provider Revenue shall exclude revenue derived by
the Service Provider solely in connection with activities other than
Offering(s), whereas advertising or sponsorship revenue derived in
connection with any Offering(s) shall be treated as provided in
paragraphs (2) and (4) of this definition.
(4) For purposes of paragraph (1) of this definition,
advertising or sponsorship revenue shall be reduced by the actual
cost of obtaining that revenue, not to exceed 15%.
(5) In instances in which a Service Provider provides an
Offering to End Users as part of the same transaction with one or
more other products or services that are not Licensed Activities,
then the revenue from End Users deemed to be recognized by the
Service Provider for the Offering for the purpose of paragraph (1)
of this definition shall be the revenue recognized from End Users
for the bundle less the standalone published price for End Users for
each of the other component(s) of the bundle; provided that, if
there is no standalone published price for a component of the
bundle, then the Service Provider shall use the average standalone
published price for End Users for the most closely comparable
product or service in the U.S. or, if more than one comparable
exists, the average of standalone prices for comparables.
(6) In the case of a Mixed Service Bundle, the revenue deemed to
be recognized from End Users for the Offering for the purpose of
paragraph (1) of this definition shall be the greater of--
(i) The revenue deemed to be recognized pursuant to paragraph
(5) of this definition; and
(ii) Either--
(A) In the case of a Mixed Service Bundle that either has
750,000 subscribers or other registered users, or is reasonably
expected to have 750,000 subscribers or other registered users
within 1 year after commencement of the Mixed Service Bundle, 40% of
the standalone published price of the licensed music component of
the bundle (i.e., the Permanent Downloads, Ringtones, Locker
Service, or Limited Offering); provided that, if there is no such
standalone published price for the licensed music component of the
bundle, then the average standalone published price for End Users
for the most closely comparable licensed music component in the U.S.
shall be used or, if more than one such comparable exists, the
average of such standalone prices for such comparables shall be
used; and further provided that in any case in which royalties were
paid based on this paragraph (6)(ii)(A) due to a reasonable
expectation of reaching 750,000 subscribers or other registered
users within 1 year after commencement of the Mixed Service Bundle
and that does not actually happen, applicable payments shall, in the
accounting period next following the end of such 1-year period,
retroactively be adjusted as if paragraph (6)(ii)(B) of this
definition applied; or
(B) Otherwise, 50% of the standalone published price of the
licensed music component of the bundle (i.e., the Permanent
Downloads, Ringtones, Locker Service, or Limited Offering); provided
that, if there is no such standalone published price for the
licensed music component of the bundle, then the average standalone
published price for End Users for the most closely comparable
licensed music component in the U.S. shall be used or, if more than
one such comparable exists, the average of such standalone prices
for such comparables shall be used.
Sound Recording Company means a person or entity that:
(1) Is a copyright owner of a sound recording embodying a
musical work;
(2) In the case of a sound recording of a musical work fixed
before February 15, 1972, has rights to the sound recording, under
17 U.S.C. chapter 14, that are equivalent to the rights of a
copyright owner of a sound recording of a musical work under title
17, United States Code;
(3) Is an exclusive Licensee of the rights to reproduce and
distribute a sound recording of a musical work; or
(4) Performs the functions of marketing and authorizing the
distribution of a sound recording of a musical work under its own
label, under the authority of the Copyright Owner of the sound
recording.
Standalone Non-Portable Subscription Offering--Mixed means a
Subscription Offering through which an End User can listen to sound
recordings either in the form of Eligible Interactive Streams or
Eligible Limited Downloads but only from a non-portable device to
which those Eligible Interactive Streams or Eligible Limited
Downloads are originally transmitted.
[[Page 54484]]
Standalone Non-Portable Subscription Offering--Streaming Only
means a Subscription Offering through which an End User can listen
to sound recordings only in the form of Eligible Interactive Streams
and only from a non-portable device to which those Eligible
Interactive Streams are originally transmitted while the device has
a live network connection.
Standalone Portable Subscription Offering means a Subscription
Offering through which an End User can listen to sound recordings in
the form of Eligible Interactive Streams or Eligible Limited
Downloads from a portable device.
Stream means the digital transmission of a sound recording of a
musical work to an End User--
(1) To allow the End User to listen to the sound recording,
while maintaining a live network connection to the transmitting
service, substantially at the time of transmission, except to the
extent that the sound recording remains accessible for future
listening from a Streaming Cache Reproduction;
(2) Using technology that is designed such that the sound
recording does not remain accessible for future listening, except to
the extent that the sound recording remains accessible for future
listening from a Streaming Cache Reproduction; and
(3) That is subject to licensing as a public performance of the
musical work.
Streaming Cache Reproduction means a reproduction of a sound
recording embodying a musical work made on a computer or other
receiving device by a Service Provider solely for the purpose of
permitting an End User who has previously received a Stream of that
sound recording to play the sound recording again from local storage
on the computer or other device rather than by means of a
transmission; provided that the End User is only able to do so while
maintaining a live network connection to the Service Provider, and
the reproduction is encrypted or otherwise protected consistent with
prevailing industry standards to prevent it from being played in any
other manner or on any device other than the computer or other
device on which it was originally made.
Student Plan means a discounted Subscription Offering available
on a limited basis to students.
Subscription Offering means an Offering for which End Users are
required to pay a fee to have access to the Offering for defined
subscription periods of 3 years or less (in contrast to, for
example, a service where the basic charge to users is a payment per
download or per play), whether the End User makes payment for access
to the Offering on a standalone basis or as part of a bundle with
one or more other products or services.
Total Cost of Content or TCC means the total amount expensed by
a Service Provider or any of its Affiliates in accordance with GAAP
for rights to make Eligible Interactive Streams or Eligible Limited
Downloads of a musical work embodied in a sound recording through
the Service Provider for the Accounting Period, which amount shall
equal the Applicable Consideration for those rights at the time the
Applicable Consideration is properly recognized as an expense under
GAAP. As used in this definition, Applicable Consideration means
anything of value given for the identified rights to undertake the
Licensed Activity, including, without limitation, ownership equity,
monetary advances, barter or any other monetary and/or nonmonetary
consideration, whether that consideration is conveyed via a single
agreement, multiple agreements and/or agreements that do not
themselves authorize the Licensed Activity but nevertheless provide
consideration for the identified rights to undertake the Licensed
Activity, and including any value given to an Affiliate of a Sound
Recording Company for the rights to undertake the Licensed Activity.
Value given to a Copyright Owner of musical works that is
controlling, controlled by, or under common control with a Sound
Recording Company for rights to undertake the Licensed Activity
shall not be considered value given to the Sound Recording Company.
Notwithstanding the foregoing, Applicable Consideration shall not
include in-kind promotional consideration given to a Sound Recording
Company (or Affiliate thereof) that is used to promote the sale or
paid use of sound recordings embodying musical works or the paid use
of music services through which sound recordings embodying musical
works are available where the in-kind promotional consideration is
given in connection with a use that qualifies for licensing under 17
U.S.C. 115.
Sec. 385.3 Late payments.
A Licensee shall pay a late fee of 1.5% per month, or the
highest lawful rate, whichever is lower, for any payment owed to a
Copyright Owner and remaining unpaid after the due date established
in 17 U.S.C. 115(c)(2)(I) or (d)(4)(A)(i), as applicable and
detailed in part 210 of this title. Late fees shall accrue from the
due date until the Copyright Owner receives payment, except that
where payment is due to the mechanical licensing collective under 17
U.S.C. 115(d)(4)(A)(i), late fees shall accrue from the due date
until the mechanical licensing collective receives payment.
Sec. 385.4 Recordkeeping for promotional or free trial non-royalty-
bearing uses.
(a) General. A Licensee transmitting a sound recording embodying
a musical work subject to section 115 and subparts C and D of this
part and claiming a Promotional Offering or Free Trial Offering zero
royalty rate shall keep complete and accurate contemporaneous
written records of making or authorizing Eligible Interactive
Streams or Eligible Limited Downloads, including the sound
recordings and musical works involved, the artists, the release
dates of the sound recordings, a brief statement of the promotional
activities authorized, the identity of the Offering or Offerings for
which the zero-rate is authorized (including the internet address if
applicable), and the beginning and end date of each zero rate
Offering.
(b) Retention of records. A Service Provider claiming zero rates
shall maintain the records required by this section for no less time
than the Service Provider maintains records of royalty-bearing uses
involving the same types of Offerings in the ordinary course of
business, but in no event for fewer than five years from the
conclusion of the zero rate Offerings to which they pertain.
(c) Availability of records. If a Copyright Owner or agent
requests information concerning zero rate Offerings, the Licensee
shall respond to the request within an agreed, reasonable time.
Subpart B--Physical Phonorecord Deliveries, Permanent Downloads,
Ringtones, and Music Bundles
Sec. 385.10 Scope.
This subpart establishes rates and terms of royalty payments for
making and distributing phonorecords, including by means of Digital
Phonorecord Deliveries, in accordance with the provisions of 17
U.S.C. 115.
Sec. 385.11 Royalty rates.
(a) Physical phonorecord deliveries and Permanent Downloads. For
every physical phonorecord and Permanent Download the Licensee makes
and distributes or authorizes to be made and distributed, the
royalty rate payable for each work embodied in the phonorecord or
Permanent Download shall be either 9.1 cents or 1.75 cents per
minute of playing time or fraction thereof, whichever amount is
larger.
(b) Ringtones. For every Ringtone the Licensee makes and
distributes or authorizes to be made and distributed, the royalty
rate payable for each work embodied therein shall be 24 cents.
(c) Music Bundles. For a Music Bundle, the royalty rate for each
element of the Music Bundle shall be the rate required under
paragraph (a) or (b) of this section, as appropriate.
Subpart C--Eligible Interactive Streaming, Eligible Limited
Downloads, Limited Offerings, Mixed Service Bundles, Bundled
Subscription Offerings, Locker Services, and Other Delivery
Configurations
Sec. 385.20 Scope.
This subpart establishes rates and terms of royalty payments for
Eligible Interactive Streams and Eligible Limited Downloads of
musical works, and other reproductions or distributions of musical
works through Limited Offerings, Mixed Service Bundles, Bundled
Subscription Offerings, Paid Locker Services, and Purchased Content
Locker Services provided through subscription and nonsubscription
digital music Service Providers in accordance with the provisions of
17 U.S.C. 115, exclusive of Offerings subject to subpart D of this
part.
Sec. 385.21 Royalty rates and calculations.
(a) Applicable royalty. Licensees that engage in Licensed
Activity covered by this subpart pursuant to 17 U.S.C. 115 shall pay
royalties therefor that are calculated as provided in this section,
subject to the royalty floors for specific types of services
[[Page 54485]]
described in Sec. 385.22, provided, however, that Promotional
Offerings, Free Trial Offerings, and certain Purchased Content
Locker Services shall instead be subject to the royalty rates
provided in subpart D of this part.
(b) Rate calculation. Royalty payments for Licensed Activity in
this subpart shall be calculated as provided in this paragraph (b).
If a Service Provider includes different Offerings, royalties must
be calculated separately with respect to each Offering taking into
consideration Service Provider Revenue and expenses associated with
each Offering.
(1) Step 1: Calculate the all-in royalty for the Offering. For
each Accounting Period, the all-in royalty for each Offering under
this subpart shall be the greater of the applicable percent of
Service Provider Revenue, as set forth in table 1 to this paragraph
(b)(1), and the result of the TCC Prong Calculation for the
respective type of Offering, as set forth in table 2 to this
paragraph (b)(1):
Table 1 to Paragraph (b)(1)
----------------------------------------------------------------------------------------------------------------
Royalty year 2018 2019 2020 2021 2022
----------------------------------------------------------------------------------------------------------------
Percent of Service Provider Revenue....... 11.4 12.3 13.3 14.2 15.1
----------------------------------------------------------------------------------------------------------------
Table 2 to Paragraph (b)(1)
------------------------------------------------------------------------
Type of offering TCC prong calculation
------------------------------------------------------------------------
Standalone Non-Portable Subscription The lesser of 22% of TCC for
Offering--Streaming Only. the Accounting Period and 50
cents per subscriber per
month.
Standalone Non-Portable Subscription The lesser of 21% of TCC for
Offering--Mixed. the Accounting Period and 50
cents per subscriber per
month.
Standalone Portable Subscription The lesser of 21% of TCC for
Offering. the Accounting Period and 80
cents per subscriber per
month.
Bundled Subscription Offering.......... 21% of TCC for the Accounting
Period.
Free nonsubscription/ad-supported 22% of TCC for the Accounting
services free of any charge to the End Period.
User.
Mixed Service Bundle................... 21% of TCC for the Accounting
Period.
Purchased Content Locker Service....... 22% of TCC for the Accounting
Period.
Limited Offering....................... 21% of TCC for the Accounting
Period.
Paid Locker Service.................... 20.65% of TCC for the
Accounting Period.
------------------------------------------------------------------------
(2) Step 2: Subtract applicable Performance Royalties. From the
amount determined in step 1 in paragraph (b)(1) of this section, for
each Offering of the Service Provider, subtract the total amount of
Performance Royalty that the Service Provider has expensed or will
expense pursuant to public performance licenses in connection with
uses of musical works through that Offering during the Accounting
Period that constitute Licensed Activity. Although this amount may
be the total of the Service Provider's payments for that Offering
for the Accounting Period, it will be less than the total of the
Performance Royalties if the Service Provider is also engaging in
public performance of musical works that does not constitute
Licensed Activity. In the case in which the Service Provider is also
engaging in the public performance of musical works that does not
constitute Licensed Activity, the amount to be subtracted for
Performance Royalties shall be the amount allocable to Licensed
Activity uses through the relevant Offering as determined in
relation to all uses of musical works for which the Service Provider
pays Performance Royalties for the Accounting Period. The Service
Provider shall make this allocation on the basis of Plays of musical
works or, where per-play information is unavailable because of bona
fide technical limitations as described in step 4 in paragraph
(b)(4) of this section, using the same alternative methodology as
provided in step 4.
(3) Step 3: Determine the payable royalty pool. The payable
royalty pool is the amount payable for the reproduction and
distribution of all musical works used by the Service Provider by
virtue of its Licensed Activity for a particular Offering during the
Accounting Period. This amount is the greater of:
(i) The result determined in step 2 in paragraph (b)(2) of this
section; and
(ii) The royalty floor (if any) resulting from the calculations
described in Sec. 385.22.
(4) Step 4: Calculate the per-work royalty allocation. This is
the amount payable for the reproduction and distribution of each
musical work used by the Service Provider by virtue of its Licensed
Activity through a particular Offering during the Accounting Period.
To determine this amount, the result determined in step 3 in
paragraph (b)(3) of this section must be allocated to each musical
work used through the Offering. The allocation shall be accomplished
by dividing the payable royalty pool determined in step 3 for the
Offering by the total number of Plays of all musical works through
the Offering during the Accounting Period (other than Plays subject
to subpart D of this part) to yield a per-Play allocation, and
multiplying that result by the number of Plays of each musical work
(other than Plays subject to subpart D of this part) through the
Offering during the Accounting Period. For purposes of determining
the per-work royalty allocation in all calculations under this
paragraph (b)(4) only (i.e., after the payable royalty pool has been
determined), for sound recordings of musical works with a playing
time of over 5 minutes, each Play shall be counted as provided in
paragraph (c) of this section. Notwithstanding the foregoing, if the
Service Provider is not capable of tracking Play information because
of bona fide limitations of the available technology for Offerings
of that nature or of devices useable with the Offering, the per-work
royalty allocation may instead be accomplished in a manner
consistent with the methodology used for making royalty payment
allocations for the use of individual sound recordings.
(c) Overtime adjustment. For purposes of the calculations in
step 4 in paragraph (b)(4) of this section only, for sound
recordings of musical works with a playing time of over 5 minutes,
adjust the number of Plays as follows:
(1) 5:01 to 6:00 minutes--Each Play = 1.2 Plays.
(2) 6:01 to 7:00 minutes--Each Play = 1.4 Plays.
(3) 7:01 to 8:00 minutes--Each Play = 1.6 Plays.
(4) 8:01 to 9:00 minutes--Each Play = 1.8 Plays.
(5) 9:01 to 10:00 minutes--Each Play = 2.0 Plays.
(6) For playing times of greater than 10 minutes, continue to
add 0.2 Plays for each additional minute or fraction thereof.
(d) Accounting. The calculations required by paragraph (b) of
this section shall be made in good faith and on the basis of the
best knowledge, information, and belief at the time payment is due,
and subject to the additional accounting and certification
requirements of 17 U.S.C. 115(c)(2)(I) and (d)(4)(A)(i) and part 210
of this title. Without limitation, statements of account (where
applicable) shall set forth each step of the calculations with
sufficient information to allow the assessment of the accuracy and
manner in which the payable royalty pool and per-play allocations
(including information sufficient to demonstrate whether and how a
royalty floor pursuant to Sec. 385.22 does or does not apply) were
[[Page 54486]]
determined and, for each Offering reported, also indicate the type
of Licensed Activity involved and the number of Plays of each
musical work (including an indication of any overtime adjustment
applied) that is the basis of the per-work royalty allocation being
paid.
(e) Computation of subscriber months in TCC Prong Calculation.
In connection with the TCC Prong Calculation in step 1 in paragraph
(b)(1) of this section for an Accounting Period, to the extent
applicable, the total number of subscriber-months for the Accounting
Period shall be calculated, taking all End Users who were
subscribers for complete calendar months, prorating in the case of
End Users who were subscribers for only part of a calendar month,
and deducting on a prorated basis for End Users covered by an
Offering subject to subpart D of this part. The product of the total
number of subscriber-months for the Accounting Period and the
specified number of cents per subscriber shall be used as the
subscriber-based component (if any) in step 1 for the Accounting
Period.
Sec. 385.22 Royalty floors for specific types of Offerings.
(a) In general. The following royalty floors for use in step 3
of Sec. 385.21(b)(3)(ii) shall apply to the respective types of
Offerings.
(1) Standalone Non-Portable Subscription Offering--Streaming
Only. Except as provided in paragraph (a)(4) of this section, in the
case of a Subscription Offering through which an End User can listen
to sound recordings only in the form of Eligible Interactive Streams
and only from a non-portable device to which those Streams are
originally transmitted while the device has a live network
connection, the royalty floor is the aggregate amount of 15 cents
per subscriber per month.
(2) Standalone Non-Portable Subscription Offering--Mixed. Except
as provided in paragraph (a)(4) of this section, in the case of a
Subscription Offering through which an End User can listen to sound
recordings either in the form of Eligible Interactive Streams or
Eligible Limited Downloads but only from a non-portable device to
which those Streams or Eligible Limited Downloads are originally
transmitted, the royalty floor is the aggregate amount of 30 cents
per subscriber per month.
(3) Standalone Portable Subscription Offering. Except as
provided in paragraph (a)(4) of this section, in the case of a
Subscription Offering through which an End User can listen to sound
recordings in the form of Eligible Interactive Streams or Eligible
Limited Downloads from a portable device, the royalty floor is the
aggregate amount of 50 cents per subscriber per month.
(4) Bundled Subscription Offering. In the case of a Bundled
Subscription Offering, the royalty floor is the aggregate amount of
25 cents per month for each Active Subscriber.
(b) Computation of royalty floors. For purposes of paragraph (a)
of this section, to determine the royalty floor, as applicable to
any particular Offering, the total number of subscriber-months for
the Accounting Period shall be calculated by taking all End Users
who were subscribers for complete calendar months, prorating in the
case of End Users who were subscribers for only part of a calendar
month, and deducting on a prorated basis for End Users covered by an
Offering subject to subpart D of this part, except in the case of a
Bundled Subscription Offering, subscriber-months shall be determined
with respect to Active Subscribers. The product of the total number
of subscriber-months for the Accounting Period and the specified
number of cents per subscriber (or Active Subscriber, as the case
may be) shall be used as the subscriber-based component of the
royalty floor for the Accounting Period. A Family Plan shall be
treated as 1.5 subscribers per month, prorated in the case of a
Family Plan subscription in effect for only part of a calendar
month. A Student Plan shall be treated as 0.50 subscribers per
month, prorated in the case of a Student Plan End User who
subscribed for only part of a calendar month.
Subpart D--Promotional Offerings, Free Trial Offerings and Certain
Purchased Content Locker Services
Sec. 385.30 Scope.
This subpart establishes rates and terms of royalty payments for
Promotional Offerings, Free Trial Offerings, and certain Purchased
Content Locker Services provided by subscription and nonsubscription
digital music Service Providers in accordance with the provisions of
17 U.S.C. 115.
Sec. 385.31 Royalty rates.
(a) Promotional Offerings. For Promotional Offerings of audio-
only Eligible Interactive Streams and Eligible Limited Downloads of
sound recordings embodying musical works that the Sound Recording
Company authorizes royalty-free to the Service Provider, the royalty
rate is zero.
(b) Free Trial Offerings. For Free Trial Offerings for which the
Service Provider receives no monetary consideration, the royalty
rate is zero.
(c) Certain Purchased Content Locker Services. For every
Purchased Content Locker Service for which the Service Provider
receives no monetary consideration, the royalty rate is zero.
(d) Unauthorized use. If a Copyright Owner or agent of the
Copyright Owner sends written notice to a Licensee stating in good
faith that a particular Offering subject to this subpart differs in
a material manner from the terms governing that Offering, the
Licensee must within 5 business days cease Streaming or otherwise
making available that Copyright Owner's musical works and shall
withdraw from the identified Offering any End User's access to the
subject musical work.
Dated: July 3, 2023.
-----------------------------------------------------------------------
David P. Shaw,
Chief Copyright Royalty Judge
-----------------------------------------------------------------------
David R. Strickler,
Copyright Royalty Judge
-----------------------------------------------------------------------
Steve Ruwe,
Copyright Royalty Judge
Approved by:
-----------------------------------------------------------------------
Carla D. Hayden,
Librarian of Congress.
[FR Doc. 2023-14925 Filed 8-9-23; 8:45 am]
BILLING CODE 1410-72-P
