[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 II





Library of Congress





-----------------------------------------------------------------------





Copyright Royalty Board





-----------------------------------------------------------------------





37 CFR Part 385





Determination 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. 25973).
---------------------------------------------------------------------------

    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 Johnson.
---------------------------------------------------------------------------

    \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 
Johnson:
    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