Music Choice v. Copyright Royalty Board

United States Court of Appeals
FOR THE DISTRICT OF COLUMBIA CIRCUIT

Argued September 15, 2014 Decided December 19, 2014

No. 13-1174

MUSIC CHOICE,
APPELLANT

v.

COPYRIGHT ROYALTY BOARD,
APPELLEE

SIRIUS XM RADIO INC. AND SOUNDEXCHANGE, INC.,
INTERVENORS

Consolidated with 13-1183

On Appeals From a Final Determination
of the Copyright Royalty Judges

Matthew S. Hellman argued the cause for appellant
SoundExchange, Inc. With him on the briefs were David A.
Handzo, Michael B. DeSanctis, and Erica L. Ross.
 2
Paul M. Fakler argued the cause for appellant Music
Choice. With him on the briefs was Martin F. Cunniff.

Sonia K. McNeil, Attorney, U.S. Department of Justice,
argued the cause for appellee. With her on the brief were
Stuart F. Delery, Assistant Attorney General, and Mark R.
Freeman, Attorney. Scott R. McIntosh, Attorney, entered an
appearance.

Todd D. Larson argued the cause for intervenor Sirius
XM Radio Inc. With him on the brief was R. Bruce Rich.

Paul M. Fakler and Martin F. Cunniff were on the brief
for intervenor Music Choice in support of appellee.

Matthew S. Hellman, David A. Handzo, Michael B.
DeSanctis, and Erica L. Ross were on the brief for intervenor
SoundExchange, Inc. in support of appellee.

Before: SRINIVASAN, Circuit Judge, EDWARDS and
SENTELLE, Senior Circuit Judges.

Opinion for the Court filed by Senior Circuit Judge
SENTELLE.

SENTELLE, Senior Circuit Judge: In 2013, the Judges of
the Copyright Royalty Board issued a determination setting
royalty rates and defining terms for statutorily defined
satellite digital audio radio services (SDARS) and preexisting
subscription services (PSS). SoundExchange, an organization
that collects and distributes royalties to copyright owners,
appeals the Judges’ determination, arguing that the Judges
arbitrarily set SDARS and PSS rates too low.
SoundExchange also contends that the Judges erred in
defining “Gross Revenues” and eligible deductions for
 3
SDARS. Music Choice, a PSS that provides music-only
television channels, also appeals the determination, arguing
that the Judges arbitrarily set PSS rates too high. Concluding
that the Judges acted within their broad discretion and on a
sufficient record, we affirm the Copyright Royalty Judges’
determination of royalty rates and terms for both SDARS and
PSS.

I. BACKGROUND

A. Statutory and Regulatory Framework

Statutory law creates two types of copyrights in musical
recordings. First, 17 U.S.C. § 106(4) covers the underlying
“musical work” and protects the owner’s exclusive right to
perform the work in public. See SoundExchange, Inc. v.
Librarian of Congress, 571 F.3d 1220, 1222 (D.C. Cir. 2009)
(citing 17 U.S.C. § 106(4)). Broadcast of a musical work is a
performance of the work and therefore requires a license from
the copyright owner. Id. Second, since 1972, the law has also
protected a limited copyright in a “sound recording,” the
musical work as preserved in a recording medium. The law,
however, did not recognize an exclusive right in the public
performance of a sound recording until 1995. Id. As we
noted in SoundExchange, the 1995 amendments to the
Copyright Act afford the owner of a copyright of a sound
recording “the narrow but exclusive right ‘to perform the
copyrighted work publicly by means of a digital audio
transmission.’” Id. (quoting §§ 106(6), 114(d)). When
Congress recognized this exclusive right, it also enacted a
detailed statutory scheme providing for the administration of
this protected right, codified in Title 17 of the United States
Code. See 17 U.S.C. §§ 114, 801–804. The statutory scheme
requires “certain digital music services . . . to pay recording
companies and recording artists when they transmit[] sound
 4
recordings.” Recording Industry Ass’n of America v.
Librarian of Congress (“R.I.A.A.”), 176 F.3d 528, 530 (D.C.
Cir. 1999).

The statute provides for the appointment of three
Copyright Royalty Judges by the Librarian of Congress. 17
U.S.C. § 801(a). If the owners of sound recording copyrights
are unable to negotiate a mutually acceptable royalty with
digital music services, the statute empowers the Judges to set
“reasonable rates and terms of royalty payments.” Id.
§ 114(f)(1)(A).

The statute mandates that the rates “shall be calculated to
achieve the following objectives”:

(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.
 5
(D) To minimize any disruptive impact on the
structure of the industries involved and on
generally prevailing industry practices.

Id. § 801(b)(1).

The statute includes a special compulsory statutory
license for the benefit of preexisting subscription services and
preexisting satellite digital audio radio services to protect the
investment of noninteractive services that had come into
existence before the recognition of the digital performance
right. Id. § 114(f)(1); see also H.R. Rep. No. 105-796, at 80–
81 (Conf. Rep.). The statute directs the Copyright Royalty
Judges to “make determinations and adjustments of
reasonable terms and rates” for preexisting services based on
the enumerated policy factors set forth above. 17 U.S.C.
§ 801(b)(1). As to newer noninteractive services, the Judges
are to determine rates that “most clearly represent the rates
and terms that would have been negotiated in the marketplace
between a willing buyer and a willing seller.” Id.
§ 114(f)(2)(B); see also R.I.A.A., 176 F.3d at 533 (discussing
difference between statutory licenses for preexisting services
and newer noninteractive services).

B. The Proceedings Below

Then-current SDARS and PSS rates were scheduled to
expire in 2013. In January 2011, the Copyright Royalty
Board scheduled a proceeding to establish PSS and SDARS
royalty rates and terms for the years 2013 through 2017. See
Determination of Rates and Terms for Preexisting
Subscription Services and Satellite Digital Audio Radio
Services, 78 Fed. Reg. 23,054, 23,054 (Apr. 17, 2013) (“Final
Determination”), as corrected, 78 Fed. Reg. 31,842 (May 28,
2013) (codified at 37 C.F.R. § 382.1 et seq.).
 6
SoundExchange, Music Choice, and Sirius XM successfully
petitioned to participate. Thirteen months of discovery and
motion practice culminated in a 19-day administrative trial, at
which the Judges heard from 32 fact and expert witnesses. In
April 2013, the Judges issued their Final Determination.

1. Setting the SDARS Rate

At the time of the proceeding, the current SDARS rate
was 8% of gross revenues, established by the Judges in the
last preceding ratemaking and affirmed by this court against
challenge by SoundExchange. SoundExchange, 571 F.3d at
1223–25. In the current proceeding, SoundExchange
proposed rates beginning at 12% in 2013, rising to 20% in
2017. Sirius XM, the only satellite provider currently subject
to the rate, proposed rates in the 5% to 7% range. Final
Determination, 78 Fed. Reg. at 23,061.

In proceedings to determine rates for the digital
performance of sound recordings, the parties unsurprisingly
introduced evidence of royalty agreements covering
analogous services such as webcasting, interactive
subscription rates, or non-preexisting, noninteractive services.
Sirius XM supported its rate proposal with evidence of direct
license agreements between Sirius XM and independent
record labels for the performance of sound recordings. Final
Determination, 78 Fed. Reg. at 23,061–62.

SoundExchange based its rate proposal on its expert
witness’s analysis of interactive streaming agreements.
Interactive services, in contrast to satellite radio and
preexisting subscription services, allow an end user to hear a
particular song on demand, and do not benefit from a
compulsory license. See 17 U.S.C. §§ 114(d)(2)–(3); H.R.
Rep. No. 105-796, at 87–88 (Conf. Rep.). SoundExchange’s
 7
expert witness, Dr. Janusz Ordover, examined “seven market
agreements for digital music between certain interactive
subscription services that stream music over the Internet and
each of the four major record labels” which included royalty
rates ranging from 50% to 70%. Final Determination, 78
Fed. Reg. at 23,062. Ordover then adjusted these rates to
account for “the fact that the Sirius XM satellite radio service
. . . transmits both music and non-music content,” as well as
the differences between satellite radio and interactive
subscription services. Id. at 23,063. These adjustments
yielded a rate of 22.23%. Id.

The Judges found Sirius XM’s direct license agreements
to be comparable to a degree, but after identifying certain
weaknesses, concluded that the top range of those
benchmarks (7%) set the lower bound of reasonable rates. Id.
at 23,063–65. The Judges found SoundExchange’s
benchmarks less helpful. They were not persuaded that Dr.
Ordover properly accounted for the differences between the
benchmark agreements and the rights and parties at issue in
the SDARS proceeding. They were also concerned by the
“yawning gap,” id. at 23,066, between the current SDARS
rate and the interactive services benchmarks. The Judges
concluded that SoundExchange’s adjusted benchmark of
22.23% “can be viewed as no more than the upper bound of
the zone of reasonableness, although it is a bound that the
Judges have little confidence in.” Id.

Left with a large divide between Sirius XM’s 7% and
SoundExchange’s 22.23%, the Judges considered three
interim “guideposts” to help determine the reasonable rate.
The Judges looked at SoundExchange’s proposed statutory
SDARS rates, which began at 12%; the prevailing SDARS
rate of 8%; and the unadjusted benchmark rate of 13%
determined in the prior round of ratemaking. The Judges then
 8
analyzed the parties’ benchmarks and the interim guideposts
in light of the Section 801(b) factors. The Judges found a
downward adjustment appropriate to account for Sirius XM’s
investment in satellite infrastructure. Id. at 23,068–71. Based
on this analysis, the Judges arrived at an SDARS rate of 11%.
In order to avoid disruption, the Judges adopted a staggered
schedule beginning at 9% in 2013 and increasing by .5%
annually until achievement of 11% in 2017. Id. at 23,071.

2. Defining “Gross Revenues” and Eligible
Deductions for SDARS

In the same proceedings, the Judges considered the
definition of “Gross Revenues” and deductions applicable to
the SDARS. Final Determination, 78 Fed. Reg. at 23,071–
75. The Copyright Act employs a percentage-of-revenue
metric for calculating licensing fees. “Gross Revenues”
represents the revenue base against which the percentage rate
is applied in order to calculate the total royalty obligation.
The Copyright Royalty Judges promulgate regulations to
define the scope of “Gross Revenues” for each type of
service. See 37 C.F.R. § 382.11. The copyright user may also
take deductions from its total royalty obligation to offset
separate payments and non-compensable revenue. For
example, a copyright user may deduct from its total payments
the cost of separate direct-licensing agreements, such as the
direct licenses Sirius XM had reached with independent
labels. See Final Determination, 78 Fed. Reg. at 23,071–73.

In defining “Gross Revenues,” the Judges allowed Sirius
XM to exclude revenues received for “[c]hannels,
programming, products and/or other services offered for a
separate charge where such channels use only incidental
performances of sound recordings.” 37 C.F.R. § 382.11.
Sirius XM offers several subscription packages, including a
 9
bundled “Select” package with music and non-music
channels, and a “talk-only” package with exclusively non-
music channels. The Judges allowed Sirius XM to exclude
from its royalty base revenue from talk-only packages and
advertising on non-music channels. Final Determination, 78
Fed. Reg. at 23,071–72.

The Judges also allowed Sirius XM, after calculation of
its total revenue royalty obligations including deduction of the
above items from Gross Revenues, to then deduct revenues
attributable to its use of sound recordings created on or before
February 14, 1972. The Judges reasoned that since federal
copyright protection does not extend to pre-1972 sound
recordings, such recordings are outside the federal statutory
license and revenue associated with such recordings may be
deducted. Final Determination, 78 Fed. Reg. at 23,073.

3. Setting the PSS Rates

Both parties requested PSS rates that would drastically
depart from the then-prevailing rate of 7.5%. SoundExchange
requested rates that would begin at 15% in 2013, and rise to
45% in 2017. Id. at 23,056. SoundExchange based its rate
proposal on its analysis of “over 2,000 marketplace
agreements, representing a variety of rights licensed.” Id. at
23,057. Music Choice requested a rate of 2.6%, which it
based on the rates it pays performing rights societies (such as
ASCAP and BMI) for the use of copyrighted musical works.
Id. at 23,056–57.

The Judges “conclude[d] that neither Music Choice’s nor
SoundExchange’s proffered rate guidance provide[d] a
satisfactory benchmark upon which they [could] rely to
determine the sound recording performance rates for” PSS.
Id. at 23,058. Lacking guidance from the parties, the Judges
 10
considered the then-current 7.5% rate, which had been
determined by settlement negotiation, in light of the record
and the Section 801(b) factors. The Judges concluded that
“nothing in the record persuade[d] [them] that 7.5% . . . is too
high, too low or otherwise inappropriate.” Id. The Judges
ultimately set PSS rates at 8.5%, with an upward adjustment
to account for Music Choice’s planned channel expansion.
The rate would start at 8% in 2013 and increase to 8.5% for
2014 through 2017. Id. at 23,061.

II. ANALYSIS

Our review of the decisions of the Copyright Royalty
Judges is deferential. 17 U.S.C. § 803(d)(3) expressly adopts
the standard of review set forth in the Administrative
Procedure Act, 5 U.S.C. § 706. We are to hold unlawful and
set aside a decision of the Copyright Royalty Judges “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
[Judges] have no basis in the record.” SoundExchange, 571
F.3d at 1223 (internal citations omitted) (quoting 5 U.S.C.
§ 706(2)(A)). Further, we have previously noted that we are
especially deferential to the Judges of the Copyright Royalty
Board for three distinct reasons:

First, the [Board] is required “to estimate the
effect of the royalty rate on the future of the
music industry,” which requires a “forecast of
the direction in which the future public interest
lies . . . based on the expert knowledge of the
agency.” Second, the agency has “legislative
discretion in determining copyright policy in
order to achieve an equitable division of music
industry profits between the copyright owners
and users.” Finally, “the statutory factors pull
 11
in opposing directions, and reconciliation of
these objectives is committed to the [agency]
as part of its mandate to determine
‘reasonable’ royalty rates.” [S]ee Fresno
Mobile Radio, Inc. v. FCC, 165 F.3d 965, 971
(D.C. Cir. 1999) (“When an agency must
balance a number of potentially conflicting
objectives . . . judicial review is limited to
determining whether the agency’s decision
reasonably advances at least one of those
objectives and its decisionmaking process was
regular[.]”).

Id. at 1223–24 (quoting R.I.A.A., 662 F.2d at 8–9) (other
citations omitted). Applying that standard to our review of
the record before us, we find no basis to set aside the decision
of the Copyright Royalty Board.

A. SoundExchange’s Challenge to the
Copyright Royalty Judges’ Final
Determination

SoundExchange appeals from the Judges’ setting of
royalty rates and terms for both satellite digital audio radio
services and preexisting subscription services, contending that
the Judges’ action was arbitrary and capricious and resulted in
unlawfully low rates for both services. Upon review, we
uphold the rates as to both categories.

1. The SDARS Royalty Rate

SoundExchange argues that the Copyright Royalty
Judges’ setting of rates for satellite digital audio radio
services was arbitrary, capricious, and not supported by the
written record. SoundExchange primarily challenges the
 12
Judges’ rejection of SoundExchange’s proposed benchmarks
and the Judges’ subsequent reliance on interim guideposts to
help determine the reasonable rate. We hold that the Judges
acted within their broad discretion, and on the basis of a
sufficient record, when they discounted SoundExchange’s
benchmarks and considered interim guideposts. The Judges
could properly consider the then-current 8% rate as an interim
guidepost. The Copyright Act directs the Judges to make
“adjustments” to the prevailing rate, 17 U.S.C. § 801(b)(1),
and allows them to consider “prior determinations,” id.
§ 803(a)(1). The Judges could also consider the lowest of
SoundExchange’s proposed royalty rates (12%) as a rate, in
the record of the current proceeding, that could fall within the
zone of reasonableness. The 13% guidepost warrants greater
discussion.

a. Reliance on 13% Rate as “Extra-
Record”

In contending that the Copyright Royalty Judges acted
arbitrarily and capriciously, and failed to make a
determination supported by the record, SoundExchange
focuses on the Judges’ use of the 13% rate. During the last
round of ratemaking, the Judges looked to “comparable
marketplace royalty rates as ‘benchmarks,’ indicative of the
prices that prevail for services purchasing similar music
inputs for use in digital programming.” Determination of
Rates and Terms for Preexisting Subscription Services and
Satellite Digital Audio Radio Services (“SDARS-I”), 73 Fed.
Reg. 4080, 4088 (Jan. 24, 2008). In that prior determination,
the Judges “considered the record evidence reflecting various
experts’ opinions and concluded that a rate equal to 13% of
[satellite radio] gross revenue, as proposed by
SoundExchange, ‘marks the upper boundary for a zone of
reasonableness for potential marketplace benchmarks from
 13
which to identify a rate that satisfies’ the objectives in § 801.”
SoundExchange, 571 F.3d at 1222–23 (quoting SDARS-I, 73
Fed. Reg. at 4094). After considering the Section 801(b)
factors, the Judges set the SDARS rate at 8%. SDARS-I, 73
Fed. Reg. at 4097–98.

SoundExchange characterizes the 13% rate as an obsolete
benchmark divorced from the proceedings below, and argues
that considering the 13% rate violated the Copyright Act’s
requirement that the record support the determination. 17
U.S.C. §§ 803(a)(1), (c)(3). SoundExchange states that
“[t]here is no dispute that the 13% benchmark was not part of
the record in this proceeding,” SoundExchange’s Opening
Brief (“SX Br.”) 18, and that the Judges’ unexpected reliance
on the 13% rate deprived SoundExchange of “the chance to
respond to material central to the tribunal’s decision,” id. at
20 (emphasis in original).

SoundExchange’s argument is unpersuasive. The Judges
did not consider the 13% rate as a current marketplace
benchmark. Rather, they considered it as one of several
guideposts in light of the fact that the Judges had previously
derived “the prevailing statutory rate of 8%” by “adjust[ing]
down from a 13% rate . . . based on the fourth Section 801(b)
factor.” Final Determination, 78 Fed. Reg. at 23,066
(explaining prior ratemaking). Thus, the 13% rate did not
represent an extra-record market rate, but represented a
component of a prior determination. Consideration of the
13% rate as a guidepost is consistent with the Copyright Act’s
contemplation that the Judges would make “adjustments” to
prevailing rates, and that they could consider “prior
determinations.” 17 U.S.C. §§ 801(b)(1), 803(a)(1). The
Copyright Act permitted the Judges to consider the 13% rate,
and there was no undue surprise resulting from the Judges’
consideration of it.
 14

b. Reliance on 13% Rate as “Stale”
and “Obsolete”

SoundExchange also argues that the Judges erroneously
rejected current benchmark data in favor of a “stale” and
“obsolete” benchmark. SX Br. 20–23. In addition to
contending that the 13% rate was not a part of the record,
SoundExchange contends that reliance on the old 13%
benchmark was arbitrary “given that the Judges were
presented with current data,” i.e., Ordover’s benchmark
analysis. Id. at 21 (emphasis in original).

We hold that the Judges did not err in relying on the 13%
rate, and that the 13% rate did not represent a “stale,”
“outdated,” or “obsolete” benchmark. SoundExchange’s
argument rests on the erroneous assertion that the Judges
treated the 13% rate as a current market benchmark. See SX
Br. 30–33. As we stated above, the Judges did not consider
the 13% rate as a current benchmark. Instead, they
considered the 13% rate as a component of a prior
determination, in order to bridge the gap between Sirius XM’s
highest benchmark and SoundExchange’s lowest benchmark.
See Final Determination, 78 Fed. Reg. at 23,066. Thus, the
Judges did not reject a current benchmark in favor of a “stale”
benchmark. They found serious problems with
SoundExchange’s benchmark, partially credited it, and used
permissible indicia of reasonableness to help fix the rate
between 7% and 22.23%.
 15
c. Rejection of SoundExchange’s
Benchmarks

SoundExchange further contends that the Judges
arbitrarily rejected Ordover’s benchmark analysis, SX. Br.
25–33, and “if the Judges were not fully persuaded by the
current benchmarks, they should have adjusted those
benchmarks based on record evidence,” id. at 24. We
disagree. The Judges were within their broad discretion to
discount Ordover’s benchmarks and look elsewhere for
guidance. The Judges adequately considered, and explained
their dissatisfaction with, the proposed benchmarks. Final
Determination, 78 Fed. Reg. at 23,062–71. While the Judges
might have made further adjustments to Ordover’s
benchmarks to render them useful, see id. at 23,090–92
(dissenting opinion of Copyright Royalty Judge Roberts), the
Judges were not required to do so. The mandate to issue
determinations “supported by the written record,” 17 U.S.C.
§ 803(c)(3), does not hamstring the Judges when neither party
proposes reasonable or comparable benchmarks. The Act
expressly allows the Judges also to consider prevailing rates
and prior determinations. Id. at §§ 801(b)(1), 803(a)(1).
Having considered and discounted SoundExchange’s
proposed benchmarks, the Judges did not act arbitrarily when
they looked to the 13% unadjusted benchmark from the prior
determination as an interim guidepost.

SoundExchange also maintains that, despite the Judges’
references to SoundExchange’s adjusted benchmark rate of
22.23% as representing the top end of reasonable rates, the
13% rate arbitrarily capped the zone of reasonableness. SX
Br. 30–33. We disagree. Even if the 13% rate marked the
top end of a “zone of reasonableness,” this is no reason to
overturn the Judges’ determination. Nothing requires the
Judges, if they choose to use a zone of reasonableness, only to
 16
use market benchmarks to set the upper and lower boundaries.
The Copyright Act permits, but does not require, the Judges to
use market rates to help determine reasonable rates. See
SoundExchange, 571 F.3d at 1224 (“[T]he agency [is] under
no obligation to choose a rate derived from a market-based
approach.”); R.I.A.A., 176 F.3d at 532–33 (“RIAA’s claim
that the statute clearly requires the use of ‘market rates’ is
simply wrong.” (emphasis in original)). Other potentially
reasonable rates, such as the prevailing statutory rate, may
also bound the zone. Having explained their dissatisfaction
with SoundExchange’s benchmarks, and free to consider the
13% rate as a component of a prior determination, the Judges
would not have erred had they used the 13% rate to cap the
top end of a zone of reasonable rates.

The Judges acted within their discretion when, after
identifying weaknesses with the proposed benchmarks, they
employed interim guideposts to determine a reasonable rate.
Thus, we affirm the Judges’ determination of Section 114
rates for satellite digital audio radio services.

2. SoundExchange’s Challenge to the
Definition of “Gross Revenues” and
Allowance for Deductions for SDARS

SoundExchange contends that the Judges acted arbitrarily
by allowing Sirius XM to exclude from “Gross Revenues”
revenue attributable to non-music programming, and deduct
from its total royalty obligations revenue attributable to pre-
1972 sound recordings. We will uphold the Judges’ definition
of “Gross Revenues” and allowance for a pre-1972 sound
recording deduction as reasonable exercises of the Judges’
broad discretion.
 17
a. Exclusion for Non-Music
Programming

SoundExchange argues that the Judges’ definition of
“Gross Revenues” arbitrarily double discounted for the value
of non-music programming. SX Br. 33–34. SoundExchange
contends that the benchmarks offered by Sirius XM and
SoundExchange, and the resulting statutory royalty rate,
already discounted for the value of Sirius XM’s non-music
programming. SoundExchange maintains that since the
royalty percentage rate fully discounts for the value of Sirius
XM’s non-music programming, allowing a separate deduction
from the royalty base effectively gives Sirius XM the same
discount twice.

We disagree. There was no such double discounting.
The Judges “agree[d] with Sirius XM’s counter argument that
Dr. Ordover’s modeling allocated revenues for both the music
and non-music programming for Sirius XM’s standard
‘Select’ package, but that allocation in no way relates to the
separately priced non-music packages offered by Sirius XM
that are the subject of the exemption.” Final Determination,
78 Fed. Reg. at 23,072 n.45 (internal quotation marks
omitted). Under this reasonable interpretation of the
evidence, neither the benchmarks nor the Judges’ royalty rate
took into account revenue from separately priced non-music
packages. Instead, the benchmarks and percentage rate
accounted for the value of non-music programming within
Sirius XM’s bundled “music and talk” packages. It was
perfectly reasonable, then, for the Judges to define “Gross
Revenues” to exclude revenue solely attributable to non-
music programming.
 18
b. Deduction for Pre-1972
Performances

SoundExchange argues that the Judges arbitrarily
allowed Sirius XM to deduct from its royalty obligations
revenue attributable to the use of pre-1972 sound recordings.
While there is no federal copyright protection for the
performance of sound recordings created on or before
February 14, 1972, see Sound Recording Amendment, Public
Law 92-140, 85 Stat. 391 (1971), the extent of state law
protection is the subject of ongoing litigation, see, e.g., Order
Granting Pl.’s Mot. Sum. J., Flo & Eddie Inc. v. Sirius XM
Radio, Inc., No. CV 13-5693, 2014 WL 4725382 (C.D. Cal.
Sept. 22, 2014).

SoundExchange presents two alternative arguments,
depending on whether there is state law protection for pre-
1972 recordings. First, SoundExchange maintains that if
there is no state law protection for pre-1972 recordings, the
benchmark rates already account for the diminished value of
those recordings. In the agreements SoundExchange cited,
the buyers purchased rights to play all of the record labels’
sound recordings; the agreements did not distinguish between
pre-1972 and post-1971 recordings when determining the
percentage rate or royalty base. SoundExchange argues that
the market takes pre-1972 recordings into account when
setting royalty rates; reducing the royalty base, then, would be
redundant. SX Br. 38.

This first argument is unavailing. The agreements cited
by SoundExchange were either discounted by the Judges or
not moved into evidence before them. See Final
Determination, 78 Fed. Reg. at 23,064 (eschewing reliance on
Last.FM agreement, which SoundExchange cites to support
its argument regarding pre-1972 recordings (SX Br. 38)). In
 19
contrast, the direct license agreements Sirius XM entered into
evidence allow Sirius XM to adjust its revenue base
downward to account for the use of pre-1972 recordings. See
Sirius XM Dir. Ex. 7 at ¶ 2(a)(ii)(F)(4). The record evidence
does not compel the conclusion that the benchmark rates, or
the Judges’ statutory royalty rate, already took into account
the copyright status of pre-1972 recordings. Thus, the Judges
did not arbitrarily “double discount” for pre-1972 recordings.

Alternatively, SoundExchange argues that if there is state
law protection for pre-1972 recordings, there is no need for a
separate deduction. In such a case, Sirius XM would need to
directly license those performances from rights holders. The
existing direct license carve out would allow Sirius XM to
deduct the costs of these state rights licenses.
SoundExchange maintains that any separate deduction would
be superfluous, as the direct licensing deduction would
account for any revenue associated with the use of pre-1972
recordings. SX Br. 38–40.

This alternative argument is also unavailing. The
amendments to the Copyright Act say nothing about state
level rights. Accordingly, we need not determine whether the
Judges could have designated the direct licensing deduction as
the vehicle for excluding revenues associated with pre-1972
works, rather than creating a separate deduction as they did.
It is clear, however, that the Copyright Act does not mandate
SoundExchange’s preferred system for accounting for pre-
1972 works. Furthermore, there is no specter of “double
deducting.” There is no indication that under the system
established by the Judges, Sirius XM could take the deduction
for pre-1972 recordings, and then take a second deduction for
the direct licensing of state level rights. Thus, the pre-1972
deduction is not “redundant” or “superfluous.” Instead, the
deduction is simply not SoundExchange’s preferred method
 20
for dealing with pre-1972 works. The Judges acted within
their broad discretion when permitting a separate deduction
for revenues associated with the use of pre-1972 works.

3. SoundExchange’s Challenge to the
PSS Royalty Rate

SoundExchange argues that the Copyright Royalty
Judges’ setting of rates for preexisting subscription services
was arbitrary, capricious, and not supported by the written
record. Upon review, we will uphold the Judges’ setting of
PSS rates.

a. Reliance on the Prevailing
Settlement Rate

SoundExchange claims that the Judges acted arbitrarily,
and failed to issue a determination supported by the record, by
rejecting record evidence and relying on the existing 7.5%
rate—a rate, SoundExchange contends, for which no party
advocated and no evidence supports. SoundExchange notes
that this rate “‘is the product of settlement negotiations that
occurred in SDARS I between Music Choice and
SoundExchange.’” SX Br. 41 (quoting Final Determination,
78 Fed. Reg. at 23,058). In support of this proposition,
SoundExchange notes that it submitted as benchmarks “over
2,000 marketplace agreements, representing a variety of rights
licensed.” Final Determination, 78 Fed. Reg. at 23,057.
SoundExchange argues that the Judges arbitrarily rejected
these more recent data points in favor of the “outdated”
settlement rate. It 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
 21
cannot properly be said to be a market benchmark rate.” Id. at
23,058. The statute requires the Judges to issue
determinations “supported by the written record.” 17 U.S.C.
§ 803(c)(3). SoundExchange also argues that simply reciting
that “nothing in the record persuades the Judges” that the
prevailing rate is unreasonable, id., does not show that 7.5% is
reasonable, or that it is supported by the written record.

SoundExchange’s argument is unavailing. First, whether
the prevailing rate represents a “market benchmark” is not
determinative. As we explained above, nothing in the statute
requires the Judges to rely on market rates or agreements
when setting Section 114 rates. See SoundExchange, 571
F.3d at 1224; R.I.A.A., 176 F.3d at 532–34. The Judges acted
within their broad discretion when they rejected Music Choice
and SoundExchange’s benchmarks. The Judges did not reject
the proposed benchmarks without any consideration, but
offered a reasoned explanation as to why the kinds of rights
covered by the benchmark agreements were not sufficiently
comparable to the digital performance rights covered by
Section 114. Final Determination, 78 Fed. Reg. at 23,057–
59. The Judges were under no obligation to salvage
benchmarks they found to have fundamental problems.

Further, given 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 make
“adjustments” to the prevailing rate, 17 U.S.C. § 801(b)(1),
and consider “prior determinations,” id. § 803(a)(1), and rates
established “under voluntary license agreements,” id.
§ 114(f)(1)(B). The Judges sufficiently explained their
rejection of the parties’ benchmarks and how the prevailing
rate was reasonable given the Section 801(b) factors. Final
Determination, 78 Fed. Reg. at 23,058–59. The Copyright
 22
Royalty Judges’ use of the 7.5% rate as a starting point in its
analysis was not arbitrary or capricious.

b. Agency Precedent

SoundExchange also argues that the Judges violated
agency precedent. SoundExchange contends that agency
precedent required the Judges to follow a two-step process:
First, the judges should set a range of reasonable rates based
on marketplace benchmarks; second, they should find the
proper rate from within that range (or make adjustments to
those benchmarks) by applying the Section 801(b) factors.
SX Br. 40–42.

This argument is unpersuasive. We have previously
rejected the notion that the rate-maker “must first determine
the range of market rates that are appropriate and then select a
rate from within the range of fair market rates that meets the
objectives of § 801(b)(1)(A)–(D).” R.I.A.A., 176 F.3d at 532–
33. The Copyright Act does not “clearly require[] the use of
‘market rates.’” Id. at 533 (emphasis in original). Instead,
“‘reasonable rates’ are those that are calculated with reference
to the four statutory criteria.” Id. As we have made clear
both above and in an earlier decision, in a Section 114
proceeding, “the agency [is] under no obligation to choose a
rate derived from a market-based approach.”
SoundExchange, 571 F.3d at 1224. We thus hold that
SoundExchange’s agency precedent argument fails.

B. Music Choice’s Challenge to the Copyright
Royalty Judges’ Final Determination

Music Choice also argues that the Copyright Royalty
Judges’ setting of PSS rates was arbitrary, capricious, and not
supported by the written record. Music Choice argues that the
 23
Judges set PSS rates too high. In support of its request for a
PSS rate of 2.6%, Music Choice introduced agreements that it
had reached with ASCAP and BMI, leading performance
rights organizations, for the use of copyrighted musical works
and compositions in its residential audio service. Music
Choice represented that it pays ASCAP and BMI each 2.5%
of Gross Revenues. Music Choice then introduced evidence
that international jurisdictions value sound recordings and
musical works similarly. Final Determination, 78 Fed. Reg.
at 23,055–58.

1. Rejection of the Musical Works
Benchmark

Music Choice argues that the Judges’ rejection of the
musical works benchmark was contrary to precedent, and that
the Judges had not adequately explained their reasons for
departing from this precedent. The Copyright Act requires
the Judges to follow enumerated precedent, including prior
determinations and interpretations of the Librarian of
Congress. 17 U.S.C. § 803(a)(1). Music Choice contends
that in the initial PSS rate determination, the Librarian held
that the musical works rates set the upper bound of reasonable
rates for PSS. 1 Music Choice’s Opening Brief (“MC Br.”)
19–22; Determination of Reasonable Rates and Terms for the
Digital Performance of Sound Recordings, 63 Fed. Reg.
25,394 (May 8, 1998). Music Choice maintains that the
Judges violated the precedent, established by the Librarian in
the first PSS ratemaking, when they did not rely on musical
works rates as persuasive benchmarks. We disagree.

1
At that time, the Copyright Act conferred ultimate authority on
the Librarian of Congress to set rates and terms (based on
recommendations made by copyright arbitration royalty panels and
the Registrar of Copyrights). See R.I.A.A., 176 F.3d at 530–31.
 24

The Librarian did not determine, as a matter of law, that
future rate-makers must begin with the musical works rate.
See Final Determination, 78 Fed. Reg. at 23,055. To the
extent the PSS determination is precedential, the Judges have
adequately explained their departure. Cf. Intercollegiate
Broad. Sys., Inc. v. Copyright Royalty Bd., 574 F.3d 748, 762
(D.C. Cir. 2009) (“The Judges are free to depart from
precedent if they provide reasoned explanations for their
departures.”). The Judges stated that Music Choice “fail[ed]
to place” the initial PSS determination “in its historical
context.” Final Determination, 78 Fed. Reg. at 23,055. “The
Librarian had before him for consideration only the musical
works fees and the Music Choice partnership license
agreement. The Judges have more evidence in this
proceeding upon which to base a decision.” Id. Thus, the
Judges properly distinguished the Librarian’s initial PSS
determination from the proceedings below.

Music Choice maintains that even if the Librarian’s PSS
determination is not binding precedent, the Judges erred when
they rejected the musical works rate evidence in the record.
MC Br. 28–31. We disagree. The Judges acted well within
their broad discretion when they eschewed reliance on the
musical works rates. The Judges explained that the “musical
works market involves different sellers (performing rights
societies versus record companies) selling different rights”
than the sound recording rights at issue in this case. Final
Determination, 78 Fed. Reg. at 23,058. The Judges cited
other instances where the agency rate-maker rejected reliance
on musical works benchmarks. See id. at 23,058 n.16. The
Judges did not err when they found Music Choice’s evidence
from foreign jurisdictions unpersuasive. The Judges noted
that in an earlier proceeding, they discounted the significance
of how foreign jurisdictions treat different types of rights,
 25
finding that “comparability is a much more complex
undertaking in an international setting than in a domestic one.
There are a myriad of potential structural and regulatory
differences whose impact has to be addressed in order to
produce a meaningful comparison.” Mechanical and Digital
Phonorecord Delivery Rate Determination Proceeding, 74
Fed. Reg. 4510, 4522 (Jan. 26, 2009) (quoted in Final
Determination, 78 Fed. Reg. at 23,058). Music Choice failed
to address those differences, and thus failed to make the
foreign jurisdictions’ treatment of musical works meaningful.
The Judges did not summarily reject Music Choice’s
proffered musical works rates, but offered a reasoned
explanation for finding them not comparable and
unpersuasive.

2. Reliance on the Prevailing Settlement
Rate

Music Choice also argues that the Judges erred in relying
on the prevailing 7.5% settlement rate. For the reasons
discussed above, in connection with SoundExchange’s similar
contention, this argument fails. Music Choice raises the
additional point that the then-prevailing rate was “driven
solely by the disparate impact of rate litigation costs on a
small company like Music Choice.” MC Br. 36. Music
Choice maintains that it did not settle because it thought that
7.5% resembled a hypothetical market rate, or was otherwise
fair. Instead, it agreed to the rate entirely to avoid the
overbearing expenses of rate litigation, expenses that a well-
funded entity like SoundExchange could bear much better.
Id. at 36–38.

We hold that the Judges did not err when relying on the
settlement rate. The Judges conceded that the settlement rate
does not represent a market rate. Final Determination, 78
 26
Fed. Reg. at 23,058. But again, 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.” R.I.A.A., 176 F.3d at 533. The Act authorizes the
Judges to consider rates set “under voluntary license
agreements.” 17 U.S.C. § 114(f)(1)(B). 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.

3. Application of Section 801(b) Factors

Music Choice argues that the Judges erroneously failed to
make several downward adjustments based on faulty
interpretations and applications of the Section 801(b) factors.
MC Br. 43–53. Given the very broad discretion afforded to
the Judges in weighing these predictive and policy-laden
factors, see SoundExchange, 571 F.3d at 1223–24, we
conclude that the Judges did not exercise their discretion in an
arbitrary or capricious manner. The Judges’ upward
adjustment, based on Music Choice’s planned channel
expansion from 46 to 300 channels, warrants greater attention.
Under the second Section 801(b) factor, providing fair return
and fair income, the Judges found a 1% upward adjustment
appropriate to compensate for an expected increased use of
copyrighted works. Final Determination, 78 Fed. Reg. at
23,059–60.

Music Choice argues that there was no record evidence
that an upward adjustment was warranted based on Music
Choice’s planned channel expansion. MC Br. 40–42. Music
Choice maintains that the Section 114 license is a public
performance license, not a use license, and that mere
transmission without a corresponding listener does not
 27
constitute a performance. MC Reply 20 (citing United States
v. ASCAP, 627 F.3d 64, 73 (2nd Cir. 2010)). Thus, Music
Choice argues, the fact that it will transmit additional
channels does not necessarily mean that there will be
additional performances of sound recordings. Music Choice
posits a scenario in which the channel expansion does not
draw in additional listeners; there could be the same number
of listeners, but spread among more channels. Music Choice
argues that the Judges lacked any evidence that the channel
expansion would lead to increased listenership, and thus erred
when they adjusted the rate upward based on this
consideration. MC Br. 40–42; MC Reply 19–23.

Music Choice’s argument fails. The Judges acted
reasonably when they inferred that the channel expansion
would lead to increased performances of copyrighted works.
The Copyright Act empowers the Judges to “predict the future
course of the music industry.” SoundExchange, 571 F.3d at
1225. “The ‘arbitrary and capricious’ standard is particularly
deferential in matters implicating predictive judgments.”
Rural Cellular Ass’n v. F.C.C., 588 F.3d 1095, 1105 (D.C.
Cir. 2009). Given the broad discretion afforded to the Judges
in making predictive judgments, the Judges acted on
sufficient evidence, and not arbitrarily, when they determined
that Music Choice’s planned channel expansion warranted an
upward adjustment.

We also question Music Choice’s reliance on U.S. v.
ASCAP, 627 F.3d 64 (2nd Cir. 2010), for the proposition that
there cannot be proof of additional performances without
proof of additional listeners. U.S. v. ASCAP concerned
“whether a download of a digital file containing a musical
work constitutes a public performance of that musical work.”
Id. at 68. The Second Circuit held that digital downloads are
not such a performance because downloads “are not . . .
 28
contemporaneously perceived by the listener.” Id. at 73.
“The downloaded songs are not performed in any perceptible
manner during the transfers; the user must take some further
action to play the songs after they are downloaded.” Id. The
Second Circuit distinguished downloads from radio
broadcasts and digital streaming transmissions; in the latter
group, there is a performance “because there is a playing of
the song that is perceived simultaneously with the
transmission.” Id. at 74. Music Choice’s service is analogous
to radio and streaming broadcasts, as the listener
contemporaneously perceives the playing of a song while
Music Choice transmits it. Music Choice has not persuaded
us, through its citation to U.S. v. ASCAP, that the Copyright
Royalty Judges, in making their predictive judgment, required
proof of additional listeners and could not consider an upward
adjustment based on Music Choice’s planned channel
expansion.

Concluding that the Copyright Royalty Judges’ Final
Determination is supported by the record, and is neither
arbitrary nor capricious, we affirm the Judges’ determination
of Section 114 royalty rates and terms for preexisting
subscription services.

III. CONCLUSION

We have repeatedly recognized that the Copyright Act
gives the Judges of the Copyright Royalty Board broad
discretion to set rates and terms for compulsory licenses of the
digital performance of sound recordings. See, e.g.,
SoundExchange, 571 F.3d at 1223–24. The Act requires the
Judges to weigh competing, policy-laden, and forward-
looking factors when determining reasonable rates. Id. We
hold that the Copyright Royalty Judges did not exercise their
broad discretion in an arbitrary or capricious manner when
 29
setting royalty rates and terms for satellite digital audio radio
services and preexisting subscription services. The Judges
reasonably explained their determinations, and based those
determinations on substantial evidence. The Judges did not
rely on impermissible or “extra-record” factors. The
determination of the Judges of the Copyright Royalty Board
as to royalty rates and terms for satellite digital audio radio
services and preexisting subscription services is therefore
affirmed.

So ordered.