Music Choice v. Copyright Royalty Board

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). SoundEx-change, 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 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 *1004 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 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. (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 noninter-active 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, 1998 U.S.C.C.A.N. 639, 656-67 .(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 noninter- *1005 active 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.). 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 SoundEx-change. SoundExchange, 571 F.3d at 1223-25. In the current proceeding, Soun-dExchange 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, 1998 U.S.C.C.A.N. 639, 662-64 (Conf.Rep.). SoundExchange’s expert witness, Dr. Jan-usz 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 *1006 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 SoundEx-change’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 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” apd 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 “[cjhannels, 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 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 associat *1007 ed 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%. SoundExehange requested rates that would begin at 15% in 2013, and rise to 45% in 2017. Id. at 23,056. SoundExehange 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 “conelude[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 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^] [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.” Sou ndExchange, 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 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 SoundExehange appeals from the Judges’ setting of royalty rates and terms *1008 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 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 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 “adjusting] down from a 13% rate ... based *1009 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. 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, SoundEx-change contends that reliance on the old 13% benchmark was arbitrary “given that the Judges were presented with current data,” ie., 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. SoundEx-change’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%. c. Rejection of SoundExchange’s Benchmarks SoundExchange further contends that the Judges arbitrarily rejected Ord-over’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 Soun-dExchange’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 *1010 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 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. 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. *1011 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 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 *1012 method 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. 8. 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 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 SoundEx-change’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 Royalty Judges’ use of the 7.5% rate as a *1013 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)(l)(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.” SoundEx-change, 571 F.3d at 1224. We thus hold that SoundExehange’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 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. *1014 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 28,055. “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, 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 Soun-dExchange’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 Fed.Reg. at 23,058. *1015 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 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 *1016 Circuit held that digital downloads are not such a performance because downloads “are not ... 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 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. Footnotes [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.