2013-08657. Determination of Rates and Terms for Preexisting Subscription Services and Satellite Digital Audio Radio Services

AGENCY:

Copyright Royalty Board, Library of Congress.

ACTION:

Final rule and order.

SUMMARY:

The Copyright Royalty Judges are announcing their final determination of the rates and terms for the digital transmission of sound recordings and the reproduction of ephemeral recordings by preexisting subscription services and preexisting satellite digital audio radio services for the period beginning January 1, 2013, and ending on December 31, 2017.

DATES:

Effective date: April 17, 2013.

Applicability date: The regulations apply to the license period January 1, 2013, through December 31, 2017.

ADDRESSES:

The final determination also is posted on the Copyright Royalty Board Web site at http://www.loc.gov/​crb.

FOR FURTHER INFORMATION CONTACT:

Gina Giuffreda, Attorney Advisor. Telephone: [phone]. Telefax: [phone].

SUPPLEMENTARY INFORMATION:

I. Introduction

The Copyright Royalty Judges (“Judges”) convened this rate determination proceeding in accordance with 17 U.S.C. 803(b) and 37 CFR 351. On January 5, 2011, the Judges published in the Federal Register a notice announcing commencement of this proceeding with request for Petitions to Participate in this proceeding. The purpose of the proceeding is to determine the rates and terms of royalty payments payable by Preexisting Subscription Services (“PSS”) and Satellite Digital Audio Radio Services (“SDARS”) under the Copyright Act, 17 U.S.C. 112 and 114. The rates and terms set in this proceeding apply to the period January 1, 2013, to December 31, 2017. Having carefully considered the relevant law and the evidence received in this proceeding, the Copyright Royalty Judges determine that the appropriate Section 114(f)(1) rates for the PSS are 8% of Gross Revenues for 2013 and 8.5% for 2014 through 2017. The Section 114(f)(1) rates for Sirius XM are 9% of Gross Revenues for 2013, 9.5% for 2014, 10.0% for 2015, 10.5% for 2016, and 11.0% for 2017.

A. The 2012 Proceeding ^[1]

The following entities filed Petitions to Participate and were the only remaining, non-settling participants at the time of hearing: SoundExchange, Music Choice, and Sirius XM. On May 25, 2012, the participants submitted a stipulation in which they agreed to the proposed Section 112 license rates and terms.

On June 5, 2012, the remaining participants in the proceeding commenced the direct case relating to Section 114 rates and terms. The Judges heard the rebuttal case beginning August 13, 2012. All parties presented evidence in the form of written testimony, live testimony, documentary evidence,^[2] and oral argument by counsel. Participants also designated background testimony from the last rate determination relating to SDARS and PSS. The parties submitted written proposed Findings of Fact and Conclusions of Law and responses to the same. On October 16, 2012, all parties presented closing argument. In all, the Judges heard evidence and oral argument for a period of 19 days. The parties presented 32 fact and expert witnesses.

The Judges make this Final Determination of Rates and Terms pursuant to 17 U.S.C. 803(c)(2) and 37 CFR Part 353. After evaluating the evidence to determine a range of reasonable royalty rates based on market benchmarks, the Judges subjected those presumed rates to the policy analysis required by 17 U.S.C. 801(b) of the Act.

On December 14, 2012, the Judges issued to the parties their Initial Determination. Pursuant to 17 U.S.C. 803(c)(2) and 37 CFR Part 353, SoundExchange and Sirius XM each filed a motion for rehearing. The Judges requested responses from the parties regarding each of the motions. Order Requesting Responses to Motions for Rehearing, Docket No. 2011-1 CRB PSS/Satellite II (Jan. 8, 2013). SoundExchange and Sirius XM each filed timely responses. After reviewing both motions and the responses thereto, the Judge denied both motions for rehearing. Order Denying Motions for Rehearing, Docket No. 2011-1 CRB PSS/Satellite II (Jan. 30, 2013). As explained in the January 30, 2013, Order, the Judges determined that none of the grounds set forth in the motions constituted the type of exceptional case—namely, (1) an intervening change in controlling law, (2) the availability of new evidence, or (3) a need to correct a clear error or prevent manifest injustice—warranting a rehearing. Id.

The Judges agreed with the parties, however, that clarification was needed in order to prevent “an unintended double exclusion” from Gross Revenues for the Direct License Share in § 382.12(d) and the Pre-1972 Recording Share in § 382.12(e). Id. at 5. After reviewing the respective proposals of SoundExchange and Sirius XM, the Judges adopted Sirius XM's proposal, finding that “Sirius XM's approach adequately addresses SoundExchange's double credit concern and in a way that may help to ensure a more accurate reflection of the legal status of the pre-1972 recordings with respect to the licenses at issue in this proceeding.” Id. Consequently, in this Final Determination, the Judges adopt Sirius XM's proposed language which will appear as § 382.12(d)(4): “No performance shall be credited as an Internet Performance of a Directly-Licensed Sound Recording under this section if that performance is separately credited as an Internet Performance of a Pre-1972 sound recording under paragraph (e)(1) of this section.”

B. Prior Proceedings

For the current licensing period, the Judges adopted agreed royalty rates for the PSS and made a determination of applicable rates for SDARS after a contested hearing. The Judges caused the prior SDARS determination [hereinafter SDARS-I] to be published in the Federal Register [hereinafter FR ] at 73 FR 4080 (Jan. 24, 2008).

The Judges' predecessors considered the reasonable rate standard and the Section 801(b)(1) policy factors in three prior proceedings: a Section 116 jukebox rate adjustment by the Copyright Royalty Tribunal (“Tribunal”); a Section 115 mechanical rate adjustment, also by the Tribunal; and a proceeding under the Copyright Arbitration Royalty Panel (“CARP”) system administered by the Librarian of Congress (“Librarian”) for preexisting subscription services under Section 114(f)(1)(B), the same section involved in this proceeding. Participants sought judicial review of all three prior determinations. A fuller history of prior proceedings and the outcomes and resolutions of those proceedings is included in SDARS-I. See 73 FR 4080, 4082-4085 (Jan. 24, 2008).

In Recording Indus. Ass'n of America v. Copyright Royalty Tribunal, 662 F.2d 1 (DC Cir. 1981), the U.S. Court of Appeals for the DC Circuit discussed its judicial review standard. The DC Circuit concluded that

To the extent that the statutory objectives [set forth in Section 801(b)] determine a range of reasonable royalty rates that would serve all these objectives adequately but to differing degrees, the Tribunal is free to choose among those rates, and courts are without authority to set aside the particular rate chosen by the Tribunal if it lies within a “zone of reasonableness.”

Id. at 9 (footnotes omitted).

In 1993, Congress replaced the Tribunal with the CARP system. In 1995, Congress passed the Digital Performance Right in Sound Recordings Act of 1995, creating the Section 114 digital performance right license that is the subject of this proceeding. The Copyright Royalty Distribution and Reform Act of 2004 established the Copyright Royalty Judges as a decision-making body in the Library of Congress. The Judges follow relevant precedent of the Tribunal and CARP system and strive to adopt reasonable royalty rates that satisfy the policy objectives set forth in Section 801(b). To determine rates, the Judges begin with an analysis of proposed market benchmarks, if any, and voluntary license agreements as described in Section 114(f)(1)(B), and the participants' supporting testimony. The Judges then measure the rate or range of rates that process yields against the statutory policy objectives to reach a determination of rates and terms.

II. The Standard for Determining Royalty Rates

Section 801(b)(1) of the Copyright Act provides that the Judges shall “make determinations and adjustments of reasonable terms and rates of royalty payments” for the statutory licenses set forth in Sections 114(f)(1) (“digital performance license”) and 112(e) (“ephemeral license”) of the Act. These licenses contain similarities and important differences in their standards for setting royalty rates. Both require the determination of reasonable rates and terms. The digital performance license requires that the rates (but not the terms) be calculated to achieve the following objectives:

• To maximize the availability of creative works to the public.
• 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.
• 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.
• To minimize any disruptive impact on the structure of the industries involved and on generally prevailing industry practices.

17 U.S.C. 801(b)(1).^[3]

The participants in this proceeding reached agreement on the rates and terms of the Section 112 license prior to the hearing; consequently, the focus of this determination is the application of the Section 801(b) factors to Section 114 rates. In SDARS-I, the Judges set forth in great detail the historical treatment of these factors by the Tribunal and the Librarian in his administration of the CARP system. See, SDARS-I, 73 FR at 4082-84. Consideration of this history produces the following approach.

[The Judges] shall adopt reasonable royalty rates that satisfy all of the objectives set forth in Section 801(b)(1)(A)-(D). In doing so, [they will] begin with a consideration and analysis of the [market] benchmarks and testimony submitted by the parties, and then measure the rate or rates yielded by that process against the [Section 801(b)] statutory objectives to reach [a] decision * * *.

The issue at hand is whether these policy objectives weigh in favor of divergence from the results indicated by the benchmark marketplace evidence.

Id. at 4084, 4094 (citations omitted).^[4]

In this proceeding, Music Choice argues that the Judges must consider an additional factor, applicable only to the PSS rate. Music Choice parses the Librarian's PSS determination [hereinafter, PSS-I], 63 FR 25394 (May 8, 1998), and Section 803(a)(1) to conclude that the Judge's benchmark analysis must begin with the current royalty fees paid by Music Choice to the performing rights societies (ASCAP, BMI and SESAC) for musical works. Music Choice contends that the Librarian's use of the musical works benchmark in 1998 mandates that the Judges must use that same benchmark in these proceedings in the absence of a better, comparable benchmark. See Music Choice PCL ¶ 53.

The Judges reject Music Choice's argument for several reasons. First, Music Choice does not, and cannot, cite any statutory license rate proceeding in which the adjudicator found that factual marketplace observations in a particular royalty rate proceeding must be given a priori consideration in a subsequent proceeding. Second, in the PSS-I decision, the Librarian did not rely solely upon the musical works benchmark, but instead relied upon some unspecified combination of factors. See PSS-I, 63 FR at 25410. Even if the Judges were inclined to accord some precedential value to the musical works benchmark from PSS-I in this proceeding, the Judges cannot discern the degree to which that benchmark influenced or altered the Librarian's decision.

Third, Music Choice fails to place the PSS-I decision in its historical context. 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.

Therefore, in this proceeding, the Judges consider the musical works evidence offered by Music Choice not as binding precedent but as evidence offered in the normal course, along with all other current evidence, not giving the musical works benchmark any preference as a starting point for, default position in, or other limitation on a proper evaluation of all of the benchmark evidence.

III. Determination of the Royalty Rates

The Judges have considered carefully the relevant law and the evidence received in this proceeding. Based upon that evidence and law, and for the reasons detailed in the following discussion, the Judges have determined applicable royalty rates for the licensing period January 1, 2013, through December 31, 2017, for the only existing SDARS, Sirius XM, and for the PSS.^[5]

IV. The Section 112 Ephemeral License

With respect to the Section 112(e) ephemeral license, the parties submitted a joint stipulation. SoundExchange and Music Choice ask for continued application of the language of 37 CFR 382.2(c), which requires a minimum fee advance payment of $100,000 per year, payable no later than January 20 of each year, with royalties accruing during the year recoupable against the advance. Joint Stipulation at 2-3 (May 25, 2012). SoundExchange and Sirius XM ask that the same minimum fee proposal apply to Sirius XM. Id.

All parties agree that the value accorded the Section 112 license is combined with that of the Section 114 license and that the value is allocated 5% to the Section 112 license and 95% to the Section 114 license, consistent with the current regulations applicable to webcasters, broadcasters, SDARS, and new subscription services. See 37 CFR 380.3, 380.12, 380.22; 382.12; and 383.3. The parties submitted no other evidence on either the minimum fee or the Section 112(e) license fee allocation; consequently, the Judges approve and adopt the respective minimum fees and Section 112(e) royalty rates for PSS and SDARS as set forth in the Joint Stipulation.

V. The Section 114 Digital Performance License

With respect to the royalty rates for the Section 114 digital performance license, Music Choice requests a rate of 2.6% of Gross Revenues, applicable to each of the years in the licensing period. SoundExchange requests the following percentage of Gross Revenues rates for the PSS: 15% for 2013; 20% for 2014; 25% for 2015; 35% for 2016; and 45% for 2017. Second Revised Proposed Rates and Terms of SoundExchange, Inc., at 6 (Sept. 26, 2012). Both SoundExchange and Music Choice ask that the definition of Gross Revenues, currently set forth in 37 CFR 382.2(e), apply in the new licensing period.

A. Section 114 Rates for PSS

Since 1998, when the decision in PSS-I established the initial royalty rates, the PSS have paid a fee based on a percentage of Gross Revenues ^[6] as defined by regulation. Neither Music Choice nor SoundExchange proposes altering this rate structure for the 2013-17 license term, nor do they propose changes to the Gross Revenues definition. As discussed in detail below, however, SoundExchange requests that the Judges add an adjustment to the percentage-of-revenue metric to address what it perceives as a deliberate reduction in revenues paid to Music Choice for its residential audio service by certain cable operators that are co-owners (partners) of Music Choice.

The rates the Judges establish under Section 114(f)(1) for the digital performance of sound recordings must be calculated to achieve the objectives set forth in Section 801(b)(1)(A) through (D) of the Act. Where the determination standard is reasonable rates calculated to achieve the Section 801(b)(1) factors, the Judges have found market benchmarks, if any, to be a useful starting point. See SDARS-I, 73 FR at 4088; Phonorecords I, 74 FR 4510, 4517 (Jan. 26, 2009). As discussed below, the parties disagree about what constitutes the most appropriate benchmark to guide the Judges in determining a reasonable rate.

1. Music Choice's Proposed Musical Works Benchmark for PSS Rates

As discussed above, Music Choice argues that the annual royalties it pays to the three performing rights societies (ASCAP, BMI, and SESAC) for the right to perform musical works to subscribers of its residential audio service is, by virtue of the Librarian's determination in PSS-I, a precedential benchmark in this proceeding. Although the Judges reject the PSS-I benchmark as a precedent in this proceeding, they nevertheless weigh whether the rates are a useful benchmark in this proceeding.

Music Choice represents that it pays ASCAP and BMI each 2.5% of gross revenues attributable to residential service each and pays an annual flat fee to SESAC that amounts to approximately [REDACTED] of net revenue, for a total of [REDACTED].^[7] Del Beccaro Corrected WDT at 21-22, MC 17, MC 18 and MC 19, PSS Trial Ex. 1. Music Choice submits that this rate (i.e., [REDACTED]) represents the upper bound of a reasonable royalty rate for the Section 114 and Section 112 licenses.

Two pieces of evidence, in Music Choice's view, corroborate use of musical works licensing rates as a benchmark. First, Music Choice observes equivalence between the fees for the performance of sound recordings and musical works in Canada and the United Kingdom. Music Choice cites four decisions of the Canadian Copyright Board, involving licensing fees for commercial radio, cable television, satellite music services and radio services of the Canadian Broadcasting Corporation (“CBC”). According to Music Choice, in those decisions, the Canadian Board found that royalty rates for sound recordings and musical compositions have equivalent value. See, e.g., Del Beccaro Corrected WDT at MC 6 at 30-33 (commercial radio) and MC 7 at 14 (cable television), PSS Trial Ex. 1.^[8] Moreover, Music Choice represents that in the United Kingdom, sound recording royalty rates for commercial broadcasting services are less than those for musical works. Id. at 19. According to Music Choice, if Music Choice's service were transmitted through cable in the U.K., Music Choice would pay 5.25% of 85% of gross revenues for the musical works performance right, but would pay only 5% of 85% of gross revenues for the sound recording performance right. Id. at 19, MC 11. Music Choice represents that the U.K. Copyright Tribunal has found the same equivalence. Id. at 19-20 & MC 12, ¶ 53.

Music Choice further asserts that the validity of the proposed musical works benchmark to set rates in this proceeding is corroborated by an economic model called the Asymmetric Nash Bargaining Framework (“Nash Framework”) offered by Dr. Crawford.^[9] Dr. Crawford uses the Nash Framework to determine potential outcomes that could occur in hypothetical negotiations between record labels and PSS providers. Crawford Corrected WDT at 12, PSS Trial Ex. 4. According to Dr. Crawford, as a non-cooperative bargaining model, the Nash Framework is designed to yield predictions about how outcomes are determined when firms negotiate; that is, how two firms would split the surplus of their interaction (i.e., revenues minus costs) in a hypothetical negotiation. Id. at 16. Three factors (the Nash factors) are analyzed to determine the split: (1) The combined agreement surplus; ^[10] (2) each firm's “threat point”; ^[11] and (3) each firm's bargaining power. Id.

Dr. Crawford's stated goal in applying the Nash Framework was to first establish the Nash factors for the hypothetical market (the sale of rights between one record company and one PSS provider) and compare them to the Nash factors in the actual musical works market (the sale of rights between the three performing rights societies and one PSS provider). Id. at 18. Dr. Crawford determined that the combined agreement surplus in the hypothetical PSS market was the total profits that the PSS provider earned before paying the royalty for digital performance rights. Id. at 45.

Dr. Crawford determined that in the hypothetical market, the threat point for a PSS provider would be zero because, in the absence of an agreement, the PSS provider could not offer music and therefore could not earn a surplus. Id. at 19. He determined, however, that the threat point for a hypothetical record company would be negative because the failure to reach an agreement would have negative implications for the record company in other, non-PSS markets. Specifically, a record company's failure to reach an agreement with a PSS provider could diminish that record company's sales of compact disks, because, according to Music Choice, there is a significant promotional benefit to the record company from the PSS.^{[12] [,]  [13]} Id.

With respect to the last Nash factor, bargaining power, Dr. Crawford assumed it to be equal based on his assessment of Music Choice's existing technology platform and contract, which, he contended, cannot be easily replaced or replicated, and his observations of Music Choice's bargaining efforts for sound recording performance rights with respect to music videos. Id. at 15, 22.

Applying the Nash factors to the existing market for the PSS musical works performance right, Dr. Crawford determined that the threat point for a PSS provider would again be zero. He determined that the “threat point” for the performing rights society would be negative due to the predicted loss of promotional value from the PSS. Id. at 28. Dr. Crawford again assumes equal bargaining power between the PSS provider and the performing rights society, based largely upon his observations that the two possess equal patience in their negotiations. Id. at 29. Dr. Crawford opines that his analysis of the Nash factors in the hypothetical musical works market indicates that there should be a 50/50 split of the surplus in hypothetical negotiations between the PSS and the performing rights society, the same conclusion he reached with respect to the hypothetical market for the rights in this proceeding. Because of the similarities between the Nash factors in the PSS hypothetical market and the market for musical works, Dr. Crawford concludes that the musical works market makes for a good benchmark for the hypothetical sound recording performance right market at issue in this proceeding. Id. at 30.

Dr. Crawford also proffered a surplus splitting analysis which he asserted helped to corroborate the reasonableness of Music Choice's rate proposal. This analysis involved Music Choice's own operating profits to estimate how PSS profits would be allocated between a Licensee and Licensor in the PSS market. Crawford Corrected WDT at 43, PSS Trial Ex. 4. Dr. Crawford adjusted Music Choice's 2006-2010 operating profit to remove the actual royalty paid by Music Choice for sound recording performance rights. He then applied the capital asset pricing model ^[14] to derive an expected rate of return on Music Choice's assets. He determined that that hypothetical rate of return would be 8.33%. Id. at Appendix B.3, B.4. He then multiplied the 8.33% rate by Music Choice's average operating profits to determine cost of capital. He then subtracted cost of capital from the royalty-adjusted operating profits to derive the residual profits for each year. Id. at 47. According to Dr. Crawford, this calculation showed that Music Choice's cumulative returns in excess of its cost of capital, but before payment of sound recording royalties, would amount to 3.05% of Music Choice's 2006-2010 royalties. Id. He then applied a range of allocations for the hypothetical cumulative returns of between 20% and 80%. This calculation yielded a range of royalties from 0.61% to 2.43%. Id. at 48.

2. SoundExchange's Proposed Marketplace Agreements Benchmark for PSS Rates

In an effort to frame a zone of reasonable rates, SoundExchange offers rates from over 2,000 marketplace agreements, representing a variety of rights licensed. SoundExchange witness, Dr. George Ford, observes that PSS like Music Choice have certain distinctive features that make it difficult to identify a suitable benchmark market. 6/18/12 Tr. 2814:9-20 (Ford). First, according to Dr. Ford, Music Choice does not sell its service directly to subscribers, but rather to cable television operators who then bundle the Music Choice programming with a package of television programming for ultimate sale to subscribers. Music Choice is, therefore, an intermediary between cable operators and their subscribers, unlike any of the digital music services the Copyright Royalty Judges have previously dealt with. Ford Second Corrected WDT at 12-13, SX Trial Ex. 79; 6/18/12 Tr. 2810:20-2811:4 (Ford). Second, Music Choice's service is almost always bundled with a hundred or more channels of video and is almost never sold on a stand-alone basis. Ford Second Corrected WDT at 13, SX Trial Ex. 79. According to Dr. Ford, this bundling makes it difficult to determine the specific consumer value for Music Choice's programming alone. Id.

Given these difficulties, Dr. Ford uses an all-inclusive approach of examining royalty rates for different digital music markets: portable and non-portable interactive subscription webcasting, cellular ringtones/ringbacks, and digital downloads. Id. at 15-16, Table 1. According to Dr. Ford, most of the over 2,000 licensing agreements he examined across these markets calculate royalties based on a “greater of” (sic.) methodology that includes a per-play royalty fee, a per-subscriber fee, and a revenue-based fee. Id. at 13 n.21. Dr. Ford analyzed only the revenue-based fees. He contended that doing so makes his results conservative because either of the other two payment metrics could, under certain circumstances, result in a larger total royalty fee than the revenue-based calculation. 6/18/12 Tr. 2861:3-13 (Ford). According to Dr. Ford, his analysis of the agreements showed a percentage-of-revenue rate of 70% for digital downloads, 43% to 50% for ringtones/ringbacks, and 50% to 60% for portable and non-portable interactive subscription webcasting, respectively. Ford Second Corrected WDT at 15-16, Table 1, SX Trial Ex. 79. According to Dr. Ford, SoundExchange's rate proposal for PSS comports well with the range established by these agreements, in that it rises above the lowest average rate (43%) only in the last year of the licensing term. Therefore, according to Dr. Ford, SoundExchange's proposal can “be presumed to be a reasonable proxy for a market outcome.” Id. at 16; see also 6/18/12 Tr. 2831:8-15 (Ford).

3. Analysis and Conclusions Regarding the Proposed Rate Guidance

Based upon the evidence put forward in this proceeding, the Judges conclude that neither Music Choice's nor SoundExchange's proffered rate guidance provides a satisfactory benchmark upon which they can rely to determine the sound recording performance royalty rates for the PSS for the upcoming license period. The parties' proposals are so far apart, and both so far from the current rate, that they cannot even be said to describe a “zone of reasonableness.” The only remaining guidance the Judges have upon which to base the new rates is the current royalty rate of 7.5% of PSS Gross Revenues. This rate approximates the middle of the wide spectrum proposed by the parties. It is the rate against which the Judges will test the Section 801(b) policy factors.

a. Music Choice's Proposed Musical Works Guidance

Having rejected Music Choice's argument that the musical works benchmark utilized by the Librarian of Congress in PSS-I is binding precedent in this proceeding,^[15] the Judges examine the proposed benchmark on its own merits and find it lacks comparability to the target market. Dr. Crawford, who advocates the appropriateness of the musical works rates as a benchmark for the PSS rates, acknowledges that a benchmark market should involve the same buyers and sellers for the same rights. Crawford Corrected WDT at 24, PSS Trial Ex. 4. However, the musical works market involves different sellers (performing rights societies versus record companies) selling different rights. See SDARS-I, 73 FR at 4089. The fact that a PSS needs performing rights to musical works and sound recordings to operate its service does not make the rights equivalent, nor does it say anything about the relative values of those rights.^[16]

Music Choice's reliance on foreign rates to support its proffer of the musical works guidance is unpersuasive. The Judges have considered before the significance of foreign countries' treatment of the licensing of exclusive rights granted by copyright. In the proceeding to set rates and terms for the compulsory license to reproduce musical compositions under Section 115 of the Copyright Act, certain participants offered evidence of license rates in the U.K., Canada and Japan. See Phonorecords I, 74 FR 4510, 4521 (Jan. 26, 2009). In rejecting the foreign rates as comparable benchmarks, the Judges stated 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.” Id. at 4522. Neither Mr. Del Beccaro nor Dr. Crawford even attempts an analysis or discussion of the intricacies of Canadian and U.K. markets for performance rights for musical works and sound recordings, and Music Choice itself concedes that particular license rates in Canada and Europe “do not necessarily determine what the specific market rate in the United States should be for the sound recording right.” Music Choice PFF ¶ 135.

Likewise, the Judges are not persuaded that Dr. Crawford's application of the Nash Framework provides corroboration. The Nash Framework is a theoretical concept whose goal is to evaluate how the surplus from a hypothetical transaction might be divided between negotiating parties. Even assuming that the Nash Framework has predictive value in some real-world contexts, Music Choice provided no data to support the theoretical approximations in the market for any intellectual property rights, much less those that the Judges are charged with evaluating. Therefore, the Judges find that the Nash Framework is not useful corroborating evidence.^[17]

b. SoundExchange's Marketplace Agreements Guidance

The Judges do not endorse the music service benchmarks offered by SoundExchange and supported by Dr. Ford as persuasive benchmarks. Typically the volume (over 2,000) of marketplace agreements that Dr. Ford examined for music products and services would be a sufficiently deep sample set to provide a useful framework for a marketplace benchmark. The four markets Dr. Ford examined, however—portable and non-portable subscription interactive webcasting, ringtones/ringbacks, and digital downloads—involve the licensing of products and rights separate and apart from the right to publicly perform sound recordings in the context of this proceeding. The buyers are different from the target PSS market; thus, the key characteristic of a good benchmark—comparability—is not present.

The Judges agree with Dr. Ford's observations that Music Choice has several distinct features, such as its intermediary role between cable systems and subscribers and the bundling of Music Choice's services with multiple channels of video and other non-music programming, which significantly dim the possibility of market comparators. In the absence of some rational, reasoned adjustment to make the music agreements data more comparable to the PSS market, the Judges find its probative value in this proceeding of only marginal value.

c. The Prevailing Statutory Rate

The Judges are left, therefore, with a consideration of the existing 7.5% royalty rate which is the product of settlement negotiations that occurred in SDARS-I between Music Choice and SoundExchange but is a rate for which neither party advocates. Although it is a rate that was negotiated in the shadow of the statutory licensing system and cannot properly be said to be a market benchmark rate, nothing in the record persuades the Judges that 7.5% of Gross Revenues, as currently defined, is too high, too low or otherwise inappropriate. Accord, Phonorecords I, 74 FR at 4522.

1. Application of Section 801(b) Factors

Based on the record evidence in this proceeding, the Judges have determined that the benchmark evidence submitted by Music Choice and SoundExchange has failed to provide the means for determining a reasonable rate for the PSS, other than, perhaps to indicate the extreme ends of the range of reasonable rates. The testimony and argument of Music Choice demonstrates nothing more than to show that a reasonable rate cannot be as low as the rates (i.e., [REDACTED] of Music Choice's revenues) paid by Music Choice to the three performing rights societies for the public performance of musical works. The benchmark testimony of SoundExchange is of even lesser value. The proposed rate of 15% for the PSS for the first year of the licensing period, deemed reasonable by Dr. Ford (at least in the beginning of the licensing period), stands as the upper bound of the range of reasonable rates. Within that range is the current 7.5% rate. On the record before us, the Judges are persuaded that the current rate is neither too high, too low, nor otherwise inappropriate, subject to consideration of the Section 801(b) factors discussed below.

a. Maximize Availability of Creative Works

To argue for an adjustment in its favor under the first Section 801(b) factor, Music Choice touts that it is a music service that is available in over 54 million homes, with 40 million customers using the service every month. 8/16/12 Tr. 3878:3 (Del Beccaro); 6/11/12 Tr. 1462:5-11, 1486:19-1487:2 (Del Beccaro). According to Music Choice, channel offerings have increased through the years, and they are curated by experts in a variety of music genres. Del Beccaro Corrected WDT at 3, 24, PSS Trial Ex. 1. Music Choice also highlights recent developments in technology that enable Music Choice to display original on-screen content identifying pertinent information regarding the songs and artists being performed. Id. at 24, MC 23; Williams WDT at 12, PSS Trial Ex. 3; 6/11/12 Tr. 1461:14-1462:1, 1491:2-12 (Del Beccaro). According to Music Choice, these elements, along with certain promotional efforts that Music Choice makes on behalf of artists, support a downward adjustment in the rates. In any event, an upward adjustment in the rates, argues Music Choice, would not affect the record companies' bottom-line because PSS royalties are not a material revenue source for record companies. Music Choice PFF ¶¶ 409-417.

SoundExchange submits that a market rate incorporates considerations under the first Section 801(b) factor, citing the decision in SDARS-I, and that if PSS rates turn out to be too high and drive Music Choice from the market, presumably consumers will shift to alternative providers of digital music where higher royalty payments are more likely for record companies. Ford Second Corrected WDT at 19-21, SX Trial Ex. 79.

The current PSS rate is not a market rate, so market forces cannot be presumed to determine the maximum amount of product availability consistent with the efficient use of resources. See SDARS-I, 73 FR 4094. However, the testimony demonstrates that Music Choice has not, under the current rate, reduced its music offerings or contemplated exiting the business; in fact, it will be expanding its channel offerings in the near term. Del Becarro Corrected WDT at 3, 24, PSS Trial Ex. 1; see also 6/11/12 Tr. 1460:21-1461:1 (Del Beccaro). The Judges find no creditable evidence in the record to suggest that the output of music from record labels has been impacted negatively as a result of the current rate. The record shows no persuasive evidence that a higher PSS royalty rate would necessarily result in increased output of music by the record companies, nor that a lower rate would necessarily further stimulate Music Choice's current and planned offerings. In sum, the policy goal of maximizing creative works to the public is reasonably reflected in the current rate and, therefore, no adjustment is necessary.

b. Afford Fair Return/Fair Income Under Existing Market Conditions

Music Choice submits that the Judges need not worry about the impact of a low royalty rate on the fair return to record companies and artists for use of their works because royalties from the PSS market are so small as to be virtually inconsequential to companies whose principal business is the sale of CDs and digital downloads. Music Choice PFF ¶¶ 420-430. With respect to Music Choice's ability to earn a fair income, however, Music Choice argues that it is not profitable under the current 7.5% rate. Mr. Del Beccaro testified that its average revenue per customer for its residential audio business has been on the decline since the early 1990s, down from $1.00 per customer/per month to [REDACTED] per customer/per month currently. Del Beccaro Corrected WDT at 40, PSS Trial Ex. 1. He further testified that after 15 years of paying a PSS statutory rate between 6.5% and 7.5% Music Choice has not become profitable on a cumulative basis and is not projected to become so within the foreseeable future. Id. at 42. Music Choice represents that it has a cumulative loss at the end of 2011 of [REDACTED], projected to grow to [REDACTED] in 2012 and continue to increase throughout the 2013-17 license period. Del Beccaro Corrected WRT at MC 69 at 1 and MC 70 at 1, PSS Trial Ex. 21. These losses lead Music Choice to conclude that it has not generated a reasonable return on capital under the existing rates. Music Choice PFF ¶¶ 442-43.

Music Choice's claims of unprofitability under the existing PSS rate come from the oblique presentation of its financial data and a combining of revenues and expenses from other aspects of its business. The appropriate business to analyze for purposes of this proceeding is the residential audio service offered by Music Choice, the subject of the Section 114 license. Music Choice, however, reports costs and revenues for its residential audio business with those of its commercial business, which is not subject to the statutory license. This aggregation of the data, which Music Choice acknowledges cannot be disaggregated, see 6/11/12 Tr. 1572:3-1576:2 (Del Beccaro), masks the financial performance of the PSS business. As a consolidated business, Music Choice has had significantly positive operating income between 2007 and 2011 and made profit distributions to its partners since 2009. Ford Amended/Corrected WRT at SX Ex. 362-RR, p. 3 (PSS_002739), SX Trial Ex. 244; SX Trial Ex. 64 at 3 (PSS_002715); SX Trial Ex. 233 at 3 (PSS_366020). Dr. Crawford's effort to extract costs and revenues from this data for the PSS service alone for use in his surplus analysis cannot be credited because of his lack of familiarity with the data's source. 6/13/12 Tr. 1890:15-1891:10 (Crawford).^[18] The Judges find no persuasive evidence to suggest that Music Choice has not operated successfully and received a fair income under the existing statutory rate.^[19]

With respect to fair return to the copyright owner, the Judges' examination is whether the existing statutory rate has produced a fair return with respect to the usage of sound recordings. During the current licensing period, Music Choice provided 46 channels of music programming. Music Choice plans to expand the number of music channels it provides dramatically in the coming licensing term, however, up to 300 channels by the first quarter of 2013. Del Beccaro Corrected WDT at 3-4, PSS Trial Ex. 1; 6/11/12 Tr. 1490:8-16 (Del Beccaro). This expansion will result in a substantial increase in the number of plays of music by Music Choice, even if the ultimate listenership intensity of its licensees' subscribers cannot be measured. Music Choice provided no evidence, however, to suggest that the planned expansion in usage would result in increased revenues to which the statutory royalty rate is to be applied. Indeed, Music Choice has declared itself to be in a mature market with no expectation of increasing profits. 8/16/12 Tr. 3855:17-3856:7 (Del Beccaro).

Music Choice presented no evidence to suggest that copyright owners would be compensated for the increased usage of their works. Dramatically expanded usage without a corresponding expectation of increased compensation suggests an upward adjustment to the existing statutory rate is warranted. Measurement of the adjustment is not without difficulty because any downstream increases in listenership of subscribers as a result of additional music offerings by Music Choice cannot be readily predicted. It is possible that listenership overall may remain constant despite the availability of several additional music channels. It is more likely, however, that Music Choice would not make the expansion, and incur the additional expense of doing so, without reasonable expectation that subscribers or advertisers would be more attracted to the expanded offerings, although the Judges have no evidence to suggest that the net increase in listenership (or advertising revenue) would be anything more than modest.

SoundExchange refers to prior rate decisions and the application of the fair return/fair income factor by the Judges and their predecessors. SoundExchange asserts that the Judges are looking for a fair return/fair income result that is consistent with reasonable market incomes. SX PFF at ¶ 491, citing SDARS-1, 73 F.R. 4080, 4095 (Jan. 24, 2008). Referring to testimony by Messrs. Ciongoli and Van Arman, SoundExchange emphasizes how vital statutory royalty income is to copyright owners—both the record labels and the artists, whose share SoundExchange distributes directly. See 6/13/12 Tr. 2138:5-2142:9 (Ciongoli), Van Arman WDT at 4, SX Trial Ex. 77. Although the income from any one statutory license may not be great, SoundExchange cites the aggregate value of income from all of the statutory licenses as vital to the industry. With respect to fair income to the rights user, SoundExchange points to the profit on the consolidated financial statements of Music Choice over the past five years, 2007-2011.

The balance of fair return and fair income appears to have been maintained at the current PSS rates. This factor does not argue in favor of drastic cuts or increases in the current rate. Music Choice's planned increase in usage, however, argues in favor of an increase in the rates going forward to fairly compensate the licensors for the additional performances.

The Judges determine, therefore, that a 1% upward adjustment of the benchmark (from 7.5% to 8.5% of Gross Revenues), phased in during the early part of the licensing period, is appropriate to serve the policy of fair return/fair income.

c. Weigh the Relative Roles of Copyright Owners and Copyright Users

This policy factor requires that the rates the Judges adopt reflect the relative roles of the copyright owners and copyright users in the product made available with respect to relative creative contribution, technological contribution, capital investment, cost, risk, and contribution to the opening of markets for creative expression and media for their communication. Music Choice argues that its creative and technological contributions, and capital investments, outweigh those of the record companies. First, Music Choice touts the graphic and informational improvements made to its on-screen channels, noting that what were once blank screens now display significant artist and music information. According to Music Choice, costs for these improvements have exceeded [REDACTED]. Del Beccaro Corrected WDT at 31-32, PSS Trial Ex. 1. Second, Music Choice offers increases in programming, staff size and facilities, along with enhancements to product development and infrastructure. Music Choice estimates that costs for these improvements have exceeded [REDACTED]. Id. Regarding costs and risks, Music Choice points to its lack of profitability and the exit of other PSS from the market as evidence of its continued risk and limited opportunity for profit. Music Choice PFF ¶¶ 512-520. Finally, with respect to opening new markets, Music Choice touts the PSS market itself for which it remains the standard-bearer in disseminating music to the public through cable television. Id. at ¶ 523.

SoundExchange offers little more on the third Section 801(b) factor beyond Dr. Ford's contention that he saw no evidence to support that Music Choice makes contributions to creativity or availability of music that are beyond those of the music services he included in his benchmarks, and therefore, according to Dr. Ford, the third factor is accounted for in the market. Ford Second Corrected WDT at 21, SX Trial Ex. 79; 6/18/12 Tr. 2849:10-16 (Ford).

In considering the third factor, the Judges' task is not to determine who individually bears the greater risk, incurs the higher cost or makes a greater contribution in the PSS market, and then make individual up or down adjustments to the selected rate based upon some unspecified quantification. Rather, the consideration is whether these elements, taken as a whole, require adjustment to the Judges' selected benchmark rate of 7.5%. Upon careful weighing of the evidence, the Judges determine that no adjustment is necessary. Music Choice's investments in programming offerings, staff, and facilities, and other related products and services are no doubt impressive, but they have been accomplished under the current rate. As discussed above, Music Choice has already begun to expand its channel offerings and has allocated greater financial resources to its residential audio business. All of these undertakings, plus the investments made and costs incurred to date have been made under the existing rate, and the Judges have no persuasive evidence to suggest that these contributions have not been accounted for in the current rate. On the other side of the ledger, SoundExchange has not offered any persuasive evidence that the existing rate has prevented the music industry from making significant contributions to or investments in the PSS market or that those contributions are not already accounted for in the current rate. Therefore, no adjustment is warranted under this factor.

d. Minimize Disruptive Impact

Of the four Section 801(b) factors, the parties devoted most of their attention to the last one: Minimizing disruption on the structure of the industries and on generally prevailing industry practices. This is perhaps not surprising, given the role this factor played in SDARS-I in adjusting the benchmark rates upon which the Judges relied to set the royalty fees. See SDARS-I, 73 FR at 4097-98. Because the Judges have identified as reasonable the rate for PSS currently in place, the Judges' analysis of the disruption factor is confined to that rate.

SoundExchange argues that the current rate is disruptive to the music industry. Dr. Ford testified that “the current practice of applying an exceedingly low rate to deflated revenues is disruptive of industry structure, especially where there are identical services already paying a higher rate.” Ford Second Corrected WDT at 23, SX Trial Ex. 79. This results, according to Dr. Ford, in a tilting of the competitive field for music services in favor of Music Choice, thereby disrupting the natural evolution of the music delivery industry. Dr. Ford, however, concedes that the PSS market has unique and distinctive features that distinguish it from other types of music services, thereby substantially reducing the likelihood that the PSS and other music services would be viewed as substitutes for one another. Further, Dr. Ford failed to present any empirical evidence demonstrating a likelihood of migration of customers from music services paying higher royalty fees to the PSS as a result of his perceived royalty imbalance. Dr. Ford's conclusion that the current rate paid by the PSS for the Section 114 license has caused a disruption to the music industry (or would likely do so in the upcoming license period) is mere conjecture.

Music Choice also contends that the current rate is disruptive. The Judges find its argument weak and unsubstantiated. The test for determining disruption to an industry, announced by the Judges in SDARS-I, is whether the selected rate directly produces an adverse impact that is substantial, immediate, and irreversible in the short-run. SDARS-I, 73 FR at 4097. The current rate has been in place for some time and, despite Music Choice's protestations that it has never been profitable, it continues to operate and continues to increase its expenditures by expanding and enhancing its services in the face of the supposedly disruptive current royalty rate. Music Choice's argument that DMX's bankruptcy and Muzak's decision to limit its participation in the PSS market are evidence of the onerous burden of the current rate are without support. Music Choice has failed to put forward any evidence demonstrating a causal relationship between the actions of those services and the current PSS royalty rate. In sum, the Judges are not persuaded by the record testimony or the arguments of the parties that the current PSS rate is disruptive to a degree that would warrant an adjustment, either up or down.

2. The Judges' Rate Determination for PSS

In light of the Judges' analysis of the Section 801(b) factors, the Judges set forth the following PSS rates: for 2013: 8.0%; for 2014: 8.5%; for 2015: 8.5%; for 2016: 8.5%; and for 2017: 8.5%.

The Judges have chosen to phase-in the increase over the first two years of the license period to moderate any potential negative impact the rate increase might have on the PSS. Should Music Choice alter its anticipated usage under the statutory license in the future, such evidence can be taken into account in a future rate proceeding; however, the Judges received no evidence that suggests that Music Choice's channel line-up, once expanded in 2013, will shrink considerably during the license period.

In addition to proposing rates, SoundExchange raises an additional matter. Though not technically a rate, nor strictly an amendment of the Gross Revenues definition as it applies to PSS, SoundExchange requests a means for capturing revenues from cable systems that are owners of equity or capital interests in Music Choice who do not engage in arm's length transactions with Music Choice for its product offerings. Second Revised Proposed Rates and Terms of SoundExchange, Inc., at 6-7 (Sept. 26, 2012). Put another way, SoundExchange seeks to capture any price breaks that Music Choice offers its affiliates for the Music Choice service.

The proposed price adjustment for affiliated cable systems would be calculated by multiplying the total number of subscribers for the month for each affiliated cable system by the average per-subscriber royalty payment of the five largest paying unaffiliated cable systems that provide the Music Choice service. These adjustments would then be added to Music Choice's Gross Revenues. In support of its “Non Arm's Length Transaction” adjustment for affiliated cable systems, Dr. Ford testified that a straight percentage-of-revenue metric would not adequately account for the situation where Music Choice offers per-subscriber rate discounts to its cable partners. 8/20/12 Tr. 4216:21-4217:8 (Ford). According to Dr. Ford, over half of Music Choice's non-partner cable systems pay approximately [REDACTED] per subscriber per month in licensing fees to Music Choice, whereas the partner cable systems pay only [REDACTED] per subscriber per month. Ford Amended/Corrected WRT at 5, SX Trial Ex. 244.

The Judges are not persuaded that a “Non Arm's Length Transaction” adjustment is warranted. It is not surprising that the affiliated cable operators, which in most instances have more subscribers than the non-affiliated systems, would be able to negotiate lower per-subscriber licensing fees due to their ability to deliver more subscribers to the service. Therefore, the differences in subscriber fees between affiliates and non-affiliates could be unrelated to the operator's status vis-à-vis Music Choice. Further, the affiliated cable systems represent a third of Music Choice ownership whereas Music Choice's record company partners own one quarter of the company. 6/11/12 Tr. 1454:16-22 (Del Beccaro). Therefore, it is not unreasonable to assume that the record label owners would serve as a counterweight to the affiliated cable systems. Therefore, the Judges conclude that on the current record, any influence on subscriber rates from the competing stakeholders of Music Choice, if any, would be a wash.

B. Section 114 Royalty Rates for SDARS

SoundExchange proposes the following percentage of revenue rates for SDARS: 12% for 2013; 14% for 2014; 16% for 2015; 18% for 2016; and 20% for 2017. Second Revised Rates and Terms of SoundExchange, Inc., at 2 (Sept. 26, 2012). Sirius XM counters with a proposed royalty rate in the range of 5% to 7% of Sirius XM's monthly U.S. gross revenues. Proposed Rates and Terms of Sirius XM Radio, Inc., at 4 (Sept. 26, 2012).

1. Sirius XM's Proposal

a. Direct License Benchmark

Beginning in 2010, Sirius XM commenced a coordinated effort to negotiate sound recording performance rights directly with individual record labels. Sirius XM first attempted to engage the four major record companies in discussions but was unsuccessful. Id.; 6/7/12 Tr. 669:8-672:9, 713:3-11, 714:11-715:4 (Frear); 6/11/12 Tr. 1347:7-21, 1348:20-1349:4 (Karmazin). Sirius XM then enlisted Music Reports, Inc. (“MRI”) to formulate and execute a direct licensing strategy with as many independent record labels as possible. Together, Sirius XM and MRI developed the terms and conditions of a template Direct License, key provisions of which include:

• A pro rata share of 5%, 6%, or 7% of gross revenues, defined by reference to 37 CFR 382.11;
• A grant of rights to Sirius XM to operate all of its various services (satellite radio plus other services such as webcasting);
• “Additional functionality” granted to Sirius XM, including elimination of the Section 114 license sound recording performance complement, which allows Sirius XM to play more music from a particular artist in a given period of time;
• Direct, quarterly payment of 100% of the royalties to the record label;
• Payment of advances to the 5 largest record labels; and
• The possibility, but not the promise, of increased play on Sirius XM's music services.

Gertz Corrected WDT at 8-11, SXM Dir. Trial Ex. 14; Gertz Revised WRT at 2, SXM Reb. Trial Ex. 8; 6/8/2012 Tr. 986:20-987:5 (Blatter). Sirius XM executed the first Direct Licenses in August of 2011 and by the time of the closing of testimony in this proceeding, Sirius XM had Direct Licenses with 95 independent record labels that set a royalty rate of between 5% and 7% of gross revenues, depending on the particular agreement. 8/13/12 Tr. 3015:16-20 (Frear); 8/15/12 Tr. 3679:22-3680:1 (Gertz).

b. The Noll Analysis

Sirius XM's expert economist, Dr. Roger Noll, contends that the 95 Direct Licenses are the best benchmark for SDARS rate setting in this proceeding because, unlike in SDARS-I, the Judges now have direct evidence of competitively negotiated marketplace rates for the exact service at issue in this proceeding. Noll Revised Amended WDT at 7, 11, 33-36, SXM Dir. Trial Ex. 1. Dr. Noll testified that the Direct Licenses are representative, for benchmarking purposes, of the types of sound recordings available across the industry, including those distributed by major record labels. Id. at 39-45; see also 6/5/12 Tr. 261:6-262:14 (Noll)(contending that the 95 Direct Licensors as a group offer a scope of sound recordings comparable to those not so licensed).

Dr. Michael Salinger, another Sirius XM expert economist, concludes that the fact that 95 record companies accepted the Direct License offer suggests that the current 8% statutory rate is, if anything, above the competitive rate for sound recordings. Salinger Corrected WRT at 13, SXM Reb. Trial Ex. 9. Further, Sirius XM argues that the number of Direct Licenses undoubtedly would have been higher but for the efforts of SoundExchange, the American Association of Independent Musicians and others to undermine and interfere with its Direct License Initiative.^[20] See, e.g., Sirius XM PFF¶¶ 116-120.

Dr. Noll asserts that license agreements between major record labels and certain customized non-interactive webcasters provide marketplace evidence of rates that corroborate the 5%-7% rates achieved in the Direct Licenses. Noll Revised Amended WDT at 16, SXM Dir. Trial Ex. 1. Focusing principally on the sound recording rights agreements between the digital music service Last.fm and the four major record labels,^[21] Dr. Noll determined that for its non-interactive subscription streaming service, Last.fm agreed to pay:

• [REDACTED]
• [REDACTED]
• [REDACTED]

Id. at 76-79 (footnote omitted), Tables 2.1-2.1c and Appendices E-H.^[22]

Using the rates gleaned from the Last.fm agreements for the non-interactive subscription streaming service, which he deemed to be the most similar to Sirius XM's satellite radio service in terms of functionality, Dr. Noll computed a hypothetical royalty rate by multiplying the Last.fm percentage-of-revenue rates [REDACTED] by the implicit per-subscriber price for Sirius XM's music channels ($3.00-$3.45). Dr. Noll then divided the resulting per subscriber monthly fee by Sirius XM's average revenue per user ($11.38) to express the hypothetical royalty rates as a percentage of revenue. Id. at 15; 6/5/12 Tr. 285:7-293:9 (Noll). This yielded an average royalty rate as a percentage of Sirius XM music channel revenue of 6.76%. Id. at 90; 6/5/12 Tr. 293:5-9 (Noll). Because this hypothetical rate fit squarely within the 5%-7% rate range of the Direct Licenses, Dr. Noll opines that the Last.fm agreement rates are corroborative of the rates contained in the Direct Licenses. He further concludes that the range of rates in the Direct Licenses represent the upper end of a reasonable royalty rate because the customized, non-interactive Last.fm services offer greater functionality and sound quality than the channels offered by Sirius XM. Id. at 14-16; 6/5/12 Tr. 292:2-14 (Noll).

2. SoundExchange Proposed Benchmarks

SoundExchange's expert economist, Dr. Janusz Ordover, offers a principal benchmark, and two alternatives, based upon his examination of 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. Dr. Ordover chose interactive subscription services because of his belief that they represent voluntary transactions in a competitive marketplace free of regulatory overhang. He also opined that such transactions provide sufficient information based on multiple buyer/seller interactions, are not distorted by the exercise of undue market power on either the buyer's or seller's side, and involve digital music services that are similar to Sirius XM. 6/14/12 Tr. 2359:11-2360:9, 2256:13-2261:3 (Ordover).

a. Ordover's Interactive Streaming Benchmark

Dr. Ordover derived his principal interactive streaming benchmark by determining the percentage of revenues that streaming services paid to the major record labels pursuant to their respective agreements. He then multiplied that percentage by an estimated retail price for a hypothetical music-only satellite radio service. See generally Ordover Third Corrected/Amended WDT at 18-25, SX Trial Ex. 74. Beginning with data from July 2010, he derived the effective percentage of revenue paid by each interactive service by taking the amount of royalty fees paid to the record companies and dividing it by each service's gross subscription revenues. 6/14/12 Tr. 2274:10-16 (Ordover). In other words, Dr. Ordover relied on royalty payments the labels reportedly received under the agreements rather than the percentage-of-revenue rates specified in the agreements which contained “greater of” royalty formulations.^[23] 6/14/12 Tr. 2274:22-2275:1, 2363:14-2364:7 (Ordover). In calculating actual licensing fees paid, Dr. Ordover used gross subscription revenues of the interactive services without any deductions or carve-outs. Ordover Third Corrected/Amended WDT at 19, SX Trial Ex. 74. Examining the agreements, he determined that the annual payments as a percentage of gross subscription revenues of the services ranged from 50% to 70%, and tended to cluster in a range of 60% to 65%. Id. at 19-21; 6/14/12 Tr. 2275:4-12 (Ordover).

Dr. Ordover then attempted to account for the fact that the Sirius XM satellite radio service, unlike interactive subscription services, transmits both music and non-music content by reducing the percentage-of-revenue rate from the interactive subscription agreements by half. He chose this reduction percentage principally based upon his observation of the identical $9.99 retail prices offered by Sirius XM for non-music and mostly music stand-alone subscriber packages. The result was a percentage-of-revenue rate range of between 30% and 32.5%. Dr. Ordover proposed this range as a benchmark for the SDARS rates for the 2013-17 statutory licensing period. Ordover Third Corrected/Amended WDT at 17, SX Trial Ex. 74.^[24]

b. Ordover's Content Comparability Adjustment

Dr. Ordover offered an alternative approach that involved an examination of per-subscriber royalty rates from interactive music streaming subscription services in an effort to adjust for the differences in service attributes between satellite radio and interactive subscription services. He first determined an unweighted average monthly royalty of $5.95 per subscriber (monthly licensing fees paid divided by monthly subscriber counts) for interactive subscription services. He then adjusted this fee by the ratio of the retail price of a hypothetical music-only satellite radio service (50% of the $12.95 subscription price for the Sirius XM Select programming package ^[25] ) to the retail price for interactive subscription services ($9.99). Ordover Third Corrected/Amended WDT at 30-31, SX Trial Ex. 74. This percentage, when applied to the average per-subscriber royalty paid by interactive services ($5.95), yields $3.86 for the hypothetical music-only satellite radio service. Dividing this number by the $12.95 Sirius XM subscription price provides a percentage-of-revenue rate of 29.81%. Id. at 32.

c. Ordover's Interactivity Adjustment

Dr. Ordover's second alternative approach attempts to adjust for the presence of interactivity alone in the rates yielded by his primary benchmark under the assumption that interactivity is the material difference between interactive subscription services and satellite radio. Ordover Third Corrected/Amended WDT at 33, SX. Trial Ex. 74. To derive the value of interactivity, he compared the retail prices for interactive music streaming services with the retail prices for non-interactive music streaming services. He determined that interactive music streaming services are uniformly priced at $9.99 per month, while non-interactive services prices averaged $4.86. Id. at 31-32, Table 4 and 33-34, Table 5.^[26] Dr. Ordover then used the ratio to adjust the average per-subscriber royalty paid by interactive services ($5.95) to calculate an equivalent payment for satellite radio. This calculation yielded a percentage-of-revenue royalty rate of 22.32% for Sirius XM, which Dr. Ordover concludes represents the lower bound of a reasonable royalty rate. 6/14/12 Tr. 2282:12-2283:22, 2334:8-11 (Ordover).^[27] Dr. Ordover offered no alternative that attempted to account for the combination of content and interactivity differences.

3. Analysis and Conclusions Regarding the Proposed Benchmarks

For the reasons stated herein, the Judges determine that an analysis of the benchmark evidence presented in this proceeding establishes that reasonable royalty rates for the use of sound recordings under the Section 114 statutory license cannot be lower than 7%, the upper bound of the range of rates of the Direct Licenses. The Judges find that Dr. Ordover's proposed benchmark rates of between 30%-32.5% are beyond the zone of reasonableness, given that they are four times greater than the rate of 8% that the Judges set four years ago in SDARS-I and are based on a limited data set of questionable comparability to the target market. As a result, the Judges are left with no acceptable benchmark by which to mark an upper bound for a zone of reasonableness. The Judges rely, therefore, on data points such as the lowest rate proposed by SoundExchange and the unadjusted upper bound in SDARS-I to guide the determination of what the upper bound should be in this proceeding.

a. Analysis of Sirius XM's Proposed Direct License Benchmark

The Direct Licenses that Sirius XM proposed as the foundation for a benchmark have the surface appeal of a comparable benchmark because they involve the same sellers and buyers as the target market. A closer examination, however, reveals the weaknesses of the Direct Licenses as a data set. First, the direct licensors represent a sliver of the universe of rights holders for sound recordings: 95 of over 20,100 rights holders to which SoundExchange distributes payments. See Bender WDT at 4, SX Trial Ex. 75; 8/13/12 Tr. 3015:16-20 (Frear). They also represent a subset of the 691 independent labels that Sirius XM approached in the first instance. Ordover Amended WRT at 4 n.8, SX Trial Ex. 218; SX Trial Ex. 301. Sirius XM opined that the number of Direct Licenses would likely have been substantially higher but for the alleged interference of SoundExchange and others who purportedly attempted to discourage record labels from negotiating with Sirius XM. The Judges are not persuaded by the evidence in the record that SoundExchange's alleged actions materially frustrated Sirius XM's efforts to execute Direct License agreements. Therefore, the Judges must evaluate the Direct Licenses for what they are, which is to say, a very small subset of the sound recording market.^[28]

The Direct Licenses do not include any of the major record labels whom, by virtue of the depth and breadth of their music catalogues, make up a critical portion of the sound recording market. Dr. Noll's observation that the works licensed by the Direct Licensors represent the kinds of sound recordings performed on Sirius XM does not diminish the importance of the catalogues of the major labels. It would be difficult to imagine a successful SDARS service that did not have access to the types of recordings that the major labels possess. The “representativeness” of the sound recordings contained in the catalogues of the Direct Licensees does not equate to their popularity, an essential ingredient to Sirius XM's music offerings. 6/7/12 Tr. 836:17-22 (Gertz)(“Sirius XM is very hits driven, and they want to have the most successful service they can, so they're going to use what's popular.”). Nevertheless, the rates the Judges set must evaluate the universe of sound recordings available for licensing under the statute. The vast majority of those sound recordings would not currently be considered to be “popular,” although many might have qualified as “hits” in their day or are viewed as popular in a particular genre.

Furthermore, the Judges note that the additional considerations and rights granted in the Direct Licenses that are beyond those contained in the Section 114 license weaken the Direct Licenses' comparability as a benchmark. For example, the Direct Licenses provide for payment of 100% of the royalties to the Direct Licensors, 6/6/12 Tr. 341:10-342:3 (Noll), thereby avoiding the statutory apportionment of 50% to record companies and 50% to artists and performers.^[29] See 17 U.S.C. 114(g). Certain of the Direct Licenses, in particular those of the larger independent labels, provide for cash advances and accelerated royalty payments, considerations that also are not provided for under the statutory license. See, e.g., Gertz Revised WRT at SXM Reb. Ex. 8, pp. 3-4 and SXM Reb. 23, pp. 3-4, SXM Reb. Trial Ex. 8. In addition, Sirius XM absorbs all of the administrative costs of the licensing process under the Direct Licenses, which, under the statutory license, are borne by the copyright owners, artists and performers. Eisenberg Amended/Corrected WRT at SX Ex. 313-RR, SX Trial Ex. 245. With respect to rights granted under the Direct Licenses, Sirius XM receives a waiver of the sound recording complement of the statutory license and the ability to perform the works of the Direct Licensors on other services not covered by the statutory license.

Dr. Noll's analysis does little to address the Judges' concerns regarding the Direct Licenses. Dr. Noll contends that the fact the Direct License rates are lower than the current 8% statutory rate is explained by a “demand diversion effect.” In other words, Dr. Noll posits that record labels engage in price competition aimed at increasing their market share through increased plays on Sirius XM, thereby reducing the royalty rates demanded, which reflects what would happen in the market as a whole in the absence of a statutory rate. Noll Revised Amended WDT at 36-38, SXM Dir. Trial Ex. 1.

Dr. Noll's demand diversion theory, however, has limited explanatory power. It may well be that independent record labels took the Direct License offer because of the valuable non-statutory benefits discussed above, and there is testimony in the record to this effect. See, e.g., SX Trial Ex. 317 at SXM-CRB_DIR_00079565; 8/20/12 Tr. 4156:5-4157:3 (Powers). Further, independent labels may have a greater incentive than majors to secure performances of their works on services such as Sirius XM, which would increase the attractiveness of a Direct License relationship. Powers WRT at 4, SX Trial Ex. 243; Eisenberg Amended/Corrected WRT at SX Ex. 329-RR at SXM_CRB_DIR_00042287, SX Trial Ex. 245 (email from MRI to independent label emphasizing that a Direct License offers the possibility of increased airplay). Although major labels also must compete with other majors and with independent labels for airplay, none was apparently so motivated by that concern to negotiate separately with Sirius XM. Therefore, the differing motivations of the “sellers” in the proposed Direct License benchmark suggest a weakness regarding comparability to the target market.

Dr. Noll's benchmark analysis, whether considered as corroboration of the rates in the Direct Licenses or standing alone, contains significant flaws. His reliance on the Last.fm agreements with the four major record labels, which provide the critical data to his calculations, is valid to the extent that the agreements are shown to be representative of non-interactive subscription webcasting services. See SDARS-I, 73 FR at 4090. Two of the agreements, however, have expired and are no longer in effect. Ordover Amended WRT at 25, SX Trial Ex. 218. Last.fm now pays those record companies at the statutory webcasting rate, which is not a market rate. 8/14/12 Tr. 3308:8-20, 3317:10-16 (Ordover). Even if the Last.fm agreements were the most representative of webcasting services—and Dr. Noll has not demonstrated that they are—the Judges would not be inclined to accept them as fully comparable to the SDARS business without a persuasive adjustment to account for the functional differences between webcasting and satellite radio. Dr. Noll offered none.

The Judges also have reservations about Dr. Noll's determination of $3.00-$3.45 as the implicit monthly market price for Sirius XM's music channels.^[30] Dr. Noll identified three methods for determining the implicit price. The first is the average retail price of $3.15 taken from Last.fm's and Pandora's non-interactive subscription services. Noll Revised WRT at Table 1, SXM Reb. Trial Ex. 6. As with Last.fm, there is no adjustment to account for functional differences between the Pandora webcasting service and satellite radio, whose primary use is in the automobile.

Dr. Noll's second method is to derive a market price for Sirius XM using a survey conducted by Sirius XM's witness Professor John Hauser that attempts to measure the value of music to Sirius XM subscribers. Professor Hauser posited an anchor price for the Sirius XM service to his survey respondents, and then randomly removed features (such as lack of commercials, quality of sound, etc.) to determine how much the respondents would be willing to pay for the service after each feature is removed. Hauser Corrected WRT at 20-22, SXM Dir. Trial Ex. 24. After averaging the results, he determined that subscribers place an average value on Sirius XM's music channels of $3.24. Id. at Appendix G. Professor Hauser's survey is of limited value. By design, the higher number of features or attributes of the Sirius XM service included in the survey, the lower the estimated value of any given service. This feature of the survey produces anomalous outcomes, such as survey results showing that subscribers would pay a certain amount for ubiquitous station availability, premium sound quality and absence of commercials, all without any programming content. See Ordover Amended WRT at 35, SX Trial Ex. 218.

Third, Dr. Noll sought to calculate the cost of inputs necessary for delivery of Sirius XM's programming via satellite and its subsidization/installation of radio receivers in automobiles (described as “unique” costs to the satellite radio service), to then deduct those costs from gross revenues, and allocate the remaining revenue between music and non-music content. Noll Revised Amended WDT at 81-83, 85, SXM Dir. Trial Ex. 1. After making these calculations, Dr. Noll credited 55.1%, or $3.45, to music channels. Id. at 88 and Table 3. Sirius XM contends that including the unique delivery costs and investments of its service is appropriate in Dr. Noll's calculation. Sirius XM cites to major record company agreements with Cricket and MetroPCS (mobile service providers that bundle telephone service and interactive music service into a single package) that reflect that a percentage royalty rate for music must be reduced by a commensurate proportion to reflect revenue collected for the non-music portion of the bundled service. Sirius XM PFF ¶¶169-173.

SoundExchange's expert economist, Dr. Thomas Lys, explained, however, that because most of the unique costs that Dr. Noll allocated are relatively fixed, the per-subscriber amounts vary inversely with the number of subscribers. Lys WRT at 57, SX Trial Ex. 240. Dr. Noll performed his calculation of costs using 2010 data, but had he used subscriber numbers for the years thereafter, which have continued to increase, and are anticipated to increase further in the coming licensing term, the analysis would show lower unique costs per subscriber and a higher value of music. Id. The dependency of Dr. Noll's methodology on timing and the number of subscribers undermines its reliability for quantifying what the unique costs are likely to be in the coming rate term. Id. at 58.

Sirius XM's analogy to the bundled services of Cricket and MetroPCS is inapposite. Unlike those services, the success of Sirius XM is dependent upon its access to music. 6/14/12 Tr. 2270:7-2271:15 (Ordover); see also 6/5/12 Tr. 235:6-10 (Noll)(“It's a bundle of services, it's a distribution system, a bunch of nonmusic content and a bunch of music content, all of which are essential. And you pull the plug on any one of them, and the whole thing collapses.”); 6/11/12 Tr. 1431:10-17 (Karmazin). The value of Sirius XM's satellite radio service is the bundling of music and non-music content with its delivery platform, and Sirius XM has failed to present convincing evidence that its delivery platform and non-music content, alone, present a viable business.^[31]

In sum, these concerns, coupled with those surrounding the Direct Licenses themselves, show weaknesses in the proposed Direct License benchmark that diminish its usefulness. Therefore, the Judges find that the 7% rate, which represents the high end of the Direct License rates, represents the lower bound of a zone of reasonableness. The Judges believe that a rate any lower, given the prevailing statutory rate, would more than likely be overly influenced by the particular terms of the Direct License agreements, which are not part of the statutory license.^[32]

b. Analysis of SoundExchange's Proposed Interactive Subscription Services Benchmark

The Judges have determined in the past that the interactive subscription service market has characteristics reasonably similar to those of the SDARS market. SDARS-I, 73 FR at 4093. Moreover, Dr. Ordover's proposed interactive subscription service benchmark in this proceeding analyzes certain useful data sets that make it difficult to dismiss the proposed benchmark outright. For example, the agreements Dr. Ordover examined represent a relevant data source from which to consider the outcomes of marketplace negotiations. That being said, the Judges do not find that the market for interactive subscription streaming services as characterized by Dr. Ordover in this proceeding offers a foundation to support a comparable benchmark from which to begin an analysis of reasonable rates for SDARS for the upcoming license period.

For example, the rights licensed by interactive subscription services are not the same as those by non-interactive services such as the SDARS, and the Judges did not find Dr. Ordover's efforts to adjust for the differences to be helpful. Dr. Ordover attempted to account for these differences by offering two alternative approaches, both of which seek to enhance the comparability of this proposed benchmark to the SDARS market. As discussed above, his first alternative approach attempts to adjust for service content between the two markets. The Judges doubt whether this approach adequately adjusts the interactive subscription service market to account for differences in attributes and functionality between that market and satellite radio. Dr. Ordover's second alternative approach attempts to adjust for interactivity. Ordover Third Corrected/Amended WDT at 33-34, SX Trial Ex. 74. The Judges found this effort to be somewhat more pertinent.

Nevertheless, the Judges find that the differences between Sirius XM and the “buyers” in the proposed benchmark severely constrain the usefulness of the proposed Ordover benchmark. Dr. Ordover's proposed interactive subscription streaming service benchmark was based on licensing fees paid to the four major record labels for 2011 by seven internet streaming services and for one-half of 2012 for some of those services. Ordover Third Corrected/Amended WDT at 19-21, SX Trial Ex. 74.

Dr. Ordover characterizes the streaming services as “well-established services like Microsoft Zune, Napster, and Rhapsody, and newer market entrants like Rdio and MOG.” Id. Dr. Ordover concedes, however, that in October 2011, Rhapsody announced that it was acquiring Napster. Id. at 19, n.16. Notably, in 2012 Microsoft ceased offering Zune as a stand-alone service and rolled it into its XBOX service suite. See http://www.xbox.com/​en-S/​Live/​Partners/​Zune. In addition, one of the services upon which Dr. Ordover based his proposed benchmark, Slacker Premium, was not introduced until May 2011, so not even a full year's payment data was available for that service.

The royalty implications of these details are uncertain, but these details about the proposed benchmark market underscore the fluid nature of the subscription streaming market and the difficulty of generalizing the royalty obligations of a market based on a few quarters worth of payment data for a handful of services. In short, the interactive subscription service market upon which Dr. Ordover relied is in a constant state of flux. No single buyer or group of buyers in that market seems comparable to Sirius XM in terms of its name recognition and status as the sole provider of satellite radio service. Therefore, the Judges believe that Sirius XM likely would have been in a preferential bargaining position to the interactive subscription service providers and may have negotiated very different rates as a result. The Judges do not believe that Dr. Ordover accounted for this difference.

Whereas the Judges criticized the Direct License benchmark data set for lacking one or more major record labels, the proposed Ordover benchmark also lacks the balance of representing a broader subset of record labels. Although Sirius XM's service may be “hit” driven, it features a broad range of music offerings that span several decades and several genres. Indeed, the Judges suspect that much of the value of the Sirius XM service as opposed to broadcast terrestrial radio and other competitors is that Sirius XM plays a greater range of music, much of which may be licensed by non-major labels. Therefore, by focusing on the catalogues that the major record labels possess, although a crucial component of Sirius XM's service, the proposed Ordover benchmark overlooked a subset of the entire universe of sound recordings for which the Judges must set a rate in this proceeding. The Judges believe that these comparability differences may help to explain why the rates in the subscription services market are so much higher than those in the Direct Licenses, although other factors are also at play.

The yawning gap between the current rate of 8% and the highest rates proposed by Dr. Ordover raises additional concerns about the proposed Ordover benchmark. Indeed, the Judges find that the rates Dr. Ordover calculated based on his proposed principal benchmark (30%-32.5%) and his first alternative adjustment (29.81%) are so much higher than the current statutory rate that they are outside the zone of reasonableness. The rate that Dr. Ordover derives from his second alternative adjustment (22.32%), while suggesting a more reasonable alternative, 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.

As a result, after analyzing the proposed benchmarks, both of which are flawed, the Judges are left with a zone of reasonableness with a floor of 7% and an upper bound that can be no more than 22.32%. The Judges are also informed by SoundExchange's proposed rates for SDARS, which start at 12% for 2013. Presumably, SoundExchange would not have proposed this entry rate if it did not believe it to be reasonable. Lastly, the Judges consider the prevailing statutory rate of 8%, which the Judges adjusted down from a 13% rate in SDARS-I based on the fourth Section 801(b) factor. SDARS-I, 73 FR at 4093-4098. With these guide posts in mind, the Judges analyze the Section 801(b) factors.

4. Application of Section 801(b) Factors

The Copyright Act requires that the Judges establish rates for the Section 114 license that are reasonable and calculated to achieve the four specific policy objectives set forth in Section 801(b) of the Copyright Act. In analyzing the Section 801(b) factors the Judges determine whether adjustments to the rate indicated by marketplace benchmarks, if any, are warranted and, if so, whether there is sufficient evidence in the record to support such adjustments. SDARS-I, 73 FR at 4094 (Jan. 24, 2008). The absence of solid empirical evidence that might suggest a difference between the benchmark and target markets cautions against the need for an adjustment. Id. at 4094-4095. In SDARS-I, the Judges determined that no adjustment was warranted for the first three factors but that a downward adjustment was warranted for the fourth factor—minimization of disruptive impact—for reasons discussed in section (d) below. SoundExchange argues that no adjustment is warranted in the current proceeding. SX PFF ¶ 497. Sirius XM contends, however, that an analysis of the Section 801(b) factors “counsels setting a royalty rate at the low end of the range of reasonable rates.” Sirius XM PFF ¶ 227.

a. Maximize the Availability of Creative Works

Sirius XM contends that a downward adjustment from the benchmark rate is warranted with respect to the first Section 801(b) factor—maximizing the availability of creative works to the public.^[33] Sirius XM PFF ¶ 227. To support its contention, Sirius XM argues that the term “availability” in this factor encompasses both the incentive to produce creative products and the delivery of those products to consumers. Id. at ¶ 228. Sirius XM states that its service enhances the delivery and availability of sound recordings by: “providing an uninterrupted nationwide broadcast of unparalleled breadth and depth; exposing listeners to music that is not played elsewhere; and creating original music programming to promote artists * * *.” Id. at ¶ 230. Sirius XM also contends that, unlike its service, which it posits promotes phonorecord sales, internet subscription services, which formed the basis of the proposed Ordover benchmark, show no such promotional value, and in fact, may cannibalize phonorecord sales. Id. at ¶¶ 253-254 and 257-260.

Much of the evidence that Sirius XM presented to show the promotional effect of Sirius XM's service on phonorecord sales consists of testimony detailing record labels' efforts to get their artists airplay on Sirius XM and elsewhere. See, e.g., id. at ¶ 253 (“SoundExchange's witness Darius Van Arman, co-owner of several independent record labels, conceded that `one of the goals of [his labels'] promotional activities [is] to get [his] artists airplay * * * includ[ing] airplay on Sirius XM.”). It is not surprising that record labels seek airplay for the artists they represent. Nor would it be surprising to learn that increased airplay on Sirius XM can enhance phonorecord sales. Those facts alone, even if assumed to be true, would not provide the type of substantial empirical evidence that might support a downward adjustment from the rates most strongly suggested by the evidence in the record.

As SoundExchange notes, “Sirius XM's case attempting to connect Sirius XM airplay with sales of sound recordings consists of less than ten pieces of anecdotal evidence over a five-year period.” SX RFF at ¶ 228 (emphasis in original). The Judges agree with SoundExchange that Sirius XM provides insufficient probative evidence upon which the Judges could make a meaningful assessment of the relative promotional value of Sirius XM's service vis á vis interactive internet subscription services. See 10/16/2012 Tr. 4874:16-18 (Sirius XM closing argument by Mr. Rich, noting the anecdotal nature of Sirius XM's promotional evidence).

The Judges are also unpersuaded by Sirius XM's assertion that the purported lack of promotional value of interactive internet services should warrant a downward adjustment from a marketplace benchmark rate upon which the Judges might rely. Evidence to support Sirius XM's contention that interactive internet services are substitutional and may cannibalize phonorecord sales is sparse. In this regard, the Judges place little credence on blanket statements such as that by Dr. Noll that “there's no question” that interactive subscription services have no promotional impact on record sales, because “[o]n demand services let customers play a specific recording on request, allowing the same control over play sequence that customers have in playing recordings from personal libraries.” Sirius XM PFF at ¶ 257, quoting Noll Revised Amended WDT at 22 and 6/5/12 Tr. 227:16-228:16 (Noll).^[34]

Dr. Noll's statement is more descriptive of the nature of interactive services than supportive of the claim that those services increase substitution. Even if the Judges were to take at face value Dr. Noll's implication that a subscriber's ability to play a particular track on demand discourages the subscriber from purchasing that track, that fact alone would not address the more general notion regarding the relative promotional value (or lack thereof) of interactive internet streaming services. Promotional value can extend far beyond a service's impact on a single track by a single artist; it may extend to an artist's entire catalogue as well as to related artists or genres.

With respect to the potential substitutional effect of interactive internet streaming services, Sirius XM references certain industry projections that assume a certain rate of cannibalization for such services. Sirius XM PFF ¶¶ 259-260. Even if these rates were assumed to be reasonable projections for this type of service, they are not supportive of a downward adjustment from a marketplace-influenced benchmark rate because they are already taken into account in determining the royalty rates that the services pay. See, e.g., id. at ¶ 260; PSS Ex. 8 at 3 (SX02 00027594); 6/13/12 Tr. 2061:16-2062:3 (Bryan) (Warner Music Group's estimate of potential substitution effect of Spotify's service). In sum, the Judges find no probative evidence to warrant an adjustment from a marketplace-derived benchmark rate under this factor.

b. Afford Fair Return/Fair Income Under Existing Market Conditions

With respect to the second Section 801(b) factor—affording a fair return to the copyright owner and fair income to the copyright user—the Judges find that little has changed since SDARS-I, in which the Judges determined that no adjustment from the benchmark rate was warranted.

1. The Parties' Contentions

SoundExchange argues that no adjustment to a marketplace-derived benchmark rate is warranted “unless there is a clear showing that the benchmark rates were elevated by the exercise of monopoly power.” SX PFF ¶ 504. SoundExchange contends that no such monopoly power was shown with respect to the marketplace benchmark that SoundExchange proposes or with respect to “other non-statutory distribution channels.” Id. SoundExchange contends that “any downward adjustment would amount to a `subsidy' for Sirius XM, which would provide the company with an unwarranted competitive advantage relative to rival distributors of music content, and also dilute the incentives for the creation of new works and for the efficient transmission of music through new and emerging channels.” Id.; Ordover Third Corrected/Amended WDT at 10, SX Trial Ex. 74. SoundExchange also stresses the growing importance to artists and record labels of digital income streams as sales of physical products decline. SX PFF ¶ 528.

For its part, Sirius XM contends that “the implementation of this factor requires assessing whether the royalty rate allows both the buyer (Sirius XM) and the sellers (the record labels) to recover their costs, including the financial cost of capital used to make investments.” Sirius XM PFF ¶ 263. Sirius XM states that those costs must be measured cumulatively and not as a “snapshot of annual operating costs.” Id. While Sirius XM concedes that the company has shown a “recent trend of profitability,” it contends that “it will be years before Sirius XM recoups all of its losses from the last two decades; thus, any increase to those costs, such as an increase in the SoundExchange royalty rate, will only lengthen the time it takes to recoup these losses and directly interfere with Sirius XM's ability to achieve a fair return on its investments.” Id. at ¶ 265.

2. The Judges' Analysis

In SDARS-I, the Judges stated:

Affording copyright users a fair income is not the same thing as guaranteeing them a profit in excess of the fair expectations of a highly leveraged enterprise. Nor is a fair income one which allows the SDARS to utilize its other resources inefficiently. In both these senses, a fair income is more consistent with reasonable market outcomes.

73 FR 4095 (footnote omitted).

In the absence of substantial evidence in the record to the contrary, any marketplace benchmark rate that guides the selection of rates will encompass such a return because it represents the best evidence of reasonable market outcomes. In this proceeding, the Judges find the proposed Direct License benchmark provides useful guidance for setting the lower bound of a zone of reasonable rates. The Judges find no probative evidence, however, to suggest that that rate should be adjusted under this factor. Presumably, being marketplace-inspired, the rate already reflects a fair income and a fair return.

SoundExchange stresses the growing importance of digital revenue streams for copyright owners, a trend that was certainly in play during the SDARS-I proceeding. The Judges find no material change in that trend in the current record that would warrant an upward adjustment from a marketplace-derived benchmark rate. In turn, Sirius XM stresses the importance of the rate on the timing of Sirius XM's return to profitability. The SDARS made similar points in the SDARS-I proceeding. Sirius XM's current trend toward profitability and its ability to pass on at least a portion of the rate increase to its subscribers^[35] suggests that the prevailing statutory rate—which was informed by a marketplace benchmark and which is within the zone of reasonableness the Judges establish in the current proceeding—did not hinder Sirius XM's ability to earn a fair income.

In this proceeding, the Judges set the lower bound of the zone of reasonableness at 7% based on marketplace outcomes. Therefore, the Judges are confident that fair income and returns are reflected by that rate. In SDARS-I the Judges found that the same was true with respect to the statutory rate of 8%, as well as the 13% rate to which the Judges applied the Section 801(b) factors to derive the 8% rate. The current record does not support a change in that conclusion. Therefore, the Judges find no justification in this proceeding for an adjustment either up or down pursuant to the second Section 801(b) factor for rates in a range of 7% to 13%.

c. Weigh Relative Roles of Copyright Owner and Copyright User

1. The Parties' Contentions

According to Sirius XM, “when using the royalty rate paid by an Internet-based music service as a benchmark for setting royalty rates for Sirius XM, one must first identify the contributions that are unique to Sirius XM * * * and then compare these contributions to those made by the Internet-based services that are being proposed as a benchmark.” Sirius XM PFF ¶ 277, referencing Noll Revised Amended WDT at 25-26, SXM Dir. Trial Ex. 1; 8/14/2012 Tr. 3463:13-3464 (Noll). In this regard, Sirius XM notes that it has spent over $10 billion in creating and supporting its service and that those costs have not yet been recovered. Sirius XM PFF ¶ 278. According to Sirius XM, these “massive contributions only continue to increase, and far outweigh those made by Internet-based services that serve as benchmarks for setting royalty rates for Sirius XM.” Id. See also id. at ¶ 295 (quoting Professor Noll and Mr. Karmazin for the proposition that Sirius XM's costs in developing its system far exceed those of the internet streaming companies). Sirius XM also points to its payments to automakers to encourage them to include Sirius XM's service in the vehicles they make, payments which, according to Sirius XM, the internet-based streaming services do not make. Sirius XM PFF ¶¶ 294, 296.

Sirius XM also contends that the record industry does not incur any additional incremental cost in making digital sound recordings available to Sirius XM. Id. at ¶¶ 279, 303. Sirius XM contends that, as a result of its contributions to its service, it should receive a downward adjustment from the benchmark rate, to the extent that rate is based on an internet streaming benchmark. Id. at ¶ 297-298 (“simply applying the percentage-of-revenue rate paid by benchmark Internet-based music services to the full revenues of Sirius XM without adjustment [to either the rate or the revenue base] would fail to recognize Sirius XM's relative contribution and `would effectively give record labels a share of revenues that have nothing to do with the sound recording rights they are licensing.' ”) Id. at ¶ 298, quoting Salinger Corrected WRT ¶ 18, SXM Reb. Trial Ex. 9.

For its part, SoundExchange stresses the risks and costs the record labels incur in making, promoting and distributing music. For the perspective of a major record label, SoundExchange's evidence consists largely of testimony from UMG's Mr. Ciongoli who detailed UMG's costs in finding, developing, and marketing artists. SX PFF ¶¶ 535-542. In addition, SoundExchange presented the testimony of Mr. Van Arman who discussed the costs and efforts that independent labels typically incur in finding and promoting artists. Id. at ¶ 544.

2. The Judges' Analysis

The Judges' task with respect to the Section 801(b) factors is to determine whether the record presents solid empirical evidence of a difference between the benchmark market, if any, and the target market that would warrant an adjustment in the rate most strongly suggested by the evidence.

In SDARS-I, the Judges found that

[C]onsidering the record of relevant evidence as a whole, the various sub-factors identified in this policy objective may weigh in favor of a discount from the market rate because of the SDARS' demonstrated need to continue to make substantial new investments to support the satellite technology necessary to continue to provide this specific service during the relevant license period. However, inasmuch as we find this issue is intimately intertwined with evidence impacting our consideration of the fourth 801(b) policy objective (i.e., minimizing any disruptive impact on the structure of the industries involved), we will treat the effect of this particular matter as part of our consideration of the fourth policy objective.

73 FR 4096.

With this exception, we found no other rationale for an adjustment either up or down from the benchmark rates based on this factor. Id. at 4096-4097.

In the current proceeding, in setting the lower bound of the zone of reasonableness the Judges were guided by the Direct Licenses Sirius XM negotiated with certain independent records labels. Deriving the upper bound of the zone of reasonableness has proved to be more problematic. The Judges conclude that the upper bound cannot be above the 22.32% rate from Dr. Ordover's second alternative approach. Moreover, the Judges are confident that the current statutory rate of 8% is within the zone of reasonableness. The Judges also are informed by the presumed reasonableness of the 12% rate that SoundExchange proposed for 2013 and by the 13% benchmark rate that served as a benchmark in SDARS-I. Given the Judges' relative confidence in the reasonableness of the 7% to 13% range, consideration of the third Section 801(b) factor is directed at that range.

Since the Direct License benchmark involves the same buyer (Sirius XM) and the same sellers (record labels) as the buyers and sellers in this proceeding—and they negotiated over the same rights set—the Judges find that the buyers and sellers in the benchmark market sufficiently replicate those in the target market.^[36] With respect to the rights for which Sirius XM and the independent labels negotiated, evidence in the record indicates that the Direct Licensors granted broader rights than just the public performance rights that are the subject of the Section 114 compulsory license and that Sirius XM offered incentives beyond those that would be available through the compulsory licensing scheme. See, e.g., SX PFF ¶¶ 386-400 (citing defrayed administrative costs, the ability of Sirius XM to play more of an artist's works over a given time, and direct payment of the artists' share to the independent labels, among others, as key differences between the Direct Licenses and the compulsory licenses that might warrant a lower effective royalty rate).

The Judges acknowledge the differences between the compulsory license and the Direct Licenses, but view many of those differences as more a matter of administrative convenience than of differences in the substantive rights of the parties with respect to the public performance right. For example, if an artist is entitled to a certain share of profits from the sale of his or her records, that right is not diminished by the fact that the share is paid by SoundExchange or by the label with which the artist has signed. As far as the non-administrative differences (e.g., waiver of the statutory restriction on the number of times an artist's works may be played over a given time), it may well be that the benefits inure equally to both Sirius XM and the artists represented by the independent labels, many of whom may value broader exposure in lieu of statutory restrictions on the amount their works may be played.^[37] Therefore, the rights that the parties negotiated for in the benchmark market are reasonably comparable to those in the target market and no upward adjustment from the lower bound benchmark rate is warranted based on the third Section 801(b) factor.

With respect to the upper bound of the zone of reasonableness, which is informed by licenses between major record labels and certain interactive streaming services, the sellers are acceptably comparable to the target market (i.e., record labels). Although evidence in the record addresses the costs UMG incurs generally as a major record label and the costs Mr. Van Arman's independent labels incur in developing the artists that they sign, no substantial empirical evidence addresses the unique costs and contributions that the record labels make with respect to providing their recordings to Sirius XM. Because the Judges conclude that the sellers in the proposed benchmark market that guided the upper bound of the zone of reasonable rates are comparable to those in the target market, their contributions, risks, and costs are presumed to already be incorporated into the rates that set the upper bound. Therefore, the sellers' contributions in the target market do not indicate that an adjustment from the bounds of the zone of reasonableness is warranted.

Determining the comparability between the buyers that yielded the upper bound requires a comparison between Sirius XM and the internet streaming services that are the buyers in the proposed Ordover benchmark market. As the Judges recognized in SDARS-I, Sirius XM has demonstrated the need to continue to make substantial new investments to support the satellite technology necessary to continue to provide its specific service during the relevant license period. The Judges have no substantial evidence in the record, however, that would lead to a conclusion that the internet-based streaming services have ongoing distribution system costs anywhere near those of Sirius XM. According to Mr. Karmazin, Sirius XM anticipates investing more than [REDACTED] to maintain, upgrade, and, where necessary, replace its technological infrastructure during the 2013-17 licensing period. Karmazin WDT at 4, SXM Dir. Trial Ex. 19. A large portion of the system's costs relate to Sirius XM's satellites. According to Sirius XM, over the past six years, the company has spent approximately $1.5 billion replenishing satellites. Sirius XM PFF ¶ 289; Meyer WDT at 23-24, SXM Dir. Trial Ex. 5. A satellite's useful life is between 12 and 15 years. Meyer WDT at 24, SXM Dir. Trial Ex. 5. Sirius XM expects that its newly replenished satellite networks will maintain its services through 2020. Sirius XM PFF ¶ 291; Meyer WDT at 24, SXM Dir. Trial Ex. 5.

Such substantial financial outlays are unique to Sirius XM, which has developed a proprietary music distribution system, rather than use the existing internet framework, as the services in the proposed Ordover benchmark market have done. Although the costs of developing and launching the current generation of satellites has already been sunk, it is not unreasonable for Sirius XM to expect to recoup a certain amount of those costs over the expected useful life of the satellites. Moreover, the costs of maintaining the current satellites and planning and developing the new generation of satellites will require additional, substantial costs over the license period. See Meyer WDT at 24, SXM Dir. Trial Ex. 5. In light of the substantial evidence in the record of the unique and substantial financial costs that Sirius XM has incurred and anticipates incurring over the license period to maintain and upgrade its distribution system, the Judges find that the most appropriate rate for the current license period will be somewhat below the 12%-13%, which the Judges are reasonably confident represents the top of the zone of reasonableness. Therefore, the rates that the Judges announce in this determination for the SDARS reflect a downward adjustment from the 12%-13% range based upon the third Section 801(b) factor.

d. Minimize Disruptive Impact

Although the rate the Judges set in this proceeding is just one component that will impact the future of Sirius XM and the copyright owners, the rate could be considered disruptive (and thereby warrant an adjustment) if the unadjusted rate “directly produce[d] an adverse impact that is substantial, immediate, and irreversible in the short-run because there is insufficient time for either the SDARS or the copyright owners to adequately adapt to the changed circumstances 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.” 73 FR 4097. In SDARS-I, the Judges found that a downward adjustment from the upper boundary of the marketplace benchmark was justified on two grounds: (1) The SDARS' were not sufficiently profitable and did not have a sufficiently broad subscriber base to sustain an immediate rate increase from a range of 2.0%-2.5% to 13% of revenues and (2) a 13% rate would potentially constrain the SDARS' ability to undertake satellite investments planned for the license period, which, if delayed, could disrupt the SDARS' consumer service. Id.

1. The Parties' Contentions

In determining whether the fourth factor warrants an adjustment in the current proceeding, Sirius XM invites the Judges to consider Sirius XM's “tumultuous financial history” as well as the “increasing risks the Company is likely to face in the coming license term.” Sirius XM PFF ¶ 304. When so considered, Sirius XM argues that it “faces a threat of disruption that is `equal to or even greater than the one it faced at the time of the last rate proceeding.' ” Id., quoting Stowell WDT ¶ 41, SXM Dir. Trial Ex. 18. Sirius XM details its near-brush with bankruptcy in 2008 after Sirius and XM merged and its ultimate deal with Liberty Media Corporation to avert a bankruptcy filing. Frear WDT at 3-5, SXM Dir. Trial Ex. 12. It also notes its achievement of profitability in 2010. Id. at 7.^[38] While acknowledging that its recent performance is “encouraging,” Sirius XM points out that it has cumulative net operating losses of $8 billion, which it has incurred over the past two decades. Sirius XM notes that any increases in its costs will lengthen the time it takes to recoup these losses. Id. at 7-8. Nevertheless, Sirius XM anticipates that its adjusted earnings before depreciation and amortization (“EBITDA”) for 2012 will be $860 million on revenues of $3.3 billion, which should allow Sirius XM to return capital to its investors. Id. at 14 and n.11.

Notwithstanding its anticipated profitability for 2012, Sirius XM notes that it is still in a financially tenuous position in the longer term. Threats that Sirius XM anticipates to its continued financial success include: its increasing dependence on the automobile industry; competitive threats from internet-based providers and “connected-car” technology; significant debt on Sirius XM's balance sheet; and the continuing risk that Sirius XM could lose access to the credit markets to refinance its debt. Sirius XM PFF ¶¶ 322-328; Frear WDT at 21-22, SXM Dir. Trial Ex. 12. Although Sirius XM has been able to offset a portion of recent rate increases by passing some of those costs on to customers, continuing to do so in the future, Sirius XM contends, will run the risk of en masse subscriber defections. Id. at n.17.

On the other hand, SoundExchange recommends that the Judges use a more circumspect approach to applying the fourth factor. SoundExchange proffers that “the fourth policy objective should be limited to a temporary facilitation of the ability of nascent and emerging services to gain consumer acceptance and potentially achieve an efficient scale of operation * * *. [O]nce a company achieves a material presence in the marketplace, as Sirius XM indubitably has, use of the fourth policy factor to reduce market-based rates should be considered only with extreme caution, and should never be used to shield the service at issue from the full rigors of vigorous marketplace competition.” SX PFF ¶¶ 550-551, citing Ordover Third Corrected/Amended WDT at 5-6, SX Trial Ex. 74.^[39]

SoundExchange points to Sirius XM's stronger financial position as evidence that no downward adjustment is warranted for this license period. To support its position SoundExchange relies in part on testimony from Professor Lys who noted,

[S]ince the merger, Sirius XM has experienced steady growth in both the number of subscribers and subscriber revenues while at the same time experiencing cost reductions. As a result, the company has achieved sustainable and growing profitability. Further, in contrast to 2009, when the company restructured its debt, Sirius XM's credit ratings and the underlying financial metrics related to its debt have improved substantially.

Lys Corrected WDT at 8, SX Trial Ex. 80.

Professor Lys further represented that in the third quarter (Q3) of 2011, Sirius XM had adjusted EBITDA ^[40] of $197 million, up 16% year-over-year for the same quarter. Id. at 19, citing “Sirius XM Radio Inc., Q3 2011 Earnings Call,” Capital IQ, November 1, 2011, p. 3.

Professor Lys also stressed Sirius XM's dramatic turn-around in free cash flow.^[41] In 2008, Sirius XM had negative free cash flow of over $550 million. Id. at 20. By 2009, Sirius XM's free cash flow had turned to a positive $185 million. By 2010, that number had reached $210 million. Sirius XM projected that its free cash flow for 2011 would reach $400 million, a 90% increase over 2010, and would continue to grow in 2012. Id., quoting “Sirius XM Radio Inc., Q3 2011 Earnings Call,” Capital IQ, November 1, 2011, p. 3.

2. The Judges' Analysis

In analyzing whether the fourth Section 801(b) factor warrants an adjustment (either up or down) to the rates delineating the zone of reasonableness, the Judges must examine the same set of circumstances that informed the Judges' analysis in SDARS-I. In SDARS-I, the Judges found that a downward adjustment from the upper boundary of the marketplace benchmark was justified because: (1) The SDARS were not sufficiently profitable and did not have a sufficiently broad subscriber base to sustain an immediate rate increase from a range of 2.0%-2.5% to 13% of revenues (a potential five-fold or six-fold increase) and (2) a 13% rate would potentially constrain the SDARS' ability to undertake satellite investments planned for the license period, which, if delayed, could disrupt the SDARS' consumer service. 73 FR at 4097. Neither of those justifications is present to the same degree in the current record.

Sirius XM is now in a far better financial position than either Sirius or XM was as a stand-alone company in 2007, the first year of the current licensing period. In 2007, Sirius and XM had combined revenues of $2.1 billion and combined adjusted EBITDA of negative $565 million. SX PFF ¶ 556; SX Trial Ex. 16 at SXM_CRB_DIR_00021681 (p.15). By year-end 2012, Sirius XM's revenues are expected to be $3.4 billion and its EBITDA is expected to be approximately $900 million. SX Trial Ex. 217 at 7. In 2007, the SDARS free cash flow was negative $505 million. In 2012, by contrast Sirius XM's free cash flow is expected to rise to $700 million. SX PFF ¶ 556, citing SX Trial Ex. 16 at SXM_CRB_DIR_00021682 (p.16); SX Trial Ex. 217 at 7; see also Lys Corrected WDT at 18-21, SX Trial Ex. 80.

Another key indicator of potential financial strength—net increase in subscribers—indicates that Sirius XM is much stronger than either of the SDARS were in 2007. In 2007, Sirius and XM had 17.3 million subscribers. By 2012, that number for Sirius XM had risen to 23.5 million. SX PFF ¶ 554. Sirius XM was able to increase its net subscribers notwithstanding a period of relatively weak car sales—a key driver of new subscribers—and despite passing on a portion of the royalty rate increase to its subscribers. Id. and Frear WDT at 21 n.17, SXM Dir. Trial Ex. 12.

In percentage terms, the gap between the prevailing rate of 8% and the 12%-13% range that guides the upper bound of the zone of reasonableness in the current proceeding is much narrower than the comparable gap presented in SDARS-I. Indeed, an increase to 12% or 13% would be less dramatic in percentage terms than was the increase the Judges adopted in SDARS-I, which Sirius XM has managed to sustain.^[42]

The Judges also find that the second rationale found in SDARS-I for a downward adjustment, potential constraint on Sirius XM's ability to undertake satellite investments during the license period, is a less pressing issue than it was during the prior licensing period. As discussed above, Sirius XM expects that its newly replenished satellite networks will maintain its services through 2020. Sirius XM PFF ¶ 291; Meyer WDT at 24, SXM Dir. Trial Ex. 5. Although there will be ongoing costs of maintaining its existing satellites—costs that justified a downward adjustment under the third factor—no substantial evidence in the record supports a downward adjustment based on Sirius XM's need to replace its existing satellites during the current licensing period. Therefore, neither of the circumstances that justified a downward adjustment under the fourth factor in SDARS-I is currently present.

The Judges find no new circumstances that would warrant a downward adjustment under the fourth factor. Sirius XM's primary contention for a downward adjustment under the fourth factor centers largely on the competitive landscape, in particular, the threat of new technologies such as the connected car technology. Sirius XM PFF ¶¶ 325-328.^[43] While emerging technologies can dramatically change the competitive landscape from one licensing period to the next, the Judges find insufficient evidence in the record to suggest that connected car technology or any of the other emerging competitive threats discussed during the proceeding warrants a downward adjustment under the fourth factor during the current licensing period.

SoundExchange has not alleged a disruption to the copyright owners as a result of the prevailing rate and the Judges are not inclined to lower that rate. Therefore, the Judges find no evidence that an increased rate, albeit one that is lower than that proposed by SoundExchange, would warrant an upward adjustment. Therefore, the Judges find no adjustment warranted under the fourth Section 801(b) factor.

5. Conclusions Regarding Section 114 Rates

After reviewing the Section 801(b) factors in light of the zone of reasonable rates that has 7% as its floor and 12%-13% as its most likely ceiling, the Judges find that the most appropriate rate for SDARS for the 2013 to 2017 licensing period is 11% of Gross Revenues. To minimize any potential disruptive impact of the rate increase, the Judges phase it in over the license period. Consequently, the Judges set forth the following SDARS rates: for 2013: 9.0%; for 2014: 9.5%; for 2015: 10.0%; for 2016: 10.5%; and for 2017: 11.0%.

VI. Definition of Gross Revenues and Deductions

A. Definition of Gross Revenues

1. SoundExchange's Proposal

The revenue base against which the adopted royalty rates would be applied is a matter of considerable disagreement between the parties. Sirius XM requests continuance of the current definition of Gross Revenues in 37 CFR 382.11, arguing that it properly identifies only those revenues that are related to the provision of statutorily licensed sound recordings. See Sirius XM PFF ¶ 416. SoundExchange favors a considerable expansion of the revenue base, stating that its proposed changes would conform the royalty base to the economics underlying the percentage-of-revenue royalty rates within the benchmarks offered in this proceeding, and would make the Gross Revenues definition easier to administer and less susceptible to interpretation and manipulation. Bender WDT at 12, SX Trial Ex. 75.

SoundExchange's proposed Gross Revenues definition ^[44] is based upon an interpretation of the Judges' use in SDARS-I of Dr. Ordover's adjusted interactive subscription service benchmark to establish the upper boundary of reasonable royalty rates for an SDARS service. SDARS-I, 73 FR at 4093-4094. The use of that benchmark, in SoundExchange's view, demonstrates an intention of the Judges to use total subscription revenue as the base against which the royalty rates should apply. Bender WDT at 6, SX Trial Ex. 75. Applying this assumption to the revenues reported by Sirius XM from 2007 through the third quarter of 2011, SoundExchange concludes that Sirius XM has paid roughly 16%-23% less in total royalty fees than intended. Id. at 6-7. SoundExchange contends that this purported revenue shortfall is due to revenue exclusions that Sirius XM makes under the current Gross Revenues definition which, SoundExchange argues, should not be allowed to continue in the new licensing period.

SoundExchange is particularly critical of a provision of the current definition that allows Sirius XM to deduct revenues received for “[c]hannels, programming products and/or services offered for a separate charge where such channels use only incidental performances of sound recordings.” 37 CFR 382.11 (paragraph (3)(vi)(B) of Gross Revenues definition). According to SoundExchange, this deduction is unwarranted for the new licensing period, because the rates that Dr. Ordover proposed on behalf of SoundExchange and those that Dr. Noll proposed on behalf of Sirius XM reflect that roughly half of the value of Sirius XM's SDARS service is derived from its music programming and roughly half from its non-music programming. According to SoundExchange, the reduction for non-music programming permitted in the current Gross Revenues definition is already built into the proposed rates and should not be further reduced from revenue. SX PFF ¶¶ 839, 845-846.

Further, SoundExchange charges that exclusion of revenue derived from non-music channels encourages manipulation to reduce the royalty base in unprincipled ways. For example, Sirius XM theoretically could disaggregate its bundled subscription price of $14.49 per month into a music package valued at $3.00 and a non-music package valued at $11.49. According to SoundExchange, such a disaggregation would result in no additional Sirius XM revenues from the separate packages, but would enable Sirius XM to reduce substantially its royalty obligation. SX PFF ¶¶ 852-853. SoundExchange suggests that its proposed Gross Revenues definition would help prevent such accounting gimmicks.

SoundExchange also proposes to eliminate from the current Gross Revenues definition a provision that authorizes an exclusion from revenues received from channels and programming that are licensed outside the Sections 112 and 114 licenses, which includes pre-1972 recordings. 37 CFR 382.11 (paragraph (3)(vi)(D) of Gross Revenues definition). Dr. Lys testified that Sirius XM excludes between 10% and 15% of its subscription revenue from the royalty base for performances of pre-1972 recordings, thereby reducing its royalty obligation. Lys WRT at 54, SX Trial Ex. 240. Yet, SoundExchange contends that Sirius XM has not identified the process it uses to identify pre-1972 recordings, or how it calculates the deduction it takes.

SoundExchange's proposed Gross Revenues definition also would eliminate five other exclusions from revenues permissible under the current regulations. First, under the proposed definition, revenues Sirius XM receives from its webcasting service, which are currently linked to the SDARS satellite radio subscription, would be included in the proposed new Gross Revenues definition. Second, revenues attributable to data services, such as Sirius XM's weather and traffic services which can be purchased on a stand-alone basis but are more commonly offered to SDARS subscribers at a discount, would be included in the proposed new Gross Revenues definition. Third, revenue attributable to equipment sales or leases used to receive or play the SDARS service, would be included. Fourth, the current exclusions for credit card fees and bad debt expense would be eliminated. Fifth, fees that Sirius XM collects for various activities related to customer account administration, such as activation fees, invoice fees, swap fees, and certain early termination fees, would be included in Gross Revenues. Bender WDT at 16-17, SX Trial Ex. 75. According to Sirius XM, the elimination of these exclusions could expand its annual revenue base, against which royalties are calculated, by over $300 million. Frear WRT at 7, SXM Reb. Trial Ex. 1.

2. Analysis and Conclusions

In SDARS-I, the parties were at loggerheads over the definition of Gross Revenues, with SoundExchange favoring an expansive reading to include “all revenue paid or payable to an SDARS service that arise from the operation of an SDARS service.” See 73 FR at 4087 (citation omitted). The SDARS in turn argued for adoption of the existing Gross Revenues definition for PSS. Id. With one exception, the Judges adopted the SDARS proposal. See id. SoundExchange's new proposal is again a request for an expansive reading of gross revenues. In its effort to respond to the Judges' criticism of its SDARS-I proposal as possessing “scant evidentiary support,” SoundExchange attempts to demonstrate how Sirius XM has under-reported revenues in the current licensing period. Alternatively, SoundExchange attempts to demonstrate how Sirius XM might manipulate its revenue base to lower its royalty obligation. With the exception of revenue deductions for privately licensed and pre-1972 sound recordings discussed, infra, both of SoundExchange's arguments lack merit.

SoundExchange's argument that Sirius XM has paid roughly 16%-23% less in royalties in the current license period than was intended under the current Gross Revenues definition depends on the assumption that there is a direct link between that definition and the adjusted interactive subscription service benchmark that SoundExchange's expert economist, Dr. Janusz Ordover, presented in SDARS-I. SoundExchange reasons that because the Ordover benchmark was fashioned from interactive service license agreements that generally provided for inclusion of mostly all of subscriber revenue, it must be the case that the Judges intended the SDARS-I rates to apply to total subscription revenues. See Bender WDT at 6-7, SX Trial Ex. 75.

The presumed linkage between the benchmark and the Gross Revenues definition is not supported by the SDARS-I decision for at least two reasons. First, the Ordover benchmark was only one factor the Judges used to establish a zone of reasonable royalty rates, marking the upper boundary. Second, the Judges never adopted the revenue definitions contained in the subscription service license agreements that Dr. Ordover proffered. The Gross Revenues definition the Judges adopted in SDARS-I was quite different from those contained in the Ordover benchmark license agreements.

In defining Gross Revenues, the Judges plainly stated that it was their intention to unambiguously relate the fee charged for a service that an SDARS provided to the value of the sound recording performance rights covered by the statutory licenses. SDARS-I, 73 FR at 4087. This relationship is especially important where, as here, the Judges adopt a percentage of revenue rather than a per-performance rate. The license agreements used in the SDARS-I Ordover benchmark do not provide for this connection between revenue and value under the statutory licenses. In sum, SoundExchange's perceived linkage between the current Gross Revenues definition and the SDARS-I Ordover benchmark favoring inclusion of total subscriber revenues is simply not there.

In the alternative, SoundExchange argues that an expansive revenue base is easier to administer and reduces the chances for manipulation. Dr. Lys testified that, from an accounting perspective, it is preferable to base contracts on a financial definition that is clear-cut to administer and easy to audit, and that a revenue definition that is all-inclusive satisfies this preference. Lys WRT at 53, SX Trial Ex. 240. While this may be true, the Judges are driven by the admonition in SDARS-I to include only those revenues related to the value of the sound recording performance rights at issue in this proceeding. SDARS-I, 73 FR at 4087. The Judges are satisfied that the exclusions permitted in the current Gross Revenues definition remain proper. However, if any party were to present evidence to demonstrate conclusively that one or more of the exclusions facilitates manipulation of fees for the sole purpose of reducing or avoiding Sirius XM's statutory royalty obligation, then an amendment or elimination of the exclusion might be warranted. SoundExchange has failed to meet this burden and has only offered speculation as to how Sirius XM might manipulate its revenue to reduce its royalty obligation. With the exception of the two deductions discussed below, SoundExchange has failed to present persuasive evidence that would warrant the changes it proposes to the calculation of gross revenues.^[45]

B. Deductions for Directly Licensed and Pre-1972 Recordings

Separate from the issue of exclusions from the Gross Revenues definition, the Judges examine the impact on the royalty calculus of the performance by Sirius XM of sound recordings that it has directly licensed from record labels. To broadcast the music offered by the Direct Licensors, Sirius XM can rely on the agreed terms instead of paying under the compulsory statutory licenses. Sirius XM also performs other sound recordings that are exempt from the compulsory licenses in the Copyright Act (i.e., pre-1972 recordings).

1. Directly Licensed Recordings

Both the Section 112 and Section 114 licenses recognize and permit the licensing of sound recordings through private negotiation. 17 U.S.C. 112(e)(5), 114(f)(3). The parties concede that directly licensed recordings are outside and, therefore, not compensable under the statutory licenses. Sirius XM contends that an exclusion from the Gross Revenues definition is necessary for directly licensed recordings; otherwise, it contends, it would be paying twice for performing these works. Sirius XM PFF ¶ 425; Proposed Rates and Terms of Sirius XM Radio, Inc., at 3 (Sept. 26, 2012).

SoundExchange acknowledges that directly licensed recordings must be accounted for, but resists a codified deduction from the Gross Revenues definition. Instead, it proposes that the payable statutory royalty amount be determined by reducing the product of the royalty rate and Gross Revenues by a percentage approximating the value of the directly licensed usage. Second Revised Proposed Rates and Terms of SoundExchange, Inc., at 4 (Sept. 26, 2012). To determine the share of performances attributable to direct licenses, SoundExchange proposes using data from Sirius XM's Internet webcasting service for music. The following calculation would then be made:

• For each month, identify the Internet webcast channels offered by the Licensee that directly correspond to music channels offered on its SDARS that are capable of being received on all models of Sirius radio, all models of XM radio, or both (the “Reference Channels”).^[46]

• For each month, divide the Internet performances of directly licensed recordings on the Reference Channels by the total number of Internet performances of all recordings on the Reference Channels to determine the Direct Licensing Share.

Id. at 4-5 (footnote omitted).^[47]

The Judges are persuaded that directly licensed recordings should not be a part of the calculus in determining the monthly statutory royalty obligation. To include those recordings, for which Sirius XM pays under a separate contract would effectively result in a double payment for the directly licensed recordings and would discourage, if not altogether eliminate, the incentive to enter into such direct licenses. Discouraging direct licensing would be inconsistent with Section 114 which recognizes, if not encourages, private licenses. See 17 U.S.C. 114(f)(1)(B). The Judges are not persuaded, however, by Sirius XM's position that the exclusion of directly licensed recordings should be from Gross Revenues, as opposed to a deduction from the total royalty obligation.

As SoundExchange correctly points out, there is no revenue recognition associated with directly licensed recordings. Those licenses represent a cost to Sirius XM. Bender WRT at 3, SX Trial Ex. 239. Sirius XM has not proposed a revenue allocation formula between directly licensed and statutorily licensed recordings that it performs on its SDARS service; it has presented a usage deduction that it seeks to apply to its revenue base. Excluding usage of sound recordings from Gross Revenues would not comport with the Judges' preference to relate royalty fees to the value of the sound recording performance rights that give rise to the royalty obligation.

The Judges are persuaded that the proposed methodology of SoundExchange to calculate the royalty deduction for directly licensed recordings (i.e. the “Direct License Share”) is the superior approach because it would allow Sirius XM to determine the percentage reduction for directly licensed recordings based upon the number of plays of those recordings compared to total plays. Despite the Judges' requests, Sirius XM and its contractor, Music Reports, Inc., were incapable of providing the Judges with accurate data as to the identity and volume of directly licensed recordings on the SDARS service. SX PFF ¶¶ 883, 886-888. Reasonable accuracy and transparency are required for calculation of the Direct License Share, and SoundExchange has demonstrated that its proposed use of the Sirius XM webcasting service as a proxy satisfies these requirements. The Judges adopt SoundExchange's Direct License Share approach.^[48]

2. Pre-1972 Recordings

The performance right granted by the copyright laws for sound recordings applies only to those recordings created on or after February 15, 1972. Sound Recording Amendment, Public Law 92-140, 85 Stat. 391 (1971). Sirius XM broadcasts pre-1972 recordings on its SDARS service and, in the present license period, excludes a portion of revenues from its Gross Revenues calculation for such use. The current Gross Revenues definition does not expressly recognize such an exclusion, which is not surprising given that there is no revenue recognition for the performance of pre-1972 works. In taking the exclusion, Sirius XM apparently relies upon the provision of the current Gross Revenues definition that permits an exclusion for programming that is exempt from any license requirement. See 37 CFR 382.11 (paragraph (3)(vi)(D) of Gross Revenues definition).

Dr. Lys testified that the deduction is between 10% and 15% of subscription revenue, a figure that Sirius XM did not dispute. Lys WRT at 54, SX Trial Ex. 240. Sirius XM requests that the Judges amend the current Gross Revenues definition to provide that its “monthly royalty fee shall be calculated by reducing the payment otherwise due by the percentage of Licensee's total transmission of sound recordings during the month that are exempt from any license requirement or separately licensed.” Proposed Rates and Terms of Sirius XM Radio Inc. at 3 (Sept. 26, 2012).^[49]

As with directly licensed works, pre-1972 recordings are not licensed under the statutory royalty regime and should not factor into determining the statutory royalty obligation. But, for the same reasons discussed in relation to the Direct Licenses, revenue exclusion is not the proper means for addressing pre-1972 recordings. Rather, the proper approach is to calculate a deduction from the total royalty obligation to account for performances of pre-1972 recordings. The question then becomes how to calculate the correct deduction. Sirius XM did not offer any evidence as to how it calculated its current deduction, or how it identified what recordings performed were pre-1972, other than the obtuse assertion of Mr. Frear that the lawyers talked to the finance team to assure a proper deduction. 8/13/12 Tr. 3125:3-3126:3 (Frear). To be allowable, a deduction for pre-1972 recordings must be precise and the methodology transparent. The Judges, therefore, adopt the same methodology applied to determining the Direct License Share, utilizing as a proxy the Sirius XM webcasting data with the accompanying restriction, to pre-1972 recordings. To be eligible for deduction of the Pre-1972 Recording Share, Sirius XM must, on a monthly basis, identify to SoundExchange by title and recording artist those recordings for which it is claiming the deduction.

VII. Terms

The Judges now turn to the terms necessary to effectuate payment and distribution. The Judges' mandate is this regard is to adopt terms that are practical and efficient. See SDARS-I, 73 FR at 4098. In general, the Judges seek, where possible, consistency across licenses to promote efficiency and minimize costs in administering the licenses. However, this goal is not overriding. See Webcasting III, 76 FR 13026, 13042 (Mar. 9, 2011).

A. Collective

SoundExchange requests to be retained as the sole collective for the collection and distribution of royalties paid by the PSS and SDARS under the Sections 112 and 114 licenses for the license period 2013-17. The PSS and SDARS do not oppose SoundExchange's request. Therefore, SoundExchange will serve as the collective for the 2013-17 license period.

B. Terms Relating to PSS

SoundExchange proposes a number of substantive and nonsubstantive changes to the current regulations dealing with PSS. Music Choice opposes the changes, some in general terms and some specifically.

1. Reorganizing Definitions

SoundExchange proposes collecting applicable PSS definitions in one place for the convenience of the users of the definitions. SX PFF ¶ 906. We believe this proposal, which is nonsubstantive and which Music Choice does not appear to oppose specifically, will enhance the utility of the rules and therefore is adopted.

2. Relocating the Statement of Account Requirement

SoundExchange proposes relocating the statement of account requirement in current § 382.4(b) and to adapt it to include the enumerated data elements from the SDARS regulations. SX PFF ¶ 908; Bender WDT at 22, SX Trial Ex. 75. Music Choice appears not to specifically oppose this change. This change would promote clarity of what information is required in a statement of account and is consistent with the SDARS license. Therefore, the Judges adopt it to enhance the utility and uniformity of the rules across licenses.

3. Applying Late Fee to Late Statement of Account

SoundExchange proposes applying a late fee to a late statement of account, to make the PSS rules consistent with the SDARS and webcasting regulations. SX PFF ¶ 907. Music Choice opposes this proposed change, contending that SoundExchange has not provided any evidence suggesting that Music Choice or Muzak has ever failed to submit a statement of account in a timely manner. Music Choice PFF ¶ 603. The Judges previously imposed a late fee for late statements of account despite the service's record of timely submitted statements of account, reiterating the importance of the timely submission of statements of account to the quick and efficient distribution of royalties. SDARS-I, 73 FR at 4100; see also, Webcasting II, 72 FR at 24107. The Judges adopt the proposed late fee under the same reasoning. The late fee adopted today is consistent with the one imposed on webcasters and the SDARS. See SDARS-I, 73 FR at 4100.

4. Clarifying Unclaimed Funds Provisions

SoundExchange proposes to conform and cross-reference regulations dealing with the use of funds where the Collective is unable to locate a copyright owner or performer who may be entitled to those funds. SX PFF ¶ 909. Because this proposal is unopposed and would enhance the clarity of the rules the Judges adopt it. On a related matter, the Judges encourage SoundExchange to provide greater clarity on what date it uses as the “date of distribution” for unclaimed funds. See, e.g., Proposed 37 CFR 382.8.

5. Changing Confidentiality Provisions

SoundExchange proposes changes to the confidentiality provisions in § 382.5 to make the PSS regulations consistent with those for Business Establishments. SX PFF ¶ 910. Music Choice specifically opposes this proposal because it believes it could cause Music Choice's confidential information to be provided to its competitors. Music Choice PFF ¶ 609. In light of Music Choice's specific opposition and SoundExchange's inadequate justification for adopting the rule the Judges refrain from adopting this proposed change.

6. Conforming Audit Processes

SoundExchange proposes conforming the PSS audit provisions in current §§ 382.5(f) (Verification of statements of account) and 382.6(f) (Verification of royalty payments) with those applicable to SDARS and webcasters. SX PFF ¶ 911. SoundExchange does not support, however, maintaining consistency across licenses with respect to the cost shifting provisions (currently 5% variance for PSS and 10% variance for SDARS and webcasters) and would prefer no change to one that would include cost shifting conformity to the higher variance standard. Music Choice specifically opposes the proposed changes, arguing, among other things, that it would permit SoundExchange to use auditors that are employees or officers of a sound recording owner or performing artists, the objectivity of which might be suspect. Music Choice PFF ¶ 611-613. Given Music Choice's concerns, which SoundExchange has not adequately addressed, and SoundExchange's own reluctance to adopt conforming provisions unless the provisions maintain differences in cost shifting (or the lower variance standard), the Judges refrain from adopting the proposed changes to the auditing provisions. To the extent that one or more of the parties have specific or general concerns about the auditing process with respect to this or other licenses that the Judges administer, the Judges would welcome guidance that might serve to enhance the fairness and efficiency of the process.

7. Technical and Conforming Changes

SoundExchange also proposes a number of technical and conforming changes, which the Judges adopt as proposed with one exception. SX PFF ¶ 912. In particular, the Judges adopt SoundExchange's proposal to relocate the provision regarding retention of records from its current location in § 382.4(f) relating to confidential information to newly renumbered § 382.4(e), which now houses the terms of the license. However, the Judges decline to adopt the proposed language because it would be a substantive change for which SoundExchange provides no justification. Therefore, the language in new § 382.4(e) remains the same as that currently found in § 382.4(f).

C. Terms Relating to SDARS

SoundExchange proposes a number of substantive and nonsubstantive changes to the current regulations dealing with SDARS.^[50] Sirius XM opposes the changes, some in general terms and some specifically.

1. Deleting Residential Subscriber Concept

SoundExchange proposes deleting the concept of “residential” SDARS subscriber in §§ 382.11 and 382.12 of the current regulations. The concept appears in the definitions of “Gross Revenues” and “Residential.” SoundExchange contends that the concept is a “confusing artifact of” a comparable term used in the PSS regulations. SX PFF ¶ 900. SoundExchange argues that the SDARS service is not primarily residential in terms of being delivered to homes and the term “residential subscriber” simply means a subscriber and, therefore, the term “residential” adds no value to the definition and creates the possibility for confusion. Id., citing Bender WDT at 20, SX Trial Ex. 75. Although Sirius XM broadly opposes adopting the changes to terms SoundExchange proposes, it does not expressly state its reasons for opposing this particular change. Given the broad analysis of the definition of “Gross Revenues” the Judges have undertaken in this determination, we are mindful of SoundExchange's concern that potentially modifying that definition with the preceding term “residential” in current § 382.12 could cause unnecessary confusion. Therefore, the Judges adopt SoundExchange's proposal to delete the definition of “residential” from current § 382.11 and the reference to “residential” in current § 382.12(a).

2. Eliminating the Handwritten Signature Requirement for Statements of Account

SoundExchange proposes that the Judges eliminate a requirement in current § 382.13(e)(3) that the signature on a statement of account be handwritten. SX PFF ¶ 901. SoundExchange contends that the current requirement hinders SoundExchange in its ability to automate the process of “ingesting statements of account and reports of use,” which would help reduce transaction costs. Id. Sirius XM does not appear to oppose the request. Although the Judges rejected such a request in Webcasting III due largely to the proposal's inconsistency with certain agreements we adopted in connection with the Webcasting III determination,^[51] the Judges find no such inconsistency here. See Bender WDT at 20-21, SX Trial Ex. 75. In light of desirability of managing administrative costs and the apparent lack of opposition to the proposal, the term is adopted as proposed.^[52]

3. Applying Late Fee to Late Statement of Account

SoundExchange proposes to amend the current late fee requirements for statements of account, 37 CFR 382.13(d), by conforming the language to that adopted by the Judges in Webcasting III, which SoundExchange contends would eliminate confusion that could result from the current language. See SX RFF ¶¶ 360-363. Sirius XM objects to this proposed change arguing that the Judges' justification for adoption of a late fee in SDARS-I—to “aid the efficient distribution of royalties”—no longer exists now that Sirius XM is the only SDAR and its statement of account provides no “additional information that would impact SoundExchange's ability to distribute Sirius XM royalties.” Sirius XM PFF ¶ 449. In addition, Sirius XM proposes that the regulations be amended to “ensure that a single late fee is to be charged in a given reporting period only in the case of a late payment.” Id.

In adopting a late fee for late statements of accounts in SDARS-I, the Judges explained that assessment of an additional late fee for a late statement of account would occur only when the royalty payment and statement of account were submitted separately and both were late; otherwise, a single late fee of 1.5% would cover both the late payment and statement of account when they were submitted together. See SDARS-I, 73 FR at 4100. The Judges find nothing in the record before us that would justify a change to this position; therefore, the Judges decline to adopt Sirius XM's proposal. The Judges adopt SoundExchange's proposed language because it eliminates any inconsistency in the current language. The proposed language provides consistency in the late fee provisions applicable to webcasters and the PSS.

VIII. Final Determination

This Final Determination sets rates and terms for Section 112 and Section 114 royalties to be paid by PSS and SDARS for the compulsory ephemeral license and digital performance license, respectively. The Register of Copyrights may review the Judges' final determination for legal error in resolving a material issue of substantive copyright law. The Librarian shall cause the Judges' final determination for the two subject licenses for the period January 1, 2013, through December 31, 2017, and any correction therefor by the Register, to be published in the Federal Register no later than the conclusion of the 60-day review period.

So ordered.

Suzanne M. Barnett,

Chief Copyright Royalty Judge.

Richard C. Strasser,

Copyright Royalty Judge.

Dated: February 14, 2013.

Dissenting Opinion of Copyright Royalty Judge Roberts

Judge Roberts, concurring with respect to the terms of payment for the PSS and SDARS, and dissenting with respect to the analysis and determination of the royalty rates.

I concur with the analysis and determination of the terms for making royalty payments under the statutory licenses at issue in this proceeding, but not the definition of Gross Revenues to be applied to SDARS in determining the royalty fee, which I do not consider to be a term of payment.^[53] I dissent, however, from the majority's evaluation and analysis of the evidence, and determination of rates for the PSS and SDARS. Rather than engage in a point-by-point discussion and disagreement with the majority's opinion, I set forth below—complete and original to me—what I believe is the proper analysis of the marketplace evidence submitted by the parties for the PSS and SDARS rates, the application of the Section 801(b) factors, and the determination of the rates, including the Gross Revenues definition for SDARS.

I. Introduction

A. Subject of the Proceeding

This is a rate determination proceeding convened by the Copyright Royalty Judges under 17 U.S.C. 803(b) et seq., and 37 CFR part 351 et seq., to establish rates and terms for the digital performance of sound recordings by preexisting subscription services (“PSS”) and preexisting satellite digital audio radio services (“SDARS”) for the license period 2013 through 2017. The Digital Performance Right in Sound Recordings Act of 1995, as amended by the Digital Millennium Copyright Act of 1998, grants to sound recording copyright owners an exclusive right to publicly perform sound recordings by digital audio transmission, subject to the statutory licenses set forth in 17 U.S.C. 112(e) and 114(f)(1) of the Copyright Act. The rates and terms set forth in this Determination are for these statutory licenses.

B. Parties to the Proceeding

The parties to this proceeding are: (1) SoundExchange, Inc. (“SoundExchange”); (2) Music Choice, Inc. (“Music Choice”); and (3) Sirius XM, Inc. (“Sirius XM”). SoundExchange is a Section 501(c)(6) nonprofit performance rights organization that collects and distributes royalties payable to performers and sound recording copyright owners for the use of sound recordings over satellite radio, the Internet, wireless networks, and cable and satellite television networks via digital audio transmissions. Bender WDT at 2, SX Trial Ex. 75. Music Choice (formerly Digital Cable Radio Associates) provides residential music service to subscribers of cable television. Del Beccaro Corrected WDT at 3, PSS Trial Ex. 1. Sirius XM provides satellite radio service broadcasts of music and non-music content on a subscription-fee basis throughout the continental United States. Meyer WDT at 4, SXM Dir. Trial Ex. 5.

C. Procedural History

On January 5, 2011, the Copyright Royalty Judges issued a Notice announcing commencement of this proceeding and requesting the submission of Petitions to Participate. 75 FR 455. Petitions to Participate were received and accepted from the above-described parties. When the negotiation period provided by 17 U.S.C. 803(b)(3) failed to yield any agreements, the Judges called for the submission of written direct statements, which were received by the November 29, 2011 deadline. Hearings on the written direct testimony were conducted from June 5, 2012 through June 18, 2012. Eight witnesses presented testimony on behalf of SoundExchange, three on behalf of Music Choice, and nine on behalf of Sirius XM.

On July 2, 2012, the participants filed their written rebuttal statements. Witness testimony in the rebuttal phase began on August 13, 2012, and concluded on August 23, 2012. Nine witnesses presented testimony on behalf of SoundExchange, two on behalf of Music Choice, and five on behalf of Sirius XM. After close of the rebuttal phase, the parties filed their proposed findings of fact and conclusions of law on September 26, 2012, and their reply findings and conclusions on October 12, 2012.

On October 16, 2012, the Judges heard closing arguments, wherein the record to this proceeding was closed. The record contains several thousands of pages of testimony, exhibits, pleadings, motions and orders.

II. The Standard for Determining Royalty Rates

Section 801(b)(1) of the Copyright Act, 17 U.S.C., provides that the Copyright Royalty Judges shall “make determinations and adjustments of reasonable terms and rates of royalty payments” for the statutory licenses set forth in Sections 114(f)(1) and 112(e). The Section 114(f)(1) digital performance license for the PSS and SDARS, and the Section 112(e) ephemeral license, contain similarities and important differences in their standards for setting royalty rates. Both require the determination of reasonable rates and terms; however, the digital performance license, in Section 801(b)(1), requires that the rates (but not the terms) be calculated to achieve the following objectives:

• To maximize the availability of creative works to the public.
• 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.
• 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.
• To minimize any disruptive impact on the structure of the industries involved and on generally prevailing industry practices.

17 U.S.C. 801(b)(1). The Section 112(e) ephemeral license requires the Judges to “establish rates that most clearly represent the fees that would have been negotiated between a willing buyer and a willing seller,” and further directs that:

• [T]he Copyright Royalty Judges shall base their decision on economic, competitive, and programming information presented by the parties, including—

○ whether use of the service may substitute for or may promote the sale of phonorecords or otherwise interferes with or enhances the copyright owner's traditional streams of revenue; and

○ the relative roles of the copyright owner and the transmitting organization in the copyrighted work and the service made available to the public with respect to relative creative contribution, technological contribution, capital investment, cost, and risk.

17 U.S.C. 112(e)(4). The ephemeral license requires adoption of a minimum fee for each type of service offered by a transmitting organization, while the digital performance license does not. 17 U.S.C. 112(e)(3). Both licenses provide that the Judges may consider the rates and terms of voluntary license agreements negotiated under the licenses. 17 U.S.C. 112(e)(4), 114(f)(1).

It is evident from the presentations of the parties that it is the Section 114(f)(1) license that is of the greater value and concern to their interests, as it was when the Judges last considered the two licenses in 2007. In that Determination, the Judges set forth in great detail the historical treatment of these factors by the Copyright Royalty Tribunal and the Librarian of Congress in his administration of the Copyright Arbitration Royalty Panel system, and I will not repeat them here. See, SDARS-I, 73 FR 4080, 4082-84 (Jan. 24, 2008). Consideration of this history produced the following approach:

[T]he path for the Copyright Royalty Judges is well laid out. We shall adopt reasonable royalty rates that satisfy all of the objectives set forth in Section 801(b)(1)(A)-(D). In doing so, we begin with a consideration and analysis of the benchmarks and testimony submitted by the parties, and then measure the rate or rates yielded by that process against the statutory objectives to reach our decision * * *.

We reject the notion, however, that Section 801(b)(1) is a beauty pageant where each factor is a stage of competition to be evaluated individually to determine the stage winner and the results aggregated to determine an overall winner. Neither the Copyright Royalty Tribunal nor the Librarian of Congress adopted such an approach. Rather, the issue at hand is whether these policy objectives weigh in favor of divergence from the results indicated by the benchmark marketplace evidence.

Id. at 4084, 4094 (citations omitted). The same approach was used by the Judges in determining royalty rates for the Section 115 mechanical license, the only proceeding involving the Section 801(b)(1) factors decided since SDARS-I. See, Phonorecords I, 74 FR 4510 (Jan. 26, 2009). None of the parties in this proceeding contend that this approach is erroneous or must be abandoned.

Music Choice, however, argues that there is an additional factor that must be considered by the Judges, applicable only to the PSS rate, that operates as a limitation on the Judges' consideration of the benchmark evidence. After a lengthy discussion of the Librarian of Congress's PSS-I determination, and citation to the 17 U.S.C. 803(a)(1) proscription that the Judges must act on the basis of prior determinations of the Librarian of Congress, Music Choice contends that the Librarian's use of the musical works benchmark (i.e., the royalty rates paid by Music Choice to the performing rights societies—ASCAP, BMI and SESAC) in 1998 operates as legal precedent in this proceeding and must “be used in the absence of any better comparable benchmark.” Music Choice PCL ¶ 53. Thus, under Music Choice's formulation, the Judges' benchmark analysis must begin with the current royalty fees paid by Music Choice to the performing rights societies musical works and, in the absence of a superior benchmark, employ this benchmark for framing the applicable PSS royalty fee.

I reject Music Choice's argument for several reasons. First, Music Choice does not, and cannot, point to a single statutory license rate proceeding where a court, the Librarian or the Copyright Royalty Tribunal has ruled that a set of factual marketplace observations used by the decisionmaker in formulating a royalty fee for a particular proceeding must be given a priori consideration in a future proceeding. Second, a plain reading of PSS-I makes it clear that the Librarian did not rely solely upon the musical works benchmark, but instead relied upon some unspecified combination of that benchmark and the performance royalty rate contained in a partnership agreement between Music Choice and certain cable television operators and record companies that created Music Choice. PSS-I, 63 FR at 25410. Even if I were inclined to accord some precedential value to the musical works benchmark in this proceeding—and I am not—I could not discern the degree to which that benchmark was influenced or altered by the Librarian's inclusion of the partnership license. Id. at 25404 (“The question, however, is whether this reference point [the musical works fees paid by Music Choice] is determinative of the marketplace value of the performance right in sound recordings; and, as the Panel determined, the answer is no.”). And, third, Music Choice's argument fails to place the PSS-I decision in its historical context. All that was available to the Librarian were the musical works fees paid to the performing rights societies and the partnership license agreement, an unsurprising circumstance given the newness of the statutory license, and the digital music marketplace in general. Concluding that selection of a factual market model from 1998 somehow limits the decisionmakers' consideration of the evidence in 2012 defies logic. I consider the musical works benchmark evidence offered by Music Choice in its normal course, discussed below, but it will not be given preference as a starting point, default position, or other limitation to my evaluation of the benchmark evidence.

III. Determination of the Royalty Rates

A. Application of Section 114 and Section 112

Based upon the applicable law and relevant evidence received in this proceeding, the Copyright Royalty Judges must determine rates for the Section 114(f)(1) digital performance license for the only existing SDARS, Sirius XM, and the PSS.^[54] The Judges also must determine rates for the Section 112(e) ephemeral license for the PSS and SDARS.

With respect to the Section 112(e) license, the Judges received a joint stipulation from the parties. SoundExchange and Music Choice ask for continued application of the language of 37 CFR 382.2(c), which requires a minimum fee advance payment of $100,000 per year, payable no later than January 20 of each year, with royalties paid during the year recoupable against the advance. Joint Stipulation at 2 (May 25, 2012). SoundExchange and Sirius XM ask that the same minimum fee proposal apply to Sirius XM. Id. For the Section 112(e) license fee, all parties request that 5% of the total royalties paid by the PSS and Sirius XM be attributable to the license, consistent with the current regulations applicable to webcasters, broadcasters, SDARS and new subscription services. 37 CFR 380.3, 380.12, 380.22; 382.12; and 383.3.

I accept the stipulations of the parties regarding the Section 112(e) rates, but not for the reason set forth by the majority (i.e., nothing else in the record). The stipulations in this proceeding and, for that matter, in prior proceedings involving the Section 112(e) license, reflect the lack of marketplace evidence as to the value of the license in isolation from that of Section 114. This does not mean, however, that the Section 112(e) license is of no value because marketplace agreements package the rights conferred by the licenses together. The parties' stipulations represent a reasonable attempt to identify the value of the Section 112(e) license if it were marketed separately to copyright users, and for that reason I find the stipulations acceptable.

I now turn to what I view should be the appropriate rate structures for the Section 114(f)(1) license for the PSS and SDARS.

B. The Rate Proposals of the Parties for the Section 114 License for the PSS

Since 1998 when the decision in PSS-I established the initial royalty rates, Music Choice has paid a fee on a percentage basis of its Gross Revenues, as defined by regulation.^[55] Neither Music Choice nor SoundExchange propose altering this rate structure for the 2013-2017 license period,^[56] nor do they propose changes to the revenue definition. SoundExchange requests the following percentage rates for the PSS: for 2013: 15%; for 2014: 20%; for 2015: 25%; for 2016: 35%; and for 2017: 45%. Second Revised Proposed Rates and Terms, at 6 (Sept. 26, 2012). SoundExchange also requests an additional aspect to the percentage of revenue metric to address what it perceives as a deliberate reduction in revenues paid to Music Choice for its residential audio service by certain cable operators that are co-owners (partners) of Music Choice. For transmissions through such a partner, SoundExchange asks that the total royalty fee not be less than the product of multiplying such partner's total number of subscribers to Music Choice's programming by the average per-subscriber royalty payment that Music Choice makes for the top five highest-paying customers of Music Choice that are not its partners. Id. at 7.

Music Choice requests a percentage of Gross Revenues of 2.6%, applicable to each of the years in the licensing period.^[57] Both SoundExchange and Music Choice ask that the definition of Gross Revenues, currently set forth in 37 CFR 382.2(e), apply to the new licensing period.

C. The Rate Proposals of the Parties for the Section 114 License for SDARS

1. Proposed Rates and Structure

While SoundExchange and Music Choice are content to operate mostly the same as in prior licensing periods (with the exception of royalty rates), there is not a similar level of harmony as to the specifics of the rate structure for SDARS. SoundExchange does recommend retention of the percentage of revenue metric, but seeks expansion of the revenue base, and proposes a methodology for exclusion of the value of privately negotiated digital performance licenses from the total statutory royalty fee. Second Revised Proposed Rates and Terms at 3-5 (Sept. 26, 2012). Sirius XM favors maintenance of the current revenue definition and payment scheme. Sirius XM PFF at 203-204.

As with the PSS, SoundExchange argues for an accelerating royalty rate during the five-year license period as follows: for 2013: 12%; for 2014: 14%; for 2015: 16%; for 2016: 18%; and for 2017: 20%. Second Revised Proposed Rates and Terms, at 2 (Sept. 26, 2012). Sirius XM counters with a royalty rate in the range of 5% to 7% of Sirius XM's monthly Gross Revenues, as currently defined in 37 CFR 382.11, applicable to each month of the upcoming licensing period.

2. Proposed Definition of Gross Revenues

The revenue base against which the adopted royalty rates shall be applied is a matter of considerable disagreement between the parties. Sirius XM requests continuance of the current definition of Gross Revenues found in 37 CFR 382.11, while SoundExchange favors a considerable expansion of the revenue base. SoundExchange would like to see Gross Revenues redefined as follows:

1. Gross Revenues shall mean revenues recognized by the Licensee in accordance with GAAP from the operation of an SDARS in the U.S., and shall be comprised of the following:

i. All subscription, activation, subscription-related and other revenues recognized by Licensee from fees paid or payable by or for U.S. subscribers to Licensee's SDARS with respect to any and all services provided by the Licensee to such subscribers, unless excluded by paragraph 3 below;

ii. Licensee's advertising revenues, or other revenues from sponsors, if any, attributable to advertising on channels of Licensee's SDARS in the U.S. other than those that use only incidental performances of sound recordings, less advertising agency and sales commissions attributable to advertising revenues included in Gross Revenues; and

iii. Revenues attributable to the sale, lease or other distribution of equipment and/or other technology for use by U.S. subscribers to receive or play the SDARS service, including any shipping and handling fees therefor.

2. Gross revenues shall include such payments as set forth in paragraphs 1.i through iii of the definition of “Gross Revenues” to which Licensee is entitled but which are paid to a parent, subsidiary or division of Licensee.

3. To the extent otherwise included by paragraph 1, Gross Revenues shall exclude:

i. Royalties paid to Licensee by persons other than subscribers, advertisers and sponsors for intellectual property rights;

ii. Revenues from the sale of phonorecords and digital phonorecord deliveries sold by Licensee (but not any affiliate fees or other payments by a third party for advertising of downloads sold by a third party);

iii. Sales and use taxes;

iv. Revenues recognized by Licensee for the provision of—

A. Data services (e.g. weather, traffic, destination information, messaging, sports scores, stock ticker information, extended program associated data, video and photographic images, and such other telematics and/or data services as may exist from time to time, but not transmission of sound recording data), when such services are provided on a standalone basis (i.e. priced separately from Licensee's SDARS, and offered at the same price both to subscribers to Licensee's SDARS and persons who are not subscribers to Licensee's SDARS);

B. Channels, programming, products and/or other services provided outside of the United States; and

C. Separately licensed services, including webcasting, interactive services, transmissions to business establishments, and audio services bundled with television programming and subject to the rates provided in part 383, when such services are provided on a standalone basis (i.e. priced separately from Licensee's SDARS, and offered at the same price both to subscribers to Licensee's SDARS and persons who are not subscribers to Licensee's SDARS).

Id. at 2-3.

Jonathan Bender, CEO of SoundExchange, testified that the proposed definition will correct the inequities of the current definition and, at the same time, allow for greater ease of administration. His premise is based upon an interpretation of the Judges' use in SDARS-I of Dr. Ordover's adjusted interactive subscription service benchmark to establish the upper boundary of reasonable royalty rates for an SDARS service. SDARS-I, 73 FR at 4093. The use of that benchmark, in Mr. Bender's view, demonstrates an intention of the Judges to use total subscription revenue as the base against which the royalty rates should apply. Bender WDT at 6, SX Trial Ex. 75. Applying this assumption to the revenues reported by Sirius XM from 2007 through the third quarter of -2011, Mr. Bender concludes that Sirius XM has paid roughly 16%-23% less in total royalty fees than intended. Bender WDT at 6-7, SX Trial Ex. 75.

SoundExchange targets for elimination several exclusions from Gross Revenues permitted under the current regulations. The first is paragraph (3)(vi)(B) of the current definition, which allows Sirius XM to deduct revenues received for “channels, programming products and/or services offered for a separate charge where such channels use only incidental performances of sound recordings.” 37 CFR 382.11. This exclusion does not make sense for the new licensing period, according to SoundExchange, because the rates proposed by Dr. Ordover on behalf of SoundExchange and Dr. Noll on behalf of Sirius XM reflect that roughly half of the value of SiriusXM's SDARS service is derived from its music programming and roughly half from its non-music programming. The exclusion for non-music programming is, therefore, built into the rates and should not be double counted in revenue. SX PFF at 383 (¶ 838). Further, SoundExchange charges that exclusion from revenue of non-music channels encourages manipulation to reduce the royalty base in unprincipled ways. For example, Sirius XM could disaggregate its bundled subscription price of $14.49 per month into a music package valued at $3.00 and a non-music package valued at $11.49. Sirius XM would not recognize any additional revenues from the separate packages, but could realize substantial reductions in royalty obligation. SX PFF at 387 (¶ 854). SoundExchange urges the Judges to prevent such arbitrary actions from occurring.

SoundExchange also submits that paragraph (3)(vi)(D) of the current definition should be eliminated. That paragraph allows for deduction of revenues received from channels and programming that are licensed outside the Sections 112 and 114 licenses. 37 CFR 382.11. Dr. Lys testified that Sirius XM excludes roughly between 10% and 15% of its subscription revenue from the royalty base for performances of pre-1972 recordings to its subscribers, thereby reducing its royalty obligation. Lys WRT at 54 (¶ 119), SX Trial Ex. 240. Yet, Sirius XM has not disclosed the process it uses to identify pre-1972 recordings, or how it calculates the deduction it takes.

SoundExchange's proposed Gross Revenues definition also eliminates five other exclusions permissible under the current regulations. First, under its proposed definition, revenues received from Sirius XM's webcasting service, which are currently linked to the SDARS satellite radio subscription, would come into the SDARS revenue base. Second, data services, such as Sirius XM's weather and traffic services which can be purchased on a stand-alone basis but are more commonly offered to SDARS subscribers at a discount, would be included in the revenue base. Third, revenue attributable to equipment sales or leases used to receive or play the SDARS service, would be included. Fourth, the current exclusions for credit card fees and bad debt expense would be eliminated. And, fifth, fees collected by Sirius XM for various activities related to customer account administration, such as activation fees, invoice fees, swap fees, and in certain cases early termination fees, would be included in the revenue base. According to Sirius XM, the elimination of these deductions would, in total and based upon 2012 estimates, expand its annual revenue base, against which royalties are calculated, by over $300 million. Frear Revised WRT at ¶ 16, SX Reb. Trial Ex. 1.

3. Analysis and Conclusions

In SDARS-I, the parties were at loggerheads over the definition of revenue, with SoundExchange favoring an expansive reading to include “all revenue paid or payable to an SDARS service that arise from the operation of an SDARS service,” SoundExchange Third Amended Rate Proposal (Aug. 6, 2007) at section 38_.2(g), and the SDARS arguing for adoption of the existing Gross Revenues definition for PSS. XM Rate Proposal (Jan. 17, 2007) at section 26_.2(d); Sirius Rate Proposal (Jan. 17, 2007) at section 26_.2(d). With one exception, the Judges adopted the SDARS proposal. See SDARS-I, 73 FR 4080, 4087 (Jan. 24, 2008). SoundExchange's new proposal is again a request for an expansive reading of revenue. Its effort to respond to the Judges' criticism of its SDARS-I proposal as possessing “scant evidentiary support,” is an attempt to demonstrate how Sirius XM has under-reported revenue in the current licensing period by applying a slanted interpretation of the SDARS-I decision. Alternatively, it is an attempt to demonstrate how Sirius XM might manipulate its revenue base to lower its royalty obligation. With the exception of revenue exclusions for privately licensed and pre-1972 sound recordings discussed, infra, both of SoundExchange's arguments lack merit.^[58]

Mr. Bender's argument that Sirius XM has paid roughly 16%-23% less in royalties in the current license period than was intended by the Judges in SDARS-I is dependent upon the assumption that there is a direct link between the revenue definition and the adjusted interactive subscription service benchmark presented by SoundExchange's expert economist, Dr. Janusz Ordover. His reasoning is that because the Ordover benchmark was fashioned from interactive service license agreements that generally provided for inclusion of mostly all of subscriber revenue, it must be the case that the Judges intended the SDARS-I rates to apply to total subscription revenues. Bender WDT at 6-7, SX Trial Ex. 75. This presumed linkage between the benchmark and the revenue definition is not supported by the SDARS-I decision. The Judges never expressed an intention to adopt the revenue definitions contained in the subscription service license agreements used by Dr. Ordover; to the contrary, the Gross Revenues definition adopted in SDARS-I was quite different from those contained in the license agreements on the whole. Rather, in defining revenue, the Judges plainly stated that it was their intention to unambiguously relate the fee charged for a service provided by an SDARS to the value of the sound recording performance rights covered by the statutory licenses. SDARS-I, 73 FR at 4087. This is especially important where, as here, a proxy for use of sound recordings must be adopted because technological impediments do not permit implementation of a per-performance fee. The license agreements used in the Ordover benchmark do not provide for this connection between revenue and value under the statutory licenses, which is not surprising given that the interactive service license agreements conveyed rights beyond those granted by the Section 114 license. In sum, Mr. Bender's perceived linkage between the revenue definition and the Ordover benchmark favoring inclusion of total subscriber revenues is simply not there.

In the alternative, SoundExchange argues that an expansive revenue base is easier to administer and reduces the chances for manipulation. Dr. Lys testified that, from an accounting perspective, it is preferable to base contracts on a financial definition that is clear-cut to administer and easy to audit, and that a revenue definition that is all-inclusive satisfies this preference. Lys WRT at 54 (¶ 117), SX Trial Ex. 240. While this may be true, SDARS-I included only those revenues related to the value of the sound recording performance rights at issue in the proceeding. SDARS-I, 73 FR at 4087. I am satisfied that the exclusions permitted in the current definition of Gross Revenues remain proper. However, if evidence were presented to conclusively demonstrate that one or more of the exclusions produces or results in a manipulation of fees for the sole purpose of reducing or avoiding Sirius XM's statutory royalty obligation, then an amendment or elimination of the exclusion might be appropriate. SoundExchange has failed to meet this burden and has only offered speculation as to how Sirius XM might engage in revenue allocation to reduce its royalty obligation. With the exception of the exclusions for directly licensed and noncompensable sound recordings discussed infra, SoundExchange has failed to present persuasive evidence that any of the remaining exclusions it has targeted for elimination (fees received for non-music services,^[59] webcasting, data services, equipment sales, credit card fees and bad debt expenses, and customer account fees) have, in fact, been abused or otherwise manipulated for the sole purpose of improperly reducing Sirius XM's statutory royalty obligations.

4. Deductions for Directly Licensed and Pre-1972 Recordings

Separate from the issue of exclusions from the Gross Revenues definition addressed above, I consider the impact to the royalty calculus of the performance by Sirius XM of sound recordings that it has directly licensed from record labels (and, therefore, does not rely upon the licenses offered by Sections 112 and 114), and the performance of sound recordings not compensable under the Copyright Act (i.e. pre-1972 recordings).

a. Directly Licensed Recordings

Both the Section 112 and Section 114 licenses recognize and permit the licensing of sound recordings through private negotiation. 17 U.S.C. 112(e)(5), 114(f)(3). The parties concede, as they must, that directly licensed recordings are separate from those covered by the statutory licenses. At the outset, the parties did not address the treatment of such recordings in the context of this proceeding, presumably because they comprised only a small percentage of the total recordings performed by Sirius XM in a given period. However, due to the increasing instances of directly licensed recordings as a result of Sirius XM's Direct Licensing Initiative, discussed infra, proposals were submitted and amended up until the closing of the record. Sirius XM contends that a deduction from Gross Revenues is necessary for directly licensed recordings; otherwise, a double payment would occur for performances of these works. Sirius XM PFF ¶ 425; Proposed Rates and Terms of Sirius XM Radio, Inc. at 3 (Sept. 26, 2012). SoundExchange acknowledges that directly licensed recordings must be accounted for, but resists a deduction from Gross Revenues. Instead, it proposes that the payable statutory royalty amount be determined by reducing the product of the royalty rate times Gross Revenues by a percentage approximating the value of the directly licensed usage. Second Revised Proposed Rates and Terms, at 4 (Sept. 26, 2012). To determine the share of performances attributable to direct licenses, SoundExchange recommends using data from Sirius XM's Internet webcasting service for music. The following calculation would then be made:

• For each month, identify the Internet webcast channels offered by the Licensee that directly correspond to music channels offered on its SDARS that are capable of being received on all models of Sirius radio, all models of XM radio, or both (the “Reference Channels”).^[60]

• For each month, divide the Internet performances of directly licensed recordings on the Reference Channels by the total number of Internet performances of all recordings on the Reference Channels to determine the Direct Licensing Share.

Id. at 4-5. SoundExchange requests that, if its approach is adopted, Sirius XM be required to notify SoundExchange monthly of each copyright owner from which Sirius XM claims to have a direct license and each sound recording Sirius XM claims to be excludable. SoundExchange would then be permitted to disclose this information to confirm whether the direct license exists and the claimed sound recordings are properly excludable.

Directly licensed recordings should not be a part of the calculus in determining the monthly statutory royalty obligation. To do otherwise would effectively result in a double payment for the directly licensed recordings and would discourage, if not altogether eliminate, the incentive to enter into such private licenses, contrary to Section 114 which recognizes, if not encourages, private licenses. 17 U.S.C. 114(f)(1)(B). I am not persuaded, however, by Sirius XM's position that the exclusion of directly licensed recordings should be from Gross Revenues, as opposed to a deduction from the total royalty obligation. As Mr. Bender correctly points out, there is no revenue recognition associated with directly licensed recordings; it is a cost to Sirius XM. Bender WRT at 3, SX Trial Ex. 239. Sirius XM has not presented a revenue allocation between directly licensed and statutory licensed recordings that it performs on its SDARS service; it has presented a usage deduction that it seeks to apply to its revenue base. Excluding usage of sound recordings from Gross Revenues would not comport with the Judges' approach in SDARS-I of unambiguously relating fees received by Sirius XM to the value of the sound recording performance rights at issue in this proceeding.

I am persuaded that the proposed methodology of SoundExchange to calculate the royalty deduction for directly licensed recordings (i.e., the “Direct License Share”) is the superior approach to allowing Sirius XM to determine the percentage reduction based upon the number of plays of directly licensed recordings to total plays. Despite my request, Sirius XM and its contractor, Music Reports, Inc., were incapable of providing accurate data as to the identity and volume of directly licensed recordings on the SDARS service. See, SX PFF at 399-400. Reasonable accuracy and transparency are required for calculation of the Direct License Share, and SoundExchange has demonstrated that its proposed use of the Sirius XM webcasting service better satisfies these requirements. Use of the webcasting service also better ties the value of the sound recordings used, by measuring the listenership for each performance, than Sirius XM's proposal for measuring only individual plays.

b. Pre-1972 Recordings

The performance right granted by the copyright laws for sound recordings applies only to those recordings created on or after February 15, 1972. Sound Recording Amendment, Public Law 92-140, 85 Stat. 391 (1971). Sirius XM makes performances of pre-1972 recordings on its SDARS service and, in the present license period, excludes a percentage of revenues from its Gross Revenues calculation for such use. The current Gross Revenues definition does not expressly recognize such an exclusion, which is not surprising given that there is no revenue recognition for the performance of pre-1972 works. In taking the exclusion, Sirius XM apparently relies upon paragraph (3)(vi)(D) of the Gross Revenues definition which permits exclusion of revenue for programming exempt from any license requirement. Dr. Lys testified that the deduction is between 10% and 15% of subscription revenue, a figure that was not disputed by Sirius XM. Lys WRT at 54, SX Trial Ex. 240. Sirius XM requests that the Judges amend paragraph (3)(vi)(D) to provide that its “monthly royalty fee shall be calculated by reducing the payment otherwise due by the percentage of Licensee's total transmission of sound recordings during the month that are exempt from any license requirement or separately licensed.” Proposed Rates and Terms of Sirius XM Radio Inc. at 3 (Sept. 26, 2012).^[61]

As with directly licensed works, pre-1972 recordings are not licensed under the statutory royalty regime and should not factor into determining the statutory royalty obligation. But, for the same reasons described above, revenue exclusion is not the proper means for addressing pre-1972 recordings. Rather, the proper approach is deduction from the total royalty obligation to account for performances of pre-1972 recordings. The question then remains as to how the correct deduction should be calculated. Sirius XM did not offer any testimony as to how it calculated its current deduction, or how it identified what recordings performed were pre-1972, other than the obtuse assertion of Mr. Frear that the lawyers talked to the finance team to assure a proper deduction. 8/13/12 Tr. 3125:3-3126:3 (Frear). To be allowable, a deduction for pre-1972 recordings must be relatively precise and the methodology transparent. The same methodology applied to determining the Direct License Share—utilizing the Sirius XM webcasting—is appropriate to identify pre-1972 recordings.

D. The Section 114 Royalty Rates for PSS

Chapter 8 and Section 114(f)(1) of the Copyright Act require the Judges to determine reasonable rates and terms of royalty payments for the digital performance of sound recordings. The rates the Judges establish under Section 114(f)(1) must be calculated to achieve the objectives set forth in Section 801(b)(1)(A) through (D) of the Act. Moreover, in establishing rates and terms the Judges may consider voluntary license agreements described in Section 114(f)(1)(B).

As the Judges have done in prior rate proceedings where the determination standard is reasonable rates calculated to achieve the Section 801(b)(1) factors, consideration of marketplace benchmarks is a useful starting point. SDARS-I, 73 FR 4080, 4088 (Jan. 24, 2008); Phonorecords I, 74 FR 4510, 4517 (Jan. 26, 2009). As discussed below, the parties disagree about what constitutes the most appropriate benchmark to guide the Judges in determining a reasonable rate. Unfortunately, there are no voluntary license agreements negotiated under Section 114(f)(1)(B) for the Judges to consider, which is not surprising considering that Music Choice is the primary PSS service that continues to operate under the statutory license. Moreover, the benchmarks offered by the parties are not for similar products drawn from a marketplace in which buyers and sellers are similarly situated. I describe and discuss them below.

1. PSS Proposed Benchmarks

a. Proposed Musical Works Benchmark

As previously discussed, Music Choice argues that the annual royalties it pays to the three performing rights societies (ASCAP, BMI, and SESAC) for the right to perform musical works to subscribers of its residential audio service is, by virtue of the Librarian's determination in PSS-I, a precedential benchmark that establishes the upper boundary of reasonable rates in this proceeding. Although this contention has been rejected, supra, Music Choice offers the testimony of Mr. Del Beccaro and Dr. Crawford as corroborative of its position that the market for licensing the performance right in musical works is the most appropriate benchmark for establishing rates in this proceeding.

Music Choice pays 2.5% of revenue each to ASCAP and BMI and pays an annual flat fee to SESAC that amounts to approximately [REDACTED] of net revenue. Del Beccaro Corrected WDT at 21-22, MC 17, MC 18 and MC 19, PSS Trial Ex. 1. The sum of those licenses amounts to [REDACTED], which Music Choice submits should represent the upper bound of a reasonable royalty rate. Two pieces of evidence, in Music Choice's view, corroborate the use of musical works licensing as a benchmark. First, Music Choice observes an equivalence between the fees for the performance of sound recordings and musical works in Canada and Europe. Music Choice cites four decisions of the Canadian Copyright Board, involving licensing fees for commercial radio, cable television, satellite music services and radio services of the Canadian Broadcasting Corporation, wherein the Board found that royalty rates for sound recordings and musical compositions have equivalent value. Del Beccaro Corrected WDT at MC 6 at 30-33 (commercial radio), MC 7 at 14 (cable television), MC 8 at 50, 58 (satellite music services), MC 9 at 4, 6, 15, 17, 30 (CBC radio services), PSS Trial Ex. 1. SoundExchange's expert economist, Dr. George Ford, who recently submitted testimony before the Canadian Copyright Board, acknowledges that in Canada the musical composition and sound recording performance royalties are equal. 8/21/12 Tr. 4304:5-22 (Ford). In the United Kingdom, the sound recording performance royalty rates for commercial broadcasting services are less than those for the musical composition performance rights. Del Beccaro Corrected WDT at MC 11, PSS Trial Ex. 1. If Music Choice's service were transmitted through cable in the U.K., Music Choice would pay 5.25% of 85% of gross revenues for the musical works performance right, but would pay only 5% of 85% of gross revenues for the sound recording performance right. Id.

The second piece of evidence to corroborate use of the musical works rate as a benchmark is an economic model called the Asymmetric Nash Bargaining Framework (referred to as the “Nash Framework”) offered by Dr. Crawford. Acknowledging that a perfect benchmark does not exist to determine the PSS sound recording performance rate, Dr. Crawford uses the Nash Framework to fashion solutions to bargaining problems between bilateral monopolists, in this case record labels on the one hand and PSS providers on the other. Crawford Corrected WDT at 12, PSS Trial Ex. 4. As a non-cooperative bargaining model, the Nash Framework is designed to yield predictions about how outcomes are determined when firms negotiate; that is, how two firms would split the surplus of their interaction (i.e., revenues over costs) in a hypothetical negotiation. Id. at 16. Three factors (the Nash factors) are analyzed to determine the split: (1) the combined agreement surplus; (2) each firm's threat point; and (3) each firm's bargaining power. Id. According to Dr. Crawford, the Nash factors determine sound recording performance royalties in the following way: “The royalty received by each firm in a bargain equals its threat point plus its bargaining power times the incremental surplus.” Id. at 17. In other words, the combined agreement surplus and threat points determine the “size of the pie,” while the bargaining power determines the “split of the pie.”

Dr. Crawford's stated goal in applying the Nash Framework is to first establish the Nash factors for the hypothetical market (the sale of rights between one record company and one PSS provider) and compare them to the Nash factors in the actual musical works market (the sale of rights between the three performing rights societies and one PSS provider). Id. at 18. In the hypothetical market, Dr. Crawford determined that the threat point for the PSS provider is zero because in the absence of an agreement, it cannot offer music and therefore cannot earn a surplus. Id. at 19. He determines, however, that the threat point for a record company is negative because the failure to reach an agreement has additional implications for the record company in other, non-PSS markets. Specifically, the failure to reach an agreement with the PSS provider would have substantial adverse impacts on the record company, such as on sales of compact disks, because there is a significant promotional benefit to the record company from the PSS provider. Id. To support this contention, Music Choice offers the testimony of Damon Williams, who testified that record company executives consider Music Choice promotional and provide artists with greater exposure. Williams WDT at 4-11, MC 28, 29, 32, PSS Trial Ex. 3. Mr. Williams offers examples of how Music Choice conducts custom promotions for artists, id. at 13-20, and points to a 2005 Arbitron survey in particular that he argues confirms that Music Choice's residential audio service sells records. Id. at 13.^[62] And Mr. Williams argues that Music Choice has become more promotional since the PSS-I proceeding by virtue of the fact that it currently reaches more customers with more channels. Id. at 24.

With respect to the last Nash factor, bargaining power, Dr. Crawford assumes it to be neutral, based upon Music Choice's existing technology platform and contract, which cannot be easily replaced or replicated, and his observations of Music Choice's bargaining efforts for sound recording performance rights with respect to music videos. Crawford Corrected WDT at 15-16, PSS Trial Ex. 4.

Applying the Nash factors to the existing market for the PSS musical works performance right, Dr. Crawford determines that the threat point for a PSS provider is again zero, and is again negative for the performing rights society (ASCAP, BMI or SESAC) because the loss of promotional value from the PSS provider produces loss of profits from other markets. Id. at 28. Dr. Crawford again assumes equal bargaining power between the PSS provider and the performing rights society, based largely upon his observations that the two possess equal patience in their negotiations. Id. at 29. This results in a 50/50 split of the surplus, the same conclusion he reached with respect to the hypothetical market. Because of the similarities between the Nash factors in the hypothetical market and the market for musical works, Dr. Crawford concludes that the musical works market makes for a good benchmark for the hypothetical sound recording performance right market at issue in this proceeding. Id. at 30.^[63]

b. Proposed Alternative Surplus Splitting Analysis

As an alternative, Dr. Crawford provided a surplus splitting analysis to corroborate the reasonableness of Music Choice's rate proposal by using financial results to construct an estimate of the profits that would be shared in a royalty payment. Crawford Corrected WDT at 43, PSS Trial Ex. 4. Dr. Crawford adjusted Music Choice's 2006-2010 operating profit to remove the actual royalty paid by Music Choice for sound recording performance rights, and then applied a capital asset pricing model ^[64] to derive an expected rate of return on assets of 8.33%. Crawford Corrected WDT at Appendix B.4, PSS Trial Ex. 4. He then multiplied the 8.33% rate by Music Choice's average operating assets to determine cost of capital, and then subtracted cost of capital from the royalty-adjusted operating profits to derive the residual profits for each year. Id. at 47. This showed that Music Choice's cumulative returns in excess of its cost of capital, but before payment of sound recording royalties, amounts to 3.05% of Music Choice's 2006-2010 royalties. Id. A 50/50 split of this surplus results in a royalty payment of 1.52% of residential audio revenues. Id. at 48. He then applied a range of 20% to 80% of the expected surplus to determine a range of reasonable royalties from 0.61% to 2.43%, not to exceed the 3.05% expected surplus. Id.

2. SoundExchange Proposed Benchmarks

SoundExchange does not offer a single market benchmark to set the royalty rates to be paid by Music Choice for the sound recording performance right, and instead offers rates from over 2,000 marketplace agreements, representing a variety of rights licensed, in an effort to frame a zone of reasonable rates. Dr. Ford observes that PSS like Music Choice have certain distinctive features that make it difficult to identify a suitable benchmark market. Ford Second Corrected WDT at 12, SX Trial Ex. 79. First, Music Choice does not sell its service directly to subscribers, but rather to cable television operators who then bundle the Music Choice programming with a package of video programming for ultimate sale to subscribers. Music Choice is, therefore, an intermediary between cable operators and their subscribers, unlike any of the digital music services the Judges have previously dealt with. Ford Second Corrected WDT at 12, SX Trial Ex. 79; 6/18/12 Tr. 2810:20-2811:3 (Ford). Second, Music Choice's service is almost always bundled with a hundred or more channels of video and is almost never sold on a stand-alone basis. Ford Second Corrected WDT at 13, SX Trial Ex. 79. This makes it difficult to determine the specific consumer value for Music Choice's programming alone. Id.

Given these difficulties, Dr. Ford uses an all-inclusive approach of examining royalty rates for different digital music markets: portable and non-portable interactive subscription webcasting, cellular ringtones/ringbacks, and digital downloads. Id. at 15-16, Table 1. Most of the over 2,000 licensing agreements he examined across these markets calculate royalties based on a “greater of” methodology that includes a per-play royalty fee, a per-subscriber fee, and a revenue-based fee. For convenience, Dr. Ford analyzed only the revenue-based fees, judging his results to be conservative because the other two payment metrics might produce a larger total royalty fee than the revenue-based calculation. 6/18/12 Tr. 2861:3-13 (Ford). His results reveal a percentage of revenue rate of 70% for digital downloads, 43% to 50% for ringtones/ringbacks, and 50% to 60% for portable and non-portable interactive subscription webcasting, respectively. Ford Second Corrected WDT at 15-16, Table 1, SX Trial Ex. 79. According to Dr. Ford, the rate proposal of SoundExchange for PSS comports well with the range established by these agreements, in that it rises above the lowest average rate (43%) only in the last year of the licensing term, and therefore can “be presumed to be a reasonable proxy for a market outcome.” Ford Second Corrected WDT at 16, SX Trial Ex. 79; 6/18/12 Tr. 2831:8-15 (Ford).

3. Analysis and Conclusions Regarding the Proposed Benchmarks

Based upon the evidence put forward in this proceeding, none of the proposed benchmarks provide a satisfactory means for determining the sound recording performance royalty to be paid by Music Choice.

Turning first to Music Choice's arguments in favor of the musical works benchmark, I find them severely wanting. The fees paid to the three performing rights societies for the performance right to musical works have been offered in several other proceedings before the Judges and have been rejected consistently. Webcasting II, 72 FR 24084 (May 1, 2007); SDARS-I, 73 FR 4080 (Jan. 24, 2008); see, also Webcasting I, 67 FR 45240 (July 8, 2002)(Librarian of Congress's determination). The primary reason for the benchmark's rejection is the lack of comparability to the target market for sound recording performance rights. Dr. Crawford, who advocates the appropriateness of the musical works market, acknowledges that a benchmark market should involve the same buyers and sellers for the same rights. Crawford Corrected WDT at 24, PSS Trial Ex. 4. However, the musical works market involves different sellers (performing rights societies versus record companies) selling different rights. See SDARS-I, 73 FR at 4089. The fact that a PSS needs performance rights to musical works and sound recordings to operate its service does not make the rights equivalent, nor does it say anything about their values individually. Further, as in previous proceedings, the evidence establishes that the market commands higher royalty fees for the licensing of sound recordings than musical works. Aaron Harrison presented a chart demonstrating the different average royalty fees that Universal Music Group, one of the major record labels, receives for digital downloads, ringtones, on-demand music videos and portable subscription services, all of which are considerably higher than the fees received by the performing rights societies.^[65] Harrison Corrected WRT at 13-14, PSS Trial Ex. 32. Dr. Ford made similar observations. Ford Amended/Corrected WRT at 7, SX Trial Ex. 244. I am once again led to the conclusion that use of the musical works market as a benchmark is fraught with flaws and only indicates that a reasonable rate for sound recordings cannot be as low as the musical works rate. See, SDARS-I, 73 FR at 4090.

Music Choice's efforts to corroborate the sufficiency of the musical works benchmark with a comparison to foreign rates also are unavailing. The Judges have considered before the significance of foreign countries' treatment of the licensing of exclusive rights granted by copyright. In the proceeding to set rates and terms for the reproduction of musical compositions under the Section 115 license of the Copyright Act, certain licensees offered evidence of license rates in the U.K., Canada and Japan. See Phonorecords I, 74 FR 4510, 4521 (Jan. 26, 2009). In rejecting the foreign rate benchmarks, the Judges stated that attempts at comparison of U.S. rights with foreign rights “underline the greater concern 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.” Id. at 4522. Neither Mr. Del Beccaro nor Dr. Crawford even attempt an analysis or discussion of the intricacies of Canadian and U.K. markets for performance rights for musical works and sound recordings, and Music Choice itself concedes that particular license rates in Canada and Europe “do not necessarily determine what the specific market rate in the United States should be for the sound recording right.” Music Choice PFF ¶ 135.

Likewise, I am not persuaded that Dr. Crawford's application of the Nash Framework provides corroboration. The Nash Framework is a highly theoretical concept whose goal is to evaluate how the surplus from a transaction might be divided among participants. As Dr. Ford points out, a problem with applying the Nash Framework to a determination of a royalty rate is that a royalty does not split surplus, it splits revenues. Ford Amended/Corrected WRT at 8, SX Trial Ex. 244. An even split of surplus, as Dr. Crawford presumes from the model, does not imply an even split of revenues. Id. Further, Dr. Crawford's efforts to apply the Nash Framework to royalties to be paid by Music Choice only contemplates a two-party transaction between record labels and Music Choice, even though Music Choice is the intermediary between cable operators that actually perform the sound recordings in the output market. The presence of an intermediary disrupts and complicates the Nash analysis because it introduces an additional bargain in the output market and requires that all three bargains be considered jointly. Id. at 15. Dr. Crawford did not take this complicating factor into consideration.

I also have serious reservations concerning Dr. Crawford's assumption that the Nash factor of bargaining power is assumed to be neutral. Mr. Del Beccaro testified that Music Choice has a number of competitors in the marketplace, meaning that record companies have other alternatives for licensing their works. Del Beccaro Corrected WDT at 36-37, PSS Trial Ex. 1. This undermines Dr. Crawford's determination of the Nash Factor threat point to the surplus received by record companies in the event no agreement is reached. If record companies have other options, then the assumed zero sum effect of the bargaining agreement under the Nash Framework is violated.

Finally, Dr. Crawford places undue reliance on the perceived promotional value of Music Choice, which is central to his application of the Nash Framework. For his conclusion to be correct—that failure to reach a bargaining agreement will result in a substantial loss of record sales due to the absence of promotional value from Music Choice—he must demonstrate a causal relationship between Music Choice's promotion of sound recordings and the sale of those recordings. His evidence on this point, however, is mostly anecdotal and weak. The surveys relied upon by Mr. Williams do not confirm a causal link between listenership to Music Choice and subsequent record sales; at best, the 2005 Arbitron survey (already more than seven years old) demonstrates that there is some correlation between listenership and sales. There could be many reasons for the correlation, including the possibility that cable subscribers who listen to Music Choice are already inclined to purchase more music. For example, the 2010 Experian Simmons survey, cited by Mr. Williams, shows that Music Choice listeners are more likely than the average person to attend concerts, know what songs are in the top 10, read Rolling Stone magazine, and consume electronic and video goods at a higher rate. Williams WDT at MC 36, PSS Trial Ex. 3. Furthermore, none of the surveys cited by Dr. Crawford, including the antiquated 2006 Sony BMG Music Lab survey, offer reliable evidence as to whether Music Choice's residential audio service creates a net promotional or substitutional effect on the purchase of CDs or other music services. Without reliable data that quantifies the net effect of Music Choice, Dr. Crawford's conclusion regarding Music Choice's promotional effect is not sustainable.

I am not persuaded that Dr. Crawford's Nash Framework analysis confirms acceptance of the musical works benchmark for PSS, nor that royalty rates in the market for sound recordings is less than that for musical works. Likewise, I do not agree that Dr. Crawford's alternative surplus splitting analysis is probative. The Judges have previously found theoretical surplus splitting models to be of limited value, and Dr. Crawford's analysis is no different. See, Webcaster II, 72 FR 24084, 24092-93 (May 1, 2007); SDARS-I, 73 FR at 4092. Although Dr. Crawford claims that his 20% to 80% range of a split of 3.05% of Music Choice's 2006-2010 revenues reflects arm's length negotiations between Music Choice and record companies, he provides no market evidence to support this contention. Crawford Corrected WDT at 49-50, PSS Trial Ex. 4. There are also methodological difficulties in the manner in which Dr. Crawford examined Music Choice's historical financial data. Specifically, he included in his cost analysis those costs associated with Music Choice's music video business in addition to the costs for the residential audio business, presumably because he was told by Music Choice personnel that it was not possible to allocate expenses between the video and audio components of the company's business. 6/12/12 Tr. 1859:21-1860:21 (Crawford). The net effect of including the music video business, which has substantial costs and not much revenue, is to drive down the surplus he proposes to be split between Music Choice and record companies. Even if I were persuaded in theory by Dr. Crawford's surplus splitting analysis—and I am not—his failure to confine his cost and revenue analysis solely to the residential audio business, which is the subject of the statutory licenses in this proceeding, prohibits its usefulness.

Turning to the music service benchmarks offered by SoundExchange and supported by Dr. Ford, one is confronted with severe theoretical and structural difficulties. Although the volume (over 2,000) of marketplace agreements examined by Dr. Ford for music products and services might suggest real usefulness in a benchmark analysis, the four markets examined—portable and non-portable subscription interactive webcasting, ringtones/ringbacks, and digital downloads—involve the licensing of products and rights separate and apart from the right to publicly perform sound recordings in the context of this proceeding. Thus, the key characteristic of a good benchmark—comparability—is not present. SDARS-I, 73 FR at 4092. The buyers are different, there are different music products included (ringtones and ringbacks, digital downloads) and there are different rights licensed in the output market. Further, I do not accept Dr. Ford's contention that, as a matter of economics, it is irrelevant that different legal rights are conveyed by the benchmark agreements he examined. 6/18/12 Tr. 2819:5-10 (Ford). The agreements examined by Dr. Ford themselves suggest that the rights licensed, and the context in which they are licensed, make a great deal of importance in determining their value.

I do agree with Dr. Ford's observations that Music Choice has several distinct features, such as its intermediary role between cable systems and subscribers and the bundling of Music Choice's services with multiple channels of video and other non-music programming, that significantly dim the possibility of market comparators. This is not to say that the value of the sound recording right in the PSS market is exceedingly low, as Music Choice would have it, nor exceedingly high, as SoundExchange would have it. SoundExchange's rate proposal begins with a rate of 15% of Gross Revenues in the first year of the licensing term, which is endorsed by Dr. Ford as being within the range of reasonable rates for the PSS even though it is far lower than the average rates he determined in his benchmark analysis. For this reason, the 15% rate represents nothing more than the uppermost bound of the range of reasonable royalty rates for the PSS.

Based upon the above analysis, I am left with a consideration of the existing 7.5% royalty rate which is the product of settlement negotiations that occurred in SDARS-I between Music Choice and SoundExchange, and is a rate for which neither party advocates. Although it is a rate that was negotiated in the shadow of the statutory licensing system and cannot properly be said to be a market benchmark, nothing in the record persuades me that 7.5% of Gross Revenues, as currently defined, is either too high, too low or otherwise inappropriate. Accord, Phonorecords I, 74 FR at 4522. I now turn to the Section 801(b) policy factors.

4. The Section 801(b) Factors

Section 801(b)(1) of the Copyright Act states, among other things, that the rates that the Judges establish under Section 114(f)(1) 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 conditions; (C) to reflect the relative roles of the copyright owner and the copyright user in the product being made available to the public with respect to relative creative contribution, technological contribution, capital investment, cost, risk, and contribution to the opening of 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 practice. 17 U.S.C. 801(b)(1). Based on the record evidence in this proceeding, the benchmark evidence submitted by Music Choice and SoundExchange has failed to provide the means for determining a reasonable rate for the PSS, other than to indicate the extreme ends of the range of rates. The testimony and argument of Music Choice demonstrates nothing more than to show that a reasonable rate cannot be as low as the rates paid by Music Choice to the three performing rights societies for the public performance of musical works. The benchmark testimony of SoundExchange is of even lesser value. The proposed rate of 15% for the PSS for the first year of the licensing period, deemed reasonable by Dr. Ford (at least in the beginning of the licensing period), stands as the absolute upper bound of the range of reasonable rates. At the middle of the range is the current 7.5% rate and, based upon the record, I am are persuaded that it is neither too high, too low, or otherwise inappropriate, subject to consideration and necessary adjustment under the Section 801(b) factors discussed below.

a. Maximize Availability of Creative Works

Both SoundExchange and Music Choice presented arguments as to how their proposed benchmark rates satisfy this factor, which are not relevant given that the musical works benchmark and the Ford music service benchmarks only serve the purpose of framing the absolute lower and upper bounds of reasonable rates. Rather, it is the current 7.5% rate to which the evidence presented under this factor must be applied.

Music Choice touts that it is a music service that is available in over 54 million homes, with 40 million customers using the service every month. 8/16/12 Tr. 3878:3 (Del Beccaro); Del Beccaro Corrected WDT at 4, 26, PSS Trial Ex. 1; 6/11/12 Tr. 1462:5-11, 1486:19-1487:2 (Del Beccaro). Channel offerings have increased through the years, curated by experts in a variety of music genres. Del Beccaro Corrected WDT at 3, 24, PSS Trial Ex. 1. Recent developments in technology permit Music Choice to display original on-screen content identifying useful information regarding the songs and artists being performed at any one time. Id. at 24; Williams WDT at 12, MC 23, PSS Trial Ex. 3; 6/11/12 Tr. 1461:14-1462: 1, 1491:1-12 (Del Beccaro). According to Music Choice, these elements, along with the promotional efforts detailed above in the context of Dr. Crawford's Nash Framework analysis, support a downward adjustment in the rates. In any event, an upward adjustment in the rates, argues Music Choice, would not affect the record companies' bottom-line because PSS royalties are not a material revenue source for record companies. Music Choice PFF ¶¶ 409-417.

SoundExchange submits that a market rate incorporates considerations under the first Section 801(b) factor, citing the Judges decision in SDARS-I, and that if PSS rates turn out to be too high and drive Music Choice from the market, presumably consumers will shift to alternative providers of digital music where higher royalty payments are more likely for record companies. Ford Second Corrected WDT at 19-20, SX Trial Ex. 79.

The current PSS rate is not a market rate so that market forces cannot be presumed to determine the maximum amount of product availability consistent with the efficient use of resources. See SDARS-I, 73 FR 4094. However, the testimony demonstrates that Music Choice has not, under the current rate, reduced its music offerings or contemplated exiting the business; in fact, it will be expanding its channel offerings in the near term. There is also no evidence that suggests that the output of music from record labels has been impacted negatively as a result of the current rate. There is no persuasive evidence that a higher PSS royalty rate will necessarily result in increased output of music by the record companies (major or independent), nor that a lower rate will necessarily further stimulate Music Choice's current and planned offerings. In sum, the policy goal of maximizing creative works to the public is reasonably reflected in the current rate and, therefore, no adjustment is necessary.

b. Afford Fair Return/Fair Income Under Existing Market Conditions

Music Choice submits that the Judges need not worry about the impact of a low royalty rate on the fair return to record companies and artists for use of their works because royalties from the PSS market are so small as to be virtually inconsequential to companies whose principal business is the sale of CDs and digital downloads. Music Choice PFF ¶¶ 420-430. With respect to Music Choice's ability to earn a fair income, however, Music Choice argues that it is not profitable under the current 7.5% rate. Mr. Del Beccaro testified that its average revenue per customer for its residential audio business has been on the decline since the early 1990's, down from $1.00 per customer/per month to [REDACTED] per customer/per month currently. Del Beccaro Corrected WDT at 40, PSS Trial Ex. 1. He further testified that after 15 years of paying a PSS statutory rate between 6.5% and 7.5% Music Choice has not become profitable on a cumulative basis and is not projected to become so within the foreseeable future. Id. at 42. Cumulative loss at the end of 2011 is [REDACTED], projected to grow to [REDACTED] in 2012 and continue to increase throughout the 2013-2017 license period. Id. at 33-34; Del Beccaro Corrected WRT at MC 69 at 1, MC 70 at 1, PSS Trial Ex. 21. These losses lead Music Choice to conclude that it has not generated a reasonable return on capital under the existing rates, which it submits should be 15% in the music industry. Music Choice PFF ¶¶ 442-43.

Music Choice's claims of unprofitability under the existing PSS rate come from the oblique presentation of its financial data and a combining of revenues and expenses from other aspects of its business. The appropriate business to analyze for purposes of this proceeding is the residential audio service offered by Music Choice, the subject of the Section 114 license. Music Choice, however, reports costs and revenues for its residential audio business with those of its commercial business, which is not subject to the statutory license. This conflation of the data, which Music Choice acknowledges cannot be separated, see SX PFF at 221-222, distorts its views regarding losses. As a consolidated business, Music Choice has had significantly positive operating income during the past five years between 2007 and 2011 and has made profit distributions to its partners since 2009. Ford Amended/Corrected WRT at SX Ex. 362, p. 3, SX Trial Ex. 244; SX Trial Ex. 64 at 3; SX Trial Ex. 233 at 3. Dr. Crawford's effort to extract costs and revenues from this data for the PSS service alone for use in his surplus analysis cannot be credited because of his lack of familiarity with the data's source. 6/13/12 Tr. 1890:15-1891:10 (Crawford).^[66] Music Choice has operated successfully and received a fair income under the existing statutory rate.^[67]

With respect to fair return to the copyright owner, the examination is whether the existing statutory rate has produced a fair return with respect to the usage of sound recordings. During the period of the current rate and before, Music Choice provided 46 channels of music programming to the subscribers of its licensees. However, Music Choice is expanding the number of music channels dramatically in the coming licensing term, up to 300 channels by the first quarter of 2013. Del Beccaro Corrected WDT at 3-4, PSS Trial Ex. 1; 8/16/12 Tr. 3878:3 (Del Beccaro). This will result in a substantial increase in the number of plays of music by Music Choice, even if the ultimate listenership intensity of its licensees' subscribers cannot be measured. The expansion in usage will not be reflected in increased revenues to which the statutory royalty rate is to be applied, as Music Choice has declared itself to be a mature business with no expectation of increased future revenues for its business. As a result, copyright owners will not be compensated for the increased usage of their works, underscoring the Judges' preference for a per-performance metric for royalty determinations—which is not available here—as opposed to a percentage-of-revenue metric. Dramatically expanded usage without a corresponding expectation of increased compensation suggests an upward adjustment to the existing statutory rate. Measurement of the adjustment is not without difficulty because any downstream increases in listenership of subscribers as a result of additional music offerings by Music Choice cannot be readily determined nor predicted. It is possible that listenership overall may remain constant despite the availability of 300 music channels as opposed to only 46. However, it is more reasonable to conclude that Music Choice would not make the expansion, and incur the additional expense of doing so, without reasonable expectation that subscribers will be more attracted to and will consume more of the music offerings of Music Choice. A 2% upward adjustment, phased-in during the course of the license period as described below, is sufficient to provide copyright owners with a fair return for the increased use of sound recordings by Music Choice.

c. Relative Roles of Copyright Owners and Copyright Users

This policy factor requires that the rates adopted by the Judges reflect the relative roles of the copyright owners and copyright users in the product made available with respect to relative creative contribution, technological contribution, capital investment, cost, risk, and contribution to the opening of markets for creative expression and media for their communication. For its part, Music Choice's arguments that its creative and technological contributions, and capital investments, outweigh those of the record companies center on the same aspects of its business. First, Music Choice touts the graphic and informational improvements made to its on-screen channels, noting that what were once blank screens now display significant artist and music information. Costs for these improvements have exceeded [REDACTED]. Del Beccaro Corrected WDT at 31-32, PSS Trial Ex. 1. Second, Music Choice offers increases in programming, staff size and facilities, along with enhancements to product development and infrastructure. Costs for these improvements have exceeded [REDACTED]. Id. Regarding costs and risks, Music Choice points to its lack of profitability and the exit of other PSS from the market as evidence of its continued risk and limited opportunity for profit. Music Choice PFF ¶¶ 512-520. Finally, with respect to opening new markets, Music Choice touts the PSS market itself for which it remains the standard-bearer in disseminating music to the public through cable television. Id. at ¶ 523.

SoundExchange offers little more on the third Section 801(b) factor beyond Dr. Ford's contention that he saw no evidence to support that Music Choice makes contributions to creativity or availability of music that are beyond those of the music services he included in his benchmarks, and therefore the third factor is accounted for in the market. Ford Second Corrected WDT at 21, SX Trial Ex. 79; 6/18/12 Tr. 2849:10-16 (Ford).

In considering the third factor, the Judges' task is not to determine who individually bears the greater risk, incurs the higher cost or makes a greater contribution in the PSS market, and then make individual up or down adjustments to the selected rate based upon some unspecified quantification of these differences. Rather, the consideration is whether these elements, taken as a whole, require adjustment to the existing rate of 7.5%. Upon careful weighing of the evidence, I determine that no adjustment is necessary. Music Choice's investments in programming offerings, staff and facilities, and other related products and services are no doubt impressive, but they have been accomplished under the current rate and previous rates that are only slightly lower (the low being 6.5%). As discussed above, Music Choice has already begun to expand its channel offerings by several multiples and has allocated greater financial resources to its residential audio business. All of these undertakings, plus the investments made and costs incurred to date have been made in the shadow of the existing rate, and have not been prevented as a result of that rate. Likewise, on the other side of the ledger, SoundExchange has not offered any persuasive evidence that the existing rate has prevented the music industry from making significant contributions to or investments in the PSS market.

d. Minimize Disruptive Impact

Of the four Section 801(b) factors, the parties devoted most of their attention to the last one: minimizing disruption on the structure of the industries and on generally prevailing industry practices. This is perhaps not surprising, given the role this factor played in SDARS-I in adjusting the benchmark rates utilized by the Judges to set the royalty fees. See SDARS-I, 73 FR at 4097-98. Music Choice presents a considerable volume of testimony and argument as to why the SoundExchange proposed rates would be disruptive, if not debilitating, to its business; and SoundExchange presents testimony and argument as to why Music Choice's proposed rates would disrupt the music industry. These contentions, however, are inapposite as neither the SoundExchange nor the Music Choice benchmark analyses serve the purpose of determining a reasonable rate other than to mark the extreme ends of the boundary within which a reasonable rate can be located. Because I have identified as reasonable the rate for PSS currently in place, my analysis of the disruption factor is confined to that rate.

SoundExchange argues that the current rate is disruptive to the music industry. Dr. Ford testified that “the current practice of applying an exceedingly low rate to deflated revenues is disruptive of industry structure, especially where there are identical services already paying a higher rate.” Ford Second Corrected WDT at 23, SX Trial Ex. 79. This results, according to Dr. Ford, in a tilting of the competitive field for music services in favor of Music Choice, thereby disrupting the natural evolution of the music delivery industry. Dr. Ford, however, appears to ignore his own earlier assertions that the PSS market has unique and distinctive features that distinguish it from other types of music services, thereby substantially reducing the likelihood that the PSS and other music services are substitutes for one another. Further, Dr. Ford failed to present any empirical evidence demonstrating a likelihood of migration of customers from music services paying higher royalty fees to the PSS as a result of his perceived royalty imbalance.^[68] Dr. Ford's conclusion that the current rate paid by the PSS for the Section 114 license has caused a disruption to the music industry is mere speculation.

Music Choice also contends that the current rate is disruptive, and I likewise find its argument weak and unsubstantiated. The test for determining disruption to an industry, announced by the Judges in SDARS-I, is whether the selected rate directly produces an adverse impact that is substantial, immediate, and irreversible in the short-run. SDARS-I, 73 FR at 4097. The current rate has been in place for some time and, despite Music Choice's protestations that it has never been profitable, it continues to operate and continues to increase its expenditures by expanding and enhancing its services in the face of the supposedly disruptive current royalty rate. Music Choice's argument that DMX's bankruptcy and Muzak's decision to limit its participation in the PSS market are evidence of the onerous burden of the current rate are without support because Music Choice has failed to put forward any evidence demonstrating a causal relationship between the actions of those services and the PSS royalty rate.

In sum, I am not persuaded by the record testimony or the arguments of the parties that the current PSS rate is disruptive to a degree that necessitates an adjustment.

5. Conclusions Regarding Section 114 Rates

Upon a careful weighing of the evidence submitted by the parties, I believe that the application of the Section 801(b) factors to the rate of 7.5% of Gross Revenues requires an upward adjustment to account for the coming expanded use of music by Music Choice in the 2013-2017 licensing term. If the Judges preferred per-usage royalty metric could be applied to the PSS—which it cannot—the value of the increased usage would be captured in the metric through the measurement of listenership to the sound recordings received by Music Choice consumers through their respective cable systems. The percentage-of-revenue metric, however, will not account for the expanded use in the short term, as cable operators will continue to pay fees for the Music Choice service in approximately the same amounts, and will only increase in the long term, presumably, if the volume of cable subscribers (or per-subscriber license rates) increases significantly. The testimony, however, suggests this possibility to be unlikely, as Music Choice itself declares the PSS market mature. 8/16/12 Tr. 3855:17-3856:7 (Del Beccaro); 8/23/12 4707:8-19 (Crawford).

The following are the rates that I believe are appropriate and supported by the evidence in this proceeding: for 2013: 8.5%; for 2014: 9.0%; for 2015: 9.5%; for 2016: 9.5%; and for 2017: 9.5%.

SoundExchange raises an additional matter with respect to the total royalty obligation of the PSS. Though not technically a rate, nor strictly an amendment of Gross Revenues, SoundExchange requests a means for capturing revenues from cable systems that are owners of equity or capital interests in Music Choice who do not engage in arm's length transactions with Music Choice for its product offerings. Second Revised Proposed Rates and Terms, at 6-7 (Sept. 26, 2012). Put another way, SoundExchange seeks to capture any price break that Music Choice offers its ownership partners for the Music Choice service. The price break to a specific partner cable system would be calculated by multiplying the total number of subscribers for the month for that system by the average per-subscriber royalty payment of the five largest paying cable systems providing Music Choice that are not its partners. This reconciling for each partner cable system would then be added to Music Choice's Gross Revenues overall calculation. In support of its “Non Arm's Length Transaction” adjustment for cable partners, Dr. Ford testified that a straight percentage of revenue metric is problematic where Music Choice offers per-subscriber rate discounts to it cable partners. “I believe that, if we are going to properly compensate someone for the use of their property, we ought to be compensating them for use and not have the compensation affected by peculiar ownership structure of the entities that easily arise.” 8/20/12 Tr. 4216:21-4217:8 (Ford). Over half of Music Choice's non-partner cable systems pay approximately [REDACTED] per subscriber per month in licensing fees to Music Choice, whereas the partner cable systems pay only [REDACTED] per subscriber per month. Ford Second Corrected WDT at 5, SX Trial Ex. 244.

I am not persuaded that a “Non Arm's Length Transaction” adjustment is warranted. Implicit in Dr. Ford's observation of Music Choice's licensing of its service to its cable partners is the assumption that the partners have the ability to exert downward pressure on Music Choice revenues so as to avoid payment of music royalties and thereby boost their own bottom-lines. Such presumed use of Music Choice as a loss leader is not borne out by the evidence in this proceeding. The partnership agreements between Music Choice and its cable operators are lengthy and complicated and vary from partner to partner. It is not surprising that the partner cable operators, which are in most instances of greater size with respect to numbers of subscribers than the non-partner licensors of Music Choice's service, would be able to negotiate lower per-subscriber licensing fees due to their ability to deliver more subscribers to the service. Further, the cable partners represent only a third of Music Choice ownership, and do not appear to be able to influence rates any more than Music Choice's record company partners, who own one quarter of the company. 6/11/12 Tr. 1454:16-22 (Del Beccaro). SoundExchange's “Non Arm's Length Transaction” adjustment is founded upon inference and speculation and is not supported by the record evidence.

E. The Section 114 Royalty Rates for SDARS

As with the consideration of reasonable rates for the PSS, I begin my analysis for SDARS with the proffered benchmarks of Sirius XM and SoundExchange, respectively.

1. SDARS Proposed Benchmarks

a. The Direct Licenses

Beginning in the summer of 2010, Sirius XM commenced a coordinated effort to negotiate sound recording performance rights directly with individual record labels. 6/7/12 Tr. 669:8-672:9, 713:3-11, 714:11-715:4 (Frear). Dubbed the Direct License Initiative, Sirius XM first attempted to engage the four major record companies in discussions but was unsuccessful. Id.; 6/11/12 Tr. 1347:7-21, 1348:20-1349:4 (Karmazin). Sirius XM then enlisted the services of Music Reports, Inc. (“MRI”) to formulate and execute a direct licensing strategy with as many independent record labels as possible. Together, Sirius XM and MRI developed the terms and conditions of a Direct License, the highlights of which include:

• A pro rata share of 5%, 6%, or 7% of gross revenues, defined by reference to 37 CFR 382.11;
• A grant of rights to Sirius XM to operate all of its various services (satellite radio plus other services such as webcasting);
• “Additional functionality” granted to Sirius XM, including elimination of the Section 114 license sound recording performance complement;
• Direct, quarterly payment of 100% of the royalties to the record label;
• Payment of advances to the 5 largest record labels;
• The possibility, but not the promise, of increased play on Sirius XM's music services.

Gertz Corrected WDT ¶ 14(a), (b), SXM Dir. Trial Ex. 5. The first Direct Licenses were executed in August of 2011 and by the time of the closing of testimony in this proceeding, Sirius XM had Direct Licenses with 95 independent record labels. 8/13/12 Tr. 3015-16-20 (Frear); 8/15/12 Tr. 3679:22-3680:1 (Gertz).

Sirius XM's expert economist, Dr. Roger Noll, advises that the 95 Direct Licenses are the best benchmark for rate setting in this proceeding because, unlike in SDARS-I, the Judges now have direct evidence of competitively negotiated marketplace rates for the exact service at issue in this proceeding. Noll Revised Amended WDT at 7, 11, 33-36, SXM Dir. Trial Ex. 1. Dr. Noll testified that the Direct Licenses are representative, for benchmarking purposes, of the types of sound recordings available across the industry, including those distributed by major labels. Id. at 39-44; see also 6/5/12 Tr. 261:6-262:14 (Noll)(the 95 Direct Licensors as a group offer a scope of sound recordings comparable to those not so licensed). The fact that the Direct Licenses represent only a small percentage of market share of music available does not alter the incentive to create demand diversion, Dr. Noll opines, because the major record labels and the independent labels signed to the Direct Licenses both seek to maximize their number of plays on Sirius XM's music services. A Direct Licensor would find a 7% license rate more attractive than the current 8% statutory rate if the lower rate would cause an increase in the number of plays. Noll Revised Amended WDT at 40-41, SXM Dir. Trial Ex. 1. Dr. Michael Salinger, another Sirius XM expert economist, concludes that the fact that 95 record companies accepted the Direct License offer suggests that the current 8% statutory rate is, if anything, above the competitive rate for sound recordings. Salinger Corrected WRT at ¶ 28, SXM Reb. Trial Ex. 9. Further, Sirius XM argues that the number of Direct Licenses undoubtedly would have been higher but for the efforts of SoundExchange, the American Association of Independent Musicians and others to undermine and interfere with its Direct License Initiative. Sirius XM devoted considerable time and testimony in an effort to support this contention. See, e.g., Sirius XM PFF at 61-63.

b. The Noll Benchmark

Dr. Noll asserts that license agreements between major record labels and certain customized non-interactive webcasters provide marketplace evidence of rates that corroborate the 5%-7% rates achieved in the Direct Licenses. Noll Revised Amended WDT at 16, 72, SXM Dir. Trial Ex. 1. Focusing principally on the agreements between the digital music service Last.fm and the four major record companies,^[69] Dr. Noll determined that for its non-interactive subscription streaming service, Last.fm agreed to pay:

• [REDACTED]
• [REDACTED]
• [REDACTED]

Noll Revised Amended WDT at 76-79 (footnote omitted), Tables 2.1-2.1c and Appendices E-H, SXM Dir. Trial Ex. 1. Examining these same agreements for Last.fm's interactive on demand service—[REDACTED]—led Dr. Noll to conclude that sound recording rights owners charge [REDACTED] for non-interactive services than they do for interactive/on-demand services. Id.^[70]

Using the rates gleaned from the Last.fm agreements for the non-interactive subscription streaming service, which he deemed to be the most similar to Sirius XM's satellite radio service in terms of functionality, Dr. Noll computed his reasonable royalty fee by multiplying the Last.fm revenue rate [REDACTED] against the implicit per-subscriber price of Sirius XM's music channels ($3.00-$3.45), and then divided the resulting per-subscriber monthly fee into Sirius XM's average revenue per user ($11.38) in order to express the fee as a percentage of revenue. Id. at 15; 6/5/12 Tr. 285:7-293:9 (Noll). This yielded an average royalty rate as a percentage of Sirius XM revenue of 6.76%. Id. at 90; 6/5/12 Tr. 293:5-9 (Noll). Because this average rate fit squarely between the 5%-7% range of the Direct Licenses, Dr. Noll opines that his calculation is corroborative of the rates contained in Direct Licenses and further concludes that it represents the upper end of a reasonable royalty rate because the customized, non-interactive services he examined offer greater functionality and sound quality than the channels offered by Sirius XM. Id. at ¶¶ 14-16; 6/5/12 Tr. 292:2-14 (Noll).

2. SoundExchange Proposed Benchmarks

SoundExchange's expert economist, Dr. Janusz Ordover, offers a principal benchmark, and two alternatives, based upon his examination of market agreements for digital music between interactive subscription services streaming music and the four major record companies. Dr. Ordover chose interactive subscription services because of his belief that they represent voluntary transactions in a competitive marketplace free of regulatory overhang, provide sufficient information based on multiple buyer/seller interactions, are not distorted by the exercise of undue market power on either the buyer's or seller's side, and involve digital music services that are similar to Sirius XM. 6/14/12 Tr. 2359:11-2360:9, 2257:5-11, 2257:12-20, 2257:21-2258:2 (Ordover).

Dr. Ordover's principal benchmark is to calculate the percentage of total revenues represented by royalty payments made by interactive services to record companies, and then apply that percentage of revenue to the amount that he determined to be the retail price of a music-only satellite service in order to calculate the corresponding percentage-of-revenue for the Sirius XM service. See generally Ordover Third Corrected/Amended WDT at 18-25, SX Trial Ex. 74. Beginning with data from July 2010, he derived the effective percentage-of-revenue paid by each interactive service by taking the amount of royalty fees paid to the major record labels and dividing it by each service's gross subscription revenues. In other words, he relied on royalty payments made, rather than the percentage-of-revenue rates specified in the agreements which contained “greater of” royalty formulations.^[71] In calculating actual licensing fees paid, Dr. Ordover used gross subscription revenues of the interactive services without any deductions or carve-outs. Ordover Third Corrected/Amended WDT at 19, SX Trial Ex. 74. Examining the agreements, he determined that the annual payments as a percentage of gross revenues of the services ranged from 50% to 70%, and tended to cluster in a narrower range of 60% to 65%. 6/14/12 Tr. 2275:4-12 (Ordover); Ordover Third Corrected/Amended WDT at 19-21, SX Trial Ex. 74. Dr. Ordover then adjusted the benchmark to account for the fact that the Sirius XM satellite radio service, unlike interactive subscription services, transmits both music and non-music content. Reducing the percentage-of-revenue by half, principally based upon his observation of the identical $9.99 retail prices offered by Sirius XM for non-music and mostly music stand-alone subscriber packages, yielded rates for Sirius XM between 30% and 32.5% for the 2013-2017 statutory licensing period. Ordover Third Corrected/Amended WDT at 17, SX Trial Ex. 74.^[72]

As his first alternative benchmark, Dr. Ordover examines per-subscriber royalty rates from interactive subscription services in an effort to account for the differences in service attributes between satellite radio and interactive subscription services. He first determined an unweighted average monthly royalty of $5.95 per subscriber (monthly licensing fees paid divided by monthly subscriber counts) for interactive services, and then adjusted this fee by the ratio of the retail price of a hypothetical music-only satellite radio service (50% of the $12.95 subscription price for the Sirius XM Select programming package ^[73] ) to the retail price for interactive subscription services ($9.99). Ordover Third Corrected/Amended WDT at 30-31, SX Trial Ex. 74. This percentage, when applied to the average per-subscriber royalty paid by interactive services ($5.95), yields $3.86 for the hypothetical music-only satellite radio service. Dividing this number by the $12.95 Sirius XM subscription price provides a percentage-of-revenue rate of 29.81%. Id. at 32.

Dr. Ordover's second alternative benchmark approach attempts to adjust for the presence of interactivity alone in the rates yielded by his primary benchmark under the assumption that interactivity is the material difference between interactive subscription services and satellite radio. Ordover Third Corrected/Amended WDT at 34, SX. Trial Ex. 74. To derive the value of interactivity, he compared the retail prices for interactive music streaming services with the retail prices for non-interactive music streaming services in order to obtain a ratio. He determined that interactive music streaming services are uniformly priced at $9.99 per month, while non-interactive services prices averaged $4.86. Ordover Third Corrected/Amended WDT at 31-32 Table 4, SX Trial Ex. 74; Id. at 33 Table 5.^[74] Dr. Ordover then used the ratio to adjust the average per-subscriber royalty paid by interactive services ($5.95) to calculate an equivalent payment for satellite radio. This yielded a percentage-of-revenue royalty rate of 22.32% for Sirius XM, which Dr. Ordover concludes represents the lower bound of a reasonable royalty rate. 6/14/12 Tr. 2282:12-16 (Ordover).^[75]

3. Analysis and Conclusions Regarding the Proposed Benchmarks

The Direct Licenses offered by Sirius XM have the surface appeal of a good benchmark in that they involve the same sellers and buyers in the target market; however, a closer examination reveals that they are fraught with problems. First, they represent a sliver of the universe of rights holders for sound recordings: 95 of over 20,100 rights holders to which SoundExchange distributes payments, Bender WDT at 4, SX Trial Ex. 75, and a subset of the 691 independent labels that Sirius XM approached in the first instance. Ordover Amended WRT at 4 n.8, and 6, SX Trial Ex. 218; SX Trial Ex. 301 at 53. Much was made by Sirius XM in this proceeding that the number of Direct Licenses would have been substantially higher but for the interference of SoundExchange. It is not within the Judges' jurisdiction to determine that SoundExchange's actions amounted to legal interference with contractual relations or otherwise frustrated Sirius XM's efforts to execute more Direct License agreements. The Direct Licenses are evaluated for what they are, not for what they might have been, and what they are is a very small percentage of the sound recording market.^[76]

Second, the Direct Licenses do not include any of the major record labels whom, by virtue of their size of the music market and the popularity of their recordings, Sirius XM cannot do without. Dr. Noll's observation that the works licensed by the Direct Licensors are representative of the kinds of sound recordings available to Sirius XM in the market is beside the point, for the Judges have concluded before that sound recordings, particularly those of the major record labels, are not readily substitutional for one another, let alone with those of independent record labels. Phonorecords I, 74 FR 4510, 4519 (Jan. 26, 2009); see, generally Webcaster I, 72 FR 24084 (May 1, 2007). The “representativeness” of the sound recordings contained in the catalogs of the Direct Licensors do not equate to their popularity,^[77] an essential ingredient to Sirius XM's music offerings. 6/7/12 Tr. 836:17-22 (Gertz)(“Sirius XM is very hits driven, and they want to have the most successful service they can, so they're going to use what's popular.”).

Third, I am troubled by the additional considerations and rights granted in the Direct Licenses that are beyond those contained in the Section 114 license, thereby weakening their comparability as a benchmark. The Direct Licenses provide for payment of 100% of the royalties to the Direct Licensors, 6/6/12 Tr. 341:10-342:3 (Noll), thereby avoiding the statutory apportionment of 50% to record companies and 50% to artists and performers.^[78] 17 U.S.C. 114(g)(2). Certain of the Direct Licenses, in particular those of the largest independent labels, provide for cash advances and accelerated royalty payments, also not available under the statutory license. See, e.g., Gertz Revised WRT at SXM Reb. Ex. 23, pp. 3-4, SXM Reb. Trial Ex. 8; Gertz Revised WRT at SXM Reb. Ex. 8, pp. 3-4, SXM Reb. Trial Ex. 8. Sirius XM absorbs all of the administrative costs of the licensing process under the Direct Licenses, which under the statutory license are borne by the copyright owners, artists and performers. Eisenberg Amended/Corrected WRT at SX Ex. 313-RR, SX Trial Ex. 245. And with respect to rights granted under the Direct Licenses, Sirius XM receives a waiver of the sound recording complement of the statutory license, and the ability to perform the works of the Direct Licensors on other services not covered by the statutory license.

My concerns regarding the Direct Licenses are not cured by Dr. Noll's analyses. Dr. Noll contends that the fact the Direct License rates are lower than the current 8% statutory rate is explained by a demand diversion effect—record labels engaging in price competition aimed at increasing their market share through increased plays on Sirius XM, thereby reducing the royalty rates demanded—and represents what would happen in the market as a whole in the absence of a statutory rate. Noll Revised Amended WDT at 36-38, SXM Dir. Trial Ex. 1.^[79] His demand diversion theory, however, has limited explanatory power. It may well be that independent record labels took the Direct License offer because of the valuable non-statutory benefits discussed above, and there is testimony in the record to this effect. See, e.g., SX Trial Ex. 317 at SXM-CRB_DIR_00079565; 8/20/12 Tr. 4156:5-4157:3 (Powers). Further, independent labels have greater concerns than majors in securing performances of their works on services such as Sirius XM, increasing the attractiveness of a Direct License relationship. Powers WRT at 4, SX Trial Ex. 243; Eisenberg Amended/Corrected WRT at SX Ex. 329-RR, p. SXM_CRB_DIR_00042287, SX Trial Ex. 245 (email from MRI to independent label emphasizing that a Direct License offers the possibility of increased airplay). The incentive of increased airplay does not necessarily exist for major record labels, whose works are already performed in large numbers by Sirius XM's hits-driven programming. Harrison Corrected WRT at 9-10, PSS Trial Ex. 32.

Dr. Noll's benchmark analysis, whether considered as corroboration of the Direct Licenses or stand-alone, contains significant flaws. His reliance on the Last.fm agreements with the four major record labels, which provide the critical data to his calculations, is valid to the extent that it is representative of non-interactive subscription webcasting services. See SDARS-I, 73 FR at 4090. Two of the agreements, however, have expired and are no longer in effect. Ordover Amended WRT at 25, SX Trial Ex. 218. Last.fm now pays those record companies at the statutory webcasting rate, which is not a per se market rate. 8/14/12 Tr. 3308:8-20, 3317:10-16 (Ordover). Even if the Last.fm agreements were the most representative of webcasting services—and Dr. Noll has not demonstrated that they are—I would not be inclined to accept them as fully comparable to the SDARS business without some adjustment for the functional differences between webcasting and satellite radio. No persuasive adjustment was offered.

I also have reservations about Dr. Noll's determination of $3.00-$3.45 as the implicit monthly market price for Sirius XM's music channels.^[80] Dr. Noll identified three methods for determining the implicit price. The first is the average retail price of $3.15 taken from Last.fm's and Pandora's non-interactive subscription services. Noll Revised WRT Table 1, SXM Reb. Trial Ex. 6. As with Last.fm, there is no adjustment to account for functional differences between the Pandora webcasting service and satellite radio, whose primary use is in the automobile. The second is to derive a market price for Sirius XM using a survey conducted by Sirius XM's witness Professor John Hauser that attempts to measure the value of music to Sirius XM subscribers. Professor Hauser posited an anchor price for the Sirius XM service to his survey respondents, and then randomly removed features (such as lack of commercials, quality of sound, etc.) to determine how much the respondents would be willing to pay for the service after each feature is removed. After averaging the results, he determined that subscribers place an average value on Sirius XM's music channels of $3.24. Hauser Corrected WDT at Appendix G, SXM Dir. Trial Ex. 24. Professor Hauser's survey is of limited value. By design, the higher number of features or attributes of the Sirius XM service included in the survey, the lower the estimated value of any given service. This produces anomalous results, such as his survey showing that subscribers would pay a certain amount for ubiquitous station availability, premium sound quality and absence of commercials all without any programming content whatsoever. Ordover Amended WRT at 35, SX Trial Ex. 218.

Third, Dr. Noll sought to calculate the cost of inputs necessary for delivery of Sirius XM's programming via satellite and its subsidization/installation of radio receivers in automobiles (described as “unique” costs to the satellite radio service), to then deduct those costs from gross revenues, and allocate the remaining revenue between music and non-music content. Noll Revised Amended WDT at 81-83, 85, SXM Dir. Trial Ex. 1. After making these calculations, Dr. Noll credited 55.1%, or $3.45, to music channels. Id. at 88 and Table 3. Sirius XM contends that including the unique delivery costs and investments of its service is appropriate in Dr. Noll's calculation, and cites to major record company agreements with Cricket and MetroPCS (mobile service providers that bundle telephone service and interactive music service into a single package) that reflect that a percentage royalty rate for music must be reduced by a commensurate proportion to reflect revenue collected for the non-music portion of the bundled service. Sirius XM PFF ¶¶ 169-173. However, SoundExchange's expert economist, Dr. Thomas Lys, explained that because most of the unique costs that Dr. Noll allocated are relatively fixed, the per-subscriber amounts vary inversely with the number of subscribers. Lys WRT at 57, SX Trial Ex. 240. Dr. Noll performed his calculation of costs using 2010 data, but had he used subscriber numbers for the years thereafter which have continued to increase and are anticipated to increase further in the coming licensing term, the analysis would show lower unique costs per subscriber and a higher value of music. Lys WRT at 57, SX Trial Ex. 240. The dependency of Dr. Noll's methodology on timing and the number of subscribers undermines its reliability for quantifying what the unique costs are likely to be in the coming rate term. Id. at 58. Moreover, Sirius XM's analogy to the bundled services of Cricket and MetroPCS is inapposite. Unlike those services, the success of Sirius XM is dependent upon its access to music. 6/14/12 Tr. 2270:7-2271:15 (Ordover); see also 6/5/12 Tr. 235:6-10 (Noll)(“It's a bundle of services, it's a distribution system, a bunch of nonmusic content and a bunch of music content, all of which are essential. And you pull the plug on any one of them, and the whole thing collapses.”); 6/11/12 Tr. 1431:10-17 (Karmazin). The value of Sirius XM's satellite radio service is the bundling of music and non-music content with its delivery platform, and Sirius XM has failed to present convincing evidence that its delivery platform and non-music content, alone, present a viable business.^[81]

In sum, these concerns, coupled with those surrounding the Direct Licenses themselves, do not inspire confidence that the Direct Licenses are the best benchmark for rate setting in this proceeding. Rather, I believe that the rates between 5% and 7% contained in the Direct Licenses mark the lower boundary of the range of reasonable rates to be determined in this proceeding. The evidence presented establishes that reasonable rates cannot be lower. I now examine the benchmarks offered by SoundExchange and Dr. Ordover.

As an initial matter, the Judges have determined in the past that the interactive subscription service market is a benchmark with characteristics reasonably comparable with non-interactive SDARS. SDARS-I, 73 FR at 4093. Sirius XM, however, charges that Dr. Ordover began his analysis in the wrong place by examining rates for interactive services instead of non-interactive services. I do not agree. In saying this, I do not suggest that the market for interactive services, in and of itself, offers the best benchmark from which to begin an analysis of reasonable rates for Sirius XM's satellite radio service. Adjustments, as discussed below, are necessary for the benchmark to be at all useful. However, as a starting point, the interactive subscription service market is more illustrative of a competitive marketplace (willing buyer/willing seller) than the non-interactive subscription service market, where negotiated rates are likely influenced by the availability of the statutory licensing regime for webcasting. See Webcasting III, 76 FR 13026 (Mar. 9, 2011)(citing Noncommercial Educational Broadcasting Compulsory License, Final rule and order, 63 FR 49823, 49834 (Sept. 18, 1998))(“[I]t is difficult to understand how a license negotiated under the constraints of a compulsory license, where the licensor has no choice to license, could truly reflect `fair market value.' ”). Furthermore, the agreements examined by Dr. Ordover represent a more robust data source from which to consider the outcomes of marketplace negotiations, as opposed to Dr. Noll's confined use of only the Last.fm agreements.^[82, 83] His observation of a clustering of effective percentage of revenue rates between 60% and 65% for interactive subscription services is supported by empirical evidence and is not misleading or under inclusive.^[84] Ordover Third Corrected/Amended WDT at 21 Table 1, 26, Table 2, SX Trial Ex. 74.

I am not persuaded that Dr. Ordover's perceived “failure” to incorporate the costs of Sirius XM's satellite delivery platform renders his interactive subscription services benchmark fatally flawed or in need of adjustment. Dr. Noll asserts that the Sirius XM satellite radio service should be viewed as a bundle of three inputs—music content, non-music content, and the satellite platform for delivering the content—and attempts to separately value each component of the bundle. Noll Revised Amended WDT at 80, SXM Dir. Trial Ex. 1. Consumers do not value the satellite platform independent of the content it transmits, 6/7/12 Tr. 666:5-11 (Frear), and Sirius XM has not successfully demonstrated that the satellite platform can be unbundled and sold separately. See SDARS-I, 73 FR at 4089; see also Ordover Amended WRT at 33, SX Trial Ex. 218 (Cricket license agreements reflect that its delivery system provides services that have independent value to consumers). The value of Sirius XM's service is the end product to the consumer, as is the case with the interactive subscription service consumer, and no adjustment for the delivery mechanism is necessary.

To be sure, the rights licensed by interactive subscription services are not the same as those by non-interactive services such as the SDARS, and adjustment to the interactive benchmark is necessary to account for these differences. See SDARS-I, 73 FR at 4093. Dr. Ordover attempted to account for these differences by offering two alternative benchmarks. His first alternative begins with the average monthly per-subscriber fee paid by interactive services and reduces that fee in proportion to the ratio of the retail price of Dr. Ordover's hypothetical music-only satellite radio service to the retail price of interactive services. There are doubts as to whether this approach accurately adjusts the interactive service benchmark to account for differences in attributes and functionality between interactive subscription services and satellite radio, and SoundExchange backed away from advocacy of this model in its post-trial submissions. I focus, instead, on Dr. Ordover's second alternative approach, which begins with the average monthly per-subscriber fee paid by interactive services ($5.95) and then reduces that fee in proportion to the ratio of the average retail price of non-interactive music services to the retail price of the interactive services ($4.86/$9.99). Ordover Third Corrected/Amended WDT at 34, SX Trial Ex. 74.

It is readily apparent that Dr. Ordover's interactivity adjustment to his interactive subscription services benchmark in this proceeding is not the same as the one he performed in SDARS-I. Dr. Ordover based his adjustment in SDARS-I on per-play rates from non-interactive video streaming services, a market that both parties concede effectively no longer exists. SX RFF at 164; 8/15/12 Tr. 3573:22-3574:3 (Noll). However, I am not persuaded that the difference—using retail prices for non-interactive services in this proceeding rather than per-play rates—renders his analysis invalid. A straightforward comparison of per-play rates in the interactive and non-interactive markets would be flawed, in that it would not account for differences in intensity of use (average number of plays per subscriber) between the markets, and would involve analysis of non-interactive rates from a market subject to influences of the statutory license. Comparing retail prices between the markets, as Dr. Ordover does, is a reasonable approach as the value of interactivity to consumers will likely be reflected in retail prices. 8/16/12 Tr. 3836:5-11 (Salinger).

While I find Dr. Ordover's comparison of retail prices in the interactive and non-interactive markets conceptually sound, his analysis is not without warts. In deriving his average non-interactive service price for the five non-interactive services he examined, Dr. Ordover's averaging technique placed greater weight on the higher-priced services.^[85] A more accurate method for calculating the average price is to include a single time-frame observation—the price of a year of service—for each of the five services. This procedure reduces the average price to $4.01. Noll Revised WRT at 25, SXM Reb. Trial Ex. 6. Dr. Ordover also did not weight his average by the number of subscribers to each service to account for differences in popularity, presumably because data was not available for all five services. It exists, however, for Pandora, Last.fm and Live365. I accept Dr. Noll's weighted adjustment to $3.15 because of the unlikelihood that the other two services used by Dr. Ordover, Musicovery and Sky.fm, would significantly impact the calculation. Id. at 25-26.

In converting his price for non-interactive services to a price for Sirius XM, Dr. Ordover used the monthly price charged to subscribers for the Sirius XM Select package. Ordover Third Corrected/Amended WDT at 43, SX Trial Ex. 74. Dr. Noll suggests that using Sirius XM's Average Revenue Per User (“ARPU”) makes more sense, stating that “I doubt that Dr. Ordover disagrees that ARPU, not sticker price, is the correct basis for calculating royalties.” Noll Revised WRT at 20 n.5, SXM Reb. Trial Ex. 6. ARPU was $11.22 in the first quarter of 2011 and rose to $11.49 in the first quarter of 2012 after the Sirius XM price increase. I use $11.49 as the most current ARPU figure in the record, and the one most representative for the coming licensing term.

Making the adjustments for the price of non-interactive services and revenues for Sirius XM ^[86] yields a percentage-of-revenue rate for Sirius XM of 16.2%.^[87] Dr. Ordover opines that his second alternative benchmark generates a lower bound estimate of reasonable rates. Ordover Third Corrected/Amended WDT at 33, SX Trial Ex. 74. However, I am not confident that his benchmark fully adjusts for interactivity to the level of service offered by Sirius XM's satellite radio service. For example, Pandora and Last.fm allow more user control of content than Sirius XM. Noll Revised WRT at 27, SXM Reb. Trial Ex. 6. Musicovery allows users to create playlists within a social network, to ban songs and artists from certain customized channels, and to skip songs altogether, while Sky.fm permits caching for later listening. Id. at 28. Additionally, Dr. Ordover's use of the average per-subscriber royalty payment of $5.95, which is drawn from the 60% average royalty fee for interactive services, bakes in the interactive service royalty to his calculation by virtue of its use as a multiplier.

There are other concerns with Dr. Ordover's analysis. For example, Live 365, which charges the most of the non-interactive services that Dr. Ordover observed, offers more than 7,000 channels that are pre-programmed by independent entities and other content that does not closely resemble the Sirius XM satellite. This reduces my confidence that Dr. Ordover's 16.2% benchmark is as reliable as the one the Judges considered in SDARS-I. In sum, the 16.2% royalty rate marks the upper bound of reasonable rates in this proceeding, with the lower bound marked by the 5%-7% rates from the Direct Licenses. The appropriate royalty rates lie within this zone, identified by my Section 801(b) policy analysis described below.

4. The Section 801(b) Factors

In SDARS-I, the Judges determined that an evaluation of the marketplace evidence hued in the direction of Dr. Ordover's interactivity-adjusted interactive subscription market analysis that marked the upper bound of reasonable royalty rates in that proceeding. See 73 FR at 4094. For the reasons stated above, the market-based evidence presented in this proceeding does not weigh in favor of either SoundExchange's or Sirius XM's presentations. Rather, reasonable rates to be paid by Sirius XM for the 2013-2017 licensing period lie along the continuum of rates marked at the lower end by 5%-7% from Sirius XM's presentation and at the upper end by 16.2% by SoundExchange's presentation. Consideration of the Section 801(b) policy factors locates the appropriate royalty rates within that range.

a. Maximize Availability of Creative Works

The first policy objective set forth in Section 801(b)(1) is to “maximize the availability of creative works to the public.” 17 U.S.C. 801(b)(1)(A). Sirius XM argues that application of the first factor favors adoption of rates at the lower end of the range for three reasons.^[88] First, Sirius XM contends that its satellite radio service enhances the delivery and availability of sound recordings by providing nationwide transmissions of sound recordings not played elsewhere. Second, Sirius XM submits that royalties from the Section 114 SDARS license are too small a portion of record companies' overall revenue to be a driving force behind decisions to produce creative works. Thus, according to Sirius XM, a lower royalty rate will not reduce record companies' incentives. Third, Sirius XM argues that the promotional effects created by its artist-themed channels, special benefits and programming exert a direct promotional impact on the sale of sound recordings thereby generating revenue for rightsholders and inducing them further to create new sound recordings. Sirius XM RFF ¶ 99.

I am not persuaded that any of these reasons augurs in favor of rates at the lower end of the range of reasonable rates. While it is acknowledged that Sirius XM's signal is capable of reception in locations in the United States not served by over-the-air terrestrial broadcast radio or wireless Internet service, Dr. Noll could not estimate what percentage of the population (approximately 2% in the U.S.) in these unserved areas actually subscribes to Sirius XM's satellite radio service. Noll Revised Amended WDT at 18-21, SXM Dir. Trial Ex. 1. Even for those persons in unserved areas who do subscribe to Sirius XM, there is no evidence that this group depends upon Sirius XM in order to access music. In fact, Sirius XM's own internal survey demonstrates that subscribers who deactivate their Sirius XM service typically turn to consumption of music on CDs. SX Trial Ex. 8 at 23 (SXM_CRB_DIR_00042796).

With respect to the percentage of record company revenues represented by Sirius XM's Section 114 royalty payments, it is true that the percentages of the totals are low; nonetheless, there is testimony that the royalty payments contribute significantly to overall profitability. See, 6/13/12 Tr. 2141:1-10 (Ciongoli)(UMG); Ford Amended/Corrected WRT at 13, SX Trial Ex. 244 (Warner); PSS Trial Ex. 33 (Sony). Therefore, it cannot be said that Section 114 royalty rates—whether low or high within the range—have no impact whatsoever on record companies' incentives to create new sound recordings.

Finally, there is no objective, quantifiable evidence that Sirius XM's promotional activities with respect to its music offerings, events, and surrounding programming produce a net positive impact on record company revenues. While these activities, viewed individually, may have promotional effect on record sales, there is insufficient evidence in the record as to the overall effect of Sirius XM's satellite radio service on all streams of record company revenues from sound recordings. Indeed, Sirius XM's witness Steven Blatter conceded that his examples of on-the-air activities showed only a correlation between airplay and record sales and nothing more. 6/8/12 Tr. 1032:20-1033:7 (Blatter). It may be that Sirius XM's use of sound recordings has an overall substitutional effect upon record company revenues, as opposed to an overall promotional effect. Sufficient and creditable evidence is not present in this record to quantify the promotional/substitutional effect of Sirius XM's service.

In sum, I find that the policy goal of maximizing the availability of creative works to the public is not, due to the paucity of the evidentiary presentations, advanced by royalty rates at either the upper bound or the lower bound of the range of reasonable rates determined from my analysis of the marketplace evidence.

b. Afford Fair Return/Fair Income Under Existing Market Conditions

The second policy objective seeks “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.” 17 U.S.C. 801(b)(1)(B). SoundExchange contends that dramatic changes in the recorded music business within the last decade have placed a greater emphasis on digital exploitation of sound recordings versus physical sales, thereby increasing the importance of revenues generated by the Section 114 license. Sirius XM contends that lower royalty rates are necessary to enable it to recover the investments in its satellite business and achieve profitability.

Charles Ciongoli, Executive Vice President and Chief Financial Officer for Universal Music Group North America (“UMG”), testified that the recorded music business' new reliance on digital revenues is the result of consumers purchasing fewer physical products as they gain more widespread access to music through digital services. As a result, companies like UMG cannot rely solely on the sale of physical products or permanent downloads, as in years past, and must obtain substantial royalty revenues from “access” services, such as Sirius XM, in order to survive. Ciongoli Corrected WDT at 4-6, SX Trial Ex. 67. See also Bryan Corrected WDT at 3-4, SX Trial Ex. 66; 6/13/12 Tr. 1969:21-1970:12 (Bryan). SoundExchange submits that digital royalties are even more important for independent record companies to ensure a fair return on their efforts to develop artists in the short and long terms. Van Arman WDT at 3, SX Trial Ex. 77.

Sirius XM states that the costs of its investments in the satellite business, expenses related to research, development, and permitting, and its operating losses must be measured cumulatively, not as a snapshot of annual operating costs, in considering fair return to the user under the second Section 801(b) factor. Sirius XM PFF ¶ 263. The evidence, according to Sirius XM, demonstrates that it is a long way from earning any return on its billions of dollars of expenditures, in contrast to the record companies which have “presented no evidence that the record industry is not currently earning a fair return on its investments in the production of creative works.” Id. at ¶ 264.

Evaluating royalty rates that would enable recovery of expenditures of Sirius XM over more than a decade of operations is not required under the second Section 801(b) factor.^[89] As the Judges observed in SDARS-I, “[a]ffording copyright users a fair income is not the same thing as guaranteeing them a profit in excess of the fair expectations of a highly leveraged enterprise.” SDARS I, 73 FR at 4095 (footnote omitted). During the current five-year licensing period, Sirius XM has publicly reported in its SEC filings adjusted Earnings Before Interest, Taxes, Depreciation and Amortization (“EBITDA”) of positive $2.1 billion, and net income of positive $3.2 billion. Frear WDT at 7, SXM Dir. Trial Ex. 12 (2008-2010 results); Lys WRT at SX Ex. 231-RP, SX Ex. 232-RP, SX Trial Ex. 40 (2011 and 2012 first quarter results), SX Trial Ex. 240; SX Trial Ex. 217 (2012 second quarter results and 2012 full-year guidance). By the end of 2012 under the current 8% of Gross Revenues royalty rate, Sirius XM expects to report cumulative adjusted EBITDA of positive $2.6 billion, net income of positive $3.4 billion, and free cash flow of positive $1 billion. SX Trial Ex. 217. EBITDA results are predicted to increase in the coming years, whether the royalty rates are set beginning at 9% in 2013 and rising 1% per year to end at 13% (Morgan Stanley's “base case” scenario), or beginning at 12% in 2013 and rising by 2% per year to end at 20% (Morgan Stanley's “bear case” scenario). SXM Reb. Trial Ex. 12 at 9; SX PFF ¶ 568. In sum, I cannot discern how selection of any rate within the range of reasonable rates suggested by the marketplace evidence will fail to enable Sirius XM to earn a fair income in the upcoming licensing period.

With respect to fair return to the copyright owner, I accept the testimony of Mr. Ciongoli and others that revenues from the statutory licenses are of greater importance to record labels as a result of the changes brought about by digital distribution of music and that such revenues contribute to the overall profits. Their importance may be offset somewhat by the gains achieved by the lower costs associated with digital distribution and the efficiencies achieved by the record industry in recent years through downsizing. At best, the record testimony suggests that a royalty rate above the existing 8% of Gross Revenues will promote a fair return to copyright owners in the upcoming licensing term, but the evidence does not permit quantification of an increase with accuracy. Nevertheless, I am satisfied that the rates set forth below incorporate the policy considerations of fair return/fair income prescribed in the second Section 801(b) factor.

c. Relative Roles of Copyright Owners and User

This policy factor requires that the rates adopted reflect the relative roles of the copyright owners and copyright user in the product made available with respect to relative creative contribution, technological contribution, capital investment, cost, risk, and contribution to the opening of markets for creative expression and media for their communication. 17 U.S.C. 801(b)(1)(C). The majority of the evidence and arguments submitted by the parties on this factor can be generally described by a single inquiry: who spent more on their business? Compare, SX PFF ¶¶ 535-544 with Sirius XM PFF ¶¶ 278, 289-290, 294. Capital investments, costs and risk, however, are only part of the analysis required by the third Section 801(b) factor. Relative creative and technological contributions, as well as contributions to opening new markets must also be considered. Sirius XM contends that it has pioneered and built a complex satellite delivery system that assures uninterrupted, nationwide availability of programming content, thereby creating a satellite radio business that did not previously exist. Sirius XM PFF ¶¶ 280-286. SoundExchange counters that Sirius XM has exploited mostly existing technology, principally designed and built by WorldSpace, Boeing, PanAmSat and the United States Army. SX RFF ¶¶ 252-257.

As is stated with respect to the PSS, supra, the task is not to consider each element of the third factor separately and make unspecified, unquantified up or down adjustments to the chosen royalty rates. Rather, the task with respect to the SDARS rate is to consider the elements as a whole and determine whether such consideration warrants any directional change in the range of rates established by the evaluation of the marketplace evidence (i.e., 5%-7% on the lower end to 16.2% on the upper end). I conclude, upon careful weighing of the evidence, that the third Section 801(b) factor does not require royalty rates that hue to either end of the spectrum of reasonable rates. In fact, little has changed in the evidentiary record relevant to this factor since SDARS-I. Sirius XM continues to overstate the originality of its technological contributions, as well as its exposure to risk. Elbert Designated WRT passim, SX Trial Ex. 410. No new markets have been opened during the current licensing term, nor is there evidence suggesting that the situation will change in the upcoming term. As was the case in SDARS-I, Sirius XM and the record companies continue to invest large sums in operating and advancing their businesses, as well as developing products for the future. The evidence does not indicate that the output or efforts of either side warrants a higher or lower royalty rate.

d. Minimize Disruptive Impact

The fourth policy factor under Section 801(b) requires the Judges to determine rates that “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). The analytical framework for my evaluation of this factor is well established. A royalty rate may be considered disruptive “if it directly produces an adverse impact that is substantial, immediate and irreversible in the short-run because there is insufficient time for [the parties impacted by the rate] to adequately adapt to the changed circumstances 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.” SDARS-I, 74 FR at 4097; see also Phonorecords I, 74 FR 4510, 4525 (Jan. 26, 2009).

In SDARS-I, it was the Judges' consideration of this factor that merited the adoption of a rate below the upper bound of the zone of reasonable market rates suggested by the interactivity-adjusted Ordover benchmark (i.e., 13%). It is, therefore, not surprising that the parties have devoted most of their argument under Section 801(b) to the fourth factor. Much of this argument is inapposite here, however, because it is made in support of the parties' respective rate proposals. The task here is to evaluate rates within the 5%-7% to 16.2% zone of reasonableness and select a rate or rates, consistent with the other Section 801(b) factors, that will not cause disruption. This requires consideration of the evidence on disruption presented in this proceeding, not the evidence that was presented or evaluated by the Judges in SDARS-I.

The record in this proceeding demonstrates that Sirius XM is in a far stronger financial position than it was at the time of SDARS-I. At the end of 2007, Sirius and XM ^[90] had a total of 17.3 million subscribers. SX Trial Ex. 16 at 18 (SXM_CRB_DIR_00021683). By the end of 2012, Sirius XM has announced that, with a net increase of 1.6 million subscribers this year, it will attain 23.5 million subscribers. SX Trial Ex. 217 at 7. In 2007, Sirius and XM had combined revenue of only $2.1 billion, with combined adjusted EBITDA of negative $565 million. SX Trial Ex. 16 at p. 14-15 (SXM_CRB_DIR_00021680-81). By the end of 2012, Sirius XM has announced that its revenue will be $3.4 billion, and its adjusted EBITDA will be approximately a positive $900 million. SX Trial Ex. 217 at 7. A similar situation applies to free cash flow, rising from negative $505 million in 2007 to approximately positive $700 million in 2012. SX Trial Ex. 16 at 16 (SXM_CRB_DIR_00021682); SX Trial Ex. 217 at 7; see also Lys Corrected WDT at 18-21, SX Trial Ex. 80. In 2007, Sirius and XM faced considerable expense in the completion of their satellite builds, the failure of which, the Judges recognized, “clearly raises the potential for disruption of the current consumer service.” SDARS-I, 73 FR at 4097. Sirius XM has no plans to launch or invest in new satellites during the 2013-2017 licensing period. Lys WRT at SX Ex. 211-RP at 44, SX Trial Ex. 240; 6/6/12 Tr. 607:18-22 (Meyer).

Despite its strong financials in 2012, Sirius XM's witnesses attempt to paint a grim picture for the upcoming licensing term. David Stowell, professor of finance at Northwestern University's Kellogg School of Management, testified that Sirius XM's financial history and substantial accumulated losses evince a threat of disruption caused by higher royalty rates that is “equal to or even greater than the one it faced at the time of the last rate proceeding.” Stowell WDT at 21, SXM Dir. Trial Ex. 18. David Frear, Sirius XM's Chief Financial Officer, testified that Sirius XM's brush with bankruptcy in late 2008 (where it struggled to repay the balance due on notes that matured on February 17, 2009, until receiving a loan from Liberty Media) requires that Sirius XM maintain a cash reserve of at least $750 million to guard against future calamity. Frear WDT at 4-5, SXM Dir. Trial Ex. 12; 6/7/12 Tr. 663:17-665:2 (Frear). Mr. Frear and Mel Karmazin, the Chief Executive Officer of Sirius XM, testified that the satellite delivery infrastructure of Sirius XM radio is inherently risky and that any number of events could seriously impact its ability to deliver programming and result in large, unanticipated expense. Frear WDT at 9, SXM Dir. Trial Ex. 12; Karmazin WDT at 17, SXM Dir. Trial Ex. 19. James Meyer, Sirius XM's President of Operations and Sales, testified that current economic uncertainty can affect the purchase of Sirius XM in automobiles and increase the number of current subscribers discontinuing service (described as the “churn rate”). 6/6/12 Tr. 566:21-568:16 (Meyer); Meyer WDT at 18-19, 29-30, SXM Dir. Trial Ex. 5. And William Rosenblatt, president of GiantSteps Media Technology Strategies, along with Messrs. Meyer, Karmazin, Frear and Professor Stowell, testified that rapidly evolving Internet-based competitors, advantaged by rapidly expanding wireless broadband capabilities and the explosion of smartphone use, present a potentially great disruptive challenge to Sirius XM during the 2013-2017 licensing period. Rosenblatt Corrected WDT passim, SXM Dir. Trial Ex. 17; Meyer WDT at 7-18, SXM Dir. Trial Ex. 5; Stowell WDT at 10-11, 22, SXM Dir. Trial Ex.18; 6/11/12 Tr. 1429:6-13 (Karmazin); 8/13/12 Tr. 3042:5-3043:9 (Frear).

The problem with Sirius XM's parade of horribles is that—with one exception—it is belied by the evidence and, in most instances, by Sirius XM's own public statements. Dr. Thomas Lys, SoundExchange's expert economist, presented data projecting Sirius XM's likely future EBITDA and free cash flow (two financial measures that the Judges focused on in considering the disruption factor in SDARS-I) using forecasts from Morgan Stanley and Sirius XM's own internal projections. Morgan Stanley projects significant positive EBITDA and free cash flow for Sirius XM in the upcoming license period under varying scenarios with different royalty rates and economic conditions. Lys WRT at 21-31, SX Trial Ex. 240. Particularly relevant to the consideration of reasonable rates is Morgan Stanley's recent 2012 baseline projection which assumes that royalty rates will begin at 9% in 2013 and rise 1% per year to end at 13%. SXM Reb. Trial Ex. 12 at 9. Under this projection, Sirius XM's EBITDA will increase each year despite the increases in rates and will be higher than Sirius XM has achieved in the history of its company. Id. Furthermore, projections made using Sirius XM's own internal forecasts generally corroborate these results. Lys WRT at 11, SX Trial Ex. 240.

Sirius XM vehemently opposes consideration of either the Morgan Stanley or its own internal projections, arguing that long-term financial projections for Sirius XM are not reliable. Sirius XM RFF ¶¶ 118-129. It is certainly true that the longer the term of forecast, the lesser the degree of accuracy that can be expected in the latter portion of the term. However, Sirius XM does not and cannot contend that short-term projections, either its own or those of Morgan Stanley, are highly unreliable.^[91] Both show substantial EBITDA profitability and positive free cash flow, even under scenarios that exceed the range of reasonable rates I have identified in this proceeding. See Lys Corrected WDT at 25-28, SX Trial Ex. 80. A royalty rate can be disruptive under the fourth Section 801(b) factor if it “produces an adverse impact that is substantial, immediate and irreversible in the short-run.” SDARS-I, 73 FR at 4097. The Morgan Stanley and Sirius XM internal projections convincingly reveal that disruption to Sirius XM will not occur in the short run by royalty rates within the range of reasonable rates identified in this proceeding.^[92]

There is also another element of Sirius XM's business operation that persuades me that rates within the reasonable range I have identified in this proceeding will not be disruptive: the Music Royalty Fee. The U.S. Music Royalty Fee was adopted by Sirius XM in July 2009, with the permission of the Federal Communications Commission, as a result of the Sirius and XM merger and in response to the royalty rates adopted in SDARS-I,^[93] to pass through to subscribers Sirius XM's music royalty costs. After adopting the $1.98 per subscriber per month charge (it is currently $1.42), Mr. Karmazin informed investors that there was no “discernable impact on churn,” meaning that the overall price increase to subscribers did not impact Sirius XM's ability to retain its subscribers. Lys WRT at 33-34, SX Trial Ex. 240. Sirius XM's long-range planning documents reveal an intention for future use of the Music Royalty Fee to recoup music licensing expenses, SX Trial Ex. 9 at 6 (SXM_CRB_DIR_00031738), and neither Messrs. Karmazin nor Frear denied that the Music Royalty Fee will continue to appear on customers' bills in some amount in the upcoming 2013-2017 licensing period. Sirius XM's demonstrated ability to pass through music licensing costs to its subscribers without discernible, negative impact to its satellite radio business further belies its claims that increased royalty fees from current levels will be disruptive.^[94]

Sirius XM also posits several financial risks for the upcoming license period that it claims will be exacerbated by higher royalty rates. First, Sirius XM contends that higher royalty rates will reduce available levels of free cash flow to such an extent as to impair Sirius XM's ability to account for possible downturns in any of its key performance metrics such as churn, conversion from trial to paid subscriptions, and average revenue per user over the upcoming rate term. Sirius XM RFF ¶ 117. Unlike consideration of the Sirius XM or Morgan Stanley forecasts, which have reasonable reliability at least in the short term, Sirius XM's suggestions of possible downturns in its satellite radio business are no more than that. While downturns are possible, Sirius XM has not presented compelling testimony that any one or more events are probable and, therefore, must be considered closely.

Second, Sirius XM contends that its “brush with bankruptcy” in the aftermath of the July 2008 merger demonstrates the risk of its debt level and difficulty in accessing credit markets, all of which will be made worse by higher royalty rates. Sirius XM PFF ¶¶ 313-314. The “brush with bankruptcy” argument, however, is a red herring, as it was caused by the need to refinance during a global-wide credit crisis and had nothing to do with the Section 114 royalty rates. 8/20/12 Tr. 4040:14-4042:7 (Lys).^[95] Professor Stowell's conclusion that Sirius XM has a “realistic possibility” of defaulting on its outstanding debt in the near future is speculative and not based upon the possibility of higher Section 114 royalty rates, since the ratings agencies do not discuss such royalties as a primary risk of Sirius XM in assessing its credit quality and likelihood of default. Lys WRT at 23, SX Trial Ex. 240; 8/20/12 Tr. 4049:2-4050:13 (Lys). Moreover, credit rating agencies have repeatedly raised Sirius XM's credit rating over the last few years and believe that it has strong liquidity and ability to finance its debt. Lys Corrected WDT at 31-32, SX Trial Ex. 80; Lys WRT at 47-48, SX Trial Ex. 240.

Third, Sirius XM argues that higher royalty rates will disrupt its ability to recoup billions of dollars of accumulated losses. Sirius XM PFF ¶ 312; see also Frear WDT at 7, SXM Dir. Trial Ex. 12 (discussing decrease in Sirius and XM stock prices from 2000 to 2007). Past losses and decreases in stock prices, however, do not have relevance to a disruption analysis under Section 801(b). There is no evidence that the expenditures of prior investors will have any impact on the future decision making or operation of the company. See, e.g., 6/8/12 Tr. 1297:1-8 (Stowell) (Professor Stowell acknowledging that he did not know if any pre-2008 investors are still owners of Sirius XM stock today).

In sum, there is no persuasive evidence that an increase in royalty rates from the current level and within the range of reasonable rates identified by the analysis of market evidence will cause disruption to the operation of Sirius XM's satellite radio business in the beginning to middle of the 2013-2017 licensing period. There is, however, testimony that raises the potential for disruption in the latter portion of the licensing term. New Internet-based competitors, whose emergence is enabled by the explosion of wireless broadband capability and smartphone use, appear poised to offer the same advantages over terrestrial radio that Sirius XM once claimed only to itself, and without the expenses associated with a satellite-based delivery system. Meyer WDT at 5-11, SXM Direct Trial Ex. 5; Rosenblatt Corrected WDT at 12-14, 20-38, SXM Dir. Trial Ex. 17. Such competitors can also offer their customers the added benefits of increased customization and personalization which Sirius XM is incapable of providing on its satellite radio service. Meyer WDT at 8-9, SXM Dir. Trial Ex. 5; Rosenblatt Corrected WDT at 20-31, SXM Dir. Trial Ex. 17. Many of these competitive products are being introduced already, particularly in automobiles which lie at the core of Sirius XM's satellite radio business, and could cause disruption by 2016 or 2017. See Meyer WDT at 15 (all major car manufacturers expected to incorporate connected-car technology within the next three years), SXM Dir. Trial Ex. 5. SoundExchange counters that Sirius XM enjoys considerable advantages over Internet radio competitors, such as a head start in integration of satellite radios into the automobile dashboard, current agreements with auto manufacturers, and current limitations on network streaming technology. SX PFF ¶¶ 637, 639-640, 642, 645-648. Nevertheless, SoundExchange does acknowledge that Internet-based competitors will grow, along with Sirius XM, in the coming rate term. Id. ¶ 650.

The evidence suggests that competition from Internet-based radio, particularly in the automobile, may cause disruption to Sirius XM's business by the final two years of the upcoming licensing period. The potential for such disruption is underscored by the limitations afforded to long-term financial projections (four and five years from now), and the relative speed in technological development demonstrated in the marketplace in recent years for delivery of music. I cannot forecast with certainty the degree to which future Internet-based competition may cause disruption in Sirius XM's business and, therefore, cannot determine the amount to which royalty rates may or should be reduced to prevent such disruption. However, the potential for disruption is suggested sufficiently by the evidence and counsels against escalation of the royalty rates in the last two years of the 2013-2017 license period.

5. Conclusions Regarding Section 114 Rates

As discussed above, analysis of the market-based evidence presented in this case yields a range of reasonable royalty rates between 5%-7% on the lower end, and 16.2% on the upper end. I have analyzed and applied the Section 801(b) factors to this range of reasonable rates and conclude that only two of the factors—the second and the fourth—impact the selection of rates within the range for the upcoming 2013-2017 licensing term. The second factor (fair return/fair income under existing market conditions) suggests selection of royalty rates that are above the current 8% rate, albeit without specific quantification. The fourth factor (minimizing any disruptive impact on the structure of the industries involved and on generally prevailing industry practices) counsels against raising the royalty rates further in the final two years of the licensing term. I dissent from the rates adopted by the majority and submit that they should be as follows: for 2013: 10.0%; for 2014: 11.0%; for 2015: 12.0%; 2016: 12.0%; and for 2017: 12.0%.

Dated: February 14, 2013.

William J. Roberts, Jr.,

Copyright Royalty Judge.

List of Subjects in 37 CFR Part 382

• Copyright
• Digital audio transmissions
• Performance right
• Sound recordings

Final Regulations

For the reasons set forth in the preamble, the Copyright Royalty Judges amend 37 CFR part 382 as follows:

PART 382—RATES AND TERMS FOR DIGITAL TRANSMISSIONS OF SOUND RECORDINGS AND THE REPRODUCTION OF EPHEMERAL RECORDINGS BY PREEXISTING SUBSCRIPTION SERVICES AND PREEXISTING SATELLITE DIGITAL AUDIO RADIO SERVICES

1. The authority citation for part 382 continues to read as follows:

Authority: 17 U.S.C. 112(e), 114 and 801(b)(1).

§ 382.1

[Amended]

2. Section 382.1 is amended as follows:

a. In paragraph (a), by removing “114(d)(2)” and adding “114” in its place, and by removing “ephemeral phonorecords” and adding “Ephemeral Recordings” in it place;

b. In paragraph (c), by removing “ephemeral phonorecords” and adding “Ephemeral Recordings” in its place; and

c. By removing paragraph (d).

§§ 382.2 through 382.7

[Redesignated as §§ 382.3 through 382.8]

3. Redesignate §§ 382.2 through 382.7 as §§ 382.3 through 382.8, respectively, and add new § 382.2 to read as follows:

§ 382.2

Definitions.

For purposes of this subpart, the following definitions shall apply:

Collective is the collection and distribution organization that is designated by the Copyright Royalty Judges. For the 2013-2017 license term, the Collective is SoundExchange, Inc.

Copyright Owners are sound recording copyright owners who are entitled to royalty payments made under this subpart pursuant to the statutory licenses under 17 U.S.C. 112(e) and 114.

Ephemeral Recording is a phonorecord created for the purpose of facilitating a transmission of a public performance of a sound recording under a statutory license in accordance with 17 U.S.C. 114 and subject to the limitations specified in 17 U.S.C. 112(e).

GAAP shall mean generally accepted accounting principles in effect from time to time in the United States.

Gross Revenues. (1) Gross Revenues shall mean all monies derived from the operation of the programming service of the Licensee and shall be comprised of the following:

(i) Monies received by Licensee from Licensee's carriers and directly from residential U.S. subscribers for Licensee's programming service;

(ii) Licensee's advertising revenues (as billed), or other monies received from sponsors, if any, less advertising agency commissions not to exceed 15% of those fees incurred to a recognized advertising agency not owned or controlled by Licensee;

(iii) Monies received for the provision of time on the programming service to any third party;

(iv) Monies received from the sale of time to providers of paid programming such as infomercials;

(v) Where merchandise, service, or anything of value is received by Licensee in lieu of cash consideration for the use of Licensee's programming service, the fair market value thereof or Licensee's prevailing published rate, whichever is less;

(vi) Monies or other consideration received by Licensee from Licensee's carriers, but not including monies received by Licensee's carriers from others and not accounted for by Licensee's carriers to Licensee, for the provision of hardware by anyone and used in connection with the programming service;

(vii) Monies or other consideration received for any references to or inclusion of any product or service on the programming service; and

(viii) Bad debts recovered regarding paragraphs (1)(i) through (vii) of this definition.

(2) Gross Revenues shall include such payments as set forth in paragraphs (1)(i) through (viii) of this definition to which Licensee is entitled but which are paid to a parent, subsidiary, division, or affiliate of Licensee, in lieu of payment to Licensee but not including payments to Licensee's carriers for the programming service. Licensee shall be allowed a deduction from “Gross Revenues” as defined in paragraph (1) of this definition for affiliate revenue returned during the reporting period and for bad debts actually written off during reporting period.

Licensee means any preexisting subscription service as defined in 17 U.S.C. 114(j)(11).

Performers means the independent administrators identified in 17 U.S.C. 114(g)(2)(B) and (C), and the parties identified in 17 U.S.C. 114(g)(2)(D).

Qualified Auditor is a Certified Public Accountant.

4. Revise newly redesignated § 382.3 to read as follows:

§ 382.3

Royalty fees for the digital performance of sound recordings and the making of ephemeral recordings by preexisting subscription services.

(a) Commencing January 1, 2013, and continuing through December 31, 2017, the monthly royalty fee to be paid by a Licensee for the public performance of sound recordings pursuant to 17 U.S.C. 114 and the making of any number of Ephemeral Recordings to facilitate such performances pursuant to 17 U.S.C. 112(e) shall be a percentage of monthly Gross Revenues resulting from residential services in the United States as follows: for 2013, 8%; and for 2014 through 2017, 8.5%.

(b) Each Licensee making digital performances of sound recordings pursuant to 17 U.S.C. 114 and Ephemeral Recordings pursuant to 17 U.S.C. 112(e) shall make an advance payment to the Collective of $100,000 per year, payable no later than January 20th of each year. The annual advance payment shall be nonrefundable, but it may be counted as an advance of the section 112 royalties due and payable for a given year or any month therein under paragraph (a) of this section; Provided, however, that any unused portion of an annual advance payment for a given year shall not carry over into a subsequent year.

(c) The royalty payable under 17 U.S.C. 112(e) for the making of phonorecords used by the Licensee solely to facilitate transmissions for which it pays royalties as and when provided in this subpart shall be included within, and constitute 5% of, the total royalties payable under 17 U.S.C. 112(e) and 114.

(d) A Licensee shall pay a late fee of 1.5% per month, or the highest lawful rate, whichever is lower, for each payment or statement of account, or either of them, received by the Collective after the due date. Late fees shall accrue from the due date until payment and the statement of account are received.

5. Revise newly redesignated § 382.4 to read as follows:

§ 382.4

Terms for making payment of royalty fees and statements of account.

(a) Payment to the Collective. A Licensee shall make the royalty payments due under § 382.3 to the Collective.

(b) Timing of payment. A Licensee shall make any payments due under § 382.3 on a monthly basis on or before the 45th day after the end of each month for that month.

(c) Statements of Account. Licensees shall submit monthly statements of account on a form provided by the Collective. A statement of account shall contain the following information:

(1) Such information as is necessary to calculate the accompanying royalty payments;

(2) The name, address, business title, telephone number, facsimile (if any), electronic mail address and other contact information of the person to be contacted for information or questions concerning the content of the statement of account;

(3) The signature of a duly authorized officer or representative of the Licensee;

(4) The printed or typewritten name of the person signing the statement of account;

(5) The date of signature;

(6) The title or official position held in relation to the Licensee by the person signing the statement of account;

(7) A certification of the capacity of the person signing; and

(8) A statement to the following effect:

I, the undersigned officer or representative of the Licensee, have examined this statement of account and hereby state that it is true, accurate, and complete to my knowledge after reasonable due diligence.

(d) Distribution of royalties. (1) The Collective shall promptly distribute royalties received from Licensees to Copyright Owners and Performers, or their designated agents, that are entitled to such royalties. The Collective shall be responsible only for making distributions to those Copyright Owners, Performers, or their designated agents who provide the Collective with such information as is necessary to identify the correct recipient. The Collective shall distribute royalties on a basis that values all performances by a Licensee equally based upon the information provided under the reports of use requirements for Licensees contained in § 370.3 of this chapter.

(2) If the Collective is unable to locate a Copyright Owner or Performer entitled to a distribution of royalties under paragraph (d)(1) of this section within 3 years from the date of payment by a Licensee, such royalties shall be handled in accordance with § 382.8.

(e) Retention of records. Both Licensees and the Collective shall maintain books and records relating to the payment of the license fees in accordance with generally accepted accounting principles for a period of three years after the end of the period for which the payment is made. These records shall include, but are not limited to, the statements of account, records documenting an interested party's share of the royalty fees, and the records pertaining to the administration of the collection process and the further distribution of the royalty fees to those interested parties entitled to receive such fees.

6. Newly redesignated § 382.5 is amended as follows:

a. By revising the section heading;

b. In paragraph (a), by removing “which has been” and by removing “§§ 382.5 and 382.6” and adding “§§ 382.6 and 382.7” in its place;

c. By removing paragraphs (b), (c), and (f);

d. By redesignating paragraphs (d) and (e) as paragraphs (b) and (c), respectively;

e. In the introductory text of newly redesignated paragraph (b), by adding “subject to an appropriate confidentiality agreement and” after “be”;

f. By revising newly redesignated paragraphs (b)(1) and (b)(3);

g. In newly redesignated paragraph (b)(2), by removing “qualified auditor” and adding “Qualified Auditor” in its place, by removing “copyright owner or performing artist” and adding “Copyright Owner or Performer” in its place, by removing “copyright owners” and adding “Copyright Owners” in its place, and by removing “payments.” and adding “payments; and” in its place; and

h. In newly redesignated paragraph (c), by removing “(d)” and adding “(b)” in its place.

The revisions read as follows:

§ 382.5

Confidential information.

* * * * *

(b) * * *

(1) Those employees, agents, consultants and independent contractors of the Collective who are engaged in the collection and distribution of royalty payments hereunder and activities directly related hereto, who are not also employees or officers of a sound recording Copyright Owner or Performer, and who, for the purpose of performing such duties during the ordinary course of employment, require access to the records; and

* * * * *

(3) Copyright Owners and Performers whose works have been used under the statutory licenses set forth in 17 U.S.C. 112(e) and 114 by the Licensee whose Confidential Information is being supplied, or agents thereof provided that the only confidential information that may be shared pursuant to this paragraph (b)(3) are the monthly statements of account that accompany royalty payments.

* * * * *

7. Newly Redesignated § 382.6 is amended as follows:

a. In paragraph (c), by removing “with” and adding “to” in its place, by removing “parties' ” and adding “party's” in its place, by removing “served” and adding “delivered” in its place, and by removing “on the party” and adding “to the party” in its place;

b. In paragraph (d), by adding “from the date of completion of the verification process” after “years”;

c. In paragraph (e), by removing “auditor” and adding “and Qualified Auditor” in its place;

d. By revising paragraph (f); and

e. In paragraph (g), by removing “copyright owners” and adding “Copyright Owners” in its place.

The revision reads as follows:

§ 382.6

Verification of statements of account.

* * * * *

(f) Costs of the verification procedure. The interested party or parties requesting the verification procedure shall pay all costs of the verification procedure, unless an independent and Qualified Auditor concludes that during the period audited, the Licensee underpaid royalties by an amount of five (5) percent or more; in which case, the service that made the underpayment shall bear the costs of the verification procedure.

* * * * *

8. Newly redesignated § 382.7 is amended as follows:

a. In paragraph (c), by removing “with” and adding “to” in its place, by removing “parties' ” and adding “party's” in its place, by removing “interest” and adding “intent” in its place, by removing “served” and adding “delivered” in its place, and by removing “on the party” and adding “to the party” in its place;

b. In paragraph (d), by adding “after completion of the verification process” after “years”; and

c. In paragraph (e), by removing “auditor” and adding “and Qualified Auditor” in its place; and

d. By revising paragraph (f).

The revision reads as follows:

§ 382.7

Verification of royalty payments.

* * * * *

(f) Costs of the verification procedure. The interested party or parties requesting the verification procedure shall pay for all costs associated with the verification procedure, unless an independent and Qualified Auditor concludes that, during the period audited, the Licensee underpaid royalties in the amount of five (5) percent or more, in which case, the entity that made the underpayment shall bear the costs of the verification procedure.

* * * * *

9. Newly redesignated § 382.8 is amended as follows:

a. By revising the section heading;

b. By removing “copyright owner or performer” and adding “Copyright Owner or Performer” in its place;

c. By removing “date of distribution” and adding “date of the last distribution from the royalty fund at issue” in its place; and

d. By removing “this period” and adding “the three-year claim period” in its place.

The revision reads as follows:

§ 382.8

Unclaimed funds.

* * * * *

§ 382.10

[Amended]

10. Section 382.10 is amended as follows:

a. In paragraph (a), by removing “2007” and adding “2013” in its place and by removing “2012” and adding “2017” in its place;

b. In paragraph (b), by removing “112” and adding “112(e)” in its place; and

c. In paragraph (c), by adding “voluntary” before “license agreements”.

11. Section 382.11 is amended as follows:

a. In the definition of “ Collective”, by removing “2007-2012” and adding “2013-2017” in its place and by removing “period” and adding “term” in its place;

b. In the definition of “ Copyright Owners”, by removing “114(f)” and adding “114” in its place;

c. By adding in alphabetical order a definition for “ Directly-Licensed Recording”;

d. In the definition of “ Ephemeral Recording”, by removing “114(f)” and adding “114” in its place;

e. In paragraph (1)(i) of the definition of “ Gross Revenues”, by removing “residential”;

f. In paragraph (3)(vi)(D) of the definition of “ Gross Revenues”, by removing “ephemeral recordings” and adding “Ephemeral Recordings” in its place.

g. By adding in alphabetical order a definition for “ Pre-1972Recording”;

h. By removing the definition for “ Residential”; and

i. In the definition of “ Term”, by removing “2007” and adding “2013” in its place and by removing “2012” and adding “2017” in its place.

The additions read as follows:

§ 382.11

Definitions.

* * * * *

Directly-Licensed Recording is a sound recording for which the Licensee has previously obtained a license of all relevant rights from the Copyright Owner of such sound recording.

* * * * *

Pre-1972 Recording is a sound recording fixed before February 15, 1972.

* * * * *

12. Section 382.12 is revised to read as follows:

§ 382.12

Royalty fees for the public performance of sound recordings and the making of ephemeral recordings.

(a) In general. The monthly royalty fee to be paid by a Licensee for the public performance of sound recordings pursuant to 17 U.S.C. 114(d)(2) and the making of any number of Ephemeral Recordings to facilitate such performances pursuant to 17 U.S.C. 112(e) shall be a percentage of monthly Gross Revenues as follows: for 2013, 9.0%; for 2014, 9.5%; for 2015, 10.0%; for 2016, 10.5%; and for 2017, 11.0%, except that the royalty fee so determined may be reduced by the Direct License Share or the Pre-1972 Recording Share as described in paragraphs (d) and (e), respectively, of this section.

(b) Ephemeral recordings. The royalty payable under 17 U.S.C. 112(e) for the making of phonorecords used by the Licensee solely to facilitate transmissions for which it pays royalties as and when provided in this subpart shall be included within, and constitute 5% of, the total royalties payable under 17 U.S.C. 112(e) and 114.

(c) Ephemeral recordings minimum fee. Each Licensee making Ephemeral Recordings pursuant to 17 U.S.C. 112(e) shall make an advance payment to the Collective of $100,000 per year, payable no later than January 20th of each year. The annual advance payment shall be nonrefundable, but it shall be considered as an advance of the Ephemeral Recordings royalties due and payable for a given year or any month therein under paragraphs (a) and (b) of this section; Provided, however, that any unused annual advance payment for a given year shall not carry over into a subsequent year.

(d) Direct license share. The percentage of monthly Gross Revenues royalty fee specified in paragraph (a) of this section may be reduced by a percentage as set forth in this paragraph (referred to herein as the “Direct License Share”).

(1) Subject to paragraph (d)(3) of this section, for each month, the Direct License Share is the result of dividing the Internet Performances of Directly-Licensed Recordings on the Reference Channels by the total number of Internet Performances of all sound recordings on the Reference Channels.

(2) For purposes of paragraph (d)(1) of this section:

(i) A “Performance” is each instance in which any portion of a sound recording is publicly performed to a listener within the United States by means of a digital audio transmission or retransmission (e.g., the delivery of any portion of a single track from a compact disc to one listener) but excluding an incidental performance that both:

(A) Makes no more than incidental use of sound recordings including, but not limited to, brief musical transitions in and out of commercials or program segments, brief performances during news, talk and sports programming, brief background performances during disk jockey announcements, brief performances during commercials of sixty seconds or less in duration, or brief performances during sporting or other public events; and

(B) Other than ambient music that is background at a public event, does not contain an entire sound recording and does not feature a particular sound recording of more than thirty seconds (as in the case of a sound recording used as a theme song).

(ii) The “Reference Channels” are Internet webcast channels offered by the Licensee that directly correspond to channels offered on the Licensee's SDARS that are capable of being received on all models of Sirius radio, all models of XM radio, or either or both, and on which the programming consists primarily of music.

(3) A Direct License Share adjustment as described in paragraph (d) of this section is available to a Licensee only if—

(i) The Reference Channels constitute a large majority of the music channels offered on the Licensee's SDARS and are generally representative of the music channels offered on the Licensee's SDARS; and

(ii) The Licensee timely provides the relevant information required by § 382.13(h).

(4) No performance shall be credited as an Internet Performance of a Directly-Licensed Sound Recording under this section if that performance is separately credited as an Internet Performance of a Pre-1972 sound recording under paragraph (e)(1) of this section.

(e) Pre-1972 Recording Share. The percentage of monthly Gross Revenues royalty fee specified in paragraph (a) of this section may be reduced by a percentage as set forth in this paragraph (referred to herein as the “Pre-1972 Recording Share”).

(1) Subject to paragraph (e)(3) of this section, for each month, the Pre-1972 Recording Share is the result of dividing the Internet Performances of Pre-1972 Sound Recordings on the Reference Channels by the total number of Internet Performances of all sound recordings on the Reference Channels.

(2) For purposes of paragraph (e)(1) of this section:

(i) A “Performance” is each instance in which any portion of a sound recording is publicly performed to a listener within the United States by means of a digital audio transmission or retransmission (e.g., the delivery of any portion of a single track from a compact disc to one listener) but excluding an incidental performance that both:

(A) Makes no more than incidental use of sound recordings including, but not limited to, brief musical transitions in and out of commercials or program segments, brief performances during news, talk and sports programming, brief background performances during disk jockey announcements, brief performances during commercials of sixty seconds or less in duration, or brief performances during sporting or other public events; and

(B) Other than ambient music that is background at a public event, does not contain an entire sound recording and does not feature a particular sound recording of more than thirty seconds (as in the case of a sound recording used as a theme song).

(ii) The “Reference Channels” are Internet webcast channels offered by the Licensee that directly correspond to channels offered on the Licensee's SDARS that are capable of being received on all models of Sirius radio, all models of XM radio or both, and on which the programming consists primarily of music.

(3) A Pre-1972 Recording Share adjustment as described in paragraph (e) of this section is available to a Licensee only if—

(i) The Reference Channels constitute a large majority of the music channels offered on the Licensee's SDARS and are generally representative of the music channels offered on the Licensee's SDARS; and

(ii) The Licensee timely provides the relevant information required by § 382.13(h).

13. Section 382.13 is amended as follows:

a. By revising paragraphs (c) and (d);

b. In paragraph (e)(3), by removing “handwritten”; and

c. By adding paragraph (h).

The revisions and addition read as follows:

§ 382.13

Terms for making payment of royalty fees and statements of account.

* * * * *

(c) Monthly payments. A Licensee shall make any payments due under § 382.12 on a monthly basis on or before the 45th day after the end of each month for that month. All payments shall be rounded to the nearest cent.

(d) Late payments and statements of account. A Licensee shall pay a late fee of 1.5% per month, or the highest lawful rate, whichever is lower, each for any payment or statement of account, or either of them received by the Collective after the due date. Late fees shall accrue from the due date until payment and the statement of account are received by the Collective.

* * * * *

(h) Notification of exclusions. (1) As a condition to a Licensee's taking a Direct License Share adjustment as described in § 382.12(d), by no later than the due date for the relevant payment under paragraph (c) of this section, the Licensee must provide the Collective a list of each Copyright Owner from which the Licensee claims to have a direct license of rights to Directly-Licensed Recordings that is in effect for the month for which the payment is made, and of each sound recording as to which the Licensee takes such an adjustment (identified by featured artist name, sound recording title, and International Standard Recording Code (ISRC) number or, alternatively to the ISRC, album title and copyright owner name). Notwithstanding § 382.14, the Collective may disclose such information as reasonably necessary for it to confirm whether a claimed direct license exists and claimed sound recordings are properly excludable.

(2) As a condition to a Licensee's taking a Pre-1972 Recording Share adjustment as described in § 382.12(e), by no later than the due date for the relevant payment under paragraph (c) of this section, the Licensee must provide the Collective a list of each Pre-1972 Recording as to which the Licensee takes such an adjustment (identified by featured artist name, sound recording title, and International Standard Recording Code (ISRC) number or, alternatively to the ISRC, album title and copyright owner name).

§ 382.14

[Amended]

14. Section 382.14 is amended as follows:

a. In the introductory text of paragraph (d), by adding “, subject to an appropriate confidentiality agreement,” after “limited”;

b. In paragraph (d)(1), by removing “, subject to an appropriate confidentiality agreement,”;

c. In paragraph (d)(2), by removing “, subject to an appropriate confidentiality agreement,”;

d. In paragraph (d)(3), by removing “114(f)” and adding “114” in its place and by removing “, subject to an appropriate confidentiality agreement,” each place it appears; and

d. In paragraph (d)(4), by removing “114(f)” and adding “114” in its place.

15. Section 382.15 is amended by revising paragraph (g) to read as follows:

§ 382.15

Verification of royalty payments.

* * * * *

(g) Costs of the verification procedure. The Collective shall pay all costs associated with the verification procedure, unless it determines that the Licensee underpaid royalties in an amount of 10% or more, in which case the Licensee shall, in addition to paying the amount of any underpayment, bear the reasonable costs of the verification procedure.

16. Section 382.16 is amended by revising paragraph (g) to read as follows:

§ 382.16

Verification of royalty distributions.

* * * * *

(g) Costs of the verification procedure. The Copyright Owner or Performer requesting the verification procedure shall pay all costs associated with the procedure, unless it is finally determined that the Licensee underpaid royalties in an amount of 10% or more, in which case the Collective shall, in addition to paying the amount of any underpayment, bear the reasonable costs of the verification procedure.

Dated: February 14, 2013.

Suzanne M. Barnett,

Chief Copyright Royalty Judge.

Approved by:

James H. Billington,

Librarian of Congress.

Footnotes

1.  During the course of the proceeding, Chief Judge Sledge and Judge Wisniewski retired. Judge Sledge retired in April 2012 before the start of oral testimony. The Librarian of Congress appointed his successor, Chief Judge Barnett, in April 2012. Judge Wisniewski retired on August 31, 2012, after the conclusion of oral testimony; the Librarian appointed an interim Copyright Royalty Judge, Judge Strasser, on September 17, 2012, pending the appointment of Judge Wisniewski's successor.

2.  The Judges did not consider all of the offered testimony. Ruling on motions to strike or exclude, the Judges edited or excluded testimony during the course of the hearing. This determination is based solely on the evidence the Judges admitted.

3.  The ephemeral license requires the Judges, among other things, to “establish rates that most clearly represent the fees that would have been negotiated between a willing buyer and a willing seller.” 17 U.S.C. 112(e)(4). The ephemeral license requires adoption of a minimum fee for each type of service offered by a transmitting organization.

4.  The Judges followed the same approach in determining royalty rates for the Section 115 mechanical license, the only proceeding involving the Section 801(b)(1) factors decided since SDARS-I. See, Phonorecords I, 74 FR 4510 (Jan. 26, 2009). None of the parties in this proceeding contend that this approach is erroneous or must be abandoned.

5.  The PSS are Music Choice and Muzak. Muzak's PSS service is, apparently, only a small part of its business, and it did not participate in this proceeding. Digital Music Express, Inc., which was a PSS in SDARS-I, ceased operation. 6/11/12 Tr. 1469:14-1470:6 (Del Beccaro).

6.  The current regulation defining Gross Revenues for PSS is set forth in 37 CFR 382.2(e). As discussed infra, the Judges are adopting SoundExchange's proposal to house all PSS definitions in a single location; consequently, the PSS definitions will be located in a new § 382.2.

7.  The definitions of “revenues” used to calculate the different musical works royalties are not revealed in the evidence, nor is the “revenue” to which the [REDACTED] could be applied for comparison.

8.  SoundExchange's expert economist, Dr. George Ford, who recently submitted testimony before the Canadian Copyright Board, acknowledged that in Canada the musical composition and sound recording performance royalties are equal. 8/21/12 Tr. 4304:5-22 (Ford).

9.  Dr. Crawford concludes that his economic model confirms that the sound recording performance royalty rate for PSS should be less than its musical works rate. 6/12/12 Tr. 1803:11-1804:20 (Crawford); see also Crawford Corrected WDT at 6, 23, 25, 30, PSS Trial Ex. 4.

10.  “Combined Agreement Surplus” is the revenue a PSS would earn in the market for the PSS provider when an agreement is reached with the record label less the costs net of the digital performance right in sound recordings. Crawford Corrected WDT at 16, PSS Trial Ex. 4. “Surplus” is the payment a good or service can command beyond its cost of production. Id. at n.33.

11.  “Threat point” is the amount a firm would earn in the absence of an agreement. Crawford Corrected WDT at 16, PSS Trial Ex. 4.

12.  To support its contention that PSS are promotional for record company artists, Music Choice offered the testimony of Damon Williams, who testified that record company executives consider Music Choice promotional because they provide artists with greater exposure. Williams WDT at 4-13, MC 28, MC 29, MC 32, PSS Trial Ex. 3. Mr. Williams argues that Music Choice has become more promotional since the PSS-I proceeding because it currently reaches more customers with more channels. Id. at 24.

13.  Dr. Crawford discounted the promotional value of Music Choice because he could not quantify it. Crawford Corrected WDT at 45, PSS Trial Ex. 4.

14.  Under this model, a firm's cost of capital is based on the expected return to induce investment. Crawford Corrected WDT at ¶ 167, PSS Trial Ex. 4.

15.  See supra at Section II.

16.  The fees paid to the performing rights societies for the performance right to musical works have been offered in non-PSS proceedings and have been rejected. See Webcasting II, 72 FR 24084, 24094-24095 (May 1, 2007); SDARS-I, 73 FR 4080, 4089-4090 (Jan. 24, 2008) and Webcasting I, 67 FR 45240, 45246-45247 (July 8, 2002) (Librarian of Congress's determination).

17.  The Judges understand that Judge Roberts in his dissent provides a more spirited rejection of the probative value of the Nash Framework as proffered in this context. The Judges concur with his assessment, but believe, as a threshold matter, that the Nash Framework, without real-world data to support its predictive capacity, is unworthy of further consideration.

18.  Much was made in the hearing and in closing arguments regarding Dr. Crawford's supposed use of audited financial data and Dr. Ford's use of unaudited financial data in an effort to examine costs and revenues of the PSS service vis-à-vis Music Choice's other non-PSS services. The Judges see no superiority to either data set as presented in this proceeding.

19.  It is improbable that Music Choice would continue to operate for over 15 years with the considerable losses that it claims.

20.  Sirius XM devoted considerable energy in this proceeding to discovery and presentation of evidence regarding actions by SoundExchange and its member record labels relating to the Direct Licensing initiative. Sirius XM contends that it would have been able to present a much greater number of Direct Licenses but for the interference of SoundExchange. In a rate determination proceeding, the Judges cannot adjudicate claims of tortious interference with contractual relations or business expectancies. Indeed, Sirius XM never presented such claims to the Judges for adjudication. Those claims can only be adjudicated in a court of competent jurisdiction. Had Sirius XM been able to make a sufficient showing that actions by SoundExchange were in fact interfering with the validity of this rate determination proceeding, then the Judges would have had to decide what effect, if any, such interference might have had on the validity of these rate proceedings. The Judges allowed evidence in this proceeding only to determine whether, and to what extent, any activity by either party might have skewed the evidence upon which the Judges must rely.

21.  Dr. Noll also examined similar agreements between major labels and the music services Slacker and Turntable.

22.  Examining these same agreements for Last.fm's interactive on-demand service—[REDACTED]—led Dr. Noll to conclude that sound recording rights owners charge [REDACTED] for non-interactive services than they do for interactive/on-demand services. Dr. Noll also found similar rate differentials in the [REDACTED]. Noll Revised Amended WDT at 76-79, Tables 2.2-2.2d and Appendices I-K, SXM Dir. Trial Ex. 1.

23.  The “greater of” metric is an amount per play, an amount per subscriber, or a percentage of the service's revenues. 6/14/12 Tr. 2261:7-2262:4 (Ordover).

24.  Dr. Ordover's mathematical calculation is as follows: He took the $12.95 Sirius XM subscription price, and then multiplied that by 50% to obtain the music portion of the subscription price of $6.475. He then multiplied the music-only satellite radio subscription price by 60% to 65% (his effective percentage of royalty derived from the interactive subscription service agreements) to obtain the music royalty of $3.88 to $4.21. Finally, he divided those numbers by the Sirius XM subscription price for the Select programming package to obtain the 30% to 32.5% range. 8/16/12 Tr. 3794:13-3795:9 (Salinger).

25.  The current price for this service is $14.49. Ordover Third Corrected/Amended WDT at 31 n.33, SX Trial Ex. 74.

26.  Dr. Ordover did not provide a weighted average of the non-interactive service prices because he concluded that he did not have reliable data, nor did he include, at the Judges' invitation, ad-supported non-interactive services in his calculation, deciding that such services would add undue complexity to his methodology. Ordover Amended WRT at 38-39, SX Trial Ex. 218.

27.  This rate was calculated by multiplying the interactivity ratio of .4865 ($4.86/$9.99) by the average per-subscriber royalty payment of $5.95, yielding an equivalent satellite radio payment of $2.89. The $2.89 per-subscriber rate was then divided by the $12.95 monthly charge for the Sirius XM Select satellite radio package, resulting in the percentage of revenue rate of 22.32%.

28.  Dr. Ordover estimated that the works licensed under the Direct Licenses represent no more than 2%-4% of the total number of works performed by Sirius XM. Ordover Amended WRT at 4-5, SX Trial Ex. 218; 6/6/12 Tr. 308:3-5 (Noll).

29.  The Judges recognize that direct payment to the Direct Licensors does not relieve them of their royalty obligations to their artists and performers; however, receipt of 100% of the royalties upfront is clearly attractive to certain record labels and was a selling point in negotiations with independent record labels. Powers WDT at 4-5, SX Trial Ex. 243.

30.  The implicit monthly price is applied to the effective percentage of revenue rate of [REDACTED] from the Last.fm agreements that serve as the numerator in Dr. Noll's calculation.

31.  Likewise, Sirius XM has failed to demonstrate that it could successfully substitute away to other providers of music. If that were the case, Sirius XM could have operated its business under the Direct Licenses, for example, and avoided participation in this proceeding altogether.

32.  SoundExchange contends that the Judges should completely discount the Direct Licenses because they were negotiated under the shadow of the statutory rate which was sure to influence the rates the parties agreed to as well as their willingness to negotiate at all. SX PFF ¶ 371. Although the Judges considered that fact when determining the amount of weight that the Direct License benchmark received, the Judges question whether any agreements regarding sound recording rights could be purely market-based given the current statutory framework. With that understanding, the Judges do not have the luxury of ignoring record evidence of the contemporaneous results of arm's length negotiations between the same buyers and sellers and rights involved in the market for which the Judges are charged to determine a reasonable rate that will remain reasonable for the next five years, no matter how many weaknesses those results might exhibit.

33.  SoundExchange notes that “[t]here are no sound economic reasons to adjust market-based rates because of this statutory objective.” SX PFF ¶ 502. Other than its affirmation that the revenue from the SDARS is important to record labels, SX PFF ¶ 515, SoundExchange directs us to no evidence in the record that would warrant an upward adjustment in the rate that is most strongly indicated by the totality of the evidence in the proceeding based on the first Section 801(b) factor. Therefore, the Judges limit the discussion in this section to Sirius XM's arguments about why a downward adjustment is warranted under this factor.

34.  Dr. Noll concedes that “there's no published academic research on this issue, and, indeed, there's not enough data available for me to undertake such a research project.” 6/5/12 Tr. 227:22-228:3 (Noll). Moreover, he concedes that even the industry studies that have been done are ambiguous in their conclusions. Some think there's a substitution effect, some think there isn't. On balance, it seems to be the case that issue is unresolved, but there's no—there's no question that you wouldn't say it's a promotional effect, like satellite radio or terrestrial radio. It's either nothing or it's a substitution effect.

Id. at 228:8-16.

35.  See, e.g., 8/13/2012 Tr. 3049:8-16 (Frear).

36.  The Judges acknowledge that the sellers in the Direct Licenses represent a small subset of independent labels and exclude major labels, which, according to Sirius XM, were unwilling to enter direct license negotiations with Sirius XM. Sirius XM PFF ¶ 47-48. Nevertheless, the depth and breadth of the labels that signed Direct Licenses with Sirius XM strongly suggest that the relative role of the independent labels that entered the Direct Licenses with Sirius XM is, for the limited purpose of analyzing the third Section 801(b) factor, sufficiently comparable to that of independent labels generally and to that of the major labels. See, e.g., Sirius XM PFF ¶¶ 91-103; Noll Revised Amended WDT at 39-44 (record labels that signed Direct Licenses included: One of Billboard's top five independent labels for eight of the past nine years; labels that represented Grammy Award-winning and Grammy Award-nominated artists in multiple categories; a label that amassed more than 20 number one albums on Billboard's kids' album chart; the world's largest independent classical music label with a repertoire of over 2500 titles; a label with three songs on the Contemporary Christian top 10 during a period in 2012 and eight of the 50 spots on the Christian songs chart in 2012; a label representing an artist with the number one Billboard Heatseeker album in 2011; a label representing an artist with three Gold records; and a label that released the comedy albums of five-time Grammy Award-winning comedian George Carlin), SXM Dir. Trial Ex. 1. That being said, the absence of a major record label in the Direct License agreements supports the Judges' earlier conclusion that rates below 7% are below the lower bound of the zone of reasonableness.

37.  To the extent that the rights between the Direct License benchmark and target market vary in material respect, the Judges reflected such variances in the decision to adopt the upper end (i.e., 7%) of the range of rates for the benchmark.

38.  Sirius XM attributes its current profitability largely to its one-time merger-related cost cuts. Frear WDT at 8, SXM Dir. Trial Ex. 12. It is also noteworthy that Sirius XM reduced its programming costs by renegotiating its agreements with several high-profile content providers. Id. at 8-9. Sirius XM expects, however, that its operating costs will increase over the licensing period. Id. at 20. According to Sirius XM, optimistic public statements made by Sirius XM's management should be discounted as “puffery.” Sirius XM PFF ¶ 312 and n.66.

39.  Although the Judges agree that the fourth factor should be applied with caution (as should all of the factors), Section 801(b)(1)(D) of the Copyright Act does not restrict its application to “nascent and emerging services” and the Judges are unwilling to read such limitations into it. Nevertheless, the Judges are cognizant of SoundExchange's concern that the fourth factor not be applied in a way that would shield service providers from a competitive marketplace.

40.  EBITDA stands for Earnings, Before Interest, Taxes, and Depreciation Allowance.

41.  Free cash flow measures cash generated by a company that is not needed to fund the current period's operations or to reinvest in the firm's future operations. Lys Corrected WDT at 18-19 n. 94, citing Brealey, Richard, Stewart Myers and Franklin Allen, Principles of Corporate Finance, 2006, p. 997.

42.  The Judges are much less confident that the same could be said for an increase to the 22%-32.5% that the proposed Dr. Ordover benchmark might suggest. As a result, the Judges draw additional comfort that the upper bound of the zone of reasonableness is closer to 12%-13% than it is to 22.32%, which the Judges conclude the upper bound of the zone of reasonableness can be no more than. See infra at Section V.B.3.b.

43.  Sirius XM's contentions that it faces risks of disruption due to its increasing reliance on the OEM distribution channels (i.e., the automobile market) (Sirius XM PFF ¶¶ 322-324), or potential risks from relying on a satellite infrastructure (Sirius XM PFF ¶¶ 316-317), or risks posed by macroeconomic conditions (Sirius XM PFF ¶¶ 318-320) are too speculative to suggest the type of substantial, immediate and irreversible adverse impact that would warrant a downward adjustment of the benchmark rate under the fourth factor.

44.  SoundExchange's proposed Gross Revenues definition is set forth at Second Revised Proposed Rates and Terms of SoundExchange, Inc., at 2-3 (Sept. 26, 2012).

45.  In SDARS-I, the Judges expressly recognized an exclusion from Gross Revenues for so-called non-music services, characterized as “channels, programming, products and/or other services offered for a separate charge where such channels use only incidental performances of sound recordings.” SDARS-I, 73 FR 4102 (citing 37 CFR 382.11, definition of Gross Revenues). The Judges did so because this exclusion “unambiguously relat[ed] the fee to the value of the sound recording performance rights at issue * * *” Id. at 4088. SoundExchange argues that if the current exclusion is allowed to continue, it would result in a double deduction from Sirius XM's royalty obligation because Dr. Ordover's proposed marketplace benchmarks exclude the value of non-music services, and Dr. Noll's proffered benchmarks attempt to account for the fact that roughly half of Sirius XM's service is non-music. SX PFF ¶¶ 842-846. The Judges agree 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.” Sirius XM RFF ¶ 167. The Judges stress, however, that the exclusion is available only to the extent that the channels, programming, products and/or other services are offered for a separate charge.

46.  SoundExchange conditions the availability of its approach on the presumption that the music channels on the Internet service remain representative of the music channels offered on the SDARS service.

47.  SoundExchange requests that, if its approach is adopted, Sirius XM be required to notify SoundExchange monthly of each copyright owner from which Sirius XM claims to have a direct license and each sound recording Sirius XM claims to be excludable. SoundExchange would then be permitted to disclose this information to confirm whether the direct license exists and the claimed sound recordings are properly excludable.

48.  In doing so, the Judges also accept SoundExchange's request that the Direct License Share deduction only be available if the music channels available on Sirius XM's Internet webcast service remain representative of the music channels offered on the SDARS services.

49.  This language is the same that Sirius XM proposes be applicable to directly licensed sound recordings.

50.  Sirius XM proposes to amend current § 382.13(c) in a manner that would ensure that Sirius XM would not have to report either actual total performances of sound recordings or Aggregate Tuning Hours as required under the applicable notice and recordkeeping requirements in Part 370. Sirius XM PFF ¶ 413; Proposed Rates and Terms of Sirius XM Radio Inc., at 5 (Sept. 26, 2012). SoundExchange's proposed reply findings do not address this suggested change. The Judges decline to adopt this proposal because it appears for the first time in Sirius XM's Proposed Findings of Fact “without any citation to the record or any substantive explanation as to why such a change is needed or what benefits would result from its adoption.” Webcasting III, 76 FR at 13043 (Mar. 9, 2011).

51.  76 FR at 13045.

52.  SoundExchange also proposes a number of minor technical and conforming changes, which, it represents, are unopposed. SX PFF ¶¶ 902-903. Because these proposed changes promote efficiency, the Judges adopt them.

53.  I do concur with the decision to maintain the current definition of Gross Revenues for the PSS for the upcoming licensing term.

54.  The PSS are Music Choice and Muzak. Muzak's PSS service is, apparently, only a small part of its business, and it did not participate in this proceeding. Digital Music Express, Inc., which was a PSS in PSS-I, ceased operation in 2000.

55.  The current regulation defining Gross Revenues for PSS is set forth in 37 CFR 382.2(e).

56.  The Judges have stated a decided preference for per-performance royalty rates for statutory licenses, rather than rates as a percentage of revenue, because that metric most unambiguously relates the fee to the value of the right licensed. 73 FR at 4087. We adopted percentage-of-revenue royalty rates in SDARS-I, however, because of intractable problems associated with measuring usage and listenership to performances of sound recordings. Id. at 4088. These problems continue to exist with respect to the PSS and SDARS, and the parties agree to a percentage-of-revenue royalty fee for both Section 114 licenses. Given their agreement, and lack of evidence as to an alternative, I adopt that metric.

57.  I note that these percentage rates are quite similar to the maximum rates proposed by the services and record company copyright owners in PSS-I. See PSS-I, 63 FR 25394, 25395 (May 8, 1998)(2.0% by the services and 41.5% by the record companies).

58.  Curiously, SoundExchange does not argue for expansion of the Gross Revenues definition for PSS, a definition which has existed since the first royalty term for the PSS digital performance license. If SoundExchange's broadened revenue definition for SDARS were acceptable, it would result in the two types of services licensed under Section 114(f)(1) calculating their royalties against radically different revenue bases.

59.  In SDARS-I, the Judges expressly recognized an exclusion from Gross Revenues for so-called non-music services, characterized as “channels, programming, products and/or other services offered for a separate charge where such channels use only incidental performances of sound recordings.” SDARS-I, 73 FR 4102 (citing 37 CFR 382.11, definition of Gross Revenues). The Judges did this because the exclusion “unambiguously relat[ed] the fee to the value of the sound recording performance rights at issue * * *.” Id. at 4088. SoundExchange argues that if the exclusion is allowed to continue, it will amount to a double deduction from Sirius XM's royalty obligation because Dr. Ordover's marketplace benchmarks exclude the value of non-music services, and Dr. Noll adjusted his proferred benchmarks to account for the fact that roughly half of Sirius XM's service is non-music. SX PFF ¶¶ 842-846. I agree 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.” Sirius XM RFF ¶ 167. The exemption, however, should be available only to the extent that the channels, programming, products and/or other services are offered for a separate charge.

60.  SoundExchange conditions the availability of its approach on the presumption that the music channels on the Internet service remain representative of the music channels offered on the SDARS service.

61.  This is the same language that Sirius XM proposes also be applicable to directly licensed sound recordings.

62.  He also cites a 2008 survey by OTX, a 2012 survey by Experian Simmons, a 2004 Arbitron survey, a 2011 Ipsos OTX MediaCT survey, and a 2006 Sony BMG MusicLab survey.

63.  Dr. Crawford also concludes that the marketplace for musical works royalties might be greater than the sound recording marketplace because the performing rights society loses less than a record company in the absence of an agreement. Crawford Corrected WDT at 18, 29, PSS Trial Ex. 4.

64.  Under this model, a firm's cost of capital is based on the expected return to induce investment. Crawford Corrected WDT at ¶ 167, PSS Trial Ex. 4.

65.  Music Choice's criticisms of the Harrison chart—that it omits synchronization and master use licenses, encompasses wholesale payments rather than specific rates, and involves some agreements that convey additional rights—do not detract from the conclusion that overall the royalty fees paid for sound recordings are typically significantly higher than those paid for musical works.

66.  Much was made at trial and in closing arguments regarding Dr. Crawford's supposed use of audited financial data and Dr. Ford's use of unaudited financial data in an effort to examine costs and revenues of the PSS service vis-à-vis Music Choice's other non-statutory offerings. I see no superiority to either data set, as both contain their own difficulties.

67.  It would be surprising, if not improbable, that Music Choice would be able to operate a PSS service for over 15 years with a statutory royalty between 6.5% and 7.5%, with the considerable losses that it claims, and nonetheless continue to operate, let alone intend to expand its current operation.

68.  I note that DMX's exit from the PSS market in 2000 offers an opportunity to examine how the departure of a PSS impacts consumer choices and their consumption of music, but no such analyses were presented in this proceeding.

69.  Dr. Noll also examined agreements involving the music services Slacker and Turntable.

70.  Dr. Noll also found similar splits in [REDACTED] agreements. Id. at Tables 2.2-2.2d and Appendices I-L.

71.  The “greater of” metric is an amount per play, an amount per-subscriber, and a percentage of the service's revenues. 6/14/12 Tr. 2261:7-2262:4 (Ordover).

72.  Dr. Ordover's mathematical calculation is as follows: He took the $12.95 Sirius XM subscription price, and then multiplied that by 50% to obtain the music portion of the subscription price of $6.475. He then multiplied the music-only satellite radio subscription price by 60% to 65% (his effective percentage-of-royalty derived from the interactive subscription service agreements) to obtain the music royalty of $3.88 to $4.21. Finally, he divided those numbers into the Sirius XM subscription price for the Select programming package to obtain 30% to 32.5%. 8/16/12 Tr. at 3794:13-3795:9 (Salinger).

73.  The current price for this service is $14.49. Ordover Third Corrected/Amended WDT at 31 n.33, SX Trial Ex. 74.

74.  Dr. Ordover did not provide a weighted average of the non-interactive service prices because he concluded that he did not have reliable data, nor did he include, at my invitation, ad-supported non-interactive services in his calculation, deciding that such services would add undue complexity to his methodology. Ordover Amended WRT at 33, SX Trial Ex. 218.

75.  This was calculated by multiplying the interactivity ratio of .4865 ($4.86/$9.99) to the average per-subscriber royalty payment of $5.95, yielding an equivalent satellite radio payment of $2.89. The $2.89 per-subscriber rate was then divided by the $12.95 monthly charge for the Sirius XM Select satellite radio package, resulting in the percentage-of-revenue rate of 22.32%.

76.  I note, further, that the works licensed under the Direct Licenses represent no more than 2%-4% of the total number of works performed by Sirius XM. Ordover Amended WRT at 4-5, SX Trial Ex. 218; 6/6/12 Tr. 308:3-5 (Noll).

77.  Dr. Noll's citation to Direct Licensors' catalogues containing Broadway recordings, three former hit singles, and the recordings of George Carlin, as confirmation of the popularity of the works of the Direct Licensors overall, is not persuasive.

78.  I recognize that direct payment to the Direct Licensors does not relieve them of their royalty obligations to their artists and performers; however, receipt of 100% of the royalties upfront is clearly attractive to certain record labels and was a selling point in negotiations with independent record labels. Powers WDT at 4-5, SX Trial Ex. 243.

79.  Dr. Noll also offers his demand diversion theory as an explanation as to why SoundExchange allegedly attempted to interfere with Sirius XM's Direct License Initiative.

80.  The implicit monthly price is applied to the effective percentage-of-revenue rate of [REDACTED] from the Last.fm agreements that serve as the numerator in Dr. Noll's calculation.

81.  Likewise, Sirius XM has failed to demonstrate that it could successfully substitute away to other providers of music. If that were the case, Sirius XM could have operated its business under the Direct Licenses, for example, and avoided participation in this proceeding altogether.

82.  Dr. Noll identifies the non-interactive music services offered by Pandora, whom he categorizes as the “big elephant in the room,” as highly comparable to the satellite radio service of Sirius XM. 6/5/12 Tr. 286:21-287:7 (Noll). While his comparison is to the compatible features of Pandora, the parties have interjected and argued the royalty rates paid by Pandora for its music services. I am expressly not considering the rates, terms or conditions of Pandora's royalty payments in relation to the rates in this proceeding, for to do so would violate the terms of the Webcaster Settlement Act of 2009. See 17 U.S.C. 114(f)(5).

83.  Dr. Noll also considered agreements involving Slacker and Turntable, but only used the Last.fm agreements in his analysis. As Sirius XM acknowledges, the Slacker and Turntable services are more interactive than Last.fm, thereby weakening their comparability. Sirius XM RFF ¶ 64.

84.  Sirius XM makes much of the fact that rates obtained by the major record labels have dropped almost 20% since SDARS-I and argues that this logically must mean that music is worth less than in the prior proceeding. Sirius XM PFF ¶ 339. SoundExchange counters that the reason for the 20% drop is the decline in retail prices for interactive services, which SoundExchange concludes is an indication that consumers value interactivity less than before. SX RFF ¶ 145. Neither side provided empirical evidence to prove their point, and logic does not dictate that music is of any less, or more, value as a result of this occurrence.

85.  The five non-interactive services selected by Dr. Ordover listed one retail price for two services, two retail prices for one service, and three retail prices for two services. Ordover Third Corrected/Amended WDT at ¶ 54, Table 5, SX Trial Ex. 74. The differing prices reflect differing duration commitments for subscribers.

86.  While I am adopting these adjustments to Dr. Ordover's second alternative benchmark, I underscore that I am not adopting Dr. Noll's recommended use of the Last.fm non-interactive percentage rate (26.1%) for the same reasons that a five-year old agreement with two major record labels did not make for a useful benchmark.

87.  This is calculated by multiplying the interactivity ratio of .3153 ($3.15/$9.99) to the average per-subscriber royalty payment of $5.95, yielding an equivalent satellite radio payment of $1.87. The $1.87 per-subscriber rate is then divided by Sirius XM's ARPU ($11.49), resulting in the percentage-of-revenue rate of 16.2%.

88.  SoundExchange, citing Dr. Ordover's testimony, argues that the policy considerations of the first three factors are subsumed in the marketplace benchmarks it has proffered. SX PFF ¶¶ 502-507. Since I do not accept the benchmarks of either side as determinative of the rate to which Section 801(b) is applied, other than their ability to define the range of reasonable rates, SoundExchange's argument is inapposite.

89.  Indeed, it is difficult to imagine royalty rates, other than perhaps those approaching zero, that might make more than a dent in the recovery of billions of dollars of cumulative losses.

90.  The merger of the two companies did not occur until the following year. 6/7/12 Tr. 640:15 (Frear).

91.  The record in this proceeding is replete with public statements and assertions by the executive officers of Sirius XM that the company is and will be highly successful and profitable, both in the short term and the long term. See, e.g,, Lys Corrected WDT at 8, 10-11 (quoting Mr. Karmazin from November 2011: “[W]e believe we have many, many years of subscriber growth ahead of us.”), SX Trial Ex. 80; 8/13/12 Tr. 3179:7-10 (Frear)(Sirius XM revenue not just growing, but accelerating); Lys WRT at 31 & SX 238-RP (Mr. Karmazin in an April 2012 interview with Forbes magazine: “[W]e're a very profitable, successful company. If we want a performer, we can afford to pay more than anybody else because we're making more.”), SX Ex. 226-RP (Mr. Karmazin: “Given the predictable nature of our business, we would prefer to take advantage of a prudent level of leverage, which should mean higher returns to our equity holders over time.”), SX Trial Ex. 240.

92.  I find Professor Stowell's criticisms of equity analysts' forecasts flawed and unpersuasive. He ignores substantial, published research as to the improved accuracy of projections, particularly within recent years, as well as changes implemented by the Securities and Exchange Commission to eliminate analyst bias. Lys WRT at 11-12, SX Trial Ex. 240. He also ignores recent data criticizing the performance of equity analysts that follow Sirius XM, confining his analysis to only forecasts made prior to the Sirius and XM merger. Id. at 14.

93.  Indeed, Sirius XM originally considered calling the fee the “Copyright Royalty Board Fee” instead of the Music Royalty Fee. Lys WRT at 33 n.142, SX Trial Ex. 240.

94.  SoundExchange and Sirius XM disagree as to what percentage of licensing costs are passed through to subscribers in the Music Royalty Fee. SoundExchange contends 100%, Lys WRT at 32 & SX Ex. 240-RR, SX Trial Ex. 240, while Sirius XM contends 53%. Frear Revised WRT at 16, SXM Reb. Trial Ex. 1. A substantial portion of licensing fees is passed on to subscribers in either case, ameliorating the possibility of short-term negative effects of increased Section 114 fees.

95.  The odds of another such crisis occurring during the 2013-2017 licensing period are low since that type of crisis has happened only twice in the past 80 years. 8/20/12 Tr. 4046:5-9 (Lys).