26316 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

LIBRARY OF CONGRESS per-performance royalty fees for the look to is a hypothetical marketplace, transmissions it makes on that channel free of the influence of compulsory, Copyright Royalty Board or station in excess of 159,140 ATH at statutory licenses. Web II, 72 FR 24084, the rate of $0.0017 per performance. The 24087 (May 1, 2007). The Judges ‘‘shall 37 CFR Part 380 rates for transmissions over 159,140 base their decision on economic, [Docket No. 14–CRB–0001–WR (2016–2020)] ATH per month for the period 2017 competitive[,] and programming through 2020 shall be adjusted to reflect information presented by the parties Determination of Royalty Rates and the increases or decreases, if any, in the . . . .’’ 17 U.S.C. 114(f)(2)(B) and Terms for Ephemeral Recording and general price level, as measured by the 112(e)(4) (emphasis added). Within Webcasting Digital Performance of Consumer Price Index applicable to that these categories, the Judges’ Sound Recordings (Web IV) rate year, as set forth in the regulations determination shall account for (1) adopted by this determination. whether the Internet service substitutes AGENCY: Copyright Royalty Board, The Judges also determine herein for or promotes the copyright owner’s Library of Congress. details relating to the rates for each other streams of revenue from the sound ACTION: Final rule and order. category of webcasting service, such as recording, and (2) the relative roles and minimum fee and administrative terms. contributions of the copyright owner SUMMARY: The Copyright Royalty Judges The regulatory language codifying the and the service, including creative, announce their determination of rates rates and terms of the Judges’ technological, and financial and terms for two statutory licenses determination 1 are set out below this contributions, and risk assumption. Id. (permitting certain digital performances Supplementary Information section. The Judges may consider rates and of sound recordings and the making of terms of comparable services and I. Background ephemeral recordings) for the period comparable circumstances under beginning January 1, 2016, and ending A. Purpose of the Proceeding voluntary, negotiated license on December 31, 2020. The licenses at issue in the captioned agreements. Id. The rates and terms DATES: Effective Date: This rule is proceeding, viz., licenses for established by the Judges ‘‘shall effective on May 2, 2016. commercial and noncommercial distinguish’’ among the types of services Applicability dates: These rates and noninteractive webcasting, are and ‘‘shall include’’ a minimum fee for terms are applicable to the period compulsory. Title 17, United States each type of service. Id. (emphasis January 1, 2016, through December 31, Code (Copyright Act or Act), establishes added). 2020. exclusive rights reserved to copyright B. Procedural Posture FOR FURTHER INFORMATION CONTACT: owners, including the right to ‘‘perform Following the timeline prescribed by LaKeshia Keys, Program Specialist, at the copyrighted work publicly by means the Act, the Judges published notice of 202–707–7658 or [email protected]. of a digital audio transmission.’’ See 17 commencement of this proceeding in SUPPLEMENTARY INFORMATION: The U.S.C. 106(6). The digital performance the Federal Register.2 79 FR 412 (Jan. 3, Copyright Royalty Judges (Judges) right is limited, however, by § 114 of the 2014). Twenty-nine parties in interest hereby issue their written determination Act, which grants a statutory license for filed petitions to participate in the of royalty rates and terms to apply from nonexempt noninteractive Internet proceeding.3 Ten of those petitioners January 1, 2016, through December 31, transmissions of protected works. 17 subsequently withdrew from the 2020, to digital performance of sound U.S.C. 114(d). Eligible webcasters are proceeding, the Judges rejected the recordings over the Internet by entitled to perform sound recordings petitions of three petitioners because the nonexempt, noninteractive transmission without an individual license from the services and to the making of ephemeral copyright owner, provided they pay the 2 Contemporaneously, the Judges commenced a recordings to facilitate those statutory royalty rates for the proceeding to establish rates and terms for performances. performance of the sound recordings ephemeral recording and digital performance of The rate for commercial subscription and for the ephemeral copy of the sound sound recordings by ‘‘New Subscription Services’’ (NSS). See 79 FR 410 (Jan. 3, 2014). The NSS at services in 2016 is $0.0022 per recording necessary to transmit it. 17 issue in that companion proceeding were limited to performance. The rate for commercial U.S.C. 114(f) and 112(e). Licensee NSS transmitting to residential subscribers through nonsubscription services in 2016 is webcasters pay the royalties to a a cable television provider. See 37 CFR 383.2(h). $0.0017 per performance. The rates for Collective, which distributes the funds That proceeding was resolved by negotiated agreement and the Judges published rates and terms the period 2017 through 2020 for both to copyright owners. The statutory rates for new subscription licensees at 80 FR 36927 (Jun. subscription and nonsubscription and terms apply for a period of five 29, 2015). Settlement of the cable NSS did not have services shall be adjusted to reflect the years. any effect on the Internet subscription services at increases or decreases, if any, in the The Act requires that the Judges issue in this proceeding. ‘‘shall establish rates and terms that 3 The 29 parties that filed Petitions to Participate general price level, as measured by the were: 8tracks, Inc.; AccuRadio, LLC; Amazon.com, Consumer Price Index applicable to that most clearly represent the rates and Inc.; Apple Inc.; Beats Music, LLC; Clear Channel rate year, as set forth in the regulations terms that would have been negotiated (nka iHeartMedia, Inc.); CMN, Inc.; College adopted by this determination. in the marketplace between a willing Broadcasters, Inc. (CBI); CustomChannels.net, LLC; buyer and a willing seller.’’ 17 U.S.C. Digital Media Association (DiMA); Digitally The rates for noncommercial Imported, Inc.; Educational Media Foundation; webcasters are: $500 annually for each 114(f)(2)(B). The marketplace the Judges Feed Media, Inc.; Geo Music Group; Harvard Radio station or channel for all webcast Broadcasting Inc. (WHRB); idobi Network; transmissions totaling not more than 1 The Judges proposed to the parties a Intercollegiate Broadcasting System, Inc. (IBS); reorganization of the regulations. Only one party’s Music Reports Inc.; National Association of 159,140 Aggregate Tuning Hours (ATH) (Pandora’s) proposed regulations followed the Broadcasters (NAB); National Music Publishers in a month, for each year in the rate proposed new format. The other parties submitted Association (NMPA); National Public Radio (NPR); term. In addition, if, in any month, a proposed new subparts for each type of entity. One National Religious Broadcasters Noncommercial noncommercial webcaster makes total party (SoundExchange) specifically opposed the Music License Committee (NRBNMLC); Pandora reorganization. The Judges find that reducing the Media Inc.; Rhapsody International, Inc.; Sirius XM transmissions in excess of 159,140 ATH amount of repetition in the regulations is not Radio Inc.; SomaFM.com LLC; SoundExchange, Inc. on any individual channel or station, prejudicial to SoundExchange, and in the interests (SX or SoundExchange); Spotify USA Inc.; and the noncommercial webcaster shall pay of plain language have used the new format. Triton Digital, Inc.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00002 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26317

Judges determined they lacked the 2. The Current Proceeding To motion. Having reviewed the motions, requisite substantial interest in the Adjudicate Rates and Terms written arguments, and responses, the proceeding, and the Judges dismissed The Act provides that the Judges shall Judges denied the motions for rehearing. the Petition to Participate of another make their determinations ‘‘on the basis The Judges determined that neither of party due to a procedural default.4 of a written record, prior determinations the motions presented the exceptional case required for rehearing or 1. Negotiated Settlements and interpretations of the Copyright Royalty Tribunal, Librarian of Congress reconsideration. In other words, neither a. Educational Webcasters . . .’’ and their own prior SoundExchange nor GEO established determinations to the extent those that the Determination (1) is not The Judges published notice of the determinations are ‘‘not inconsistent supported by the evidence, (2) is CBI-SoundExchange settlement in with a decision of the Register of erroneous, (3) is contrary to legal November 2014.5 The Judges received Copyrights . . .’’ 17 U.S.C. 803(a). requirements, or (4) requires the approximately 60 comments in response 10 Pursuant to 17 U.S.C. 803(b), the Judges introduction of new evidence. See 17 to the Notice. The Judges considered the conduct a hearing to create that ‘‘written U.S.C. 803(c)(2)(A); 37 CFR 353.1 and comments, some of which supported record,’’ in order to issue their 353.2. The motions did not meet the and others of which opposed the determination as required by 17 U.S.C. required standards set by statute, by proposed settlement, and concluded 801(b)(1) and 803(1). regulation, or by case law. Nevertheless, that the CBI-SoundExchange agreement To that end, non-settling parties as discussed in the order denying provides a reasonable basis to adopt its appeared before the Judges for a SoundExchange’s motion for rehearing, proposed rates and terms. On September determination hearing. At the hearing, the Judges amended certain of the 28, 2015, the Judges published amended SoundExchange, Inc. (SoundExchange), royalty terms regulations to enhance regulations substantially in conformity a member organization comprised of clarity. The Judges incorporate the 6 with the proposal. copyright owners and performing regulatory clarifications, making this Determination final and subject to legal b. Public Broadcasters artists, and the designated Collective in this proceeding, and Mr. George review by the Register of Copyrights. The NPR–CPB settlement with Johnson, dba GEO Music, represented II. Context of the Current Proceeding SoundExchange proposed creation of a the interests of licensors. Seven new Subpart D to part 380 of the licensees participated in the hearing.8 A. Prior Rate Determinations Regulations entitled Certain The hearing commenced on April 27, Congress created the exclusive sound Transmissions by Public Broadcasting 2015, and concluded on June 3, 2015. recordings digital performance Entities. IBS was the only commenting The parties submitted proposed findings copyright in 1995. See Digital party. IBS made procedural and and conclusions (and responses thereto) Performance Right in Sound Recordings substantive objections to the settlement. in writing, prior to their closing Act of 1995, Public Law No. 104–39, Notwithstanding, the Judges concluded arguments on July 21, 2015. During the 109 Stat. 336 (Nov. 1, 1995). At the same that, as the proposed settlement would hearing, the Judges heard oral testimony time, Congress limited that performance bind only the ‘‘Covered Entities,’’ i.e., from 47 witnesses, some of them for right by granting noninteractive NPR, American Public Media, Public both direct case and rebuttal testimony. subscription services a statutory license Radio International, and Public Radio The witnesses included 16 qualified to perform sound recordings by digital Exchange, and up to 530 Originating experts. The Judges admitted 660 audio transmission. In 1998, Congress Public Radio Stations as named by CPB, exhibits into evidence, consisting of created the ephemeral recording license adoption of the settlement would not over 12,000 pages of documents, and and further defined and limited the preclude the Judges’ separate considered numerous illustrative and statutory license for digital performance consideration of the concerns of IBS, demonstrative materials that focused on of sound recordings. See Digital which is not one of the ‘‘Covered aspects of the admitted evidence and Millennium Copyright Act, Public Law Entities’’ subject to the new Subpart D. the permitted oral testimony. On October 2, 2015, the Judges On December 16, 2015, the Judges 105–304, 112 Stat. 2860 (Oct. 28, 1998) published the settlement, substantially issued their Determination of Rates and (DMCA). as proposed, as a final regulation.7 Terms. Pursuant to 17 U.S.C. 803(c)(2) 1. Web I and 37 CFR part 353, SoundExchange 4 The ten parties that withdrew their Petitions to and George Johnson dba GEO Music The Copyright Office commenced the Participate were: 8tracks, Inc.; Amazon.com, Inc.; Group (GEO) filed motions for first webcasting rate determination in CMN, Inc.; CustomChannels.net, LLC; Digitally rehearing. The Judges sought responses November 1998. The resulting rates, Imported, Inc.; Feed Media, Inc.; idobi Network; published in July 2002, covered a rate Rhapsody International, Inc.; SomaFM.com LLC; to the issues raised in the and Spotify USA Inc. The three parties whose SoundExchange motion, but did not period from October 1998 through Petitions to Participate were dismissed for lacking solicit written responses to the GEO December 2002.11 Interested parties a substantial interest in the proceeding were: Music Music motion.9 NAB, Pandora, and negotiated rates and terms for 2003– Reports Inc., NMPA, and Triton Digital. The 2004, including for the first time radio Petition to Participate of AccuRadio was dismissed iHeart filed written arguments by the Judges due to a procedural default. Although responsive to the SoundExchange broadcasters with Internet simulcast they did not formally withdraw from the proceeding, Apple, Beats, and DiMA did not file 8 Harvard Radio Broadcasting, Inc. (WHRB), 10 Order Denying in Part SoundExchange’s Written Direct Statements and did not participate Intercollegiate Broadcasting System, Inc., Motion for Rehearing and Granting in Part in the hearing. Educational Media Foundation iHeartMedia, Inc., National Association of Requested Revisions to Certain Regulatory joined with NAB and appeared by and through Broadcasters (also representing the interests of Provisions (Feb. 10, 2016) and Order Denying NAB and its counsel. Educational Media Foundation), National Religious George Johnson’s Motion for Rehearing (Feb. 10, 5 79 FR 65609 (Nov. 5, 2014). Broadcasters Noncommercial Music Licensing 2016). 6 80 FR 58201 (Sept. 28, 2015). Committee, Pandora Media, Inc., and Sirius XM 11 See 67 FR 45240 (Jul. 8, 2002); see also 67 FR 7 80 FR 59588 (Oct. 2, 2015). In publishing both Radio, Inc. 78510 (allowing non-precedential, negotiated negotiated settlements, the Judges postponed the 9 Order Permitting Written Response(s) to modification of 1998–2002 rates and terms for designation of a Collective until issuance of the SoundExchange Motion for Rehearing (Revised) ‘‘small webcasters’’ under the Small Webcaster current determination. (Jan. 6, 2016). Settlement Act of 2002).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00003 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26318 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

service.12 The published webcasting noncommercial webcasters, the (CBI), Live365, Inc. (Live365), and rate determination confirmed that the minimum fee is the only royalty fee due, IBS.20 willing buyer/willing seller standard in unless the webcaster exceeds CBI’s participation was limited to its the Act is the determining standard. The established usage limits. defense of a proposed settlement it Librarian of Congress (Librarian) Intercollegiate Broadcasting System, negotiated with SoundExchange. Under determined that rate-setters must Inc. (IBS) appealed the amount of the the CBI/SoundExchange agreement, the consider the promotion/substitution and minimum fee as it applied to Judges were asked to adopt regulations relative contribution factors, although noncommercial webcasters. The U.S. that established a subcategory of they must not consider those factors Court of Appeals for the D.C. Circuit noncommercial webcasters, viz., determinative, nor are they to use those remanded the issue for further fact- noncommercial educational webcasters additional factors to adjust a rate finding.15 The Judges received further (NEWs). The Judges did so and derived from the willing buyer/willing evidence and ruled on remand to keep established the minimum fee for the seller analysis. See 67 FR 45240, 45244 the minimum fee at $500 for all educational category at the same level as (July 8, 2002). This conclusion is part of licensees. See 75 FR 56873, 56874 (Sept. every other category of webcasting the rate-setting precedent that instructs 17, 2010). IBS again appealed to the D.C. service, i.e., $500 per year for each the Judges in the current proceeding. Circuit, challenging the application of station or channel, applicable to the flat 2. Web II Determination and Appeals the minimum fee to noncommercial fee for usage. See Digital Performance and Webcaster Settlement Acts educational webcasters. The court Right in Sound Recordings and stayed the second Web II appeal Ephemeral Recordings, 76 FR 13026 In November 2004, Congress passed pending its resolution of a (March 9, 2011) (Web III). Recognizing the Copyright Royalty and Distribution constitutional question raised by IBS in the operational constraints on Reform Act of 2004 (Reform Act), which relation to the Judges’ Web III educational webcasters, the Judges also became effective in May 2005. The determination. Ultimately, the court adopted less burdensome usage Reform Act established the Copyright again remanded Web II to the Judges.16 reporting standards for the category. Royalty Judges as the institutional The Judges conducted a de novo review Educational webcasters not exceeding successor to the arbitration panel of the record and published their 159,140 Aggregate Tuning Hours (ATH) program managed by the Copyright determination on the second remand in of webcasting per month could opt for Office. The new statute continued the 2014. See 79 FR 64669 (Oct. 31, 2014). sample reporting in lieu of census extant 2004 rates through 2005 to enable IBS moved to drop its third appeal of reporting of each sound recording the newly created Copyright Royalty Web II and the court dismissed it on performance. Educational webcasters Judges program to initiate rate September 11, 2015.17 not exceeding 55,000 ATH could forego proceedings. The new statute also After the Library published the Web II reporting usage at all by paying a $100 expanded the rate period to five years.13 The Judges published the determination, Congress passed the proxy fee to defray the cost to determination from their first Webcaster Settlement Act of 2008 (2008 SoundExchange of developing proxy webcasting rate proceeding, covering WSA) and the Webcaster Settlement Act usage data. the period 2006 to 2010, on May 1, 2007 of 2009 (2009 WSA). These acts enabled For the commercial webcaster rates, (Web II).14 In Web II, the Judges webcasters to renegotiate rates and SoundExchange and Live365 each differentiated the rate structure for terms for a portion of the Web II rate proposed a per-performance rate commercial and noncommercial period and set rates for the succeeding structure. Live365 attempted to reach a webcasters. They set commercial rate period (2011–2015). Entities per-performance rate by way of a webcasters’ rates using a per- accounting for 95% of the webcasting revenue analysis, factoring in the performance structure and set royalties paid to SoundExchange webcasting services’ costs and a noncommercial webcasters’ rates as a negotiated settlements under the 2008 presumed 20% profit, and applying the 18 flat fee up to a certain usage level, after WSA and the 2009 WSA. remainder of revenue to royalties. SoundExchange approached the which the commercial rates would 3. Web III Determination and Appeals apply. See 72 FR 24084, 24096, 24097– calculation by analyzing comparable 98. In accordance with the statute, the On January 5, 2009, the Judges market ‘‘benchmark’’ agreements, with Judges established a minimum fee of commenced a proceeding to establish adjustments as necessary to account for $500 for each channel or station in rates and terms for webcasting for the differences in the services. either category. The Judges did not period January 1, 2011, through SoundExchange relied on interactive 19 differentiate the minimum fee, as they December 31, 2015 (Web III). Many services rate agreements. based it upon the cost to interested webcasters had recently The Web III Judges rejected the SoundExchange, the designated reached agreements with Live365 attempt to base rates on a Collective, to administer the license. For SoundExchange pursuant to the WSAs service’s ability to pay. Instead, the and did not participate in the Web III Judges derived the commercial 12 See 68 FR 35008 (Jun. 11, 2003) proceeding. Only three licensees did webcasting rate in Web III from a review (noncommercial webcasters’ rates, effective 1998– participate: College Broadcasters, Inc. of market benchmarks presented by 2004); 37 FR 5693 (Feb. 6, 2004) (subscription and SoundExchange. SoundExchange nonsubscription services’ and simulcasters’ rates, 15 effective 2003–04, and new subscription services’ Intercollegiate Broad. Sys., Inc. v. Copyright provided only interactive services’ rates, effective 1998–2004). Royalty Board, 574 F.3d 748, 771 (D.C. Cir. 2009). licenses as benchmarks. The Judges 16 13 Public Law 108–419, 118 Stat. 2341. In 2004, Intercollegiate Broadcasting Sys., Inc., v. adjusted those benchmarks to account Copyright Royalty Board, No. 10–1314 (D.C. Cir. the Copyright Office initiated a proceeding to adjust for significant functional differences rates and terms for the Section 114 and 112 licenses Sept. 30, 2013) (order granting joint motion for for 2005–2006 under the CARP system. Congress vacatur and remand). between interactive services and terminated this proceeding, however, and directed 17 Intercollegiate Broad. Sys., Inc. v. Copyright that the rates and terms in effect on December 31, Royalty Board, No. 14–1262 (D.C. Cir. Sept. 11, 20 As part of the Web III determination, the Judges 2004, remain in effect at least for 2005. See 70 FR 2015) (order granting joint motion to dismiss confirmed their adoption of agreed rates and terms 7970 n.2 (Feb. 16, 2005) and 70 FR 6736 (Feb. 8, appeal). for commercial broadcasters (simulcasters) 2005). 18 79 FR 23102 n.5 (Apr. 25, 2014). proposed in a settlement agreement between 14 72 FR 24084. 19 74 FR 318 (Jan. 5, 2009). SoundExchange and the NAB. 76 FR at 13027.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00004 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26319

noninteractive services subject to the to pay higher royalty rates to Majors context and without adequate input of statutory rates and terms. than to Indies.25 the parties would be capricious. IBS appealed the Web III The Act provides that the Judges must Moreover, reopening the proceeding at determination.21 The D.C. Circuit agreed differentiate rates based upon this juncture, long after the closing of with the IBS argument that the differences in the webcasting services, the record pursuant to 37 CFR 351.12, Librarian’s appointment of the Judges but is less clear on whether the Judges for further evidence and argument on under the Reform Act violated the may also establish differential rates this issue would be improper. The Appointments Clause of the based on differences among copyright Judges, therefore, do not resolve the Constitution. The D.C. Circuit severed owners as revealed by the evidence. To legal issue they referred to the Register that portion of the Reform Act that gain clarity on the latter issue, the and do not set rates in this proceeding limited the Librarian’s ability to remove Judges referred to the Register of that distinguish among classes of Judges, remanding the substantive Copyrights the novel question whether copyright owners. merits of the determination for decision the Copyright Act permits the Judges to 2. Commercial Webcasters vs. by a validly appointed panel of Judges. differentiate based on types of licensors. Noncommercial Webcasters The Librarian appointed the current After careful review, the Register Judges and they issued a determination concluded that the Judges’ question In accordance with the statutory on remand in April 2014.22 In their Web ‘‘d[id] not meet the statutory criteria for direction to ‘‘distinguish among the III Remand, the Judges relied upon the referral,’’ and declined to answer it. different types of eligible rates set forth in the WSA agreements Memorandum Opinion on Novel nonsubscription transmission services,’’ between SoundExchange and the NAB Question of Law at 7 (Nov. 24, 2015) 17 U.S.C. 114(f)(2)(A), the Judges (and and between SoundExchange and Sirius (Register’s Opinion). the Librarian of Congress before them) XM, and, to a lesser extent, Citing the fact that no party in the have recognized noncommercial SoundExchange’s benchmark analysis of proceeding had proposed a rate webcasters as a separate rate category various interactive agreements. Id. structure that differentiated among from commercial webcasters in prior IBS appealed the Judges’ remand licensors, the Register found that ‘‘such proceedings.26 The Judges deemed determination on May 2, 2014. The D.C. a structure was not understood to be a different (and lower) rates for Circuit affirmed the determination on subject of litigation.’’ Id. at 8–9. noncommercial webcasters to be August 11, 2015.23 Consequently, the Register found that appropriate because ‘‘certain ‘noncommercial’ webcasters may B. Web IV the issue was not ‘‘presented’’ in the proceeding as required by the ‘‘novel constitute a distinct segment of the When the Judges commenced the question’’ provision in 17 U.S.C. noninteractive webcasting market that present proceeding (Web IV) in January 802(f)(1)(B). Id. at 7. The Register’s in a willing buyer/willing seller 2014, they invited all potentially Opinion appears to be premised, in part, hypothetical marketplace would affected entities to consider in the on an interpretation of the D.C. Circuit’s produce different, lower rates than we presentation of their respective cases: decisions in Settling Devotional have determined . . . for Commercial (1) The pros and cons of revenue-based Claimants v. Copyright Royalty Bd., 797 Webcasters.’’ Web II Original rates, (2) the existence or propriety of F.3d 1106 (D.C. Cir. 2015), and Determination, 72 FR at 24097. price differentiation in a market in Intercollegiate Broad. Sys. v. Copyright The record in the instant proceeding which the product (digital sound Royalty Bd., 574 F.3d 748 (D.C. Cir. demonstrates some of the reasons why, recordings) can be reproduced at a near- 2009). See Register’s Opinion at 9. The in a hypothetical marketplace, a zero marginal cost, and (3) economic Register appears to interpret those cases noncommercial webcaster’s willingness variations among buyers and sellers in to pay for sound recordings would be as barring the Judges from relying on the relevant market. 24 The parties lower than a commercial webcaster’s theories ‘‘first presented in the Judges’ addressed many of these issues in their willingness to pay. For example, a determination and not advanced by any filings (including their rate proposals) noncommercial religious broadcaster and in testimony provided during the participant.’’ Id. Section 802(f)(1)(B) provides that the that streams a simulcast of its broadcasts proceeding. is prohibited under FCC regulations Register’s timely decision of a novel 27 III. Judges’ Resolution of General Issues question is binding on the Judges. from selling advertising. NRBNMLC Because the Register has declined to Ex. 7000 ¶ 18 (Emert WDT). Increased A. Rate Differentiation Internet performances are thus unlikely decide the question that the Judges to lead to increased revenue, even as 1. Majors vs. Indies referred to her in the current proceeding, however, there is no In the evidence presented during the 26 See Determination of Reasonable Rates and hearing, the Services established a decision that binds the Judges on this Terms for the Digital Performance of Sound potentially meaningful dichotomy issue. Moreover, to the extent that the Recordings and Ephemeral Recordings, 67 FR between rates they pay to Major Labels Register’s Opinion rests on an 45240, 45258–59 (July 8, 2002) (Web I); Digital interpretation of the D.C. Circuit’s Performance Right in Sound Recordings and and those they pay to independent Ephemeral Recordings, 72 FR 24084, 24097 (May 1, record companies (Indies). Put simply, application of traditional standards of 2007) (Web II Original Determination); in the marketplace, Services have agreed administrative law to particular facts, Determination of Royalty Rates for Digital that interpretation does not constitute a Performance Right in Sound Recordings and Ephemeral Recordings, 79 FR 23102, 23122 (April 21 Intercollegiate Broad. Sys., Inc. v. Copyright resolution of a ‘‘novel question 25, 2014) (Web III Remand). Royalty Bd., 684 F.3d 1332 (2012). SoundExchange concerning an interpretation of . . . 27 The NRBNMLC also highlights a number of and CBI intervened. provisions of’’ title 17 that would bind differences between broadcasters and other ‘‘pure 22 See Determination of Royalty Rates for Digital the Judges. play’’ webcasters. See, e.g., NRBNMLC PFF ¶ 33. No Performance Right in Sound Recordings and Nevertheless, the Judges acknowledge party has proposed noncommercial broadcasters as Ephemeral Recordings, 79 FR 23102 (Apr. 25, 2014) a rate category separate from other noncommercial (Web III Remand). that interpretation of the evidence out of webcasters, and the record does not provide the 23 See Intercollegiate Broad. Sys., Inc. v. Judges a sufficient basis to establish separate rates Copyright Royalty Bd., Case No. 14–1098 (Aug. 11, 25 This point is exemplified by the different for those separate categories. Consequently, the 2015). effective rates in the Pandora/Merlin Agreement differences that the NRBNMLC highlights are 24 See 79 FR 412 (Jan. 3, 2014). and the iHeart/Warner Agreement, discussed infra. irrelevant.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00005 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26320 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

they result in an increased royalty lower royalty rate in the hypothetical operate under the rates determined by burden. See 5/21/15 Tr. at 5270 market. As discussed below, based on the Judges.31 (Henes).28 the record in the current proceeding, the b. Comparable Agreements Indeed, the NRBNMLC and Judges do not believe that the NAB SoundExchange both proposed that the satisfied that burden. Therefore, the In the current proceeding, the NAB Judges adopt a different rate structure Judges do not adopt a different rate presented no benchmarks in support of for noncommercial webcasters than for structure for simulcasters than that its rate proposal, opting instead for an 32 commercial webcasters, which suggests which applies to other commercial alternative economic analysis. The to the Judges that there is continued webcasters. NAB does not, therefore, direct the support in the marketplace for a Judges to any marketplace benchmarks different rate structure for commercial a. History to demonstrate different prevailing and noncommercial webcasters. royalty rates for simulcasters than for Therefore, for all of the foregoing No prior rate determination has other webcasters. reasons, and in accordance with the treated simulcasters differently from The only agreements in the record Judges’ reasoning from Web II and Web other webcasters. In Web I, the that relate specifically to simulcasting III, the Judges adopt a separate rate Librarian, at the recommendation of the are the NAB WSA settlement agreement structure for noncommercial webcasters Register, rejected a CARP report that set and the 27 direct licenses between than the one applicable to commercial a separate rate for retransmission of iHeartMedia and independent record webcasters. radio broadcasts by a third-party labels (the iHeart/Indie Agreements). distributor, and adopted a single rate for The NAB settlement (which the NAB 3. Simulcasters vs. Other Commercial commercial webcasters. 67 FR at repudiates as a benchmark) does not Webcasters 45252.29 support the NAB proposal. The average The NAB participated in this In Web II, the Judges rejected of the settlement rates over the Web III proceeding on behalf of its member broadcasters’ arguments that rates for rate period is precisely the same as the terrestrial radio stations that simulcast simulcasting should be different from average of the rates that the Judges over-the-air broadcasts on the Internet. (and lower than) royalty rates for other determined for all other commercial 33 iHeartMedia (iHeart) also owns and commercial webcasters. webcasters in Web III. The 2015 rate operates terrestrial broadcasting stations of $0.0025 per performance is five times that simulcast, in whole or in large part, The record before us fails to persuade us the rate that the NAB proposes for the their over-the-air programming. In this that these simulcasters operate in a 2016–2020 rate period ($0.0005). proceeding, the Judges focus solely on submarket separate from and noncompetitive The Judges cannot compare the the Internet transmissions of these with other commercial webcasters. Indeed, iHeart/Indie rates directly to the NAB there is substantial evidence to the contrary broadcasters. settlement rate because they do not The NAB argues that simulcasting is in the record indicating that commercial employ a per-performance royalty rate. different from other forms of webcasters . . . and simulcasters . . . regard Instead those agreements set royalties at commercial webcasting. Given these each other as competitors in the marketplace. the record company’s pro-rata share of purported differences, the NAB Digital Performance Right in Sound [REDACTED]% of [REDACTED]. See, advocates for a separate (lower) rate for Recordings and Ephemeral Recordings, e.g., Ex. 3351 at 7–8 (Clear Channel- simulcasters than for other commercial 72 FR 24084, 24095 (May 1, 2007), aff’d RPM Entertainment License webcasters. The NAB avers that in relevant part sub nom. Intercollegiate Agreement). Without additional data simulcasting constitutes a distinct Broad. Sys. v. Copyright Royalty Bd., (e.g., iHeart’s net simulcasting revenues submarket in which buyers and sellers 571 F.3d 69 (D.C. Cir. 2009) (Web II). and the number of simulcast would be willing to agree to lower performances of recorded music), the royalty rates than their counterparts in The NAB reached a WSA settlement Judges are unable to convert the the commercial webcasting market. See with SoundExchange prior to the [REDACTED] rate into a per- NAB Proposed Rates and Terms at 2 conclusion of Web III covering the performance rate. Moreover, there is (definition of eligible transmission) remainder of the Web II rate period and insufficient evidence and economic 30 (Oct. 7, 2014). No other party’s rate all of the Web III rate period. At the analysis in the record for the Judges to proposal treats simulcasting differently request of the NAB and SoundExchange, determine whether the headline rate for from other commercial webcasting. the Judges adopted the settlement as simulcasting in the iHeart-Indie As the proponent of a rate structure statutory rates and terms binding on all agreements fully accounts for the that treats simulcasters as a separate simulcasting broadcasters. See 75 FR economic value of the licenses to the class of webcasters, the NAB bears the 16377 (April 1, 2010). Consequently, parties.34 The Judges are unable to burden of demonstrating not only that simulcasters did not participate in the simulcasting differs from other forms of Web III proceeding, in which the Judges 31 Under the NAB settlement, participating commercial webcasting, but also that it determined rates for ‘‘all other simulcasters initially paid lower per-performance commercial webcasters.’’ Although the royalty rates than those set by the Judges in Web differs in ways that would cause willing III. In later years, however, the rates increased to buyers and willing sellers to agree to a Judges did not determine separate rates levels that exceed those set by the Judges in Web for simulcasters in Web III, because the III. As a consequence, simulcasters currently pay a 28 As discussed above, SoundExchange and two Judges adopted the NAB settlement, higher royalty rate than all other commercial groups of noncommercial webcasters—CBI and simulcasting broadcasters currently pay webcasters. Since no party has asserted that simulcasters should pay a higher rate than other NPR/CPB—submitted settlement agreements different rates than webcasters that covering certain noncommercial webcasters that commercial webcasters, the Judges do not reach that establish separate, lower effective royalty rates for issue at this time. some noncommercial webcasters. The Judges 29 The Librarian also rejected arguments that 32 See discussion infra, section IV.G.2. adopted these agreements. 80 FR 58201 (Sept. 28, broadcasters who stream their own radio broadcasts 33 In both cases the average per-performance 2015); 80 FR 59588 (Oct. 2, 2015). These agreements should be treated differently from third parties who royalty rate over the 2011–2015 period is $0.00214. demonstrate that willing sellers are prepared to stream the same broadcasts. Id. at 45254. 34 For example, the agreements include payments accept royalty rates for at least some 30 The NAB Settlement rates rose from $0.0017 that are characterized as royalties for performances noncommercial webcasters that are different and per performance in 2011 to $0.0025 in 2015. 37 CFR of recorded music by means of [REDACTED]. See, lower than commercial webcasting rates. 380.12(a). e.g., IHM Ex. 3351 at 7. Since U.S. copyright law

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00006 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26321

determine on this record whether or not ii. Local Focus and Community non-music content transmitted by the iHeart-Indie agreements support the Involvement simulcasters versus the amount NAB proposal. Therefore, the Judges NAB witnesses testified that radio transmitted by other commercial find that the iHeart-Indie agreements do broadcasters focus on their local market webcasters does not support a reduced not provide adequate evidentiary both in their terrestrial broadcasts and royalty rate for simulcasters. support for the NAB’s proposed in their simulcast streams. They iv. Degree of Interactivity differential rate for simulcasters. attribute this local focus to their legal obligations under FCC regulations, 5/20/ The NAB argues that simulcasters c. NAB’s Qualitative Arguments for a should pay a lower royalty rate in Separate Rate for Simulcasters 15 Tr. at 5075 (Newberry), to the needs of their advertisers to reach customers recognition of the fact that simulcast In lieu of quantitative benchmarks, proximate to their places of business, id. transmissions are the least interactive the NAB offers several qualitative at 5077–78, and to their desire to form of webcasting. The NAB contends arguments why willing buyers and connect with their listeners and, that three SoundExchange fact sellers would agree to lower presumably, build listener loyalty. Id. witnesses—Dennis Kooker, Raymond simulcasting rates. Each argument One aspect of that local focus is Hair, and Aaron Harrison—conceded as proceeds from two basic premises: (1) involvement in, and reporting of, much in their testimony and pretrial The programming content on a activities in the community. See, eg., depositions. NAB PFF ¶¶ 114–118. simulcast stream is the same as Knight WDT ¶ 18; Dimick WDT ¶ 33. (A) Kooker Testimony programming content on terrestrial The Judges find neither record evidence Dennis Kooker, President, Global radio; and (2) terrestrial radio is nor an articulated rationale to support a Digital Business at Sony Music fundamentally different from music lower royalty rate for simulcasters based 35 Entertainment, testified that services. on the purported local focus of radio broadcasters. The Judges decline to infer statutory licensees pay for their content at i. FCC License and Public Interest compulsory rates, and as a consequence exert Requirement such a rationale. downward pressure on privately negotiated Radio broadcasters, which are iii. On-Air Personalities and Other Non- rates. One of the original justifications for Music Content allowing statutory services to pay these lower licensed and regulated by the Federal rates was that the offering under the statutory Communications Commission (FCC), are The NAB stresses the role of on-air license would provide a user experience legally required to act in the public personalities, news, weather, and other similar to terrestrial radio. Statutory services interest. See NAB Ex. 4001 ¶ 14 non-music content in cultivating the could offer channels of particular musical (Newberry WDT). By extension, this loyalty of radio listeners and genres, but the programming would be requirement distinguishes simulcasters distinguishing a radio station from its selected by the service. If listeners wanted to from other commercial webcasters. competitors. Once again, the NAB ably select their programming, they would have to The NAB’s witnesses testified demonstrated a distinction between pay for it through directly licensed services. persuasively that the public interest simulcasting and other webcasting, but SX Ex. 12 at 15 (Kooker WDT). The NAB requirement is a key consideration for failed to articulate why that distinction contends that ‘‘Mr. Kooker recognized a radio broadcasters as they conduct their supports differential royalty rates for dichotomy between service-selected business. See, e.g., 5/20/15 Tr. at 5075 simulcasters. programming, which is eligible for the (Newberry); Dimick WDT ¶ 33. What is The NAB cites a survey conducted by lower statutory rate, and listener- far less clear is the connection between Professor Dominique Hanssens that selected programming, which requires this requirement and the NAB’s concluded that 12.2% of the value that payment of a higher, directly licensed proposal that simulcasters should pay simulcast listeners derive from listening rate.’’ NAB PFF ¶ 115. lower royalty rates than other to music-formatted stations is Even accepting Mr. Kooker’s commercial webcasters. The NAB did attributable to ‘‘hosts, DJs, and other on- testimony at face value,37 it is not a not present any persuasive evidence air personalities.’’ NAB Ex. 4012 ¶ 62, concession that simulcasters should be that the public interest requirement App. 8 (Hanssens WRT); NAB Ex. 4015 charged lower rates than other would in any way affect the royalty ¶ 67, Table 5 (Katz AWRT). The NAB webcasters. It is clear in context that the rates that willing buyers and sellers presents no evidence, however, that the ‘‘dichotomy’’ that Mr. Kooker identifies would agree to in the hypothetical on-air time consumed by on-air is that established in § 114 between market. To the extent the NAB’s personalities exceeds, on a percentage interactive services, which are directly argument is that, as a matter of public basis, the value that listeners attribute to licensed, and noninteractive services, policy, radio broadcasters’ public them. By including non-music content which are subject to the statutory interest requirement justifies lower in their transmissions, simulcasters license that is the subject of this royalty rates for simulcasting, that reduce the number of performances of proceeding.38 Mr. Kooker does not state argument is without any basis in § 114. recorded music, thus reducing their that, among statutory services, some royalty obligation under a per- should pay lower rates than others confers no exclusive right of public performance by performance rate structure. The NAB based on how interactive they are. Mr. means of terrestrial radio transmissions for sound failed to present any evidence that the Kooker’s testimony does not support a recording copyright owners, the Judges would need value of non-music content is not fully conclusion that he believes simulcasters further evidence to determine whether, as an accounted for in this reduction of should pay lower rates than other economic matter, these payments should be treated, royalties.36 Absent such evidence, the at least in part, as compensation for other uses (such as [REDACTED]) covered by the agreements Judges find that the relative amount of 37 Mr. Kooker does not cite any evidence of that do require a license under copyright law. legislative history to support his conclusion that the 35 See, e.g., NAB Ex. 4002 ¶¶ 4, 11, 30–40 (Dimick 36 Were the Judges to adopt a percentage-of- similarity of noninteractive webcasting to terrestrial WDT); NAB Ex. 4009 ¶ 5 (Dimick WRT); 5/26/15 Tr. revenue rate structure, an appropriate adjustment radio was a ‘‘justification’’ for allowing statutory 5798–99 (Dimick); 5/20/15 Tr. at 5076–78, 5104 would be necessary to reflect the lower percentage services to pay lower rates. That statement is merely (Newberry); NAB Ex. 4003 ¶¶ 2, 13–26, 29 (Knight of recorded music as compared with an Internet an expression of Mr. Kooker’s lay opinion. WDT); NAB Ex. 4005 ¶ 14, 24–34 (Downs WDT); 5/ music service. As the Judges do not adopt a 38 Mr. Kooker then argues that that distinction is 21/15 Tr. at 5217–19 (Downs); NAB Ex. 4006 ¶¶ 3, percentage-of-revenue rate structure in this ‘‘rapidly disappearing’’ in the marketplace. Kooker 9–19 (Koehn WDT). proceeding, however, no adjustment is needed. WDT at 15.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00007 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26322 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

webcasters, much less support the interactivity that a service offers as a throughout the country or even conclusion that willing sellers would relevant factor in assessing the royalty overseas. See 5/14/15 Tr. at 3909–10 accept a lower rate in the hypothetical rate that a service should be required to (Peterson); 5/29/15 Tr. at 6556 (Kooker); marketplace. pay. As such, the Judges consider it Dimick WDT ¶ 12. The choices available together with the other evidence to radio listeners are more limited than (B) Hair Testimony relevant to the NAB’s arguments. those available to simulcast listeners. In his hearing testimony, Raymond Nevertheless, Mr. Harrison’s See 5/7/15 Tr. at 2522–23 (Wilcox); Hair, International President of the testimony provides little support for the 5/29/15 Tr. at 6556 (Kooker). Through American Federation of Musicians, NAB’s assertion that simulcasters aggregation sites, such as iHeartRadio confirmed that he had previously generally should be entitled to pay and TuneIn, simulcasting offers expressed 39 the opinion that services lower royalty rates than other listeners greater functionality (e.g., the with greater ‘‘functionality’’ should pay commercial webcasters. While the NAB ability to search, pause, rewind and higher rates than services with less posits that simulcasting is less record) than radio does. See 6/1/15 Tr. functionality. 4/29/15 Tr. at 806 interactive than custom webcasting, it at 7075–77 (Burress); SX Ex. 27 at 5 (Hair).40 Mr. Hair’s opinion is not has not established (or attempted to (Kooker WRT); 5/26/15 Tr. at 5840–51 authoritative in this context, however, establish) that simulcasting as a rule is (Dimick). because he represents neither the buyer materially less interactive than any These differences may affect listening nor the seller in the hypothetical other form of non-custom, habits in a way that diminishes the transaction that he describes. noninteractive webcasting, all of which promotional effect of simulcasting. This (C) Harrison Testimony would be subject to the general is supported by uncontroverted commercial webcasting rates. The evidence that radio advertisers are The strongest evidence the NAB offers statutory license is available to services generally unwilling to pay to promote on this point is Aaron Harrison’s that offer a continuum of features, their products and services on simulcast testimony. Mr. Harrison, Senior Vice including various levels of interactivity, streams, see Downs WDT ¶ 22; 5/21/15 President, Business and Legal Affairs of which are offered in a manner Tr. at 5242–43 (Downs), and record UMG Recordings, agreed with the consistent with the license. On the companies do not view simulcasting as statement ‘‘the higher the level of record before them, the Judges find little having the same promotional impact as interactivity, the higher the rate’’ support for attempting to parse the terrestrial radio.41 See 6/1/15 Tr. at 7045, because ‘‘higher levels of interactivity levels of interactivity that the various 7048, 7050 (Burress); Ex. 3242 at 20, 33 are more substitutional than less on- statutory services offer to try to cobble (Walk Deposition at 75, 129). See also demand.’’ 4/30/15 Tr. at 1101 together a customized rate structure Blackburn WRT ¶ 42 (‘‘neither (Harrison). Mr. Harrison also agreed that among categories of commercial interactive nor noninteractive services ‘‘simulcast is the least substitutional.’’ webcasters based solely on statutorily have a statistically significant Id. permissible levels of interactivity. promotional impact on users’ As a record company executive, Mr. propensity to purchase digital tracks’’) v. Promotional Effect Harrison’s testimony provides some (Ex. 24). evidence that record companies would The record of this proceeding is In short, there is no empirical be willing to accept lower royalties from replete with statements concerning the evidence in the record that simulcasting services that are less interactive, promotional value of terrestrial radio is promotional to the same degree as because those services are less likely to play for introducing new artists and terrestrial radio, and the narrative the displace sales of sound recordings. The new songs to the public and stimulating NAB puts forward to support that probative value of his evidence in sales of sound recordings. See, e.g., proposition is flawed at best. The Judges determining whether a differential rate Knight WDT ¶¶ 30–31; Dimick WDT need not, however, decide that is justified for simulcasters is limited, ¶ 43; IHM Ex. 3226 ¶ 7 (Poleman WDT); particular question in order to however. First, Mr. Harrison was 4/28/15 Tr. at 386–87, 461–62 (Kooker). determine whether simulcasters should responding to a question posed in the There appears to be consensus, or near- receive a discounted rate. Whether or abstract, rather than identifying specific consensus, on this point. not simulcasting is as promotional as transactions that he had witnessed or in The consensus breaks down, however, terrestrial radio simply is not the which he had participated. Second, Mr. when it comes to the promotional effect relevant question. The relevant Harrison stated that he was aware of no of webcasting, including simulcasting. questions are (1) whether simulcasting empirical data on the subject, and was The NAB offers a somewhat tautological is more promotional than other forms of merely testifying as to his ‘‘perception argument: Simulcasting is, by commercial webcasting and, if so, (2) from being in the industry.’’ Id. at 1102. definition, simultaneous retransmission whether such heightened promotional In sum, testimony regarding the of the content of a terrestrial radio impact justifies a discounted rate for perceptions of an industry participant broadcast over the Internet; it is, simulcasters. Assuming for the sake of carries considerably less weight than therefore, the same as radio; therefore, it argument that a promotional impact actual examples of marketplace must have the same promotional impact could justify a discounted royalty rate behavior. Nevertheless, Mr. Harrison’s as terrestrial radio. NAB PFF ¶¶ 107– for simulcasters, the NAB would be testimony carries some weight that 113; see NAB Ex. 4000 ¶ 83 (Katz WDT); hypothetical sellers view the amount of Katz AWRT ¶ 98; see also iHeartMedia 41 The NAB and iHeart repeatedly point to PFF ¶¶ 123–124. SoundExchange evidence that record company promotional 39 The earlier statement was in comments Mr. disputes this conclusion. See personnel thank music services for playing their Hair submitted on behalf of the AFM to the artists’ music to support the conclusion that such Copyright Office in connection with a study on SoundExchange PFF ¶¶ 897–938. ‘‘spins’’ are promotional. See, e.g., Emert WDT ¶ 25; music licensing issues. The comments are not a part As SoundExchange points out, there 5/13/15 Tr. at 3573 (Morris); 5/21/15 Tr. at 5165 of the record of this proceeding. are a number of differences between (Poleman); Exs. 3241, 3569, 3570, 3576, 3575, 3576, 40 Mr. Hair’s view of what constitutes terrestrial radio and simulcasting. For 3643. The Judges do not find this argument ‘‘functionality’’ is not entirely clear, however, persuasive. It is at least equally plausible that though it appears to include the ability to ‘‘hear example, terrestrial radio broadcasts are record company executives were merely displaying what I want to hear and hear it when I want to hear (as the NAB stresses) locally-focused; ‘‘common courtesy.’’ 6/1/15 Tr. at 7046–47 it.’’ Id. at 809. simulcasts, by contrast, can be accessed (Burress).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00008 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26323

required to demonstrate that such vi. Additional Considerations iHeartMedia Proposed Rates and Terms promotional effect is greater for Supporting the Same Rate for at 3. simulcasting than for other forms of Simulcasters and Other Commercial Both proposed definitions would commercial webcasting to an extent that Webcasters permit the substitution of substantial portions of the content of a broadcast would justify a lower rate for (A) Competition With Other before retransmitting it over the simulcasters. The NAB has not done so. Commercial Webcasters Internet. [REDACTED], in fact, has The licensee services introduced two Simulcasters and other commercial already developed and deployed studies in this proceeding to webcasters compete for listeners. The [REDACTED] to accomplish this demonstrate empirically that statutory record shows that Pandora, the largest substitution more easily. See 5/13/15 Tr. webcasting is promotional. Pandora commercial webcasting service, regards at 3662 (Littlejohn); see generally IHM presented a study by Dr. Stephen iHeartRadio, one of the largest services Ex. 3210 (Littlejohn WDT). Even if the McBride that examined the effect on that aggregates simulcast streams (as Judges were persuaded that simulcast sales of particular albums (in the case of well as providing a custom streaming streams bear unique characteristics that new music) or songs (in the case of service), as a competitor, and vice versa. distinguish them from other webcast catalog material) in particular See, e.g., SX Ex. 269 at 18 (including streams, the ability and demonstrated geographic regions if Pandora did not iHeart among Pandora competitors); see willingness of broadcasters to alter those play that music in that region. See generally Ex. 166 (including Pandora streams casts doubt on any proposal to generally McBride WDT (PAN Ex. among iHeart competitors). Pandora grant simulcasting lower rates than 5020). iHeartMedia presented a study by broadly includes other interactive and other commercial webcasters. Dr. Todd Kendall that examined the noninteractive streaming services, as well as terrestrial radio, as its d. Conclusion Regarding Separate Rate relationship between music purchases for Simulcasters made on certain machines (PCs) and the competitors. See Ex. 159 at 18–19. Based on the record in the current amount of time that music was streamed Internal iHeartMedia emails demonstrate [REDACTED]. See, e.g., proceeding, the Judges do not find that on those same machines. See generally Exs. 373, 1028, 1189.The mutual a separate rate category for simulcasters Kendall WRT (IHM Ex. 3148). competition between simulcasters and is warranted. The NAB’s arguments in Dr. McBride’s study concluded that other commercial webcasters is a strong favor of a separate rate category for Pandora has a positive effect on music indication that simulcasters and other simulcasters lack support in the record, sales. See McBride WDT ¶ 49. As it commercial webcasters operate in the or are otherwise unpersuasive. The bulk focused solely on the effect that same, not separate submarkets. See Web of relevant evidence in the record Pandora, a custom radio service, has on II, 17 FR at 24095. persuades the Judges that simulcasters and other commercial webcasters music sales, the McBride study reveals (B) Proposed Definitions of Simulcast nothing about the relative promotional compete in the same submarket and value of performances by simulcasters The NAB proposes to define therefore should be subject to the same as compared with other commercial ‘‘broadcast retransmissions’’ (the term rate. Granting simulcasters differential webcasters. used to denote simulcasts in the Judges’ royalty treatment would distort regulations) as follows: competition in this submarket, Dr. Kendall’s study compares the Broadcast Retransmissions means promoting one business model at the promotional effect of interactive and transmissions made by or on behalf of a expense of others. noninteractive streaming services, Broadcaster over the Internet, wireless data finding that noninteractive services networks, or other similar transmission B. Greater-of Rate Structure have a greater promotional effect. See facilities that are primarily retransmissions of In their notice commencing this Kendall WRT ¶¶ 25–29. Again, terrestrial over-the-air broadcast proceeding, the Judges inquired about however, this study fails to compare programming transmitted by the Broadcaster price differentiation in the market and through its AM or FM radio station, simulcasters with other commercial including transmissions containing (1) the desirability of using a percentage-of- webcasters. The noninteractive services substitute advertisements; (2) other revenue rate structure in lieu of, or in that were included in Dr. Kendall’s programming substituted for programming addition to, the per-performance rate study included both simulcast and non- for which requisite licenses or clearances to structure in use for the licenses at issue simulcast webcasters. See IHM Ex. 3151 transmit over the Internet, wireless data in this proceeding. Perhaps in response networks, or such other transmission (Exhibit A to Kendall WRT). to this solicitation of comment, facilities have not been obtained, (3) SoundExchange and Pandora each The Judges are well aware of substituted programming that does not proposed different greater-of rate contain Performances licensed under 17 SoundExchange’s criticisms of these structures employing a per-play rate and two studies. However, for purposes of U.S.C. 112(e) and 114, and; (4) occasional substitution of other programming that does a percentage-of-revenue rate. assessing the strength of the NAB’s not change the character of the content of the Nevertheless, all of the Services apart argument for a separate rate for transmission. from Pandora oppose adoption of this simulcasters, it suffices to note that NAB Proposed Rates and Terms at 2. two-prong approach. As discussed these studies do not even purport to iHeartMedia proposes to amend the below, after careful consideration of all answer the central question whether current definition of ‘‘broadcast rate structure proposals presented in the simulcasting has a greater promotional retransmission’’ in 37 CFR 380.11 by proceeding, the Judges find that a effect than other forms of commercial adding: greater of rate structure is not warranted webcasting. In conclusion, the record in the current rate period. [A] Broadcast Retransmission does not does not support a separate rate for cease to be a Broadcast Retransmission 1. SoundExchange’s Support for a simulcasters on the basis of any because the Broadcaster has replaced Greater-of Rate Structure purported promotional effect programming in its retransmission of the simulcasting may have. radio broadcast, so long as a majority of the In support of its proposed greater-of programming in any given hour of the radio rate structure, SoundExchange makes broadcast has not been replaced. the following arguments.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00009 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26324 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

• According to Dr. Daniel Rubinfeld royalties on the per-play basis. A further economic deficiency in this and Dr. Thomas Lys (two Rubinfeld CWDT ¶ 112.44 two-prong approach, according to the SoundExchange economic expert opposing Services, is that it utilizes a 2. The Services’ Opposition to a Greater- witness), willing buyers and willing percentage of revenue rather than a of Rate Structure sellers have demonstrated a ‘‘revealed percentage of profits. An investment preference’’ for a greater-of rate The Services that oppose the greater- that raises revenues by less than the cost structure, as evidenced by the adoption of structure in principle argue 45 that of the investment would reduce profits, of such rates in the market.42 For such a structure allocates all of the yet, under a percentage-of-revenue example, many agreements that allow downside risk to the Services alone, prong, royalty payments would rise. In for more ‘‘lean-forward’’ functionality while allocating to the record such a scenario, the ‘‘upside’’ from contain a two-pronged per-play and companies a share of potential upside increases in revenues would not revenue percentage structure like the benefits. See, e.g., Katz AWRT ¶ 140. necessarily translate into an increase in one SoundExchange proposes.43 Such misallocation of risk and reward, profits. See Katz AWRT ¶ 150. • A greater-of structure provides according to the opposing Services, not According to the opposing Services, positive economic efficiencies that only unjustifiably allows the record forty-two percent of the Majors’ benefit licensees as well as licensors. 5/ companies to free-ride on a service’s contracts examined by Dr. Rubinfeld do 5/15 Tr. 1756–58 (Rubinfeld). economic success, but also ignores the not contain a per-play prong, • In particular, the greater-of services’ downside risk that they will contradicting SoundExchange’s claim structure provides reasonable fail to execute their respective business that the market has demonstrated a compensation to the record companies models and go out of business. See, e.g., consistent ‘‘revealed preference’’ for a because: (1) The per-play prong IHM Ex. 3216 ¶ 19–26 (Pakman WDT); greater-of approach. Katz AWRT ¶ 143. 46 provides a guaranteed revenue stream, Katz AWRT ¶ 149. According to these Services, all but one especially against the vicissitudes of of the 62 ‘‘label-service pairings’’ 44 consumer demand; and (2) the SoundExchange proposed a ‘‘55% of revenue’’ identified by Dr. Lys related to rate as the second prong of its proposed greater-of interactive services, thereby further percentage-of-revenue prong allows rate structure based on Dr. Rubinfeld’s survey of the record companies to share in any revenue percentage shares contained in his contradicting SoundExchange’s claim of substantial returns generated by a interactive benchmark agreements, which identified a revealed marketplace preference for a a range between 50% and 60% of the services’ greater of rate structure. 5/4/15 Tr. Service. Rubinfeld CWDT ¶¶ 96; 100. revenues, with the majority falling between 55% • The greater-of structure benefits the and 60%. Rubinfeld CWDT ¶ 206; SX Ex. 63, App. 1474–75 (Lys). Services because the presence of the 1a (Rubinfeld CWDT App. 1a). The following The opposing Services also note that noninteractive services and/or nonsubscription the agreements entered into by percentage-of-revenue prong, on the services also have percentage-of-revenue prongs upside, allows for a lower per-play rate [REDACTED] and [REDACTED], relied that approximate the 55% rate SoundExchange has upon by Dr. Rubinfeld, were negotiated than would exist if a single-prong, per- proposed: as parts of overall interactive play rate were established, and a lower [REDACTED]’s agreements with Universal, agreements with their record company per-play rate would encourage entry Warner, and Sony for [REDACTED] Service, which purportedly does not have on-demand counterparties, and the specific services into the market by new services. functionality, has a greater-of structure with within those agreements upon which Rubinfeld CWDT ¶ 95. percentage-of-revenue shares of between Dr. Rubinfeld relies have extra-statutory • The greater-of structure would [REDACTED]%-[REDACTED]% paid by the labels. interactive functionality. See NAB PFF enable a beneficial form of price [REDACTED]’s agreements with Universal, Sony, and Warner for [REDACTED] streaming service, ¶¶ 510, 528–530, 515–518, 525–527 discrimination. All else being equal, which allegedly does not have on-demand (and citations to the record therein).47 services facing relatively low price functionality, has a greater-of structure with a pro- The opposing Services point out that rata share of [REDACTED]% of [REDACTED] elasticities (facing more inelastic the parties to the other agreements demand) would be more likely to charge premium net revenue. [REDACTED]’s free radio service has a relied upon by Dr. Rubinfeld did not higher prices, earn greater revenues and percentage-of-revenue prong in its agreement with demonstrate an expectation that the thus trigger the percentage-of-revenue [REDACTED] for a pro-rata payment of revenue prong of the greater-of formula prong. Conversely, services facing [REDACTED]% of revenue. See SX Ex. 80, SNDEX_ _ _ would ever be triggered (given the relatively high price elasticities (facing 0024312 [REDACTED] 20130101 at SNDEX0024322 ([REDACTED] Agreement). relative levels of the per-play and more elastic demand) would be more SoundExchange acknowledges that several other revenue percentage prongs). See, e.g., likely to charge lower prices, generate agreements contain a percentage-of-revenue prong PAN Ex. 5110 5/6/15 Tr. 6956–57 lower revenues and therefore pay of 45%. More particularly, the [REDACTED] agreements with [REDACTED] and [REDACTED] (Lexton). Rather, according to the have a greater-of compensation formula that opposing Services, the percentage-of- 42 SX Ex. 17 ¶ 94 (Rubinfeld CWDT); SX Ex. 14 includes a pro-rata [REDACTED]% share of ad revenue prongs were added by the ¶¶ 25–32 (Lys WDT) (94% of 62 label-service revenues for the [REDACTED] service. SX Ex. 2070 record companies merely to create pairings adopt a greater-of structure). The majority at section 1(b), p. 1 ([REDACTED] Agreement); SX (50% to 60%) of the purely interactive agreements Ex. 2071 at section 1(d), p. 2 ([REDACTED] favorable precedent for future that contain a greater-of structure utilize the same Agreement). Also, the [REDACTED] Agreement proceedings. See generally Katz AWRT two prongs that SoundExchange proposes–a per- contains a greater-of structure that includes a pro ¶ 193–196; PAN Ex. 5365 at 5–6 play rate and a percentage-of-revenue rate. rata share of [REDACTED]% of gross, non-simulcast Rubinfeld CWDT ¶ 206; SX Ex. 63 (App. 1a). webcasting revenues. SX Ex. 33 § 3(b)(2), at 15–16. (Shapiro SWRT); 5/15/15 Tr. 4025 43 See SX Ex. 2070 (the [REDACTED] Agreement 45 The NAB, iHeart, and Sirius XM raise (Lichtman); 6/2/15 Tr. 7362–63 (Cutler). § 1(b), at1); SX Ex. 2071 (the [REDACTED] additional objections to the use of a percentage-of- Consistent with this point, the opposing Agreement § 1(d), at 2; SX Ex. 33 (the [REDACTED] revenue prong as applied to simulcasters. Because Services note that: Agreement § 3(b)(2), at 15–16); IHM Ex. 3343 at 9; the Judges decline to adopt a separate rate that • There is no evidence that IHM Ex. 3365 at 11; IHM Ex. 3356 at 9–10; applies only to simulcasters they need not address Rubinfeld CWRT ¶ 87 ([REDACTED]’s agreements these additional objections. [REDACTED] has paid royalties under with [REDACTED]); SX Ex. 80; ([REDACTED] 46 These Services assert that there is no economic Agreement); SX Ex. 87 ([REDACTED] Agreement); justification for ‘‘rewarding’’ record companies for 47 With particular regard to the [REDACTED] SX Ex. 100 ([REDACTED] Agreement); IHM Ex. ‘‘incremental value that is created by the webcaster agreements, the opposing Services also note that 3476 ([REDACTED] Term Sheet); SX Ex. 100 above and beyond that created directly by the music they were global deals (rather than U.S.-only deals) ([REDACTED] Agreement); SX Ex. 80 ([REDACTED] itself,’’ an additional value that may arise from and tied rates to the sale of [REDACTED], rendering Agreement); PAN Ex. 5014 ([REDACTED] lower price elasticities not attributable to the sound those agreements inapplicable as benchmarks. Katz Agreement). recordings. See, e.g., Katz AWRT ¶ 148. AWRT ¶ 248.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00010 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26325

the percentage-of-revenue prongs of its Similarly, the Services note that Dr. This discussion is largely academic, agreements with [REDACTED] or the Rubinfeld ignored a Slacker agreement however, because, as discussed below, Indies. See NAB PFF 603 (and record with Universal, under which Slacker the Judges have determined not to adopt citations therein); and paid (until June 2014), the greater of a greater of rate structure and instead • [REDACTED] has not paid royalties [REDACTED]% of revenue (or the stated will continue the current per-play under the percentage-of-revenue prong per-play rates) for the on demand structure for commercial webcasters. of its agreement with [REDACTED]. service, but only the greater of 6/1/15 Tr. 6896–97 (Lexton).48 [REDACTED]% of revenue (or the stated 4. The Judges Reject Adoption of a Greater-of Rate Structure 3. The Services’ Opposition to the per-play rates) for Slacker’s radio service. PAN Ex. 5034 at 0022479–80; 4/ The Judges reject the proposals by Percentage of Revenue That 50 SoundExchange Proposed 30/15 Tr. 1133:6–1135:18 (Harrison). SoundExchange and by Pandora that the The Services further note that the statutory rate should contain a greater- Even assuming that a percentage-of- [REDACTED] revenue-sharing provision of structure. Rather, the Judges find that revenue prong should be included in a relied on by SoundExchange is not for the statutory rate should continue to be greater-of rate structure, the Services ‘‘[REDACTED]’s free radio service,’’ but set on a per-play basis for commercial (including Pandora) oppose the 55% rather applies only to two premium webcasters. The Judges reach this percent figure SoundExchange subscription services and specifically conclusion for several reasons, any one proposed. Their opposition is based on excludes [REDACTED]’s free offerings.51 of which the Judges find to be sufficient the following arguments: Both subscription services offer on- to reject the greater-of approach with a First, as with his per-play proposal, demand functionality, among other percentage-of-revenue prong. Dr. Rubinfeld bases his percentage-of- 52 53 interactive features. The Judges first note that none of the revenue analysis entirely on the Fourth, the Services point out that Dr. percentage-of-revenue prongs in the unsupported and economically Rubinfeld ignored the percent-of- greater-of agreements in the record has improper assumption that, in a revenue levels in the Pandora/Merlin been triggered, which may suggest that competitive market, noninteractive Agreement and the 27 agreements the parties to those agreements viewed services would pay the same between [REDACTED] and independent the per-play rate as the rate term that percentage-of-revenue rates as do labels as they related to custom would most likely apply for the length interactive services.49 (Pureplay) webcasting. Among those of the agreement. See, e.g., 6/2/15 Tr. Second, the Services assert that agreements, all but one contained an SoundExchange’s reliance on evidence 7362–63 (Cutler) (distinguishing ‘‘hard’’ alternative greater-of prong with a negotiations over the iHeart/Warner per- that the Majors were able to extract [REDACTED]% of revenue rate, far less similar supra-competitive rates from a play rate from the percentage-of-revenue than Dr. Rubinfeld’s proposed 55% rate. prong to which Warner ‘‘agreed because handful of services that are not fully on- See, e.g., PAN Ex. 5014; IHM Ex. 3343.54 demand fails to support an importation we were never really going to hit that feature anyway.’’). of the 55% revenue rate into a fully and 50 Additionally, the Services point out that effectively competitive noninteractive beginning in June 2014, Slacker and [REDACTED] Additionally, the agreements, or market. Pandora’s RPFF ¶ 227 agreed to a reduction in the on-demand percentage portions of agreements, relied upon by to [REDACTED]% in exchange for an increase in the SoundExchange in support of a greater (responding to SX PFF ¶¶ 425–430). basic radio percentage to [REDACTED]%, but the Third, the Services argue that Dr. radio service percentage-prong royalty rate therefore of rate structure, are not contained Rubinfeld inexplicably ignored an was still significantly only 64% of the rate for the within the benchmarks relied on by agreement between Slacker and Warner on demand service. PAN Ex. 5035 at 116684–87; 4/ SoundExchange. SoundExchange, 30/15 Tr. 1137:19–1140:10 (A. Harrison). through Dr. Rubinfeld, looked at for Slacker’s DMCA-compliant 51 _ See [REDACTED] Agreement, SNDEX agreements other than his benchmark noninteractive radio service that 0024312_[REDACTED] 20130101 (SX Ex. 80) at 11 requires Slacker to pay the greater of of 82 (revenue-share provisions); id. at 3 of 82 agreements to find rate structures with [REDACTED]% of revenue (or the stated (defining ‘‘Portable Service’’); [REDACTED] a percentage-of-revenue prong. In other Agreement, SNDEX0023904_[REDACTED]_ words, the agreements that per-play rates). The terms of this 20100528 (SX. Ex. 80) at 15 of 155 (defining agreement are in stark contrast to ‘‘Tethered Service’’ and ‘‘Subscription Service’’). SoundExchange contends are most Slacker’s agreement with Warner for 52 See [REDACTED] Agreement, SNDEX0023904_ reflective of the marketplace value of Slacker’s on-demand service, under [REDACTED]_20100528 (SX. Ex. 80) at 15 of 155 the copyright owners’ rights under the which Slacker pays the greater of (describing functionality of ‘‘Subscription statutory licenses do not contain a Service’’). greater of rate structure. [REDACTED]% of revenue (or the stated 53 Additionally, the Services aver that per-play rates). PAN Ex. 5222 (Nov. [REDACTED] service relied on by SoundExchange Further, for its part Pandora pointed 2013 agreement) at 16–17; see also is not DMCA compliant, and therefore is not a to the 25% revenue rate from the 5/7/15 Tr. 2495:5–2498:8 (Wilcox). noninteractive service, as SoundExchange claims. Pandora/Merlin Agreement to support a See IHM PFF ¶¶ 352–355 (and citations to the greater of rate structure. Unlike the record therein). Furthermore, the [REDACTED]% of 48 Moreover, in this vein, the opposing Services revenue share agreed to by [REDACTED] for the steered rate provision in the Pandora/ point out that [REDACTED] did not even estimate [REDACTED] service is below SoundExchange’s Merlin Agreement, however, the 25% of the potential value of the percentage-of-revenue proposed interactive-based 55% benchmark rate. revenue prong was nothing other than a prong in its agreement with [REDACTED]. Id. at According to the Services, the provisions of the 6895. [REDACTED] agreements cited in this paragraph do figurative ‘‘cut and paste’’ of the 49 Pandora’s RPFF ¶ 226 (quoting Rubinfeld not reflect a comparable ‘‘greater of compensation Pureplay percentage rate. As such, it CWDT ¶ 169 (‘‘I have assumed that the ratio of the formula,’’ as SoundExchange claims, but rather reveals nothing about whether the average retail subscription price to the per- reflect a formula whereby a per-play rate is added parties in the marketplace would agree subscriber royalty paid by the licensee to the record to a different percent-of-revenue figure. See to include such a prong in an label is approximately the same in both interactive [REDACTED] Agreement § (1)(b), at 1–2 (SX Ex. and noninteractive markets.’’)) (emphasis added). 2070) (‘‘[REDACTED]% of Net Advertising Revenue Pandora’s RPFF ¶ 226 (quoting Rubinfeld CWDT Per Play’’); [REDACTED] Agreement § 1(d), p.2 (SX iHeartMedia’s custom offering. The Services argue ¶ 169 (‘‘I have assumed that the ratio of the average Ex. 2071) (‘‘[REDACTED]% of Net Advertising that this [REDACTED]% rate should be given little retail subscription price to the per-subscriber Revenues per Play’’). weight, in that it ‘‘was only agreed to because it was royalty paid by the licensee to the record label is 54 Pandora notes one outlier, the agreement almost certainly not going to become binding approximately the same in both interactive and between [REDACTED] and iHeartMedia, that during the term of the agreement.’’ 6/2/15 Tr. noninteractive markets.’’)) (emphasis added). contains a [REDACTED]% of revenue prong for 7362:21–7363:5 (Cutler).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00011 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26326 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

agreement.55 Indeed, Dr. Shapiro revenue rate, no such evidence has been The parties’ mutual awareness proffered virtually no justification for provided to the Judges in this reconfirms the Judges’ earlier the inclusion of the percentage-of- proceeding.58 conclusion that the promotion and revenue prong in Pandora’s proposal. For these reasons, the Judges reject substitution effects on royalty rates are Relatedly, SoundExchange’s rationale the two-pronged rate proposals ‘‘baked in’’ to a negotiated license rate. in support of a greater of structure that proposed by SoundExchange and To the extent the Judges adopt a rate record companies should share in the Pandora, and shall continue the current based on benchmark evidence, it is not upside if the Services monetize their practice of setting the statutory necessary to make additional models at a faster rate is wholly webcasting rates on a per-play basis. adjustments to benchmarks to reflect the unconvincing. Absent proof that the C. Promotion and Substitution promotion and substitution factors. The per-play prong had been set too low, Judges hold in this determination, as there is no justification for assuming The Act provides, among other things, they have held consistently in the past, that the record companies should share that the Judges base their hypothetical that the use of benchmarks ‘‘bakes-in’’ in that monetization through a marketplace rates on ‘‘economic, the contracting parties’ expectations percentage-of-revenue prong in the rate competitive[,] and programming regarding the promotional and structure.56 Dr. Rubinfeld indicated that information’’ that the parties present, substitutional effects of the agreement. his ‘‘ratio equivalency’’ per-play including promotion and substitution as For the noninteractive benchmarks methodology resulted in a per-play factors that would influence rates in the upon which the Judges rely, this long- royalty payment that approximated 55% marketplace. 17 U.S.C. 114(f)(2)(B).59 standing position to deem substitution of service revenue. Successful As set forth in this determination, and promotion effects as incorporated monetization by the Services might infra, the Judges have relied upon into the agreements appears to be fully drive the percent-of-revenue certain marketplace agreements as applicable. equivalence below 55%, but there is no benchmarks for the setting of the SoundExchange disagrees, however, economic basis to support maintaining statutory rates. In prior determinations, and points, for example, to testimony that level with a separate percent-of- the Judges have concluded that from Charlie Lexton of Merlin who revenue prong.57 contracting parties, as rational economic stated that Merlin never considered the Only SoundExchange and Pandora actors, factor in the promotion and promotional or substitutional effects proposed a two-prong approach, and, as substitution effects when negotiating when agreeing to the terms of the discussed above, the Judges find their direct licenses.60 That is, parties Pandora/Merlin Agreement. 6/1/15 Tr. reasons in support of such a structure negotiating direct licenses for the 6910 (Lexton). The Judges find that such unpersuasive. Moreover, other parties performance of sound recordings on testimony is not credible and not raised numerous, valid objections to the services will be cognizant of the sufficient to support abandonment by use of a greater-of structure with a promotion and substitution effects, and the Judges of their long-standing percent-of-revenue prong. See, e.g., NAB those effects will influence the rate at treatment of promotional and Ex. 4011 (Weil WRT) (a percent-of- which they agree to a license. Witnesses substitutional issues. Indeed, the fact revenue rate would create uncertainty on both sides in this proceeding that Merlin arguably was so cavalier and controversy regarding the definition generally agree that promotion and regarding the impact of the Pandora/ and allocation of revenue). substitution effects are factored into Merlin Agreement on the positive Finally, by maintaining the statutory negotiated agreements. See, e.g., promotional effects or the negative rate as a per-play rate, the Judges are Rubinfeld CWDT ¶ 31(d); Shapiro WDT substitutional effects (to interactive acting in a manner consistent with prior at 39).61 streaming, download sales, and other decisions, consistent with 17 U.S.C. revenue channels) implies that Merlin 803(a)(1). Although new and persuasive 58 Moreover, the Judges are concerned that, given either understood the net value of these evidence could cause the Judges in the limitations of the evidence in this proceeding regarding agreements with greater of rate structures, factors to be positive or, at worst, future proceedings to consider a greater- any attempt to ‘‘mix and match’’ per-play rates with neutral. Apparently, SoundExchange of rate structure and a percent-of- percentage-of-revenue rates could cause licensors infers: ‘‘This is not to say that [Merlin] and licensees alike to experience undesirable and did not value those terms—of course it 55 When Pandora and Merlin agreed to a lower potentially destabilizing swings in anticipated did, but there was no precise calibration per-play rate through steering, they created a rate revenues and payments over the length of the that was not the higher Pureplay rate. By contrast, license. Continuation of the current per-play rate of the negotiated rate to Merlin’s view the 25% of revenue prong that they incorporated structure helps to ameliorate this concern. of the promotional and substitutional into the agreement, which equaled the Pureplay 59 In prior proceedings, the focus of the question impact of the deal.’’ SX PFF ¶ 1101. It rate, reveals nothing about any specific negotiations of substitution has been physical record sales. In strains credulity to think that Merlin between Pandora and Merlin over that term. For the current market, however, digital access through example, if Pandora and Merlin had agreed to a interactive services is a revenue stream that might was oblivious to the potential 20% or a 30% revenue prong, that fact would be affected by consumers choosing the statutory promotional and substitutional effects of perhaps have been informative of a marketplace noninteractive streaming services. To evaluate the Pandora/Merlin Agreement, yet term. interactive licenses as benchmarks for proceeded with the deal on unaltered 56 A potential rationale for the percentage-of- noninteractive services, therefore, the Judges must revenue prong is that it could offset a per-play rate look at how the latter might prove a substitute for terms. that is ‘‘too low.’’ The Judges have taken great care the former. Additionally, the Judges reject the to discount any proposed rate that they believe 60 See Web III Remand, 79 FR 23102, 23119 n.50 argument, advanced by SoundExchange, would be too low to compensate adequately the (‘‘The adoption of an adjusted benchmark approach that the Pandora/Merlin and iHeart/ licensors for the rights under the licenses. As to determine the rates leads this panel to agree with Warner Agreements are too new and discussed below, the per-play rates that the Judges Web II and Web I that such statutory considerations adopt for commercial webcasters are consistent implicitly have been factored into the negotiated untested to support the longstanding with rates negotiated in marketplace agreements. prices utilized in the benchmark agreements. Web understanding that substitution and 57 This criticism would not apply to the II, 72 FR at 24095; Web I, 67 FR at 45244.’’). promotional effects are ‘‘baked in’’ to subscription rates for noninteractive services, based 61 The more particular issue of whether benchmark agreements. An important upon Dr. Rubinfeld’s ‘‘ratio equivalency’’ model. noninteractive services substitute for interactive However, the other criticisms set forth in the text services is part and parcel of the issue of whether aspect of the benchmarking approach is are sufficient to reject the use of a greater-of rate there has been important ‘‘convergence’’ between structure with a percentage-of-revenue prong even the two types of services, discussed at length in segmentation of listeners based on their willingness for the subscription rate. connection with the evidence regarding to pay.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00012 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26327

that it credits sophisticated business types of music listening formats. She willingness to pay (WTP) for different entities that have carefully negotiated then defined the relevant population as services. See 5/29/15 Tr. 6779, 6796–98 their agreements with an understanding comprising those individuals who (Butler) (acknowledging that she did not of market forces. That is, there is a reported themselves as currently using measure WTP—including whether WTP presumption that marketplace iHeart or Pandora. For listeners who for any listener was greater than zero). benchmarks demonstrate how parties to reported using both of these services, Her survey also did not test whether the the underlying agreements commit real Ms. Butler testified that she assigned responding listeners had any knowledge funds and resources, which serve as them to either the iHeart or the Pandora of the prices of the potential substitute strong indicators of their understanding group. Id. ¶¶ 30–31. services she provided to them when of the market. If promotional or Survey respondents were asked two asking her second question. Given that substitutional effects had separate substantive questions relating to each the Judges are attempting to set rates in values that were not already reflected in service. The first question asked: this proceeding, a survey that asks those rate and play-quantity terms, Imagine you could no longer listen to ‘‘listeners’’ to rank substitute services rational commercial entities would music on iHeart [or Pandora]. Which of without providing price information identify those promotional and the following statements represents fails to provide any meaningful substitutional effects and account for what you would be most likely to do? information as to how those ‘‘listeners’’ them explicitly. • I would find a substitute for the music will act as ‘‘consumers’’ of streaming The ‘‘baked-in’’ aspect of promotional I listen to on iHeart [or Pandora] services. and substitutional effects does not • I would stop listening to music Second, Ms. Butler did not select her address the issue of whether there is a • Don’t know/unsure survey respondents in a random difference between the promotional/ Id. ¶ 38. manner, and therefore had no ability to substitutional effects of interactive The second question asked calculate margins of error or confidence services, on the one hand, and respondents who answered the first intervals for her results. See 5/29/15 Tr. noninteractive services, on the other. To question by stating they would find a 6782 (Butler). the extent the Judges rely on substitute for the music they listened to Third, Ms. Butler intentionally SoundExchange’s interactive benchmark on either Pandora or iHeart: assigned virtually all respondents who to set statutory rates in the Which of the following, if any, would reported listening to both Pandora and noninteractive market, the Judges must be your most preferred substitute for iHeart to the iHeart group only for identify and consider any difference in iHeart [Pandora]? further questioning. This caused her to the promotional/substitutional effects Id. ¶ 40. Respondents were given a list omit about 40% of actual Pandora users between these markets to determine from her results as they related to such whether to adjust the interactive of alternatives. Id. Ms. Butler’s survey found that for Pandora users, including respondents benchmark rate. who reported using Pandora daily. Id. at These potential promotional/ Pandora users, 43.3% would listen to 6789, 6806–08. substitutional effects hypothetically one of the following services: Spotify could occur in two different ways. First, (19.7%), iTunes Radio (9.7%), Amazon Accordingly, the Judges cannot and the availability of noninteractive and Rhapsody (about 4% each), Google do not rely on Ms. Butler’s survey services could cause listeners to Play and Slacker (about 2% each), and results. substitute noninteractive listening at the Beats and Rdio (about 1% each). Id. Mr. Rosin, on whose survey the expense of interactive services. Second, ¶ 48, Figure 3. For iHeart users, Ms. Services rely, conducted his survey in a noninteractive services could substitute Butler’s survey showed that 30% would manner consistent with the standards for, or promote less, the sale of sound switch to Pandora, and 23.1% would and code of ethics of the American recordings through downloads or instead listen to another service, Association for Public Opinion otherwise. To address these issues, the including Spotify (10.7%), iTunes Radio Research, a major survey research parties rely on expert witness testimony (7.5%), or Amazon, Google Play, standards organization. PAN Ex 5021 at and on the observational and anecdotal Slacker, or Rhapsody (about 1% each). 5 n.2. (Rosin WRT). Specifically, Mr. testimony of industry witnesses. The Id. ¶ 50, Figure 5. Rosin conducted a national telephone Judges find the lay testimony to be According to SoundExchange, these survey of Americans 13 years of age and unhelpful and essentially self-serving. results show that interactive services are older. Respondents were selected Rather, the Judges find this issue to be common, if not predominant, randomly, and 2,006 interviews were technical in nature, and consider the substitutes for noninteractive services, conducted via landlines and cell expert testimony, discussed below, to be and that listeners would turn to such phones. The margin of error for his ¥ the type of evidence that has the interactive services in a hypothetical results was +/ 2%, with a confidence potential to identify whether such world in which no statutory interval of 95%. Rosin WRT at 5, 7. differences exist. SoundExchange relied noninteractive services were available. The responses to Mr. Rosin’s survey upon the survey work undertaken by SX PFF ¶¶ 1130–1131. revealed, inter alia, that Sarah Butler, a Vice President at NERA The Judges have evaluated Ms. • only 1% to 1.6% of noninteractive Economic Consulting. The Services’ Butler’s survey and the criticisms by the users reported that their listening was position was supported by the survey Services, and the Judges find that there replacing listening on interactive work of Larry Rosin, President of Edison are three significant problems with Ms. services; Research. Butler’s survey that preclude its • only 3.8% of survey respondents Ms. Butler, a survey expert, designed usefulness in attempting to demonstrate would subscribe to pay for an and constructed a consumer survey to that noninteractive statutory services interactive service; identify the types of music listening substitute for interactive services. Any • only 2% of survey respondents Pandora and iHeart substituted for, in one of these problems, standing alone, were ‘‘very likely’’ to pay the market the opinion of listeners. SX Ex. 5 at 3. is sufficient to preclude the Judges’ monthly subscription rate of $9.99 for Ms. Butler gathered information from reliance on Ms. Butler’s survey. an interactive service, and only 7% on-line survey respondents on age, First, Ms. Butler’s survey fails even to were ‘‘somewhat likely’’ to subscribe at gender, and familiarity with different attempt to measure listeners’ this price point—91% were ‘‘not at all

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00013 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26328 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

likely’’ or ‘‘not very likely’’ to subscribe subscription to an interactive service, WRT App. B at 9, and he identified at that price. because interactive services do offer additional features of an on-demand Rosin WRT at 9, 12. ‘‘freemium’’ subscriptions, which begin service (Spotify) in a prior question. See Based upon these findings, Mr. Rosin as free subscriptions subject to a id., Question 7E. There is not sufficient concluded that: conversion option. The Judges find that record evidence to suggest that the 1. Most consumers are unwilling to Mr. Rosin’s language meaningfully structuring of these questions in this pay monthly subscription fees for access reinforces the different pricing and manner weakens the probative value of to streaming services. pricing strategies that exist in the Mr. Rosin’s survey and conclusions. 2. Noninteractive services like market, because FM radio is free to the Turning to the question of whether Pandora and iHeart are not close listener and on-demand services are there is a difference between the substitutes for interactive on-demand designed to obtain paying subscribers, substitution or promotion effects of services such as Spotify. whether at the outset of the subscription interactive versus noninteractive 3. Only a small market exists for paid period or by using ad-supported services with regard to music sales, the (subscription) services. services as a ‘‘freemium’’ tool to convert parties presented different empirical 4. Listeners to Pandora would not listeners into subscribers. (Indeed, analyses. otherwise be listening to interactive SoundExchange’s economic expert, Dr. iHeart relied upon the expert services. Rubinfeld, testified that he did not even testimony of Dr. Todd Kendall, who Rosin WRT at 4. use interactive ad-supported rates as a attempted to analyze the effect of The Judges find Mr. Rosin’s random benchmark because they were designed listening to online streaming on music survey to be generally credible, and as tools to convert listeners into purchases, by reviewing data from certainly more informative than the subscribers.) 10,000 personal computers over a six non-random survey work done by Ms. The Judges take note of month period. IHM Ex. 3148 ¶ 8 Butler. Most importantly, Mr. Rosin SoundExchange’s criticism of Mr. (Kendall WRT). Dr. Kendall used three treated ‘‘listeners’’ as ‘‘consumers’’— Rosin’s decision not to rotate one of his categories of monthly data for each inquiring as to their WTP rather than multiple choice answers to the question sample computer: (1) The amount of their preferences unconstrained by of what a listener would do if no free time spent listening to music; (2) the prices. SoundExchange argues that even streaming services existed. See Rosin number of digital music purchases made this price-point inquiry indicates that WRT at App. B. The choice ‘‘would you on Amazon and iTunes; and (3) the some listeners, at some lower price just listen to less music’’ was always amount of time spent visiting music points, might be somewhat likely to asked last, whereas the other three sites, such as RollingStone.com. Id. subscribe to an on-demand service. See choices (listen to free FM radio, listen ¶¶ 10, 12; see IHM Exs. 3151–3153. Rosin WRT at 10 (only 79% of to your CDs and downloads or watch He then compared the relative respondents ‘‘not at all likely’’ or ‘‘not music videos, YouTube, or Vevo) were promotional effect of fourteen on- very likely’’ to spend $4.99 per month rotated. SoundExchange notes the demand services, including Spotify, for a streaming subscription, and that presence of a potential ‘‘recency effect’’ with the relative promotional effect of percentage drops to 69% if the price is if one choice is always presented last, nine Internet radio services, including lowered to $2.99 per month). However, possibly inducing respondents to favor Pandora and iHeart. Kendall WRT ¶¶ 9, there is no dispute that subscribers that choice. Mr. Rosin acknowledged 15–17. Dr. Kendall found that a 10% constitute a minority of overall the general existence of such an effect, increase in listening to Internet radio streaming listeners (as noted infra in the 5/14/15 Tr. 3755 (Rosin), but he was associated with a statistically discussion of ‘‘Convergence’’), so it is indicated that ‘‘pinning’’ certain options significant 0.070% increase in music not particularly revealing that these in a multiple choice question was purchasing. See id. ¶ 22; IHM Exs. 3154, levels of survey respondents would necessary to enhance the respondents’ 3156–3158. Based on this finding, Dr. consider subscribing instead to an on- ability to comprehend the question. 5/ Kendall opined that noninteractive demand interactive service at various 14/15 Tr. 3743–44 (Rosin). The Judges services are 15 times more promotional lower price points.62 do not find that there was record than interactive services. Kendall WRT The Judges reject the additional evidence sufficient to find that it was ¶ 5. criticism by SoundExchange that Mr. unreasonable for Mr. Rosin, in applying There are several important flaws in Rosin should not have presented his expertise, to weigh these technical Dr. Kendall’s work, however, that specific price points to respondents, but survey issues and construct his choices render it insufficient for the Judges to rather should have asked if they were in this manner, nor do the Judges find conclude that Dr. Rubinfeld’s interactive willing to pay a ‘‘small fee’’ for that there was sufficient record evidence benchmark should be reduced to reflect interactive subscriptions. Such a vague to indicate that Mr. Rosin’s fundamental a supposed lower promotional effect. phrase would be less informative, and conclusions would have been materially Most importantly, Dr. Kendall’s more subjective, than particular price different if he had rotated that final conclusion is premised on his finding points. The Judges also reject the choice on that single question. that on the computers he analyzed criticism that Mr. Rosin should not have Finally, the Judges do not agree with individuals spent 18 times more time indicated that an alternative to SoundExchange’s criticism that Mr. listening to interactive services than to noninteractive services was to listen to Rosin’s survey is deficient because he noninteractive services. 5/12/15 Tr. ‘‘free’’ FM radio and that another failed to describe in sufficient detail the 3274 (Kendall). When listeners spend alternative was to ‘‘pay’’ for a features offered by a hypothetical on- more time on a service, that drives demand interactive subscription service down the calculation of the number of 62 Also, to the extent subscribership might in one of his questions.63 However, in purchases per hour of listening, which increase if the subscription price were lowered, that question, he specifically mentioned is the promotional effect being sought then the commensurate royalty derived by Spotify, Rhapsody, and Rdio, see Rosin by the analysis. SoundExchange’s interactive ‘‘ratio equivalency’’ SoundExchange demonstrated in its benchmark analysis (discussed infra) would likewise be reduced. Thus, these criticisms of Mr. 63 Mr. Rosin described them in Question 9A as cross-examination of Dr. Kendall that Rosin’s survey results undermine any broad use of services that allow listeners to stream music as they this extreme multiple resulted from the SoundExchange’s own interactive benchmark. choose, for access but not ownership. different methods of recording listening

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00014 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26329

time for interactive and noninteractive whether performance of sound Judges conclude that the benchmarking services. More particularly, Spotify, a recordings on Pandora have a positive process ‘‘bakes-in’’ (internalizes) these leading interactive service, is more or negative impact on sales of those necessary elements, given the assumed widely used on desktop applications, sound recordings.’’ PAN Ex. 5020 ¶ 23 rational, maximizing nature of and Pandora is more widely accessed (McBride WDT). However, whether or sophisticated business entities. through web browsers. SX Ex. 1568; 5/ not Pandora has a net promotional or Moreover, even if the Judges were to 12/15 Tr. 3305 (Kendall). Web site substitutional effect does not address attempt to ascertain whether a particular listening measurements were cut off if the issue of whether that net effect is ROI could be met by a given rate, or the listener had not interacted with the different from the net promotional/ whether a particular business model Pandora Web site. Kendall WRT ¶ 5 substitutional effect of interactive could be sustained, the present record n.14. By contrast, listening services. would preclude such an analysis. The measurements based on the use of Rather, when relying on benchmarks, Judges would require much more desktop applications simply measured the Judges deem the benchmark detailed financial and economic data the time the application was open on a agreements of rational actors to include regarding the parties’ costs and revenues user’s desktop, and otherwise not in an implicit understanding of the before attempting to make such hibernation mode, screen saver mode, or promotional and substitutional effects of determinations. some other similar mode. Id. Further, their transaction. Therefore, Dr. Further, as the Judges have previously the default setting for the Spotify McBride’s conclusions, as well as Dr. held, the statute neither requires nor application is for it to launch when the Blackburn’s criticisms of those ‘‘Music permits the Judges to protect any given computer is turned on—even if no one Sales Experiments,’’ do not affect the business model proposed or adopted by is listening. 5/12/15 Tr. 3306–07 Judges’ rate determination. a market participant. Web II, 72 FR at 24089. The Judges further noted in the (Kendall). D. Impact of Parties’ Financial Simply put, these differences in Web III Remand that any attempt by the Circumstances measuring listening time alone skew Dr. Judges to set rates with these ROI and Kendall’s analysis and results. The Services aver that the rates set in business model issues in mind would Accordingly, the Judges cannot this proceeding must be sufficiently low essentially convert this § 114(f)(2)(B) conclude from his testimony and to permit their business models to be proceeding into a classic public utility analysis that noninteractive services are profitable. See, e.g., NAB PFF ¶¶ 119– style rate-of-return hearing. 79 FR at more promotional of music sales than 149; IHM ¶¶ 245–257 (and citations to 23107. None of the parties argues that interactive services. the record therein). Reciprocally, the statutory standard permits such a With regard to the relative SoundExchange argues that the rates process, and neither the D.C. Circuit, promotional or substitutional effects of must be sufficiently high to allow the nor the Judges (or any of their interactive versus noninteractive record companies to cover their costs predecessors) have so held. streaming services on music sales, and to obtain the necessary return on SoundExchange relies on the testimony investment (ROI), plus a profit. See, e.g., E. The Effect of the Alleged ‘‘Shadow’’ of Dr. David Blackburn. Unlike Dr. SX PFF ¶¶ 165–208 (discussing costs of the Statutory Rate Kendall, he did not attempt to relate the and investments and noting (¶ 165) that The parties assert that the benchmarks amount of time spent listening to these ‘‘[t]he rates that record companies that are adverse to their positions are services to increases in purchasing receive from streaming services ha[ve] compromised by the fact that they were music. Rather, Dr. Blackburn attempted been—and over the next five years will set in the ‘‘shadow’’ of the statutory rate. to determine whether there was any continue to be—critical to [the record See, e.g., Rubinfeld CWDT ¶¶ 80–85 meaningful promotional or substitution companies’] ability to make such (statutory rate as a shadow pushing rates effect on music sales as between those recurring investments.’’); 4/30/15 Tr. down); Talley WRT at 46; Shapiro WDT who use the two different types of 972–73 (A. Harrison) (‘‘[T]he profit at 36 (statutory rate as a shadow pulling services. maximization goal is definitely . . . a rates up); 5/15/15 Tr. 3993–94 In this instance, the particulars of the top goal of the company . . . and also (Lichtman); Fischel (same). There are study are less important than the provides the incentive to create essentially two types of statutory conclusion. Dr. Blackburn opined that, music.’’). shadows noted by the parties. based on his analysis, ‘‘neither The Judges find that they do not need The first purported shadow is cast by interactive nor non-interactive services to relate the rates set in this proceeding the existing statutory rate, whether set have a statistically significant directly to the parties’ proposed in a CRB proceeding or through the promotional impact on users’ business models. Rather, the Judges’ parties’ WSA settlements. As an initial propensity to purchase digital tracks.’’ adoption of the benchmark method of matter, the Judges find that any such SX Ex. 24 ¶ 42 (Blackburn WRT). determining rates obviates the need to: ‘‘shadows’’ that could have been cast by Because Dr. Blackburn is a (1) Analyze whether the record existing statutory rates did not SoundExchange witness, and because companies’ costs require a particular meaningfully affect the effective steered the point of the present discussion is to rate to allow them to obtain an rates in the Pandora/Merlin Agreement determine whether an interactive appropriate ROI; and (2) protect or the IHeart/Warner Agreement. As benchmark rate must be lowered or particular noninteractive services whose discussed herein, those rates are below raised to reflect such differences, his business models might require a low the otherwise applicable statutory rates, conclusion fails to support any change enough rate to sustain their survival and it would be irrational for a licensor in SoundExchange’s interactive and/or growth. Benchmarks based on to accept a rate below the statutory rate benchmark for promotional or marketplace agreements, by their very when it could have rejected the direct substitutional effects. nature, reflect the parties’ need for rates deal and enjoyed the higher statutory Finally, the Judges take note of that allow them to project a sufficient rate. Also, the supposed shadow of the Pandora’s ‘‘Music Sales Experiments’’ ROI and enable them to implement their existing rate is less relevant to the conducted by its Senior Scientist, respective business models. subscription-based benchmark proffered Economics, Dr. Stephan McBride. The As with the promotional and by SoundExchange, because it is based purpose of that experiment was ‘‘to test substitutional impact of the rates, the on benchmarks that are at a further

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00015 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26330 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

remove from the statutory license. The Services counter that, although indeterminacy of Dr. Talley’s argument. Rubinfeld CWDT ¶ 18. the logic of Dr. Talley’s point may be Further, Dr. Talley’s point appears to be Dr. Shapiro argues that the statutory correct, Dr. Talley’s analysis is purely a back-door way to question both the shadow not only exceeds the theoretical and he did not examine the applicability of the benchmarks in the marketplace rate, but also acts like a evidence to determine whether his noninteractive market, as well as the ‘‘focal point,’’ or ‘‘magnet,’’ pulling a analysis was supported by the facts. In benchmarking process itself. However, freely negotiated rate higher than it particular, the Services criticize Dr. the Judges have found that the Pandora/ would be in the absence of the statutory Talley’s ‘‘shadow’’ argument because he Merlin Agreement and the iHeart/ shadow. Shapiro WDT at 36–37. assumes that the ‘‘missing dyads’’ Warner Agreement to be sufficiently However, neither Dr. Shapiro nor any would reflect a significantly different representative benchmarks (and have other expert provides a sufficiently WTP and WTA than those of the parties found that Dr. Rubinfeld’s benchmark detailed explanation as to how the who entered into agreements (e.g., the analysis is likewise representative) in statutory rate would pull up a below- Pandora/Merlin dyad and the iHeart/ particular segments of the statutory statute consensual rate that is otherwise Warner dyad). See, e.g., Pandora RPFF market. This segmented analysis mutually beneficial. Rather, the experts 96–103 (and citations to the record strengthens the representativeness of the who advance this variant of the shadow therein). Dr. Talley counters, quite benchmarks and weakens the argument simply note the existence of a correctly, that the very point of his speculative argument that ‘‘missing ‘‘focal point,’’ ‘‘magnet’’ or ‘‘anchor’’ analysis is that no negotiations or dyads’’ might tell a different story. theory in the economic literature and agreements for above-statutory rates The second shadow identified by the then posit that such an effect is present would exist because the parties would parties is cast by the statutory rate yet in the noninteractive market—without not waste their time engaging in to be established in this proceeding. The making a sufficient connection between bargaining that was made moot by the record is replete with evidence that the theory and evidence. Indeed, Dr. statutory rate. Id. at 6032–34. parties entered into various transactions Shapiro candidly acknowledged that the Dr. Talley suggests though that Dr. with the knowledge, if not the intent, focal point/magnet/anchor hypothesis is Rubinfeld’s interactive benchmark may that such agreements could be used as not an ‘‘ironclad’’ economic law. Id. at approximate the ‘‘unseen’’ 37 n. 65. In sum, the Judges do not noninteractive transactions because it is evidentiary benchmarks in this credit this conjecture as sufficient to affected less by the shadow of the proceeding. See SX PFF ¶¶ 567–570 affect their determination of the rate in statutory rate. Id. at 6036. However, that (and citations to the record therein this proceeding. argument fails to note the fundamental regarding the Pandora/Merlin On behalf of SoundExchange, Dr. distinction in Dr. Rubinfeld’s Agreement); IHM PFF ¶¶ 359–362 (and Talley asserts that the existing statutory benchmark—that it pertains to an citations to the record therein regarding rate casts a shadow so dark as to obscure upstream market for interactive Apple’s agreements with the Majors); entirely evidence of consensual licensees in which upstream demand is NAB PFF ¶¶ 456–458. Of course, a transactions that would have been derived from downstream consumers proposed benchmark is not disqualified consummated in the noninteractive who have a positive WTP for streaming because a contracting party wanted it to space, but for the statutory rate. More services. The ‘‘missing dyads,’’ so to be a benchmark. Such a desire would particularly, Dr. Talley notes that any speak, would be those in the upstream apply to otherwise proper benchmarks pairing of willing licensors and noninteractive market in which the as it would to dubious benchmarks. The licensees (‘‘dyads’’ in Dr. Talley’s ‘‘missing’’ agreements would reflect Judges analyze the proposed parlance) in which the licensee’s WTP only the downstream demand of benchmarks based on the overall factual was greater than the statutory rate, and listeners to free-to-the-listener ad- merits attendant to their formation and greater than or equal to a licensor’s supported platforms, not those dyads applicability, not based upon the ‘‘willingness to accept’’ (WTA) (also identified by Dr. Rubinfeld in the parties’ hopes or manipulations. If a above the statutory rate), would not subscription market.65 benchmark is deficient in some manner, consummate an agreement at a Relatedly, the Services also criticize the adversarial process of this consensual rate, because the buyer Dr. Talley’s argument because it fails to proceeding allows the parties to expose would always default to the lower note the potential steering, ‘‘competitive those deficiencies. statutory rate. SX Ex. 19 at 58 (Talley dynamics’’, or other interactions that The Judges agree with a particular WRT) (Concluding ‘‘in an economic would cause dyads to cluster closely. 5/ criticism made by iHeart of the shadow environment most relevant to this 19/15 Tr. 4660–61 (Shapiro). argument asserted by SoundExchange: setting, a statutory licensing option can On balance, the Judges find Dr. In the absence of the statutory shadow, crowd out negotiated transactions for Talley’s criticism, albeit rational and the antitrust policy toward the relatively high-valuing buyer-seller hypothetically correct, too untethered noninteractive streaming market could dyads while not affecting other, low- from the facts to be predictive or useful well be different. Cf. 141 Cong. Rec. S. valuing dyads. . . . [T]his crowding out in adjusting for the supposed shadow of 11,962–63 (daily ed. Aug. 8, 1995) phenomenon can generate downward the existing statutory rate. The Services’ (Letter from Assistant Attorney General statistical bias, leaving behind only a criticisms are likewise speculative, but Andrew Fois to Hon. Patrick Leahy, July subset of negotiated deals involving that simply underscores the factual 21, 1995, noting that any buyers and sellers whose valuations noncompetitive rates created by the . . . reflect[ ] prices which serve as poor strike a deal between $0.0015 and $0.0020. existence of only a single collective benchmarks for estimating the price [to However, with the statutory rate at $0.0010, the could be corrected by the ‘‘rate panel.’’). licensee would not negotiate, but would default to which] willing buyers and sellers would the lower statutory rate. Dr. Talley describes such Although that comment was made in 64 agree.) a foreclosed agreement as having been obscured by connection with the potential the shadow of the statutory rate. anticompetitive consequence of a single 64 For example, assume the statutory rate was 65 This important distinction between listeners collective, it suggests to the Judges that $0.0010. If a licensor had a WTA of $0.0015 and based on their differentiated WTP is discussed in the so-called ‘‘shadow’’ of the statutory a licensee had a WTP of $0.0020, then in the greater detail infra in connection with Dr. absence of a statutory rate, these parties would Rubinfeld’s proposed benchmark. rate offsets any potential device that

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00016 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26331

would cause rates to deviate from an Judges ‘‘shall base their decision on be said to be ‘willing’ partners to an ‘‘effectively competitive’’ level.66 economic, competitive and agreement if they are coerced to agree to Thus, to the extent the ‘‘shadow of programming information presented by a price through the exercise of antitrust law’’ has receded, it was the parties . . .’’ 17 U.S.C. 114(f)(2)(B) overwhelming market power.’’ Web II at counterbalanced by the ‘‘shadow of the (emphasis added). Accord, 17 U.S.C. 24091. Sirius XM emphasizes other statutory rate.’’ Accordingly, the 112(e)(4) (regarding ephemeral licenses). particular language from Web II, which presence of the so-called statutory Several previous decisions by the D.C. states: ‘‘An effectively competitive shadow appears to reflect a trade-off and Circuit, the Librarian, the Judges and the market is one in which super- a second-best solution, rather than a CARP (in Web I) have discussed the competitive prices or below-market distortion of an effectively competitive concept of ‘‘effective competition’’ and prices cannot be extracted by sellers or marketplace. its relationship to § 114(f)(2)(B). buyers . . . .’’ 72 FR at 24091. Additionally, the Judges’ SoundExchange and the Services The NAB emphasizes that in the consideration of the Pandora/Merlin disagree as to whether § 114(f)(2)(B) and present proceeding the Judges must Agreement and the iHeart/Warner prior decisions require the Judges to set follow these decisions because 17 U.S.C. Agreement as appropriate benchmarks a rate that reflects an ‘‘effectively 803(a)(1) expressly requires the Judges for the ad-supported (free-to-the- competitive’’ market populated by to act in accordance with the Librarian listener) market obviates the supposed willing buyers and willing sellers. of Congress’s interpretation. NAB ‘‘shadow’’ problem. In both SoundExchange argues that no authority PFFCL ¶ 689. The Services also rely on benchmarks, the rate is below the allows for such a requirement, while the a decision by the D.C. Circuit as otherwise applicable statutory rates. The Services assert that the statute and prior persuasive, if not binding precedent, statutory rates did not cast a shadow decisions require the Judges to set rates because it states that § 114(f)(2)(B) ‘‘does that negatively affected the licensors in that would be established an not require that the market assumed by those agreements because (as noted ‘‘effectively competitive’’ market.68 the Judges achieve metaphysical infra) they voluntarily agreed to rates The Services construe § 114(f)(2)(B) as perfection in competitiveness.’’ below the applicable statutory rates (in explicitly requiring the Judges to utilize Intercollegiate Broad. Sys., Inc. v. exchange for the steering of more plays), competitive information introduced in Copyright Royalty Board, 574 F.3d 748, rather than defaulting to the higher evidence to set a marketplace rate that 757 (D.C. Cir. 2009) (emphasis added). statutory rate. reflects ‘‘effective competition,’’ and to Apparently, the Services construe the Further, in the subscription market adjust an otherwise appropriate use of the adjective ‘‘metaphysical’’ to the Judges have adopted the benchmark in order to reflect ‘‘effective require, or at least suggest, that the rates SoundExchange benchmark approach, competition.’’ In support of this reflect some lesser yet nonetheless which analogizes between the position, the Services make several effective quantum of competition. interactive and noninteractive markets. principal arguments. The Services further argue that the As Dr. Rubinfeld testified, the The Services assert that prior legislative history of Section 114 reflects interactive contracts on which he relied decisional law constitutes precedent a Congressional intention for rates to be for his subscription-based benchmark that requires the Judges to set rates that set at a level that avoids ‘‘higher-than- ‘‘minimize[] the effect of the statutory are ‘‘effectively competitive.’’ They competitive prices.’’ See 141 Cong. Rec. shadow’’ because the interactive point to the most recent determination S11945–04, S11962 (1995). In similar services cannot default to the statutory by the Judges, the Web III Remand, in fashion, according to the Services, the rate. Rubinfeld CWDT ¶ 18. which the Judges approvingly cited and legislative history makes it plain that Finally, the Judges emphasize that relied upon the language in prior the willing buyer/willing seller standard they find the ‘‘shadow’’ criticism to be decisions by the Librarian in Web I and in § 114 was intended to direct the both nihilistic and self-contradictory. If the Judges in Web II regarding the need CARP (now the Judges) ‘‘to determine the ‘‘shadow’’ infects all benchmarks so to set rates under § 114(f)(2)(B) that reasonable rates and terms.’’). H.R. Rep. as to disqualify that method of rate- reflect those that would be set in an No. 105–796 at 86 (Conf. Rep.); see H.R. setting, then the parties would need to ‘‘effectively competitive market.’’ Web Rep. No. 104–274 at 22 (1995) adjust or abandon their benchmarking III Remand at 23114 n. 37. The NAB (legislative history of DPRSRA expressly strategies and develop new bases for further notes that in Web II, the Judges provides ‘‘[i]f supracompetitive rates are analysis. That could mean the wholesale held that ‘‘neither sellers nor buyers can attempted to be imposed on operators, abandonment of benchmarking, to be the copyright arbitration royalty panel replaced by a valuation approach yet to 68 As discussed in more detail in this can be called on to set an acceptable be applied and accepted in these determination, SoundExchange asserts that its rate.’’). In this regard, the Services note interactive benchmark need not be reflective of an proceedings.67 that the Department of Justice’s ‘‘effectively competitive’’ market because such a objection to an earlier draft of the requirement is not contained within section F. The Legal Issue of Whether Effective 114(f)(2)(B). SoundExchange also argues that, statute, relating to whether the record Competition Is a Required Element of assuming an ‘‘effectively competitive’’ market companies could negotiate exclusively the Statutory Rate standard is part of the statutory scheme, its through a common agent, was resolved interactive benchmark is a product of effective because the ratemaking body (now the The statutory language that includes competition. The Services argue that their the ‘‘willing buyer/willing seller respective proposed benchmarks reflect rates that Judges) could intercede and establish language also commands that ‘‘[i]n have been set in an ‘‘effectively competitive’’ reasonable rates. 141 Cong. Rec. S. determining such rates . . . the . . . market, unlike SoundExchange’s proposed 11,962–63 (daily ed. Aug. 8, 1995) interactive benchmark that is the product of a (Letter from Assistant Attorney General market lacking the necessary competitive features. 66 The issue of ‘‘effective competition’’ is iHeart and Pandora each maintains that, even Andrew Fois to Hon. Patrick Leahy, July discussed at length, infra. assuming that the statute does not contain an 21, 1995, noting that any 67 As explained elsewhere in this determination, ‘‘effectively competitive’’ market standard, their noncompetitive rates created by the the Judges have rejected the non-benchmarking respective benchmarks are nonetheless appropriate, existence of only a single collective approaches to rate setting proposed by some parties because they represent the rates to which willing in this proceeding. They were not rejected because sellers and willing buyers would agree in the could be corrected by the ‘‘rate panel.’’). they were not benchmarks, but because each was market, notwithstanding whether those rates reflect The Services also note that, in unpersuasive in its own right. ‘‘effective competition.’’ comparable circumstances, courts

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00017 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26332 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

construe ‘‘reasonable rates’’ to be those value’’ as SoundExchange notes, as ‘‘the . . . .’’ 17 U.S.C. 114(f)(2)(B) (emphasis ‘‘rates that would be set in a competitive price at which the property would added); accord, 17 U.S.C. 112(e)(4) market.’’ ASCAP v. Showtime/The change hands between a willing buyer (identical language for the setting of Movie Channel, Inc., 912 F.2d 563, 576 and a willing seller, neither being under rates for the ephemeral license). The (2d Cir. 1990); see also NAB PFFCL any compulsion to buy or sell and both D.C. Circuit has expressly noted that, by ¶¶ 706–709 (and cases cited therein); In having reasonable knowledge of this specific language, ‘‘Congress re Pandora Media, Inc., 6 F. Supp. 3d relevant facts.’’ United States v. required the Judges to follow certain 317, 353–54 (S.D.N.Y. 2014), aff’d sub Cartwright, 411 U.S. 546, 551 (1931). statutory guidelines’’ one of which is nom. Pandora Media, Inc. v. ASCAP, Fourth, SoundExchange argues that that ‘‘the Judges must ‘base [their] 785 F.3d 73 (2d Cir. 2015). statutory enactments of the fair market decision on . . . competitive ... Finally, the NAB asserts that the value test and its willing buyer/willing information presented by the parties.’ ’’ statutory histories of the DPRA and the seller component constitute adoptions Intercollegiate Broad. Sys., Inc. v. DMCA reflect a Congressional intent to of a recognized common law definition Copyright Royalty Board, 574 F.3d 748, create a three-tier performance right/rate of the test. Therefore, the common law 753 (D.C. Cir. 2009). structure, whereby: (1) Terrestrial radio meaning should prevail because it is a SoundExchange invites the Judges to continues to enjoy free access to sound ‘‘settled principle of statutory ignore this statutory directive and recordings; (2) interactive services must construction that, absent contrary judicial command. The Judges cannot. pay market-negotiated royalties in order indications, Congress intends to adopt a The parties presented the Judges with to play sound recordings on demand; common law definition of statutory voluminous evidence and testimony and (3) noninteractive services, falling terms. United States v. Shabani, 513 comprising the required ‘‘competitive between these two extremes, cannot U.S. 10, 13 (1994); see also United information’’ relating to Dr. Rubinfeld’s play sound recordings for free, shall not States v. Wells, 519 U.S. 482, 491 (1997) proposed interactive benchmark market, to be subjected to the purely market (same). the Services’ proposed noninteractive rates paid by on-demand interactive Fifth, SoundExchange points out that, benchmarks, the noninteractive market services and, instead, shall pay when Congress intends a legal standard at issue in this proceeding, and the intermediate rates set by the Judges to be based on ‘‘effective competition,’’ alleged differences and similarities (formerly the CARP arbitrators subject to it makes the point expressly and among them.69 The Judges are Librarian review). See NAB ¶¶ 678 et explicitly defines ‘‘effective commanded by the statutory language seq.; 682 et seq. (and authorities cited competition.’’ Cf. 47 U.S.C. 543(1)(1) quoted above to ‘‘base their decision’’ therein). (defining ‘‘effective competition’’ in the on precisely this sort of information, On the other hand, SoundExchange Cable Television Consumer Protection and, as Intercollegiate Broadcast System construes § 114(f)(2)(B) as precluding and Competition Act of 1992). makes plain, it would be legal error for the Judges from adjusting an otherwise Sixth, SoundExchange characterizes the Judges to ignore this statutory appropriate benchmark in order to the references to effective competition directive. reflect ‘‘effective competition.’’ In in Intercollegiate Broad. Sys. and Web I The Judges further conclude that, support of this position, as mere dicta that may be ignored by the even if the directive that they ‘‘shall’’ SoundExchange makes several principal Judges. consider competitive information could arguments. Seventh, SoundExchange asserts that be construed as ambiguous, their First, SoundExchange emphasizes any attempt to apply an ‘‘effective consideration of ‘‘competitive that the words ‘‘effective competition’’ competition’’ requirement would render information’’ is certainly a permissible, or the like are not included within the the statutory test indeterminate, reasonable, and rational application of statute. Thus, SoundExchange unworkable, and vague. SoundExchange § 114, for a number of reasons. maintains that the plain language of the notes that the Services’ economic First, the D.C. Circuit, the Librarian, statute clearly does not include such a experts acknowledged the absence of a the Judges, and the CARP have all standard. SX PCOL ¶ 21. ‘‘bright line’’ separating a market that is acknowledged that the Judges can and Second, SoundExchange relies upon a ‘‘effectively competitive’’ from one that should determine whether the proffered statement by the CARP in Web I that is not. Moreover, SoundExchange rates reflect a sufficiently competitive ‘‘the willing buyer/willing seller asserts that there is no evidence or market, i.e., an ‘‘effectively competitive’’ standard is the only standard to be testimony setting forth what the level of market. The Judges made this point applied.’’ In re Digital Performance rates would need to be in clearly in their decision in the Web III Right in Sound Recordings and SoundExchange’s proffered interactive Remand, which included a summary of Ephemeral Recordings, No. 2000–9 benchmark market, in order for it to the past decisional language regarding CARP DTRA 1&2 at 21 (Feb. 20, 2002), equate with ‘‘effectively competitive’’ the § 114 standard: appv’d and modif’d by Librarian, 67 FR rates. 45240 (July 8, 2002) (Web I). Having considered the issue and the The D.C. Circuit has held that this statutory section does not oblige the Judges to set rates SoundExchange construes this language parties’ positions, the Judges conclude by assuming a market that achieves as confirming the exclusion of the that they are required by law to set a ‘‘metaphysical perfection and ‘‘effectively competitive’’ condition rate that reflects a market that is competitiveness.’’ Intercollegiate Broad. Sys., from the ‘‘willing buyer/willing seller’’ effectively competitive. The Judges Inc. v. Copyright Royalty Board, 574 F.3d marketplace standard. reach this conclusion through a Third, SoundExchange argues that the consideration of the plain meaning of 69 The ‘‘competitive information’’ provided by the ‘‘willing buyer/willing seller’’ standard the statute, the clear statutory purpose, parties was extensive. SoundExchange and the Services provided factual and expert testimony is essentially a restatement of the applicable prior decisions, and the regarding: (1) The ‘‘upstream’’ market (in which traditional ‘‘fair market value’’ test. See relevant legislative history. streaming services acquire licenses from the record id. at 45244 (the Librarian’s Web I The Judges’ starting point is the companies); (2) the ‘‘downstream’’ market (in which decision notes that the statutory language of the statute itself. The statute streaming services may (or may not) compete with each other for listeners); (3) the horizontal standard requires rates that reflect requires that the Judges ‘‘shall base their ‘‘upstream’’ market (where the record companies ‘‘strictly fair market value’’). The decision on [inter alia] competitive . . . compete (or fail to compete) with each other; and Supreme Court has defined ‘‘fair market information presented by the parties (4) the interactions of these several markets.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00018 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26333

748, 757 (D.C. Cir. 2009). Rather, as the that an effective level of competition Living in Nicaragua v. Reagan, 859 F.2d Librarian of Congress held in Web I, the was required for the Judges to adopt 929, 938–39 (D.C. Cir. 1988). ‘‘willing seller/willing buyer’’ standard calls those benchmarks; and (2) the facts of Accordingly, although SoundExchange for rates that would have been set in a assets that the statements relating to an ‘‘competitive marketplace.’’ 67 FR at 45244– the case that demonstrated the 45 (emphasis added); see also Web II, 67 FR sufficiently competitive nature of those effectively competitive market in the at 24091–93 (explaining that Web I required benchmarks.71 That legal conclusion D.C. Circuit’s Intercollegiate Broadcast an ‘‘effectively competitive market’’ rather and that factual finding led the Judges System decision and the Librarian’s Web than a ‘‘perfectly competitive market.’’ to an application of law to fact whereby I decision were dicta, the Judges in Web (emphasis added)). Between the extremes of they concluded that the proposed II, the Web III Remand and the present a market with ‘‘metaphysically perfect benchmarks were reflective of an proceeding were all clearly able to competition’’ and a monopoly (or collusive effectively competitive market and convert such asserted dicta into binding oligopoly) market devoid of competition holdings. there exists ‘‘[in] the real world . . . a mind- therefore satisfied the § 114(f)(2)(B) boggling array of different markets,’’ standard. Specifically, the Judges held Thus, the Judges conclude that they Krugman & Wells, supra, at 356, all of which in the Web III Remand: are bound to follow the prior directives that instruct them to make certain that possess varying characteristics of a An oligopolistic marketplace rate that did ‘‘competitive marketplace.’’ approximate the monopoly rate could be the statutory rates they set are those that Web III Remand, 79 FR at 23114 n. 37. inconsistent with the rate standard set forth would be set in a hypothetical It is noteworthy that SoundExchange in 17 U.S.C. 114(f)(2)(B), as that standard has ‘‘effectively competitive’’ market. In has not characterized the Web III been set forth by the D.C. Circuit and the light of this conclusion, based on the Remand decision as dicta. Thus, even if Librarian of Congress. . . . [I]n this foregoing reasons, the remainder of the proceeding the evidence demonstrates that the prior language on which the Web III arguments are insufficient to alter the sufficient competitive factors exist to permit Judges’ decision in this regard. Remand Judges had relied was dicta, the [benchmarks] to serve as useful there is no argument that the holding in However, in the interest of benchmarks, and does not demonstrate that completeness, the Judges address other the Web III Remand was dicta. It is also the rates in the [benchmarks] approximated arguments, including those raised by the noteworthy that SoundExchange did not monopoly rates. parties, that further support their assert that the holding in Web II, that an * * * * * conclusion. excess of market power can preclude a The parties presented no evidence from The Judges agree that the legislative finding that a buyer or seller was a which the Judges could conclude . . . that history supports the conclusion that 70 SoundExchange necessarily wielded a level ‘‘willing’’ participant, was dicta. § 114 directs the Judges to set rates that In Web III, a licensee, Live365, asked of pricing power sufficient to affect the use reflect the workings of a hypothetical the Judges to reject certain of of the WSA Agreements as benchmarks. effectively competitive market. The SoundExchange’s proposed benchmarks 79 FR at 23114 (emphasis added). Thus, legislative history equates rates set that were based on the Webcaster in the Web III Remand, the Judges under the willing buyer/willing seller Settlement Act (WSA) agreement unequivocally applied the prior standard with ‘‘reasonable rates.’’ As the between SoundExchange and the NAB, pronouncements of the D.C. Circuit, the Services note, the phrase ‘‘reasonable and the WSA agreement between Librarian, and the Judges to render an rates’’ has been construed by the rate SoundExchange and Sirius XM. (The unambiguous holding: (1) Adopting a court, in an analogous context, as ‘‘rates parties to those agreements agreed to competitiveness standard; (2) applying that would be set in a competitive allow those WSA agreements to be the facts to the competitiveness market.’’ introduced as evidence in Web III.) standard; and (3) using that application The Judges are informed by the Live365 argued ‘‘the rates . . . reflect of facts to law to reach their judgment. analogous use of the willing buyer/ the monopoly power of a single seller in Alternately stated (and applying the willing seller standard in eminent those two contracts.’’ 79 FR at 23113. D.C. Circuit’s Lawson definition of dicta domain law. See, e.g., Kirby Forest Ind., The Judges rejected that argument and quoted supra), this decision regarding Inc. v. U.S., 467 U.S. 1, 10 (1984) did so by taking a ‘‘decisional path’’ of ‘‘effective competition’’ in the Web III (applying willing buyer/willing seller reasoning based on: (1) A conclusion Remand was necessary to determine an test in eminent domain valuation issue raised in the proceeding (the dispute). In such cases, the courts must 70 Not only did SoundExchange fail to assert that effectively competitive status of the consider whether to award a forced the Web III Remand decision regarding ‘‘effective WSA settlement agreements), after competition’’ was dicta, that decision could not seller the ‘‘holdout’’ value of the seller’s possibly be construed as dicta. The distinction argument and full consideration. parcel, an additional value that exists between a holding and dictum has been thoroughly Moreover, even past dicta ‘‘deserves solely because the seller’s property is a analyzed and succinctly stated: serious consideration’’ in subsequent necessary complement to the other A holding consists of those propositions along the decisions when ‘‘sufficiently properties that are needed by the chosen decisional path or paths of reasoning that persuasive.’’ U.S. v. Libby, 475 F. Supp. (1) are actually decided, (2) are based upon the facts governmental unit. As discussed in of the case, and (3) lead to the judgment. If not a 2d 73, 81 (D.D.C. 2007). Thus, detail infra, it is precisely this holding, a proposition stated in a case counts as ‘‘persuasive dictum in an important complementary oligopoly value that the dicta. early case [can] establish[ ] [a] Judges are declining to include in the M. Abramowicz and M. Stearns, Defining Dicta, principle’’ to be followed by other 57 Stan. L. Rev. 953, 961 (2005). Courts have long statutory rate based upon their analyses held that, in contrast with a ‘‘holding,’’ dicta as courts. Committee of U.S. Citizens of the parties’ benchmarks proffered in ‘‘language unnecessary to a decision, ruling on an this proceeding. Cf. Thomas Miceli and issue not raised, or [an] opinion of a judge which 71 Both Sirius XM and the NAB assert in the C.F. Sirmans, The Holdout Problem, does not embody the resolution or determination of present proceeding that those two WSA settlement the court, . . . made without argument or full agreements were not reflective of effective Urban Sprawl and Eminent Domain, 16 consideration of the point.’’ Lawson v. U.S., 176 competition, based on evidence they have J. Housing Econ. 309, 314 (2006) F.2d 49, 51 (D.C. Cir. 1949). As detailed in the text, presented in this proceeding but was not presented (‘‘complementarities among properties a consideration of the pertinent ruling in the Web in Web III. That issue is addressed infra, but, for in the assembly case that are not present III Remand and of the ultimate decision in the Web present purposes, the pertinent point is that the III Remand itself, demonstrates that the statements Judges found on the Web III record that these WSA in the individual transaction’’ are the regarding the necessary competitive state of the settlement agreements reflected an effectively consequence of ‘‘market failure,’’ market were clearly holdings rather than dicta. competitive market. economic ‘‘rent seeking’’ and generate

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00019 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26334 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

inefficient ‘‘transaction costs’’) the Judges consider ‘‘competitive marketplace. That is, to draw on a (emphasis added). information,’’ the prior judicial and classic analogy, the experts testified to The Judges are also persuaded that the administrative holdings and different aspects of the market in much structure of the Act with regard to the pronouncements, and the legislative the same manner as the several sound recording performance right—as history all combine to clearly provide proverbial blind men 73 who, after it relates to terrestrial radio, more than ‘‘indications’’ that the Judges touching but one part of an elephant, noninteractive services, and interactive must set reasonable rates that reflect were asked to describe the animal, and services—confirms the necessity of ‘‘effective competition.’’ gave starkly different descriptions based adopting an ‘‘effectively competitive’’ Third, the mere fact that, in another upon whether they had touched only standard in the rate-setting process. setting (regarding the cable television the trunk, the torso or the tail. Perhaps Copyright owners were provided a industry) Congress chose to define an even more apt analogy has been limited performance right with regard to ‘‘effective competition’’ hardly suggests made with regard to the testimony of the use of their sound recordings by that such an ‘‘effective competition’’ experts as similar to the men in another noninteractive services—something less standard does not exist in the present fable: than the purely private market-based case. Indeed, the absence of a definition, In a certain kingdom was a cave containing rate for interactive use, but clearly more combined with the requirement that the a treasure, guarded by a beast of fierce repute. than the ‘‘zero rate’’ required from Judges weigh ‘‘competitive The king wished to know the nature of the terrestrial radio. The Judges conclude information,’’ is more consistent with beast, and dispatched three of his subjects to that a rate that simply reflected or the idea that Congress intended to invade the pitch darkness of the cave and overemphasized either of the polar delegate discretion to the Judges to report. The first returned and declared that extremes would be inconsistent with the determine whether the rates they set he had felt the head of the beast, and it was three-tier structure of the statute.72 As reflected an appropriate level of toothed and maned like a lion. The second the Services note, if the Judges were competitiveness. reported that he had felt the sides of the simply to apply the competitive beast, and that it was winged and feathered Finally, the Judges reject like an eagle. The third reported that the legs dynamics of the interactive market, they SoundExchange’s assertion that there is of the beast were long and hoofed like a would be disregarding the particular no pre-existing ‘‘bright line’’ test horse. A fearsome portrait of the beast was statutory history that led to the three- sufficient to distinguish a rate which is drawn up, and all were thereafter afraid to tier rate structure. See generally, ‘‘effectively competitive’’ from one that approach the cave. Of course, in reality, the William W. Fisher III, Promises to Keep is not. The very essence of a competitive cave contained a lion, an eagle, and a horse. at 104–05 (2004) (different statutory standard is that it suggests a continuum * * * * * treatment of terrestrial radio, interactive and differences in degree rather than in Another, less allegorical, way of saying this services, and noninteractive services kind. Once again, the statutory charge is that many of the problems that the law has based upon fundamental ability and that the Judges weigh ‘‘competitive had in handling expertise in the courtroom limits regarding the performance, information’’ indicates that the Judges have sprung from a failure to examine the promotion of, and substitution for are empowered to make judgments and concept of expertise in appropriate taxonomic detail. sound recordings). decide whether the rates proposed SoundExchange’s arguments to the adequately provide for an effective level Michael Risinger, Preliminary Thoughts contrary are unavailing. First, the fact of competition. Moreover, in the present on a Functional Taxonomy of Expertise that the statute requires the Judges to case, the Judges were presented with for the Post-Kumho World, 31 Seton consider ‘‘competitive information’’ highly specific facts regarding how to Hall L. Rev. 508, 508–09 (2000). adequately rebuts SoundExchange’s use the impact of steering on rate setting This phenomenon among experts has contention that the statutory language in order to measure and account for the particular applicability to economists. does not address the issue of ‘‘complementary oligopoly’’ power of As one prominent economist has competitiveness. That provision, the Majors that serves to prevent recently written: combined with the legislative history effective competition. Rather than a single, specific model, and the prior judicial and economics encompasses a collection of administrative pronouncements make it IV. Commercial Webcasting Rates models . . . . The diversity of models in clear that the statutory language requires A. Analyses and Findings economics is the necessary counterpart to the the Judges to establish rates that are flexibility of the social world. Different social The rates proposed by the Services effectively competitive. settings require different models. Economists Second, the Judges do not find that and SoundExchange are marked by a are unlikely ever to uncover universal, the traditional fair market value test wide disparity. Although it is general-purpose models. But . . . economists have a tendency to misuse their models. permits the Judges to ignore the unsurprising that adverse parties would have strikingly different positions, what They are prone to mistake a model for the competitive status of the hypothetical model, relevant and applicable under all market in which the statutory rate is is surprising is that, despite these differences, the parties’ positions are conditions. Economists must overcome this established. As SoundExchange temptation. concedes in the very case law that it supported to a great extent (but not in all cases) by persuasive and logical Dani Rodrik, Economics Rules 5–6 quotes, the common law meaning of a (2015) (emphasis in original). Each party phrase should only prevail when economic analyses. Initially, this created a conundrum for the Judges, and its experts nonetheless invite the construing a statute ‘‘absent contrary Judges to rely on but a single economic indications.’’ Here, the requirement that because none of these persuasive and logical economic analyses could easily model—their model—as representative of the entire noninteractive market. As 72 As discussed infra, the Judges also reject rates be rejected. proposed by several of the Services that attempt to On closer inspection, however, what this determination makes clear, the use the ‘‘zero rate’’ paid by terrestrial radio as a became clear to the Judges was that the guide in this proceeding. The rejection of such reason why many of these disparate 73 The analogy is not meant to suggest that the proposals can be seen as a bookend to the Judges’ testifying experts were metaphorically blind. requirement that the statutory rate reflect effective economic analyses and models could all Indeed, they were all learned and persuasive with competition, rather than the complementary appear to be correct was that they each regard to the aspects of the market upon which they oligopoly power present in the interactive market. reflected only a portion of the opined.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00020 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26335

Judges decline that invitation. Rather, and has been amplified by technological rates by analyzing more than 80 the Judges have found that no single changes that have allowed for a greater agreements between interactive economic model—no one mythic diversity of music services. The streaming services and record beast—reigns over the noninteractive directive in § 114, instructing the Judges companies. Dr. Rubinfeld identified 60 market writ large. Rather, the evidence to establish ‘‘rates and terms,’’ that is, such agreements that contained data on and testimony reveal a marketplace for multiple rates and terms, anticipates the per-play royalty rates. 5/28/15 Tr. 6297 sound recordings that is segmented, if potential for more than one set of rates (Rubinfeld). From those 60 agreements, not fragmented. Indeed, the Judges note and terms that would have been he selected 26 that specified minimum the economic dichotomies demonstrated negotiated in the marketplace between per-play rates. Rubinfeld CWDT ¶ 205; by the evidence: various willing buyers and willing SX Ex. 59 (Rubinfeld CWDT, Exhibit • Market Segmentation by WTP sellers. Because the marketplace as 16a) (listing 26 interactive streaming Services that attract listeners who presented by the record in this service agreements). proceeding reveals important have no willingness to pay (WTP) for According to Dr. Rubinfeld, differences across these dichotomies, access to a noninteractive service, and interactive streaming service therefore who listen mainly to ad- the Judges, as required by § 114, establish rates and terms in this benchmarks are more probative in this supported services, versus services that statutory rate proceeding than they were attract relatively more listeners who proceeding that reflect those marketplace realities. in prior statutory rate proceedings due have a WTP greater than zero, and to: (1) A ‘‘convergence’’ in features that therefore can attract more subscription- B. SoundExchange’s Rate Proposal interactive and noninteractive streaming based listeners. 1. Introduction services offer to the end-user • Market Segmentation by On- (‘‘downstream’’) market; and (2) greater Demand Functionality SoundExchange proposes a single rate head-to-head competition for listeners for all commercial webcasters using a Services that meet the statutory between interactive and noninteractive greater-of structure. All commercial definition of an ‘‘interactive service’’ streaming services. Rubinfeld CWDT webcasters would pay the greater of and thus provide an on-demand ¶ 21. 55% of revenue attributable to function, i.e., that allow listeners to webcasting and the following per- select the sound recording they wish to i. Convergence of Features performance rate: hear whenever they choose, versus SoundExchange avers that the noninteractive services, that—despite SOUNDEXCHANGE PROPOSED PER- listening choices (i.e., functionality) that whatever other functionality they may PERFORMANCE RATES interactive and noninteractive streaming include—do not and cannot provide an services offer their customers are on-demand feature. Per- becoming much more similar than they • Market Segmentation by Major or Year performance were in previous years, i.e., they are Indie rate converging. See, e.g., 5/6/15 Tr. 2013 The Majors, who have the ability to 2016 ...... $0.0025 (Rubinfeld) (‘‘[C]onvergence [m]ean[s] negotiate relatively higher rates, versus 2017 ...... 0.0026 that if I’m very active in telling Pandora the Indies, who have relatively less 2018 ...... 0.0027 [a noninteractive service] what I like market power when negotiating rates. 2019 ...... 0.0028 and don’t like, the nature of the station 2020 ...... 0.0029 • Complementary Oligopoly Power can evolve in ways that can become versus Oligopoly Market Structure more similar to what I might do on SoundExchange Rate Proposal at 2–3. Spotify [an interactive service] if I were ‘‘Complementary oligopoly’’ power exercised by the Majors designed to 2. Dr. Rubinfeld’s Proposed Interactive curating my own station.’’). thwart price competition and thus Streaming Services Benchmark According to SoundExchange, the inconsistent with an ‘‘effectively In support of its proposal, increasingly similar functionality of competitive market,’’ versus the Majors’ SoundExchange relies principally on an interactive and noninteractive streaming non-complementary oligopolistic analysis undertaken by one of its services has ‘‘blurred’’ the previous structure not proven to be the economic witnesses, Dr. Daniel distinctions between them. See, e.g., SX consequence of anticompetitive acts or Rubinfeld, of rates set forth in direct Ex. 3, ¶ 13 (Blackburn WDT); SX Ex. 32, the cause of anticompetitive results. licenses from record companies to ¶ 25 (Wilcox WRT). This purported • Custom Pureplay Webcasting versus certain interactive streaming services.74 blurring has occurred, according to SoundExchange, because of Simulcasting a. Foundation for Rubinfeld’s Proposed Custom (Pureplay) noninteractive Per-Play Rates Benchmark technological evolution, marketplace services that play only sound developments, and changes in consumer Dr. Rubinfeld derived preferences. See, e.g., Kooker WDT at recordings, versus simulcasters, who SoundExchange’s proposed per-play play principally (but not exclusively) 16; SX Ex. 21 ¶ 36 (Wheeler WDT). the sound recordings and other SoundExchange asserts that, because of 74 An ‘‘interactive service’’ is defined as one that the market changes that it has materials transmitted simultaneously on ‘‘enables a member of the public to receive a terrestrial broadcast. transmission of a program specially created for the highlighted, interactive and The presence of such dichotomies is recipient, or on request, a transmission of a noninteractive webcasters alike not particularly unusual. For example, particular sound recording . . . which is selected by recognize that any given music the recipient.’’ 17 U.S.C. 114(j)(7) (emphasis added). consumer ‘‘is both a lean forward and a in Web II, the Judges noted that the A service that fails to meet the definition of an marketplace consisted of a variety of ‘‘interactive service’’ is, by default, a noninteractive lean back type of listener,’’ whose commercial actors, who had a service that may be entitled to a statutory license particular preference ‘‘depends very heterogeneous mix of features regarding if it meets all other applicable criteria, see 17 U.S.C. much on the situation and the time of 114(d)(2)(C), including adherence to the ‘‘sound day’’ and the ‘‘mood that they’re in.’’ 5/ costs, customers, business plans, and recording performance complement’’ as defined in strategies. Such a variety exists today, 17 U.S.C. 114(j)(13). 29/15 Tr. at 6570 (Kooker); Kooker

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00021 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26336 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

WRT.75 SoundExchange further notes Wood WDT); Rubinfeld CWDT ¶¶ 53– • create stations that contain only that even Pandora has recognized that 54; 5/13/15 Tr. 3444 (Herring). Further, those songs for which the listener has for 75% of music consumers it is SoundExchange notes that a Pandora indicated a preference. SX Ex. 213. important that a music service afford listener can ‘‘seed’’ multiple stations SoundExchange notes that a prime them both ‘‘effortless listening’’ and ‘‘on with various artists and sound recording catalyst for increased convergence demand music.’’ SX Ex. 269 at 17 tracks, and then influence the types of between interactive and noninteractive (Pandora Board of Directors: Strategy recordings on each station by using streaming services is the trend away Day document, Oct. 30, 2014). Pandora’s ‘‘thumbs up/thumbs down’’ SoundExchange contends that to button. PAN Ex. 5000 ¶¶ 33–34 from desktop listening toward mobile attract and retain listeners, interactive (Westergren WDT); Fleming-Wood WDT listening. For example, SoundExchange streaming services have moved beyond ¶¶ 8–9; Blackburn WDT ¶¶ 9, 12–13; points out that during the first quarter merely playing, on demand, the Rubinfeld CWDT ¶ 53; Kooker WRT of 2015, 83% of the hours streamed by recordings selected by a listener, and ¶¶ 10–11. SoundExchange continues Pandora listeners occurred through have developed and promoted curated that Pandora listeners can also skip mobile devices. 5/13/15 Tr. 3443 playlists, radio components and other songs, another form of customization. (Herring). SoundExchange asserts that lean-back methods of music delivery. Rubinfeld CWDT ¶ 53. the leading edge of this competition to Blackburn WDT ¶ 13; Wilcox WRT ¶ 25; SoundExchange also points out that ‘‘get into the car’’ by both noninteractive Kooker WRT at 14; 5/13/15 Tr. 3448–50 Sirius XM’s noninteractive steaming and interactive streaming services (Herring). To support this point, service (‘‘My Sirius XM’’) allows should hasten this trend. 5/8/15 Tr. SoundExchange introduced evidence listeners to move ‘‘sliders’’ to change the 2731–32 (Shapiro). Moreover, because and elicited testimony describing the type of music played. For example, a on-demand song selection is often various custom radio features of several listener can direct the service to play incompatible with driving (absent predominantly interactive streaming ‘‘more acoustic’’ or ‘‘more electric’’ hands-free voice controls or self-driving services, e.g., Rdio; Rhapsody; Slacker; within a particular genre. SX Ex. 232 at cars), SoundExchange opines that Beats; Amazon; Google; and Apple. See 15–21; 5/22/15 Tr. 5419–20 (Frear). interactive streaming services have SX PFF ¶ 266 (and record citations therein). SoundExchange also notes that iHeart incentives to add ‘‘lean-back’’ SoundExchange asserts that ‘‘lean has developed a custom streaming functionality, such as Spotify’s back’’ features are a significant part of service that, according to ‘‘Shuffle’’ service, to their mobile the consumer listening experience on SoundExchange, makes it ‘‘very likely’’ services. Blackburn WDT ¶ 39. some of these services. For example, that a listener who is seeking out a Based on the foregoing points, SoundExchange points out that nearly highly popular artist or song will ‘‘hear SoundExchange concludes that, [REDACTED]% of UMG’s plays on the exact song or songs he or she had notwithstanding the requirements Slacker are such programmed streams, in mind within minutes of starting the noninteractive streaming services must station.’’ Kooker WRT at 7.76 rather than the traditional on-demand meet to be eligible for the statutory plays of an interactive service. SX Ex. 25 SoundExchange also notes that the license, statutory services are ¶ 11 (Harrison WRT). SoundExchange statutory services are developing new increasingly offering enhanced notes that on Spotify, approximately functionality that would allow even functionality that ‘‘come[ ] close to [REDACTED]% of total listening to more listener control (while still replicating’’ the on-demand listening Sony’s repertoire occurs through satisfying the DMCA requirements).77 experience of interactive streaming playlists created by Spotify or other These functions purportedly would services. Rubinfeld CWDT ¶¶ 53–54; third parties (i.e., not the listener). allow listeners to: Kooker WRT ¶ 15. • Blackburn WDT ¶ 9; Kooker WDT at 16. Repeat songs, re-listen to songs As summarized by one record company SoundExchange further asserts that they’ve ‘‘thumbed up,’’ skip additional listener feature convergence is occurring witness, statutory services now ‘‘employ tracks, and create playlists of ‘‘thumbed sophisticated algorithms, user-interface from the other direction as well, with up’’ songs, SX Ex. 1678 at 8; controls, and other computer technology statutory services adding new ‘‘lean- • forward’’ options. In May 2013, ban from stations certain artists, that allow users to communicate their SoundExchange notes, Pandora, a live tracks, instrumental recordings and preferences to the service, and the noninteractive streaming service, tempos, SX Ex. 269 at 43; 5/13/15 Tr. service to customize and curate initiated its ‘‘Pandora Premieres’’ 3498–3503 (Herring); and programming tailored to the individual feature, which ‘‘allows for on-demand user.’’ Kooker WDT at 16–17. selection of certain predetermined 76 To demonstrate this point, SoundExchange introduced evidence of several experiments that SoundExchange concludes that ‘‘[i]t is albums.’’ Pan. Ex. 5002 ¶ 30 (Fleming- purported to show the high frequency with which therefore no longer just directly licensed an iHeart station played the most popular songs of interactive services that allow users to 75 ‘‘Lean-forward’’ and ‘‘lean-back’’ are not a popular artist who was used to seed a custom statutory phrases that define types of services, and station—in contrast to the uncertain song rotation select their programming. Users of the record does not reflect any precise meanings in on terrestrial radio. Kooker WRT at 7–8. In these statutory services can also lean forward the industry. Importantly, a ‘‘lean-forward service’’ experiments on iHeart’s custom radio (i.e., non- and influence what they hear.’’ SX PFF is not necessarily the same as an ‘‘interactive simulcast), a seeded popular artist, Meghan Trainor, ¶ 278 (emphases added).78 service,’’ and a ‘‘lean-back service’’ is not and her current highest selling song, would play necessarily the same as a ‘‘noninteractive service.’’ first 92% of the time. Ms. Trainor’s first or second Compare, e.g., 4/30/15 Tr. 1182–83 (A. Harrison) current highest selling song would play first 100% 78 The words ‘‘select’’ and ‘‘influence’’ as used by (‘‘on-demand services have lean-back listening of the time. In 68% of the trials in the experiment, SoundExchange and quoted in the accompanying options’’ and ‘‘statutory [noninteractive] services the seeded station played three or more of Ms. text, supra, are italicized to foreshadow the have lean-forward capabilities.’’) with 5/13/15 Tr. Trainor’s songs among the first seven songs played. important distinction in meaning between those 3396–97 (Herring) (‘‘lean-back services are radio- SX Ex. 27 at 7. words, as discussed infra, section IV.B.3.b. Suffice like services, one where you hit play and the 77 None of the parties requested that the Judges it to note at present the different meanings of these service kind of chooses for you . . . [w]hereas . . . interpret or seek an interpretation from the Register two verbs: ‘‘to select’’ means ‘‘to choose in lean-forward we consider on-demand services. So on whether any one listener feature or combination preference to another or others; pick out; to make you go into the service and you choose exactly what of features brought a particular noninteractive a choice; pick,’’ whereas ‘‘to influence’’ means ‘‘to you want to listen to.’’). service outside the scope of the statutory license. . . . affect; sway.’’ See Dictionary.com.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00022 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26337

ii. Increased Competition for Listeners [REDACTED], and [REDACTED] as hypothetical marketplace that the Judges are in the Downstream Market 79 competitors. SX Ex. 1262 at 4–11; SX tasked with replicating (i.e., statutory webcasting services and record companies, SoundExchange avers that interactive Ex. 2157 at 5. • iHeart has monitored [REDACTED] respectively) are ‘‘similar’’ to the buyers and services and noninteractive streaming sellers in his proffered benchmark. Id. services compete with each other for on its ‘‘competitor tracker’’ since ¶¶ 122(b) and 158(b). listeners. SX Ex. 269; 5/13/15 Tr. 3462 [REDACTED] first launched Absence of Statutory license test: Dr. (Herring). SoundExchange contends that [REDACTED]. SX Ex. 211 at 6. Rubinfeld contends that the hypothetical • iHeart has strategized as to how it Pandora, iHeart, and Sirius XM are all marketplace is one in which there is no could ‘‘match or beat [[[REDACTED]’s] statutory license. Id. ¶ 122(c). He opines that, keenly aware of the developing experience,’’ and listed ‘‘major roadmap among the spectrum of potential benchmarks competition from interactive services. items to deal with [REDACTED].’’ Id. at that could have been offered, a benchmark SoundExchange points to numerous 2. based upon interactive streaming services examples in the record of this purported Finally, SoundExchange notes that agreements is least likely to be influenced by the statutory license because interactive competition for listeners between Sirius XM also internally identifies interactive and noninteractive streaming services cannot default to the statutory interactive streaming services like license and therefore, according to Dr. services. [REDACTED], [REDACTED], With regard to Pandora, Rubinfeld, his proffered benchmark is an [REDACTED], [REDACTED], and appropriate replication of a market without a SoundExchange cites the following [REDACTED] as ‘‘competitors’’ for statutory license. Id. ¶ 158(c). evidence: Same rights test: Dr. Rubinfeld asserts that • Pandora’s own internal documents listeners of its noninteractive streaming service—My Sirius XM—and highlights the products sold in the hypothetical confirm that interactive services marketplace consist of a blanket license for ‘‘compete head-to-head for listener [REDACTED] as ‘‘offer[ing] the strongest the record companies’ complete repertoires hours with services that operate under competition in terms of the quality of of sound recordings, to be used in the statutory license,’’ Kooker WDT at customization.’’ SX Ex.1759 at 15; 5/22/ compliance with the DMCA requirements. Id. 16; 15 Tr. 5461–63 (Frear). Additionally, ¶ 122(d). Unlike the other three • Pandora identifies Spotify as a Sirius XM conducted a service-wide comparability tests discussed above, with survey of ‘‘competitive listening’’ in regard to the ‘‘same rights test,’’ Dr. ‘‘competitor’’ for the ‘‘consumers [it is] Rubinfeld contends that certain adjustments trying to attract to use Pandora,’’ SX Ex. which it sought input from listeners not only on streaming services like must be made to enhance the comparability 266 at 12; 5/13/15 Tr. 3483–84 of the proffered benchmark to the (Herring); [REDACTED], [REDACTED], hypothetical market. Dr. Rubinfeld asserts • Pandora identifies as ‘‘competitor [REDACTED], and [REDACTED], but that these adjustments are necessary because services’’ Spotify’s Free Mobile App also on interactive streaming services the agreements upon which his proposed (described by Pandora as ‘‘enabl[ing] [a] like [REDACTED] and [REDACTED]. SX benchmark is based provide various hybrid ‘lean-in’/‘lean-back’ experience’’) Ex. 237 at 26. functionality that is not permitted by the statutory license (i.e., ‘‘on demand’’ choice of and Beats Music (a ‘‘[p]ure on-demand Based on his proffered evidence of ‘‘convergence’’ and ‘‘downstream songs; unlimited skips; and ‘‘cached’’ service with a novel personalization downloads). Id. ¶ 158(d).80 feature’’), SX Ex. 266 at 15–21; competition,’’ Dr. Rubinfeld concluded • Pandora’s ‘‘Competitive Intelligence that agreements between interactive Therefore, according to Dr. Rubinfeld, Report’’ details the product offerings of streaming services and record ‘‘adjustments can and should be made services like Beats, Google Play, Rdio, companies were an appropriate to account for these differences when foundation upon which to base a applying the set of interactive and Spotify, SX Ex16 52; SX Ex. 2244; 81 • In 2014, Pandora briefed its marketplace benchmark for determining benchmarks.’’ Id. incoming CEO Brian McAndrews on the rates in this proceeding. 5/15/15 Tr. c. Per-Play ‘‘Ratio Equivalency’’ in ‘‘[i]ncreased competition [that] exists 1785 (Rubinfeld). Noninteractive and Interactive Markets from Apple, Google, and [other b. Comparability of Dr. Rubinfeld’s Dr. Rubinfeld ‘‘assumed that the ratio interactive] streaming services like Proffered Interactive Streaming Services of the average retail subscription price Spotify.’’ SX Ex. 2367; 5/27/15 Tr. Benchmark to the Hypothetical Market to the per-subscriber royalty paid by the 6163–65 (Fleming-Wood); and licensee to the record label is • Pandora identified Spotify, Rdio, Dr. Rubinfeld asserts that his approximately the same in both Deezer, Rhapsody, Slacker, Google, and proposed interactive streaming services benchmark satisfies the following four interactive and noninteractive markets.’’ Apple as ‘‘competitors’’ in Pandora’s Rubinfeld CWDT ¶ 169. This ‘‘ratio survey of competitors’ product strategies part-test that he contends comprises the standard that the Judges applied in the equivalency’’ is best presented by the and business models in a ‘‘Strategic following equation: Planning Overview.’’ SX Ex. 263 at 23. Web III Remand to determine the Similarly, with regard to iHeart, usefulness of a proffered benchmark: SoundExchange notes the following Willing buyer and seller test: Dr. Rubinfeld evidence of competition between contends that the rates that the Judges are interactive streaming services and required to set must be those that would have been negotiated in a hypothetical iHeart’s custom noninteractive 80 marketplace between a willing buyer and a Dr. Rubinfeld also noted that in the interactive streaming service: streaming services agreements that formed the basis • iHeart consistently identifies willing seller. Rubinfeld CWDT ¶ 122(a). Dr. of his proffered benchmark, the licensed rights do Rubinfeld opined that the interactive interactive services like [REDACTED], not consist of a blanket license for the record streaming services agreements upon which companies’ complete repertoires of sound he based his proffered benchmark are recordings. Instead, artist/labels may limit (or 79 This proceeding involves two aspects of a indicative of the results of negotiations exclude) the right to license certain content from vertical market: (1) The ‘‘upstream royalty market,’’ between willing buyers and willing sellers interactive streaming services. Id. Dr. Rubinfeld did in which record companies charge streaming because they were entered into voluntarily not offer any proposed adjustments to account for services for the right to access the record this distinction. companies’ repertoires of sound recordings; and (2) between parties who did not have the option 81 Dr. Rubinfeld made such adjustments, as the ‘‘downstream consumer market’’ in which of electing the statutory license. Id. ¶ 158(a). discussed infra. Understanding those adjustments streaming services offer music to listeners. Same parties test: Dr. Rubinfeld contends in the proper context requires a discussion of Dr. Rubinfeld CWRT ¶ 132. that the buyers and sellers in the Rubinfeld’s basic model, which follows.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00023 Fmt 4701 Sfmt 4725 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 ER02MY16.000 26338 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

Where: performance statements). Id.83 He then proffered downward ‘‘interactivity [A] = Avg. Retail Interactive Subscription added this derived per-play value to the adjustment’’ to the royalty rate for Price stated (i.e., headline) per-play rate. Dr. interactive services, which he then [B] = Interactive Subscriber Royalty Rate Rubinfeld then took an average of these applied to determine his proposed [C] = Avg. Retail Noninteractive Subscription per-play rates, weighted by revenue, id. royalty rate for noninteractive services. Price ¶ 203, to determine the interactive [D] = Noninteractive Subscriber Royalty Rate subscriber royalty rate for his interactive i. SoundExchange’s Alternative Calculation and Confirmation of Its Dr. Rubinfeld testified that this ‘‘ratio benchmark agreements. ‘‘Interactivity Adjustment’’ equivalency’’ assumption is not only Having obtained values for [A] and important, but indeed is foundational to [B], Dr. Rubinfeld was able to calculate Dr. Rubinfeld attempted to confirm his entire analysis. 5/6/15 Tr. 2026 that the direct agreements with the the reasonableness of his 2.0 (Rubinfeld).82 interactive services provided record interactivity adjustment by considering Dr. Rubinfeld calculated the companies with a minimum revenue a different method of calculating the interactive numerator and denominator share that generally ranged between 50 adjustment, undertaken by another [A] and [B], and the noninteractive percent and 60 percent of the services’ SoundExchange expert economic numerator [C], from available data in the revenues (based on the record witness, Dr. Daniel McFadden. agreements he had analyzed. Dr. company’s share of total streams), with Rubinfeld CWDT ¶¶ 171, 209. Dr. Rubinfeld did not have data to calculate the majority falling between 55 percent McFadden conducted a ‘‘conjoint the noninteractive denominator [D]— and 60 percent. Rubinfeld CWDT ¶ 206 survey’’ 86 to determine the value that i.e., the per-play ‘‘Noninteractive and, Appx. 1. Thus, given Dr. future consumers of digital streaming Subscriber Royalty Rate.’’ Therefore, Dr. Rubinfeld’s assumption that the ratios services place on various features of Rubinfeld attempted to estimate this should be equal in both markets, the those services. Dr. McFadden number by: (1) Applying the above per-play royalty rate for noninteractive determined the value that future equation; and (2) making what he services [D] (i.e., the statutory rate) consumers place on various features describes as the necessary adjustments would also have to provide record that are available on streaming services, to the rate he derives to account for companies with the same minimum such as: (1) Limited or unlimited skips; differences between the interactive and percentage of revenue out of [C] (the (2) offline listening; (3) on-demand noninteractive markets and thus satisfy average monthly retail noninteractive (desktop and mobile); (4) addition of the ‘‘same rights’’ test. subscription price). mobile service; (5) playlists (from However, Dr. Rubinfeld needed first More particularly, to determine his algorithms and ‘‘tastemakers’’); (6) to calculate [C] (the average monthly Interactive Numerator [A] (the average presence or absence of advertising; and retail noninteractive subscription price). monthly retail interactive subscription (7) catalog size between one million and Dr. Rubinfeld calculated [C]—as he had price), Dr. Rubinfeld calculated ‘‘the twenty million. SX Ex. 15 ¶ 9 calculated [A]—as a simple average of simple average of the [monthly] (McFadden WDT). the monthly subscription prices for the subscription prices for the interactive Relying upon the entire sample of services he had identified as services, which turned out to be in this respondents to Dr. McFadden’s survey, ‘‘noninteractive.’’ Because of varying case $9.86.’’ 5/5/15 Tr. 1797 Dr. Rubinfeld summed the average rates within each service (depending on 87 (Rubinfeld). willingness to pay (WTP) values for whether the average is computed using various attributes for hypothetical To determine his Interactive monthly or yearly fees), the average Denominator [B] in his ratio (the interactive and noninteractive services, ranged between $4.84 and $5.25. 5/5/15 in the following manner. interactive subscriber royalty rate), Dr. Tr. 1797 (Rubinfeld); Rubinfeld CWDT • Rubinfeld first identified the average On the interactive side, Dr. ¶ 207. Rubinfeld included the following minimum per-play rate as defined in Having calculated values for [A], [B] each of his selected interactive attributes: (1) Unlimited skips; (2) and [C], Dr. Rubinfeld thus could, and offline listening; (3) on-demand agreements. Rubinfeld CWDT ¶ 205. did, use the ratio of the interactive to Next, Dr. Rubinfeld identified the availability (desktop and mobile); (4) noninteractive subscription prices (the mobile service; (5) playlists (from various forms of non per-play ratio of [A] to [C] 84) to solve for [D] (the consideration, if any, in these algorithms and ‘‘tastemakers’’); (6) statutory noninteractive per-play royalty absence of advertising; and (7) catalog agreements, which included non- rate). Dr. Rubinfeld determined that the recoupable cash payments and size between one million and twenty ratio of the two monthly subscription million). advertising commitments with an prices ranged between 1.88 and 2.04.85 explicit financial value. Rubinfeld Dr. Rubinfeld applied what he 86 A conjoint survey creates a slate of alternative CWDT ¶ 218. To convert these lump- considered to be a reasonable and products and asks the consumer to identify which sum payments and values into per-play conservative figure within this range, product he or she most prefers. The sets of products values, Dr. Rubinfeld divided these 2.00, as a discount factor to make his are designed to realistically mimic the actual payments by the number of actual plays market process, in which a consumer is presented with and chooses among various competing (as set forth in the applicable service’s 83 If the agreements provided the record bundles of alternatives. By presenting each companies with rights that were not quantifiable consumer with several sets of choices, the 82 This ‘‘ratio equivalency’’ assumption in Dr. (e.g., data provision or equity stakes), Dr. Rubinfeld researcher can determine the relative importance Rubinfeld’s model is essentially the same as the did not account for the possible value of those and dollar value that consumers place on each of assumption made by Dr. Pelcovits on behalf of rights in his benchmark calculation. Id. the attributes. McFadden WDT ¶ 13. SoundExchange in Web II and Web III. See 84 As a basic mathematical point, if [A]/[B] = [C]/ 87 The word ‘‘average’’ is italicized in the text, Rubinfeld CWDT ¶ 207 n.124 (acknowledging that [D], then [A]/[C] = [B]/[D]. Thus, assuming Dr. supra, to presage an important element of Dr. he followed ‘‘past practices’’); 5/6/1/155 Tr. 2026– Rubinfeld’s approach was valid, he could McFadden’s results, one that he identified and 27 (confirming that his reference to ‘‘past practices’’ mathematically determine [D] (the statutory upon which one of the Services’ economic experts, referred to Dr. Pelcovits’s approach). Dr. Rubinfeld noninteractive rate) by applying the ratio of [A] to Dr. Steven Peterson, elaborated the relationship indicates, however, that his application of the [C], since he had calculated a value for [B] (the between the average WTP in Dr. McFadden’s survey interactive benchmark analysis does not suffer from interactive royalty rate). and the bimodal nature of Dr. McFadden’s WTP the defects in Dr. Pelcovits’ application of that 85 9.86/4.84 = 2.04 (rounded). 9.86/5.25 = 1.88 results. That issue is discussed further in this model in a prior proceeding. Id. at 2027–28. (rounded). determination.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00024 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26339

• On the noninteractive side, Dr. law.88 Id. ¶ 213 (Rubinfeld CWDT). determined that the Indies accounted Rubinfeld included these attributes but However, Dr. Rubinfeld understood that for an average of 24% of the streams on excluded the following features not directly-licensed interactive services, interactive services, and he weighted his offered by statutory services: (1) such as those in his proffered benchmark by assuming that this 24% Unlimited skips; (2) offline listening; benchmarks, are usually bound by figure was also applicable to the and (3) on-demand availability (desktop contract to pay royalties on pre-1972 noninteractive market. Id. ¶ 225.91 and mobile); and catalogs greater than sound recordings. Id. After applying the foregoing steps and ten million (as arguably more reflective In order to make an ‘‘apples-to- adjustments, Dr. Rubinfeld calculated of noninteractive catalog sizes in the apples’’ comparison, Dr. Rubinfeld that, for the year 2014 (the year for market). Id. therefore corrected for these differences which he had and applied data), the Rubinfeld CWDT ¶ 209, SX Ex. 56 in royalty-bearing plays in his per-play royalty rate for noninteractive (Rubinfeld CWDT Ex. 14). interactive benchmark market and the services implied by the interactive According to Dr. Rubinfeld, the statutory noninteractive market. SX Ex. benchmark equaled $0.002376, or survey results from Dr. McFadden’s 29 ¶ 214 (Rubinfeld CWRT). Applying 0.2376 cents. SX Ex. 59 (Rubinfeld conjoint survey indicated an the foregoing factors, Dr. Rubinfeld CWDT Ex. 16a). interactivity ratio of 1.90, which Dr. calculated that the ratio of (i) royalty- (C) Adjustment for 2016–2020 Period Rubinfeld noted was less than the 2.0 bearing plays in his interactive interactivity ratio calculated by Dr. benchmark market to (ii) royalty-bearing Finally, Dr. Rubinfeld determined that Rubinfeld through his own plays in the statutory noninteractive his proposed per-play rate should methodology, discussed supra. (Because market was 1.0:1.1. Accordingly, Dr. increase by a linear $0.00008 for each the interactivity ratio measures the Rubinfeld divided his per-play rate (as year in the statutory 2016–2020 period. relationship of interactive subscription calculated in the prior steps, supra) by In support of these annual increases, Dr. Rubinfeld relied upon: (1) The average prices to noninteractive subscription a factor of 1.1.89 prices, the lower 1.90 ratio would $0.00008 annual increase in rates as set 92 indicate that noninteractive (B) Adjustment for Indies in Web III; (2) his belief that there subscription prices are closer to Dr. Rubinfeld assumed that, on would be an ever-increasing interactive subscription prices, raising average, independent record companies, convergence in the retail prices of the benchmark interactive royalty rate commonly known as Indies, (i.e., those statutory and nonstatutory services; (3) as compared to Dr. Rubinfeld’s 2.0 not owned by (or by a division of) the presence of rate escalation ratio.) Accordingly, Dr. Rubinfeld Universal, Sony or Warner) would likely provisions in the iHeart/Warner concluded that Dr. McFadden’s negotiate less beneficial arrangements Agreement and the Pandora/Merlin Agreement; and (4) the presence of alternative method of calculating the with interactive services than would annual rate escalations in the Web III value of interactivity confirmed that Dr. Majors. Rubinfeld CWDT ¶¶ 220, 223. rates. Rubinfeld CWDT ¶¶ 137–141; Rubinfeld’s own 2.0 interactivity Based on this assumption, he made a PAN Ex. 5014 at 4, 5 (Pandora/Merlin adjustment was not only reasonable, but further assumption that the difference in Agreement). Thus, Dr. Rubinfeld conservative. Rubinfeld CWDT ¶ 210. the consideration received by the Majors increased his 2014 interactive and the Indies in the interactive market ii. Additional Adjustments Made by Dr. benchmark of $0.002376 by $0.00008, would be reflected completely in the Rubinfeld for a 2015 benchmark of $0.002456. assumed fact that Indies ‘‘would not That 2015 figure was again increased by The other differences between the receive any of the non per-play financial $0.00008 to reflect a rate for 2016 of interactive market and the or other unquantified consideration $0.002536 (rounded by Dr. Rubinfeld to noninteractive market that, according to major record companies receive . . . .’’ $0.0025). Dr. Rubinfeld, required further Id. ¶ 223.90 Dr. Rubinfeld then adjustment before he could determine a iii. The Interactive Rate Is an per-play royalty rate based on his 88 The Copyright Act only covers sound ‘‘Effectively Competitive’’ Benchmark interactive benchmark analysis are recordings fixed after February 15, 1972—the Rate described below. effective date of the Sound Recording Amendment, Pub. L. 92–140, 85 Stat. 391 (1971). Protection, if SoundExchange maintains that Dr. (A) Adjustment for Royalty-Bearing any, for sound recordings fixed prior to that date Rubinfeld’s interactive benchmark rate Plays (Skips and Pre-1972 Recordings) derives from state law. reflects effective competition because 89 Dr. Rubinfeld calculated the 1.1 adjustment In his analysis, Dr. Rubinfeld factor by: (i) Estimating the number of royalty- accounted for the fact that, under the bearing plays on a hypothetical service that does [REDACTED]’s [REDACTED] agreements with the statute, a ‘‘skip,’’ i.e., a song that that a not pay for skips, utilizing information about the Majors and the Indies, ‘‘the majors received listener skips after several seconds, is number of skips; the average skip length; song [REDACTED] and the indies did not.’’ SX Ex. 128 length; and ad minutes per hour, and then dividing ¶ 29 (Rubinfeld CWDT App. 2). considered a royalty-bearing play for a that number by (ii) the estimated number of royalty- 91 Dr. Rubinfeld noted that Nielsen Soundscan noninteractive service. By contrast, bearing plays as determined by analyzing Pandora’s information he possessed indicated that the interactive services, pursuant to their SEC filings. Rubinfeld CWDT ¶ 216; SX Ex. 57 independent record companies’ 2013 market share direct license agreements with record (Rubinfeld CWDT Ex. 15a); SX Ex.58 (Rubinfeld was higher—it was approximately 35%—but he CWDT Ex. 15b). chose to use the lower 24% interactive market companies, typically are permitted to 90 Apparently, Dr. Rubinfeld did not separately figure. Rubinfeld CWDT ¶ 224 and n.131 exclude from the royalty obligation at examine the Indies/Services agreements in his (continuing to rely on the 24% figure for interactive least some skips. SX Ex.17 ¶ 212 collected interactive agreements to test his plays of Indie sound recordings and noting (but not (Rubinfeld CWDT). Offsetting to some assumptions and apply the actual differences, if linking, logically or evidentially) the unsourced any, between the headline rates and other assertion that ‘‘a substantial portion of those sound extent this downward adjustment, compensation received by the Indies, on the one recordings were distributed by major labels.’’). according to Dr. Rubinfeld, was his hand, and by the Majors, on the other hand. See 92 See 37 CFR 380.3(a)(1) (setting forth Web III understanding that statutory services Rubinfeld CWDT ¶ 223 (‘‘I also assume that these rates). Although the average rate increased annually (such as Pandora and Sirius XM) independent record companies receive the same by $0.00008, the rate remained constant for 2012 per-play rates and proportionate revenue shares as and 2013 (at $0.0021) and also remained constant contend that they are not required to the majors.’’) (emphasis added). Dr. Rubinfeld later for 2014 and 2015 (at $0.0023). Thus, in 50% of the pay royalties for pre-1972 sound modified his direct testimony to note what he year-over-year changes, the Judges declined to make recordings under federal copyright described as confirmatory evidence—that in any changes in the Web III rates.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00025 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26340 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

downstream competition mitigates any ‘‘free alternatives’’ in the downstream per-play royalty rate under the Apple arguable market power record market have reduced interactive rates in licenses with Warner and Sony range companies may have in the upstream the upstream market. 5/20/15 Tr. 5049 from $0.[REDACTED] to licensing market. (However, it is worthy (Shapiro); 5/11/15 Tr. 2973 (Katz). $0.[REDACTED], the low end of which of note that SoundExchange did not SoundExchange also points to its exceeds the highest rate proposed by attempt to demonstrate that the negotiations with interactive services as SoundExchange. Id. ¶¶ 30, 42. interactive market on which it relies for evidence that the upstream interactive SoundExchange offered the Apple its benchmark is effectively competitive, market is effectively competitive. Dr. agreements as part of its rebuttal of a until its rebuttal case, after the Services Rubinfeld, described the negotiations as number of the licensee services’ had made their direct arguments as to a ‘‘real give and take,’’ where the labels criticisms of Dr. Rubinfeld’s interactive why the interactive market is not ‘‘have in mind a particular goal, but benchmark analysis. Dr. Rubinfeld effectively competitive.) In support of they have to give up something,’’ which contended that, because the its argument, SoundExchange relies on is ‘‘consistent’’ with the ‘‘view that (noninteractive) Apple agreements were the testimony of another of its economic there’s some bargaining power on the not susceptible to those criticisms, those experts, Dr. Eric Talley. part of the services.’’ 5/5/15 Tr. 1863 criticisms would be rebutted by According to Dr. Talley, rates in the (Rubinfeld). He further testified that the evidence that the royalty rates derived interactive market are constrained by possible bargaining range would at best from the Apple agreements were two factors. First, if there is an ‘‘elastic only reveal ‘‘something about the other roughly equivalent to those derived downstream demand curve’’ for an party’s willingness to pay or willingness from the interactive benchmark input (such as a sound recording), to sell.’’ Id. at 1864–65. Dr. Rubinfeld analysis. Id. ¶ 3. upstream prices for that input will be and SoundExchange reached these Specifically, Dr. Rubinfeld argued that constrained. Second, if the ‘‘expenditure conclusions based on their the following critiques that the licensee on that input versus other inputs’’—‘‘the consideration of the back and forth and services levied against his interactive cost intensity of that particular input’’— ultimate concessions record companies benchmark analysis would not apply to is proportionately significant compared make in the final agreements reached (or Apple’s agreements with the majors for to other inputs in the downstream abandoned) with Apple, Google, Beats, its noninteractive service. market, the constraint on pricing in the Spotify and Amazon. See SX PFF ¶ 471– • The majors’ repertoires are ‘‘must upstream market will be more 80 (and citations to the record therein). haves’’ for interactive services, enabling pronounced. 5/27/15 Tr. 6054–55 the majors to charge supracompetitive d. Direct Licenses for Noninteractive (Talley).93 prices. Id. ¶ 4. The majors’ repertoires According to Dr. Talley, both of these Services Corroborate Dr. Rubinfeld’s are not ‘‘must haves’’ for a factors are present here. First, high price Interactive Benchmark noninteractive service, since a elasticity exists downstream because of SoundExchange offered analyses of noninteractive service (and not its the threat from piracy and because of direct licenses between record customers) determines which songs will competition from other outlets, such as companies and several noninteractive be played. YouTube. Second, the variable costs services to corroborate its interactive • ‘‘[B]ecause noninteractive services associated with licenses are a very benchmark analysis. These include two purportedly have the ability to steer significant element of the downstream licenses from major record companies to listeners to sound recordings offered by sellers’ expenses. Thus, these elasticities Apple, Inc. (Apple) for its iTunes Radio independent music labels and away would be passed upstream. Id. at 6054– service, and several licenses for what from majors (or away from any 58. SoundExchange describes as particular major’s repertoire), record Dr. Talley then noted that his noninteractive offerings by services that label catalogs are substitutes.’’ Id. ¶ 5. theoretical modeling demonstrated that also offer interactive streaming. iTunes Radio would have the same such downstream competitive forces ability to steer listeners as any other ‘‘will cause the WBWS price to be i. Apple Agreements noninteractive service. Id. ¶ 7. tightly clustered, reducing variations SoundExchange presented evidence • ‘‘[B]ecause interactive services are due to differences in bargaining power.’’ of Apple’s license agreements with primarily subscription services, they SX Ex. 19 at 35, 44–45 (Talley WRT); Warner and Sony, respectively, for have substantially higher ARPUs than see also SX Ex. 29 ¶ 132 (Rubinfeld Apple’s iTunes Radio service. iTunes noninteractive services, which are CWRT). Radio is a streaming service that offers primarily ad-supported,’’ and would Sound Exchange notes that Dr. users the opportunity to listen to therefore pay substantially higher Talley’s assertions regarding the highly playlists selected by industry royalties. Id. at 6. iTunes Radio, by competitive state of the downstream ‘‘tastemakers,’’ as well as playlists that contrast, is a nonsubscription service market is essentially undisputed and are generated by an algorithm based that, like other noninteractive services, borne out by the evidence. See SX PFF upon a song or artist ‘‘seeded’’ by the is primarily ad-supported. Id. ¶ 7. ¶¶ 449–458 (and record citations listener (similar to Pandora’s service). Dr. Rubinfeld also offered two therein). Moreover, SoundExchange Dr. Rubinfeld described the iTunes additional reasons why the Judges notes that Drs. Shapiro and Katz Radio service as ‘‘DMCA compliant,’’ should consider the Apple agreements. acknowledged that the presence of some although he acknowledged that the First, he noted that Apple’s ‘‘unique rights granted to Apple are ‘‘not position in the marketplace’’ confers 93 Dr. Talley’s testimony describes factors identical to the statutory license.’’ substantial bargaining power in its pertinent to the economic ‘‘Hicks-Marshall’’ Rubinfeld CWRT, App. 2, ¶¶ 1–2.94 Dr. negotiations with record companies, principle, which provides that the upstream demand for a factor of production (such as sound Rubinfeld concluded that the effective tending to negate any argument based recording licenses demanded by a webcaster) is on a disparity of bargaining power ‘‘derived’’ in part from the downstream demand for 94 All testimony on the subject of iTunes Radio between licensor and licensee. Id. the finished product (such as a subscription service was taken prior to the launch of Apple Music. Second, Dr. Rubinfeld argued that the that offers such sound recordings). Further, the Consequently, the discussion of iTunes Radio in elasticity of demand downstream will be reflected this determination does not reflect any changes non-precedential language in the in the upstream demand for that factor of Apple may have made to the service as a result of agreements demonstrates that the production. that launch. parties did not expect them to be used

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00026 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26341

in this proceeding.95 As a consequence, economists’ contrast between ‘‘perfect’’ particular suppliers.’’ Shapiro WDT at he suggested that the shadow of the and ‘workable’’ competition: 11. This ability to steer toward or away statutory license may not affect the The theoretical conditions of perfect from certain suppliers is an example of Apple agreements as strongly as other competition often are not satisfied in actual price competition, according to Dr. Katz. noninteractive benchmarks (e.g., the markets .... It is thus necessary to See Katz WDT ¶ 32 (‘‘[C]ompetition Pandora-Merlin and iHeart-Warner consider markets that are competitive, but arises only when buyers have the ability agreements). Id. ¶ 8. not perfectly so. Economists have long to substitute the offerings of one seller examined this concept, beginning with for those of another. It is this possibility ii. Other Noninteractive Agreements Professor J.M. Clark, who introduced the of substitution that drives sellers to offer SoundExchange also offered Dr. concept of ‘‘workable’’ competition. higher quality and lower prices in order Rubinfeld’s analysis of record company Economists also refer to such markets as to attract buyers to themselves rather reasonably or effectively competitive. licenses to Beats Music’s ‘‘The than their rivals. Conversely, when Sentence,’’ Spotify’s ‘‘Shuffle’’ service, Katz WDT ¶ 29 (emphasis in original). buyers lack the ability to substitute Rhapsody’s ‘‘Unradio,’’ and Nokia’s Dr. Shapiro describes a ‘‘workably’’ or among the offerings of different sellers, ‘‘MixRadio’’ to corroborate its ‘‘effectively’’ competitive market as there is no competition among sellers to interactive benchmark analysis. follows: attract customers.’’) (emphasis in SoundExchange describes these services The hallmark of a workably competitive original). as noninteractive offerings, and market is regular, significant competition The Services assert that the concludes that the effective per-play among suppliers for the patronage of buyers. interactive service agreements that rates in the agreements exceed the per- . . . A market can be workably competitive SoundExchange proffers as appropriate play rate derived from Dr. Rubinfeld’s even when the products or services offered benchmarks are not the product of such benchmark analysis of interactive by different sellers are differentiated, so long an ‘‘effectively competitive’’ market. In service agreements. See Rubinfeld as no single supplier has significant support of this assertion, the Services CWRT ¶¶ 179–201. unilateral market power. Indeed, this is the advance several arguments. norm for information products such as books, First, the Services maintain that there 3. The Services’ Opposition to the video programming, or software applications. is a fundamental difference between SoundExchange Rate Proposal and the Workable competition does not require interactive and noninteractive services Judges’ Determination on the Issues marginal cost pricing or anything approaching the textbook model of perfect that precludes the former from serving a. Dr. Rubinfeld’s Interactive Benchmark competition. A market can also be workably as an ‘‘effectively competitive’’ Must Be Adjusted To Reflect Effective competitive even if it is quite concentrated, benchmark for the latter. That Competition so long as the suppliers compete regularly fundamental distinction arises, they The Services’ expert economic and energetically to win business from each aver, from the fact that a sine qua non witnesses all agreed that other. . . . In contrast, a market that is of on-demand services is that each SoundExchange’s proposed interactive monopolized or controlled by a cartel is not downstream listener chooses the artists, workably competitive. If such markets were benchmark would fail to establish rates albums, and tracks to which he or she considered workably competitive, the listens, as well as the timing and that are ‘‘effectively competitive.’’ See, concept of workable competition would lose e.g., Katz WDT ¶¶ 5, 17, 18–34; Shapiro all meaning. Likewise, a moderately or highly frequency of each play. For this reason, WDT at 3, 10–16; Fischel & Lichtman concentrated market in which the leading on-demand interactive services must AWDT ¶ 10; 5/11/15 Tr. 2799:9–16; suppliers tacitly collude is not workably always be in a position to play any 2800:3–18; 2801:9–17 (Katz); 5/8/15 Tr. competitive. For example, if the leading sound recording a listener might 2604:10–22 (Shapiro); 5/15/15 Tr. suppliers have settled into some form of demand, and the on-demand services 4094:7–19 (Lichtman); see also, e.g., coordinated interaction, e.g., by refraining therefore lack the ability to steer Shapiro WDT at10 n.11 (‘‘My approach from competing actively to poach each performances away from higher-priced here is consistent with the one taken by other’s customers, the market will fail to be labels and toward lower-cost providers. workably competitive. More generally, if the the Judges in the Web III Remand.’’). See Shapiro WRT at 23; see also Katz leading suppliers are colluding—either WDT ¶ 17 (describing buyer choice as More particularly, the Services’ expressly or tacitly—the market is not economists equate the ‘‘effectively workably competitive. the ‘‘essence of competition’’ and competitive’’ requirement as essentially opining that ‘‘[t]he creation of a rate- equivalent to the economic concept of Shapiro WDT at 10–11 (emphasis in determination process and its willing- ‘‘workable competition.’’ In its essence, original). buyer/willing-seller standard can best ‘‘[a] workably competitive market is one According to the Services’ be reconciled with economic principles not subject to the exercise of significant economists, the presence or absence of and common sense by interpreting market power.’’ Shapiro WDT at 10.96 ‘‘workable’’ or ‘‘effective’’ competition willing buyers as those who have The NAB’s economic expert, Dr. Katz, in the present case must be determined meaningful choices among competing essentially analogizes the D.C. Circuit’s by recognizing that the noninteractive sellers, rather than facing a single, all- contrast between ‘‘metaphysical’’ and services are ‘‘aggregators,’’ that is, they or-nothing offer from a monopolist.’’). ‘‘effective’’ competition to the aggregate sound recordings they have Second, the Services note that a lack licensed from record companies in the of effective competition in the upstream 95 That proposition is questionable in light of upstream market and then provide interactive market is confirmed by the other evidence of what euphemistically could be access to such licensed sound testimony of numerous SoundExchange called ‘‘strategic behavior’’ by Apple and one of the recordings to listeners in the witnesses, who conceded that the major record companies. See IHM Ex. 3517 downstream market. In such a market, licenses between record labels and on- ([REDACTED] email from [REDACTED] to [REDACTED]) (‘‘[REDACTED].’’) (emphasis added). ‘‘workable competition’’ is present, demand services are the product of a 96 See J. M. Clark, Toward a Concept of Workable according to the Services’ economists, if market devoid of any price competition Competition, 30 a.m. Econ. Rev. 241–56 (1940); ‘‘aggregators can offer attractive between record companies to obtain Jesse Markham, An Alternative Approach to the packages without the products of additional plays on on-demand services. Concept of Workable Competition, 40 a.m. Econ. Rev. 349, 349 (1950) (treating ‘‘effective particular suppliers and to the extent to See 4/28/15 Tr. 415–16 (Kooker) (Sony competition’’ and ‘‘workable competition’’ as which these aggregators can steer their has ‘‘never cut [its] price responding to synonymous). customers toward or away from a competitor’s proposal or for more

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00027 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26342 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

plays.’’); 4/30/15 Tr. 1097–99 (A. asserted to the FTC that the proposed repertoires of one label over another. Harrison) (Universal has never lowered merger would not lessen competition These provisions apply variously to a proposed rate as a consequence of because the market for interactive playlists, artist or album features, finding out that another Major was services was already not competitive. editorial content, home-page offering a lower rate, and, more broadly, Specifically, the Services point to placements, advertisements, album Universal does not take any actions to statements to the FTC by or on behalf recommendations, and/or other ways compete with Sony or Warner with of Universal: the interactive services may promote respect to services); 5/7/15 Tr. 2485–86 [REDACTED] particular content to their users. See 4/ (Wilcox) (Warner has never offered a PAN Ex. 5349 at 1–2 (Universal). 28/15 Tr. 455–56 (Kooker); 4/30/15 Tr. lower rate to an interactive service for [REDACTED] 1144–45 (Harrison); 6/2/15 Tr. 7202–05 more plays). PAN Ex. 5349 at 17 (Universal). (Harrison); 5/7/15 Tr. 2487–88, 2490–93 Third, the Services’ economists [REDACTED] (Wilcox). concluded that the reason for the The Services disagree with PAN Ex. 5025 at 2, 18 (Pomerantz). absence of price competition in the [REDACTED] SoundExchange’s assertion that upstream interactive market is that the downstream competition causes Dr. NAB Ex. 4129 at 41–2 (Rubinfeld). repertoires of each Major are [REDACTED] Rubinfeld’s interactive benchmark to ‘‘complements’’ for each other. As Dr. reflect ‘‘effective competition.’’ In fact, PAN Ex. 5025 at 18, 21 (Pomerantz); see NAB Shapiro opined: Ex. 4129 (Rubinfeld) ([REDACTED]); 5/5/15 Dr. Katz asserts that SoundExchange’s In the parlance of economics, the ‘‘must Tr. 1956–58, 1946–47 (Rubinfeld) (quoting conclusion is 180 degrees wrong: have’’ suppliers are complements, not PAN Ex. 5345 (June 22 letter to the FTC) [W]hen you have a highly competitive substitutes, because buyers need each of (‘‘[REDACTED].’’). downstream industry, there’s going to be a them and cannot substitute one for [REDACTED] smaller markup of [retail] price over cost another.... This concept is well known in PAN Ex. 5349 at 17 (Universal) (emphasis because the competitive pressures are going economics. When two essential inputs must added); see PAN Ex. 5025 at 16 to tend to drive [retail] price to cost. So what be used together, they are often referred to as ([REDACTED]). that means is . . . for any . . . license fees ‘‘Cournot Complements.’’ The evidence . . . set by the record companies, we have a shows that the repertoires of the major record Additionally, iHeart’s economic highly competitive downstream market. companies are Cournot Complements for experts, Drs. Fischel and Lichtman, There’s going to be a smaller markup. That interactive services. relied upon a [REDACTED] document then makes it profitable, more profitable to * * * * * submitted to the FTC in connection set a higher price upstream. So, actually, the The evidence shows clearly that the major with the Universal/EMI merger, more intense the competition downstream, interactive services ‘‘must have’’ the music of contrasting the ‘‘must have’’ nature of the greater the incentive to charge a high price upstream because you don’t have to each major record company to be the interactive service market with the commercially viable. The repertoires of the worry about so-called double major record companies are not substitutes more competitive noninteractive service marginalization.98 market: ‘‘[REDACTED]’’ IHM Ex. 3054 for each other in the eyes of either interactive 5/11/15 Tr. 2819 (Katz) (emphasis ¶41 n.70 (Fischel/Lichtman WRT) services or the record companies themselves. added). This means that there is no true ‘‘buyer (quoting SNDEX 0266588–665) The Services take Dr. Talley and choice’’ in this market. Thus, the market for (emphasis added). SoundExchange to task for failing to do licensing recorded music to interactive Sixth, according to the Services, the any empirical work to confirm whether services is not workably competitive.... foregoing points demonstrate that Dr. and to what extent piracy and other Rubinfeld’s proffered interactive Shapiro WRT at 15. downstream alternative music delivery Fourth, the Services note that benchmark market not only fails to be competitors may have affected upstream SoundExchange’s economic expert, Dr. competitive, but also is even worse than interactive rates. The NAB notes that Dr. Rubinfeld, did not perform any separate a market controlled by a single Talley admitted that he had performed analysis to determine whether the monopoly supplier. Shapiro WRT at 18; no empirical analysis to ascertain proffered interactive benchmark see also Katz WDT ¶¶ 41–43 (By logic whether or to what degree ‘‘downstream reflected the dynamics of a competitive first identified by Antoine Cournot in competition is, in fact, impacting the market. Rather, he assumed, i.e., he took 1838, firms offering complementary upstream negotiations’’ in the ‘‘for granted,’’ that his proffered products tend to set higher prices than interactive market. 5/27/15 Tr. 6092–93 interactive benchmark market was would even a monopoly seller of the (Talley); see id. at 6058 (‘‘I haven’t done sufficiently competitive. 5/5/15 Tr. 1922 same products, illustrating that an empirical analysis of that (Rubinfeld). suppliers of complements do not market. . . .’’). Dr. Tally further Fifth, the Services rely upon compete with one another.); PAN Ex. admitted that he had not studied either numerous statements in several 5349 at 19 (Universal White Paper to the downstream interactive service documents from SoundExchange’s own FTC explaining that ‘‘[REDACTED]’’). market or the upstream market in which principal advocates in the present case Seventh, the Services note that the the record companies license interactive that had been submitted to the Federal Majors structure their contracts with the services. Id. at 6080–83. Finally, Trade Commission (FTC) on behalf of interactive services to avoid any price Universal seeking approval of competition with the other labels and to 98 ‘‘Double marginalization’’ occurs when the Universal’s then-proposed merger with prevent the on-demand services from upstream supplier has upstream market power and EMI—subsequently approved by the attempting to steer users away from its buyer, the downstream seller, has downstream FTC and later consummated.97 These their repertoires. See 4/28/15 Tr. 441–42 market power. In that situation, ‘‘the price of the documents, according to the Services, input is marked up twice: By the upstream firm (Kooker); 4/30/15 Tr. 1142 (Aaron and, in terms of the final product price, by the reveal that Universal and its advocates Harrison); 5/7/15 Tr. 2473 (Wilcox). downstream firm.’’ W. Kip Viscusi, et al. Economics Even more particularly, the Services of Regulation and Antitrust 239 (2005). In the 97 Professor Rubinfeld acted as economic advisor note that the Majors’ agreements with absence of downstream market power on the part to UMG and EMI in relation to that transaction, and of the upstream buyers/downstream sellers, the Mr. Pomerantz, SoundExchange’s lead counsel in the leading interactive services contain upstream firms with market power can capture the this proceeding, acted as UMG’s counsel. 5/5/15 Tr. provisions that effectively prevent the full benefit of single marginalization, i.e., of price 1942–43; 1950–51 (Rubinfeld); PAN Ex. 5345 at 1. services from favoring the artists or above marginal cost.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00028 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26343

although Dr. Talley made certain market. Rather, he was simply saying which that benchmark applies: The suppositions regarding the elasticity of something is ‘‘conceivable, if you’re noninteractive subscription market.99 demand flowing from the downstream talking about hypotheticals’’ or The evidence of a range of potential market into the upstream market, the ‘‘possible,’’ which does not imply that it steering adjustments also rebuts Services note that Dr. Talley admitted is likely, or in any way true in this case. SoundExchange’s argument that the that he had not attempted to calculate See 5/11/15 Tr. 2976–78 (Katz). concept of ‘‘effective’’ or ‘‘workable’’ any elasticity of demand whatsoever, The Judges find that the impact of competition is ‘‘fuzzy’’ and that no because ‘‘within the ambit of how I was piracy and other downstream ‘‘bright line’’ can be drawn between retained as an expert, I did not view that competitors (such as YouTube) does not effectively competitive and non- as part of my charge.’’ 5/27/15 Tr. 6093 serve to promote ‘‘effective competitive rates. The Judges find that (Talley). competition’’ in any of the relevant this ‘‘line’’ needs to be drawn on a case- The Services also note that their own upstream markets, including the by-case basis, from the evidence and experts, contrary to SoundExchange’s upstream market for sound recordings testimony adduced at the hearing. Here, assertions, had not acknowledged that licensed for use by interactive the range of steering adjustments from piracy and other forms of downstream subscription services. SoundExchange, direct noninteractive licenses has been competition had or would reduce through the testimony of Dr. Talley, did introduced in evidence, steering upstream interactive rates to an note persuasively that in theory these experiments have confirmed the ‘‘effectively competitive’’ level. Rather, downstream competitors would depress reasonableness of such an endeavor and as the NAB notes, for example, Dr. Katz the upstream price. SoundExchange also expert testimony has explained how testified that even if piracy imposes correctly noted that Drs. Katz and steering is a mechanism by which to some constraint, ‘‘that doesn’t render Shapiro concurred with that theoretical offset the complementary oligopoly the market effectively competitive . . . point. However, a close reading of the power of the Majors (while not reducing it may be pressure on the monopoly testimony of Drs. Talley, Katz, and their firm-specific and copyright- price, but, nonetheless, it’s a monopoly Shapiro reveals that none of them specific market power). price.’’ 5/11/15 Tr. 2823 (Katz). As Dr. concluded that the impact of such The Services dismiss the idea that the Katz further explained, the merger downstream competition would record companies’ negotiations with submissions made by Universal argued necessarily depress any upstream price interactive services are evidence of an that the merger would lead to lower to a level that would offset the upward effectively competitive market. The prices because it would remove the pricing effect of complementary Judges agree with the Services criticism Cournot complements pricing effect oligopoly. Rather, Dr. Talley and of this assertion. As Dr. Shapiro between UMG and EMI, and that would SoundExchange invoke the vague idea explained, the mere existence of such not have been true if prices had already that any monopoly effects—after negotiations is uninformative as to been squeezed by piracy to near the assuming the upstream impact of whether the rates negotiated between competitive level: downstream competition—would be the interactive services and the Majors ‘‘benign’’ or ‘‘pedantic,’’ and Drs. Katz are competitive. Pandora PFF ¶ 237 (and [T]he parties were saying, if we’re allowed and Shapiro acknowledged only the to merge, we would find that it would citations to the record therein). increase our profits to lower our price. So hypothetical possibility that Moreover, the Services note that Dr. clearly, piracy had not pushed them down to downstream competition in some Rubinfeld conceded that the existence such a low price that going lower would circumstance could eliminate the of such negotiations is not evidence of reduce their profit. They actually say, going anticompetitive power of upstream a competitive market, because even lower would raise our profits. And what monopolists or complementary monopolists negotiate with their that’s telling you is, along with the fact that oligopolists. the other majors are must have[s] as well, is customers. See 5/28/15 Tr. 6487–88 In the present case, though, the Judges (Rubinfeld) (‘‘Q. Do firms with [that] they were actually concerned they were are not left with mere hypotheticals pricing above the monopoly level. monopoly power ever bargain with their regarding whether the anticompetitive customers? A. Yes. Q. Do firms with 5/11/15 Tr. 2825 (Katz) (citing PAN Ex. elements of the interactive market are 5025 at 22). monopoly power ever make concessions ‘‘benign’’ or ‘‘pedantic.’’ Nor are the or change their bargaining position in Additionally, the NAB, again through Judges hamstrung, as SoundExchange response to positions taken by buyers Dr. Katz, notes that identifying a suggests, by the alleged absence of with which they are dealing? A. Yes.’’). hypothetical increase in the elasticity of ‘‘bright line’’ demarcations as to when Pandora further notes that, when demand in the upstream market arising effective competition is present and questioned on this issue by the Judges, from competition in the downstream when it is not. Rather, the Judges were Dr. Rubinfeld conceded that ‘‘the fact market is not the same as identifying a presented with hard and persuasive competitive price in the upstream evidence that competitive steering has 99 It appears that SoundExchange may be making market. Thus, the Services assert that, reduced royalty rates in the an implicit argument that the rates in its interactive although Dr. Katz testified that piracy noninteractive market and would do so benchmark market have been so reduced by and other forms of downstream in the hypothetical market as well. This downstream competition that all supranormal competition could have ‘‘some sort of an evidence of steering (provided by profits have been eliminated. However, SoundExchange did not produce evidence effect, and I believe it’s in a downward Pandora and iHeart) demonstrates a sufficient to show record company profits overall to direction,’’ 5/11/15 Tr. 2973 (Katz), he measurable range of adjustment to the support such an argument. Also, as the Judges have was not opining how far such prices that would be set in a market for previously noted, and note again in this competition might have pushed down those streaming services if the services determination, the rate-setting process under section 114(f)(2)(B) is not intended to preserve any the price. They point out that, when Dr. could inject price competition via parties’ profits. Moreover, if the Judges were to go Katz noted the hypothetical possibility steering. Thus, the rate set in Dr. down that evidentiary road and base their rate that downstream competition could Rubinfeld’s upstream interactive decision on profits and reasonable rates of return, push upstream prices down to benchmark market should be adjusted to the process would in essence become a public- utility style proceeding and, as noted elsewhere in competitive levels, he was not reflect such price competition, so that it this determination, no party has suggested that suggesting that such a hypothetical is usable as an ‘‘effectively competitive’’ section 114(f)(2)(B) proceedings could be conducted circumstance exists in the interactive rate in the segment of the market to in such a manner.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00029 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26344 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

that they’re in negotiations, per se, rates in the noninteractive market and revenues to royalties in the doesn’t mean the market is would do so in the hypothetical market noninteractive market. The Services competitive. . . .’’ 5/5/15Tr. 1861–63 as well. This evidence of steering claim, however, that Dr. Rubinfeld (Rubinfeld). (provided by Pandora and iHeart) provided no economic basis for this On this issue, the Judges also agree demonstrates a measurable range of ‘‘assumption.’’ For example, the NAB with Dr. Katz, who noted that adjustment to the prices that would be asserts that Dr. Rubinfeld admitted that negotiations over price can occur set in a market for those streaming he was only ‘‘follow[ing] past practices’’ between a monopolist and its customers services if the services could inject price of Dr. Michael Pelcovits, an economic in order to facilitate price competition via steering. Thus, the rate witness for SoundExchange in Web II discrimination and increase monopoly set in Dr. Rubinfeld’s upstream and Web III. Rubinfeld CWDT ¶ 207 profits rather than to concede to more interactive benchmark market can and n.124, 5/6/15 Tr. 2026–27 (Rubinfeld). competitive prices. Specifically, Dr. should be adjusted to reflect such price This criticism was echoed by Pandora’s Katz testified: competition, in order to render it is economic expert, Dr. Shapiro, who Bargaining with your customers and usable as an ‘‘effectively competitive’’ testified ‘‘there is simply no plausible having some of the give and take can even rate in the segment of the market to economic rationale that would support be a form of price discrimination in a way which that benchmark applies—the the use of Professor Rubinfeld’s to get additional monopoly profits, so the noninteractive subscription market.100 interactivity adjustment.’’ PAN Ex. 5023 mere fact that your customer asks for The evidence of a range of potential at 29–30 (Shapiro WRT). something and you say, okay, I will give that steering adjustments also rebuts However, Dr. Rubinfeld’s oral to you, particularly if that is going to help SoundExchange’s argument that the testimony, and the testimony of the you get more money, the fact that you do that concept of ‘‘effective’’ or ‘‘workable’’ Services’ economic experts, indicated doesn’t show you lack monopoly power. It competition is ‘‘fuzzy’’ and that no shows you are economically rational. that an economic principle indeed ‘‘bright line’’ can be drawn between underlies his assumed equivalency in 5/26/15 Tr. 5715–16 (Katz). effectively competitive and non- these ratios. More particularly, Dr. The Judges reject SoundExchange’s competitive rates. The Judges find that Rubinfeld acknowledged that his ‘‘ratio argument that evidence of its this ‘‘line’’ needs to be drawn on a case- equivalency’’ was intended to create a negotiations with interactive services by-case basis, from the evidence and rate whereby every marginal increase in demonstrates that the interactive market testimony adduced at the hearing. Here, subscription revenue would result in is effectively competitive. As the Judges the range of steering adjustments from the same increase in royalty revenue, pointed out in their Commencement direct noninteractive licenses has been whether that marginal increase in Notice in this proceeding, price introduced in evidence, steering subscription occurred in the interactive discrimination is a feature of markets experiments have confirmed the market or the noninteractive market. 5/ such as sound recording markets, where reasonableness of such an endeavor, and 5/15 Tr. 1767 (Rubinfeld). This result, the marginal physical cost of licensing expert testimony has explained how Dr. Rubinfeld agreed, reflected an a sound recording is essentially zero, steering is a mechanism by which to application of rational profit and is also a relatively common feature offset the complementary oligopoly maximizing behavior by a willing seller, in many markets. 79 FR 412, 413 power of the Majors (while not reducing as explained in colloquy with the (January 3, 2014). their firm-specific and copyright- Judges: Further, the Judges cannot ignore the specific market power). testimony from several record company [THE JUDGES] witnesses, discussed in this b. Dr. Rubinfeld’s Interactive [T]hat’s an application . . . of a determination, in which they Benchmark Is Applicable Only to the fundamental economic process of profit acknowledged that they never attempted Subscription Market maximization.... [The record companies] to meet their competitors’ pricing when The Judges find that the interactive would want to make sure that the marginal negotiating with interactive services. return that they could get in each sector benchmark proposed by would be equal, because if the marginal Thus, the existence of the negotiations SoundExchange (adjusted as discussed return was greater in the interactive space noted by SoundExchange cannot in the previous section) is informative— than the noninteractive . . . you would want override this more specific testimony. but only to a particular segment of the to continue to pour resources, recordings in The Judges were presented with noninteractive marketplace. The this case, into the [interactive] space until substantial, unrebutted evidence that foundational aspect of Dr. Rubinfeld’s that marginal return was equivalent to the the interactive services market is not interactive benchmark is his assumed return in the noninteractive space. Would effectively competitive. The Services equality between two ratios: (1) that be correct? conclude from this that the interactive Subscription revenues to royalties in the [DR. RUBINFELD] services benchmarks are wholly interactive market; and (2) subscription It would. You said that just the way I uninformative with regard to the rates would like to have said it when I was teaching that subject. Yes, I agree with that. that would be negotiated in an 100 SoundExchange may be implying that the effectively competitive noninteractive rates in its interactive benchmark market have been 5/7/15 Tr. 2325 (Rubinfeld); see market. See Shapiro WRT at 47 so reduced by downstream competition that all Rubinfeld CWRT ¶ 172 (‘‘All else equal, supranormal profits have been eliminated. (explaining that Professor Rubinfeld is However, SoundExchange did not produce the interactivity adjustment sets requesting that the Judges ‘‘replicate and evidence sufficient to show record company profits statutory rates that represent the same extend the excessive royalty rates from overall to support such an argument. Also, as the fraction of subscription prices as paid interactive services market—where Judges have previously noted, and note again in this by the on-demand services. . . .’’). determination, the rate-setting process under competition is manifestly not working— section 114(f)(2)(B) is not intended to preserve any Thus, Dr. Rubinfeld’s ‘‘ratio into the market for the licensing . . . to parties’ profits. Moreover, if the Judges were to base equivalency,’’ assumes a 1:1 statutory webcasters. . . .’’). The Judges their rate decision on profits and reasonable rates ‘‘opportunity cost’’ for record disagree. of return, the process would in essence become a companies, whereby, on the margin, a public-utility style proceeding and, as noted The Services’ own evidence elsewhere in this determination, no party has dollar of revenue spent on a demonstrates persuasively that suggested that section 114(f)(2)(B) proceedings subscription to a noninteractive service competitive steering has reduced royalty could or should be conducted in such a manner. is a lost opportunity for royalties from

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00030 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26345

a dollar to be spent on a subscription to streaming service and those listeners sense that they require no out-of-pocket an interactive service. Accordingly, and who have a WTP of zero. expenses to listen to music.’’ McFadden contrary to the Services’ criticism, Dr. The most persuasive evidence on this WDT ¶ 56 (emphasis added). Dr. Rubinfeld’s ‘‘ratio equivalency’’ does point is found in the results of the McFadden then testified that his own possess an underlying economic conjoint survey conducted by a survey data confirmed ‘‘a group of rationale. SoundExchange witness, Dr. McFadden. consumers who place a high value on However, the unwarranted Dr. McFadden performed his conjoint no out-of-pocket expenses . . . who are assumptions lurking behind Dr. survey to determine the WTP of likely to remain [on] or adopt free Rubinfeld’s economic rationale were consumers who were provided with a plans.’’ Id. noted by the Services’ economic expert menu of bundled features that reflected The Judges cannot disregard this witnesses. For example, Dr. Lichtman, bundles that existed in the marketplace. bimodal chasm. Moreover, the record is an economic expert for iHeart, testified: His findings revealed the dichotomy replete with evidence corroborating this regarding the WTP of consumers of point. For example, testimony from [Dr. Rubinfeld] assum[es], I think, a perfect industry witnesses underscored the substitution . . . assumptions about noninteractive services: substitution, competition how all of these I find that consumers of streaming services unwillingness of a substantial markets interrelate.... [I]t’s intuitive. I divide between those who are willing to pay percentage of listeners to pay any price understand why he was drawn to it. It’s so for these services (and the extra features they to listen to noninteractive services. A nice to say, yes, roughly these will all be the offer) and those who are averse to paying for Sony executive testifying on behalf of same, revenue to royalty, revenue to royalty. music streaming services.... SoundExchange stated: ‘‘It’s challenging to convince a consumer to open their 5/16/15 Tr. 4043–44 (Lichtman). McFadden WDT ¶ 10 (SX Ex. 15) [sic] wallet and pay for something that Dr. Rubinfeld’s ‘‘ratio equivalency’’— (emphasis added). This dichotomy was examined in is similar to something that is available as a means toward profit to them for free. . . .’’ 4/28/15 Tr. 376– maximization—was more than a detail by another economist, Dr. Steven Peterson, who was a joint witness for 77 (Kooker). Even when the Majors theoretical abstraction. The desire of the provide incentives and disincentives to record companies to achieve such the NAB and Pandora. Dr. Peterson noted a critical bimodality in Dr. services in the form of royalty pricing parity across markets was reductions and increases, they are confirmed by a senior Warner executive McFadden’s data (consistent with Dr. McFadden’s finding) that reflected two unable to induce more than a minority who testified on behalf of of listeners to convert from a ‘‘free’’ SoundExchange: classes of listeners; those who would pay a positive sum for various features service to a paid subscription service. Our goal, aspirationally and in actual available in a noninteractive service and One of the most successful interactive results, has been a [REDACTED] percent services, Spotify, has only been able to rev[enue] share in this area generally.... those who refused to pay any money for any features. As Dr. Peterson explained, induce approximately [REDACTED]% of So we’ve been kind of struggling, if you will, its listeners to pay for a subscription to pull these business models up to what we SoundExchange and Dr. Rubinfeld rely on the average WTP among the survey streaming service. Id. at 404–05; see id. think is the level of consideration that we at 430 (Mr. Kooker acknowledging no find appropriate for essentially all of these participants (to confirm Dr. Rubinfeld’s music models, which is the [REDACTED] interactivity adjustment), but that evidence of a meaningful group of users range. So it was a combination of trying to average obscured the clear bimodality of willing to pay to subscribe to Pandora beyond those who currently subscribe). be realistic and make major progress towards Dr. McFadden’s results: our ultimate goal. Another industry witness, Aaron Dr. McFadden presents only the estimated Harrison of Universal, acknowledged 6/3/15 Tr. 7406 (Wilcox) (emphasis average willingness to pay for each feature that he had no data to support a added). addressed in his survey. However, it is conclusion that there is ‘‘some Mere assumptions as between possible to estimate each survey participant’s meaningful group of users who would willingness to pay for the features addressed interactive and noninteractive services be willing to pay to subscribe to regarding substitution, competition, in the survey. Based on the information for individual respondents, Dr. McFadden notes Pandora beyond those who already market interrelationships and the like that there is a group of users who are averse have. . . .’’ 4/30/15 Tr. 1115 (A. are inadequate, and thus limit the to paying for music streaming services.... Harrison). This was consistent with a applicable scope of Dr. Rubinfeld’s Thus, Dr. McFadden’s results are consistent broader aspect of Mr. Harrison’s ‘‘ratio equivalency’’ approach. The with the record labels’ documents that testimony, in which he noted, ‘‘the unsupported and unrealistic indicate many consumers have a low music-buying public has never been a assumptions in the ‘‘ratio equivalency’’ willingness to pay for subscription streaming huge market. . . .’’ Id. at 990. approach are considered below. services. . . . Moreover, the distribution is Pandora’s Chief Financial Officer As Dr. Lichtman noted, the ‘‘ratio bimodal, meaning it has two peaks.... similarly testified that ‘‘approximately equivalency’’ in Dr. Rubinfeld’s model [T]he average willingness to pay for a service with no ads masks the fact that there is a an 80 percent slice of the market . . . makes assumptions regarding bimodal distribution . . . of preferences over is unwilling to spend significant money substitution, and how these markets the willingness to pay for a service with no on music,’’ as reflected in ‘‘numerous interrelate. 5/6/15 Tr. 4043–44 advertisements and that the peaks occur so studies’’ [that] show that about half of (Lichtman). That is, the ‘‘ratio that consumers at the peaks have divergent Americans will never spend another equivalency’’ approach assumes that the preferences (i.e., would respond in opposite dollar and another . . . 35 percent will listeners who willingly pay for a ways) regarding a service with or without spend . . . $15 per year.’’ 5/13/15 Tr. subscription to a service have a WTP advertisements. 3553–54, 3356–57 (Herring). This equal to the WTP of those who use ad- NAB Ex. 4013 at 32–34 (Peterson portion of the dichotomized market supported (free-to-the-listener) services. CWRT) (emphasis added; footnotes comprises the core of Pandora’s However, the record evidence is omitted). customers: ‘‘[T]hat’s the group that we overwhelming that there is a sharp This point is consistent with Dr. target . . . people that aren’t going to be dichotomy between listeners who have McFadden’s own testimony, in which able to be monetized through a $10 a a positive WTP and therefore may pay he stated: ‘‘Most users regard their use month subscription or even a $5 a a subscription fee each month for a of [streaming] services as free in the month subscription but want a free lean-

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00031 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26346 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

back experience.’’ Id. at 3554. to perform a 1:1 measure of opportunity (Lichtman) (‘‘on-demand . . . [t]hat’s Accordingly, Mr. Herring noted that costs as is the case in Dr. Rubinfeld’s the key thing that makes the services 95% of Pandora’s customers listen ‘‘ratio equivalency’’ model. different, not the little features that have through the ad-supported free-to-the- Even SoundExchange acknowledges, been added. . . .’’); Fischel/Lichtman listener, and only 5% are subscribers, ‘‘directly licensed interactive services WRT ¶ 11 (‘‘Clearly, the most important which he understood to reflect ‘‘user . . . allow users to select their difference between interactive and preference’’ for ‘‘free sources,’’ rather programming . . . whereas . . . noninteractive services is . . . on- than a ‘‘bias’’ on the part of Pandora statutory services can [only] . . . demand functionality. . . .’’).101 toward ‘‘growing market share.’’ 5/13 influence what they hear. SX PFF ¶ 278 In addition to the above ‘‘bright line’’ Tr. 3435–36 (Herring). (emphases added). As a SoundExchange difference, statutory licensees are Further supporting this dichotomy economic expert witness acknowledged, subject to the various other limits from the record company perspective, the consumer who values sound imposed by the DMCA performance an internal Warner strategy document recordings highly is apt to have an complement. 5/27/15 Tr. 6136–37 noted that ‘‘[a]d-supported services have interest in particular sound recordings, (Fleming-Wood) (‘‘[P]andora adhere[s] proven to primarily be additive and to and will be more willing to pay for a to the performance complement for be targeting a different demographic subscription that allows him or her sound recordings. . .’’); see 17 U.S.C. than paid services.’’ IHM Ex. 3118 at 11; more ‘‘functionality,’’ including the 114(j)(13). Specifically, statutory see 5/7/15 Tr. 2405–06 (Wilcox) (noting ability to select songs on demand. By services cannot offer to their listeners a that Pandora weaned listeners from contrast, the more casual listener, with pre-designated song; an entire album; terrestrial radio whose listening, a number of free alternatives such as more than four songs by the same artist therefore, had not previously been terrestrial radio, lacks the same desire to or three songs from the same album in responsible for revenues that could be select a particular song at a particular any given three-hour period; caching for monetized into upstream royalties). time. See 5/4/15 Tr. 1677, 1679 off-line playback; a listener-created Expert testimony further confirmed (Blackburn) (distinguishing ‘‘music playlist played at the listener’s this dichotomy. One of aficionados’’ who ‘‘are willing to spend discretion; the rewinding or fast- SoundExchange’s own witnesses, Dr. a lot of money on music’’ and forwarding of songs; and a preview of David Blackburn, acknowledged that, at ‘‘additional functionality’’ from ‘‘people upcoming songs. 5/6/15 Tr. 2016–18; one end of the spectrum, consumers who are unwilling to pay anything for 2049; 2088–89 (Rubinfeld). were willing to pay a lot of money, and music.’’ Additional differences highlighted by at the other end of the spectrum are This undisputed distinction drives in the participants in this proceeding people who are unwilling to pay part the bimodal nature of the include: anything for music. 5/4/15 Tr. 1679 distribution between listeners with a • Pandora’s ‘‘thumbs up/thumbs (Blackburn). An expert survey witness positive WTP for streaming and those down’’ feature, which does not provide for Pandora, Larry Rosin, surveyed with a zero WTP. a listener with the ability to select the consumers and found that, annually, for actual artist or song that is played. 5/13/ any sort of music, physical or digital, c. The Irrelevance of SoundExchange’s 15 Tr. 3446–47 (Herring). 45% of respondents paid zero; 21% ‘‘Convergence’’ Argument • The increased use of mobile spent between $1 and $30, and 18% The Services dispute the assertion devices, which does not address the spent between $31 and $60. Further, that the increased overlap among the lack of convergence between the when asked if they would pay for a features of the statutory and non- essential functionalities of the two Pandora subscription if the free-to-the- statutory services constitutes a services. 5/7/15 Tr. 2304–05 listener service was discontinued, 54% convergence that is meaningful in this said it was ‘‘not at all likely’’ that they rate setting proceeding. In support of (Rubinfeld); 4/28/15 Tr. 432–33 (Kooker). would pay for a subscription, and 25% this position, the Services make several • said it was ‘‘not very likely’’ that they specific arguments. Spotify’s mobile Shuffle service, would pay for a subscription. Rosin which is not a noninteractive service i. Fundamental Differences in the WRT Figures 2 and 9 (PAN Ex. 5021); but rather has numerous on-demand Services see 5/14/15 Tr. 3727 (Rosin). Mr. Rosin features. See IHM Ex. 3371 ¶ 14 (Fischel concluded from his survey that ‘‘the The Services note a fundamental & Lichtman SWRT). majority of people are essentially . . . difference between interactive services ii. Convergence Does Not Create seeking free services.’’ Id. at 3742. and noninteractive services. They Relevant Competition Despite the overwhelming evidence of suggest a ‘‘bright line’’ difference this dichotomy in WTP, Dr. Rubinfeld’s between statutory services and non- The Services also take issue with the model is based solely on the statutory services that legally prevents notion that functional convergence is subscription platform. Thus, it is not convergence with regard to the most probative of competition relevant to this reasonable to conclude that the ratio of critical distinction, i.e., the inability of proceeding. Specifically, the Services listeners to statutory noninteractive argue: subscription rates to royalties in the • interactive market is relevant to the services to choose the exact song or The ‘‘convergence theory’’ focuses opportunity cost to a record company of playlist of songs to which they will entirely on competition between listeners who opt instead for ad- listen, as they would if accessing their services in the downstream consumer supported noninteractive listening. own music collections. 5/13/15 Tr. market, and therefore offers no insight Rather, ad-supported (free-to-the- 3445–46 (Herring) (noting this ‘‘bright into the lack of competition in the listener) internet webcasting appeals to line’’ between statutory and non- interactive upstream market that a different segment of the market, statutory service); 5/7/15 Tr. 2304–05 SoundExchange seeks to use as its compared to subscription internet (Rubinfeld) (none of Pandora’s features benchmark market. Shapiro WRT at 46– webcasting, and therefore the two ‘‘enhance the Pandora users’ ability to 101 This criticism relates to the distinction products differentiated by this attribute select a particular song for listening at between a listener’s ability to ‘‘select’’ a song and (‘‘ads and free’’ vs. ‘‘no ads and the time he or she wants to listen to a listener’s more limited ability to ‘‘influence’’ the subscription fee’’) cannot be compared it.’’); see also 5/15/15 Tr. 3397–98 song that is played, as emphasized supra, note 76.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00032 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26347

47; 5/18/15 Tr. 4469–71; 4474–75 interactive and noninteractive services. fact that the subscription-based rate (Shapiro). The services note that, on cross- derived by Dr. Rubinfeld from 2014 • The alleged convergence in the examination, Mr. Kooker admitted to a data, $0.002376, is proximate to Dr. downstream market does not address number of acts that increased the Shapiro’s high-end proposed rate for the the question of whether the upstream chances of the desired artist playing subscription market of 0.00215. When market is effectively competitive. during his experiment: (1) He created a Dr. Rubinfeld’s proposed rate is Shapiro WRT at 46. new account for the experiment, adjusted downward to reflect an • Dr. Rubinfeld failed to consider: (1) meaning Pandora had no information on effectively competitive market (as Substitution patterns among the various what tracks or types of music the creator calculated in the Rate Conclusion modes of music consumption; and (2) liked other than the ‘‘seed’’ artist (unlike section), the two rates are even more market shares in the downstream the typical Pandora listener who has proximate. Those two benchmark market. PAN Ex. 5022 at 10 (Shapiro created many stations, used the thumbs- subscription rates therefore indicate that WDT). up/thumbs-down button, skipped competition and convergence indeed do • Attempts by on-demand services to tracks, and provided Pandora a host of cause interactive and noninteractive offer some radio-like functionality do information on his/her tastes above and royalty rates to be similar in the not demonstrate competition between beyond the first ‘‘seed’’ artist); (2) he subscription market. interactive and noninteractive services indicated that the new account user was Thus, the impact of functional in the upstream market, but rather a 25-year-old female, a demographic convergence and downstream indicate only that on-demand services which Mr. Kooker admitted was competition is relevant only in the seek to ‘‘cross- over’’ and enter the specifically chosen because it was ‘‘the subscription market. Therefore, once Dr. ‘‘lean-back’’ market. 5/13/15 Tr. 3555– typical demographic, from Sony’s Rubinfeld’s benchmark is limited to the 57 (Herring). experience, that would be looking for subscription market, the Judges find that • The fact that some consumers want pop hit type of playlists’’ (and who SoundExchange’s emphasis on the both lean-back and lean-forward would then be more likely to receive functional convergence of, and functionality does not mean that each those playlists); and (3) he skipped downstream competition between, type of service is competing with the songs until he had listened to five songs, interactive and noninteractive services other. IHM RPFF ¶ 296 (and record even though he acknowledged that such is pertinent. citations therein). activity could influence Pandora’s Another important change in • When Pandora imposed listening playlist algorithms. See 5/29/15 Tr. opportunity cost arises when the caps in 2013 and 2014, it lost listeners 6589–92 (Kooker). upstream purchaser (the noninteractive to other noninteractive services, not to • iHeart’s on-demand video service webcaster in the present context) has interactive services, indicating that the represents a very minor element of total the ability to: (1) Purchase a substitute competition did not crossover into the listenership for iHeart’s service. Fischel/ input and ‘‘bypass’’ the input from the interactive market. Fischel/Lichtman Lichtman WRT ¶ 11 n.14. complementary oligopolists or WRT ¶¶ 17–18 and Exs. A & B. • ‘‘Pandora Premieres’’ is not a monopolist; and/or (2) the ability to • Statutory noninteractive services statutory feature and does not operate ‘‘use proportionately less’’ of the input compete in the market for radio pursuant to the statutory license. 5/15/ of the complementary oligopolists or listening, which is distinct from the 15 Tr. 3444 (Herring); see 5/6/15 Tr. monopolist. In the present case, both interactive market, and about 80% of 2006 (Rubinfeld). Pandora and iHeart have demonstrated music consumption in the United States • Even though noninteractive services that, by steering,102 a noninteractive occurs via ‘‘lean-back’’ radio-listening compete with interactive services ‘‘for service can: (1) Partially ‘‘bypass’’ one experience. Fleming-Wood WDT ¶ 14 music listening generally,’’ it is or more Majors and substitute an n.2; 5/27/15 Tr. 6138 (Fleming-Wood); ‘‘marginal,’’ i.e., at that line between 80 increased proportion of songs from 5/13/15 Tr. 3397–99 (Herring); Pandora percent [lean back] and 20 percent [lean Indies or other Majors; and (2) thereby Ex. 5016 ¶ 9 and Figure 2 (Herring in],’’ and the ‘‘core businesses are very reduce their ‘‘proportion’’ of purchases AWRT) (showing 76.2% of consumers different.... They’re not substitutes from higher priced Majors up to a listen to lean-back services); see Shapiro for each other.’’ 5/13/15 Tr. 3397–99 certain level. WRT at 9 & Figure 2; 5/18/15 Tr. 4478– (Herring). Another important adjustment 79 (Shapiro) (terrestrial radio, The Judges find that there is necessary to render Dr. Rubinfeld’s noninteractive webcasting and satellite significant evidence of functional ‘‘ratio equivalency’’ useful is to make radio comprise 63% of time spent convergence (up to the limits prescribed certain that the outcome does not listening to music, and interactive by the DMCA) between interactive and simply maintain or import supranormal services account for 7%). noninteractive services. Further, the prices that are the consequence of the Judges find that downstream absence of effective competition. The iii. The Supposed ‘‘Interactive’’ Features competition exists between such need to adjust for undue market power Made Available by the Noninteractive services, based on the evidence relied dates back to Web I, in which the CARP Services Do Not Demonstrate upon by SoundExchange. stated: Convergence However, such convergence and Perhaps . . . a showing that the record The Services claim that competition are swamped by the companies themselves, or even the Majors, SoundExchange misrepresents the overwhelming evidence of the could exert oligopolistic power would tempt nature of their offerings in a manner that dichotomy regarding the WTP among the panel to import a device . . . to alleviate falsely implies a convergence of features listeners. Therefore, Dr. Rubinfeld’s the market power problem. available on noninteractive services subscription-based benchmark approach Web I CARP Decision at 23 (emphasis with features available on an interactive does not demonstrate how convergence added). service. The Services make the and competition affect the relative Additionally, Dr. Rubinfeld’s model following points. royalties in the ad-supported, free-to-the treats the complementary oligopoly • The experiment that Mr. Kooker listener market. The Judges note, performed failed to demonstrate the though, that such convergence in the 102 The concept of ‘‘steering’’ is discussed at purported convergence between subscription market is suggested by the length in connection with Pandora’s rate proposal.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00033 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26348 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

pricing in the input supplier’s market as Tables 6 and 7; 5/11/15 Tr. 2870–73 offered caching as well as unlimited its potential opportunity cost. Thus, his (Katz); 5/12/15 Tr. 3123–25 (Katz).103 skips). Thus, Dr. Katz, argues, the ‘‘ratio equivalency’’ will simply sustain The Judges reject this criticism as it numerator [C] should have been whatever complementary oligopoly pertains to the narrow segment of the adjusted downward to reflect an price distortions are present in the market to which the Judges apply the additional interactivity adjustment, interactive marketplace. In the present interactive benchmark. When the which, ceteris paribus, would have case, the ability of noninteractive segment of the market at issue consists reduced the noninteractive royalty rate services to steer away from higher of willing buyers/licensees who are proposed by Dr. Rubinfeld. providing access through subscription- priced recordings and toward lower Dr. Katz correctly notes that the based listening to listeners who have a priced recordings (or threaten to do so) numerator in Dr. Rubinfeld’s so-called WTP for either interactive or ‘‘noninteractive’’ ratio contains serves as a buffer against the noninteractive services that are close revenues from services that are not supranormal pricing that arises from the substitutes, then Dr. Rubinfeld’s ‘‘ratio DMCA-compliant. Dr. Rubinfeld should impact of complementary oligopoly equivalency’’ is reasonably based on have made a further interactivity pricing that was well-documented and revenues. Dr. Katz’s critique of the adjustment to reflect whatever marginal admitted in the filings with the Federal revenue-based approach notes that Dr. value was attributable to the additional Trade Commission (FTC) by Universal, Rubinfeld failed to factor into his functionality of his stand-ins for the its economic expert and its counsel in analysis how profit, or lack thereof, to services that he used as proxies for truly connection with the Universal-EMI be realized by the noninteractive service DMCA compliant services. However, the merger. Thus, the Judges must (to would affect the royalty it would agree Judges find that, given the degree of borrow language from the CARP to pay in the hypothetical market. convergence among all services in terms decision in Web I) ‘‘import a device’’— However, in the segment of the of functionality, as discussed supra, as a steering adjustment derived from marketplace described above, a ‘‘willing it pertains to this subset of the Pandora’s benchmark, as discussed at seller’’ would not be concerned with the noninteractive market in which listeners length infra—to lower Dr. Rubinfeld’s service’s calculus of its own profits. If subscribe, the marginal additions to interactive subscription benchmark to those profits were too low to pay a functionality that Dr. Rubinfeld may reflect the effect of price competition royalty as a percentage of revenue equal have improperly captured in his to the royalties paid by the interactive and thus excise the complementary ‘‘noninteractive’’ revenue numerator do services, the ‘‘willing seller’’ simply oligopoly power and reflect an not disqualify the use of that benchmark would not supply the noninteractive in this subscription market context.105 effectively competitive noninteractive service in that hypothetical subscription subscription market. This adjustment is marketplace. That decision by the c. Failure To Rely on the Advertising- not unlike the adjustments the Judges ‘‘willing seller’’ may foreclose one or Based Noninteractive Model That make to proposed benchmarks in more services from participation in the Predominates in the Market proceedings under § 114, in that the subscription market, but, as the Judges An important and fundamental adjustment is made to align the noted in the Web II, they are not obliged problem with Dr. Rubinfeld’s analysis, benchmark rate with the statutory rate. to set the statutory rate at a level that according to Dr. Katz, lies in Dr. permits a noninteractive service to 4. Other Critiques of Dr. Rubinfeld’s Rubinfeld’s failure to acknowledge in realize any particular profit in the Interactive Benchmark his benchmark analysis that the market.104 72 FR at 24088 n. 8. advertising-based revenue model, rather a. Dr. Rubinfeld’s Use of Revenues b. Failure To Adjust for Supposed than the subscription-based revenue Instead of Service Profits ‘‘Noninteractive’’ Services Prohibited by model, is the dominant business model for noninteractive services. Katz AWRT According to Dr. Katz, the ‘‘ratio the DMCA ¶ 53 (quoting Rubinfeld CWDT ¶ 170 equality’’ assumption is also contrary to Dr. Katz further criticized Dr. (stating that Dr. Rubinfeld’s ‘‘analysis Rubinfeld’s attempt to rely on the a fundamental economic principle. The does not explicitly account for ‘free’ ad- equivalence of the aforementioned buyer, i.e., the noninteractive service, supported services.’’). Katz AWRT ¶ 55. will determine its valuation based on ratios because Dr. Rubinfeld’s This criticism was also leveled by one the profits it expects to realize from noninteractive numerator [C] is of iHeart’s economic experts, who using the input, i.e., the sound calculated from revenue received by testified, ‘‘certainly there is no basis to services that were not actually recording, not merely the revenue it may assume that subscribers are a reasonable ‘‘noninteractive,’’ but rather offered earn. Of course, the buyer’s proxy for all listeners to noninteractive functionality that rendered them non- consideration of profits necessitates the services,’’ given that subscribers account buyer’s consideration of ‘‘cost,’’ since, DMCA compliant and hence ‘‘interactive.’’ 5/16/15 Tr. 2042–50 broadly stated, profits equal revenues 105 The Judges find that such differences in (Rubinfeld) (Rhapsody unRadio offered less costs. Katz AWRT ¶¶ 50–51, 70–71; functionality are of relatively low importance in the on-demand plays, caching, and subscription market in light of the evidence of 5/11/15 Tr. 2861 (Katz). Utilizing unlimited skips, and two other services; downstream functional convergence. In this regard, Pandora’s non-license fee costs as an Slacker Radio Plus and MixRadio Plus, it is noteworthy that even Pandora’s expert Dr. example (other noninteractive services’ Shapiro (the only Service expert to propose a cost data were not readily available), separate subscription rate) has proposed a rate quite 103 Dr. Katz did not claim that his own cost similar to the rate proposed by Dr. Rubinfeld based and assuming that the non-licensing estimates or assumed equivalencies across the two on a purely subscription-based model (Those rates costs of interactive services were the markets were necessarily accurate. Rather, he are even closer to each other after an ‘‘effectively same, Dr. Katz concluded in rebuttal emphasized that his cost-based/profit-based competitive’’ steering adjustment is applied to Dr. adjustment was premised on his estimates showed Rubinfeld’s proposed subscription rate). If there that Dr. Rubinfeld’s interactivity the invalidity of Dr. Rubinfeld’s decision simply to was truly a material issue as to how WTP, adjustment would increase to 7.9 to ‘‘assume[ ] the costs were zero.’’ 5/12/15 Tr. 3123– convergence and functionality gradations impacted equalize the ratio of profits per play to 24 (Katz). royalty rates in the noninteractive subscription 104 royalties per play across the two Even in the ad-supported market, the Judges market, the Judges would have expected to see a are not setting a rate in order to provide a service much wider gulf between the SoundExchange and markets. Katz AWRT ¶¶ 74–76 and with any level of profits or revenues. Pandora subscription-based proposals.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00034 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26349

for only four percent of Pandora’s adjustment calculated by Dr. Rubinfeld ‘‘noninteractive’’ services have some listenership and zero percent of iHeart’s. when he utilized advertising revenue features that render them imperfect Fischel/Lichtman WRT ¶ 55; 5/15/15Tr. alone in his rebuttal testimony. benchmarks, then the Judges must at 3989–90 (Lichtman).106 As already noted, the Judges consider whether and how to weigh Dr. Katz also criticized Dr. Rubinfeld’s acknowledge the validity of this those imperfections. But those attempted rebuttal of this criticism. Dr. criticism by limiting Dr. Rubinfeld’s imperfections also cut in the other Rubinfeld, in rebuttal, noted that he had noninteractive benchmark analysis to direction, and indicate that the royalty estimated a 1:1.01 ratio of advertising- the segment of the market in which rates negotiated by those services reflect only revenue to royalties in the listeners are subscribers to market forces in the subscription sector, interactive service market, which he noninteractive services. Accordingly, rather than merely the statutory rates for concluded was confirmatory of there is no reason to apply this criticism DMCA-compliant noninteractive SoundExchange’s proposed rates as further to reduce the interactive services. determined by the interactive benchmark in the segment where it is subscription revenue to royalty ratio. otherwise applicable. e. Assumed Equivalence of Demand Rubinfeld CWRT ¶¶ 161–169. Elasticities in the Interactive and d. The Alleged Circularity of Dr. According to Dr. Katz, it is incorrect Noninteractive Markets Rubinfeld’s Methodology to compare only the revenues of the ad- Dr. Katz notes that Dr. Rubinfeld at supported tiers of the two types of Pandora’s economic expert, Dr. one point conceded that the ‘‘elasticities services. Rather, the proper approach, Shapiro, levies another overall criticism of demand’’ by the interactive services according to Dr. Katz, would be to of Dr. Rubinfeld’s interactive and the noninteractive services would compare the overall revenue (ad- benchmark, characterizing it as differ inter se. However, Dr. Rubinfeld supported and subscription) per play as ‘‘circular’’ and thus ‘‘uninformative.’’ failed to address or account for this between the interactive and Dr. Shapiro noted that Dr. Rubinfeld difference. Moreover, according to Dr. noninteractive services. Otherwise, asserted that the royalty rates contained Katz, Dr. Rubinfeld later equivocated as gross disparities in average revenue per in the interactive benchmark agreements to whether, in his methodology, he was play (resulting from the number of plays ‘‘can be expected to reflect the assuming an equal elasticity of demand in each model (ad-based or incremental value of the granted for both types of services. Katz AWRT functionality over-and-above what can subscription) and in revenue per play in ¶ 47; compare 5/16/15 Tr. 2029–34 with be achieved with the statutory rights.’’ each such model) would be NAB Ex. 4233. Rubinfeld CWDT ¶ 145. Thus, according camouflaged. 5/11/15 Tr. 2854–57 Given that the Judges have to Dr. Shapiro, backing out the (Katz). dichotomized between the subscription incremental value to make an When such an overall revenue and the ad-supported (free-to-the- approach was applied by Dr. Katz to the interactivity adjustment would simply return the analysis to the subscription listener) markets, the Judges do not actual service data, he found that the believe that there are any significant ratio of interactive service revenue to rates and royalties that are predicated on the existing statutory rates. uncertainties regarding the approximate noninteractive service revenue per play equivalence of the elasticities between was not 1:1, but rather 3.96:1. Katz Therefore, Dr. Shapiro criticizes Dr. Rubinfeld’s entire interactive the interactive and noninteractive AWRT ¶ 58, Table 2. This adjustment upstream markets for the right to alone would have the effect of reducing benchmarking exercise as circular, revealing nothing about the rate that acquire licenses to play sound the proposed rate derived by Dr. recordings for subscribers.107 As Dr. Rubinfeld from $0.002668 to $0.001347, would be set absent the statutory rate. Shapiro WRT at 28–29; 5/8/15 Tr. 2723– Rubinfeld testified, when the approximately a 50% reduction. Katz downstream subscription market is AWRT ¶ 59, Table 3. In similar fashion, 24 (Shapiro); accord, 5/5/15 Tr. at 4047– 48 (Lichtman) (iHeart’s’ economic competitive, the ‘‘Hicks/Marshall iHeart’s experts compared overall per relationship’’ 108 provides that if the play (or performance) data for Spotify expert noting that the noninteractive service revenue figure that is the elasticities in the downstream market and Pandora and calculated an are the same then, ceteris paribus, interactivity adjustment of 3.2, Fischel/ numerator in Dr. Rubinfeld’s noninteractive ratio is (and must be) pursuant to the Lerner Equation the Lichtman WRT ¶ 69, also reducing the mark-up of price over cost will be the rate below the rate implied by the 1.01 dependent upon the statutory rates that serve as an input cost). same in both the upstream and downstream subscription markets, 106 Dr. Rubinfeld declined to use advertising-only The Judges need to consider this interactive services as benchmarks in his original criticism in tandem with the Services’ thereby supporting Dr. Rubinfeld’s WDT. He noted that interactive services use ad- prior criticism that the so-called ‘‘ratio equivalency’’ in the subscription supported (‘‘free-to-the listener’’) alternatives as ‘‘noninteractive’’ webcasters selected by market. 5/28/15 Tr. 6310–11 tools to convert listeners into paid subscribers (the (Rubinfeld). so-called ‘‘freemium’’ model), thereby distorting Dr. Rubinfeld actually offered non- (through ‘‘upsell incentives’’) the reliability of ad- DMCA compliant features as well. supported interactive service agreements as Consequently, when Dr. Rubinfeld 107 In fact, when the dichotomy in WTP is benchmarks. Rubinfeld CWDT ¶¶ 126, 128; see also backs out the interactive value of these applied, a discussion of overall differences in Rubinfeld CWRT at 39, n128 (no ‘‘apples to apples’’ elasticities is beside the point. Elasticity measures comparison could be made between noninteractive non-DMCA compliant services (by percentage change in quantity demanded divided services, on the one hand, and, on the other, comparing the ratio of interactive to by percentage change in price. For the ad-supported interactive services that offered an ad-supported noninteractive subscription prices) he is services, the listeners have already demonstrated an (free-to-the listener) service using obtrusive not simply returning to the existing unwillingness to pay for internet webcasting. advertising as a tool to convert listeners to Economically, their demand curve is far below the subscription services.). However, in his 11th hour statutory rates, as Dr. Shapiro asserted, demand curve for subscription listeners (reflecting supplementation to his WDT, Dr. Rubinfeld because the royalty rates for those non- the differences in WTP). It is the difference in attempted to analyze certain ad-supported services, DMCA compliant services (as the location of the demand curve, not just the contained in section ‘‘III.E’’ of his CWDT, that he Services argue) are not merely difference in elasticities that is important. In the classified as more like statutory noninteractive subscriber market though, the price-elasticity of the services. The Judges’ analysis of SoundExchange’s predicated on the prior statutory rates. listeners vs. the noninteractive listeners is of some arguments relating to these so-called ‘‘III.E’’ licenses Simply put, the Services cannot have it relevance. is set forth in section IV.B.4.l.ii, infra. both ways. If Dr. Rubinfeld’s 108 See infra, note 109.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00035 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26350 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

In the present case, because: (1) The particularly relevant. Record companies Judges need not address the Services’ WTP downstream is positive (which it would want to equalize marginal returns criticisms of Dr. McFadden’s conjoint is by definition in the subscription across the interactive and noninteractive survey. market); and (2) the products are spaces, which would be accomplished h. Dr. Rubinfeld’s Equalization of the converging in terms of functionality; by focusing on subscription revenues. Number of Plays in the Interactive and and (3) an interactivity adjustment is Thus, given the Judges’ finding that the Noninteractive Markets Was applied to reflect the critical limits of market is segmented by a dichotomized Appropriate convergence (no on-demand plays on WTP, this criticism is simply not statutory services), it was not relevant to the Judges’ determination. Dr. Katz asserts that Dr. Rubinfeld unreasonable for Dr. Rubinfeld to underestimated the number of ‘‘skips’’ g. Dr. McFadden’s Survey Results Are conclude that the elasticities of demand for which an interactive service is not Unnecessary To Confirm the Value of would be approximately the same in required to pay a royalty under the Dr. Rubinfeld’s Interactivity both the interactive and noninteractive typical interactive service contracts with Adjustment, Based on the Limited subscription markets.109 However, record companies. By contrast, a Applicability of Dr. Rubinfeld’s although this likely approximate statutory service must pay a royalty for Benchmark equivalence in downstream elasticities all plays, including such ‘‘skips.’’ would tend to equalize the upstream The Services offered numerous (SoundExchange requests that the impact on the derived demand of the criticisms of Dr. McFadden’s conjoint Judges continue this requirement. See noninteractive services, it would not be survey, which was intended by SoundExchange Proposed Rates and the only factor affecting the upstream SoundExchange to confirm Dr. Terms, Attach. A at 2–3.). Dr. Rubinfeld market, i.e., the market for which the Rubinfeld’s interactivity adjustment. utilized an adjustment factor of 1.1 for Judges are setting rates. More See, e.g., Peterson Corrected WRT ¶ 110 skips, but, according to Dr. Katz, actual particularly, the inability of listeners to (survey measures potential subscribers’ data revealed in discovery demonstrated statutory services to select a particular WTP rather than actual subscription that the adjustment factor should have song combined with the noninteractive prices); 4/29/15 Tr. 924, 926, 929–33, been 1.2, a 9.1% increase in the services’ ability to (competitively) steer 936, 938 (McFadden) (survey does not adjustment that would further lower the music toward or way from record measure value of certain features); 5/22/ rate proposed by SoundExchange. Katz companies, serve to distinguish the 15 Tr. 5562–63, 5572–73, 5579–80, AWRT ¶¶ 101–102 hypothetical noninteractive 5588–89 (Hauser) (survey contains The Judges find that Dr. Rubinfeld subscription rate from the benchmark confusing feature descriptions); id. at accurately adjusted for the number of interactive subscription rate proposed 5570–71 (survey had a high participant plays across the interactive and by Dr. Rubinfeld. attrition rate, especially among noninteractive spaces. The criticism teenagers); IHM Ex. 3124 ¶ 12 (Hauser leveled by Dr. Katz focused only on the f. Failure To Use a Mix of All Interactive WRT) (survey participants were number of ‘‘skips.’’ However, Dr. Revenues (Advertising and confused by incentive alignment Rubinfeld made a further adjustment for Subscription) in the Ratios language). The Services asserted that Dr. the fact that interactive services The Services argue that Dr. Rubinfeld, McFadden’s survey would have typically paid royalties for pre-1972 rather than isolating subscription supported a rate much lower than the recordings, whereas the noninteractive revenue ratios from ad-supported ratios, benchmark rate proposed by Dr. services did not. This fact required an should have determined the value of his Rubinfeld had he corrected for Dr. increase in the noninteractive royalty interactivity adjustment by comparing McFadden’s purported errors. Fischel/ rate relative to the interactive royalty all of the actual revenue in both markets Lichtman WRT ¶ 75 and IHM Ex. 3060 rate (i.e., a smaller interactivity (i.e., a mix of subscription and (Fischel/Lichtman WRT, Ex. E.). adjustment in the denominator [D] in advertising revenue. See Katz AWRT The Judges note initially that, in this the ratios discussed in section I.A.1.c, ¶¶ 58–59 NAB PFF ¶ 368. The Judges narrow context of this subscription supra). would find that argument meritorious if market, Dr. Rubinfeld’s methodology for For example, assume there were 100 they were to attempt to apply Dr. calculating the interactivity adjustment plays in each market and in each market Rubinfeld’s ‘‘ratio equivalency’’ outside is not inappropriate. Dr. Rubinfeld 10 of those plays were pre-1972 of the subscription market. The reasonably determined the concept of a recordings. If the royalty rate criticism is inapposite, however, given ‘‘ratio equivalency’’ between revenues (assumedly) was 0.3 cents in each market, then the interactive average rate the Judges’ application of Dr. and subscription royalties in a market would be 0.3 cents. However, in the Rubinfeld’s methodology only to with both: (1) A WTP sufficient to noninteractive market, where no royalty subscription services. In the generate subscriptions in each market; was paid on the 10 pre-1972 recordings, subscription market where a positive and (2) a downstream convergence of the average royalty rate was only 0.27 WTP is self-evident from the presence of features as between the two markets, cents.112 except for the nonconvergence arising subscribers, convergence and Thus, to equalize the markets on a downstream competition are from the statutory restrictions on 110 per-play basis, the noninteractive noninteractive services. Thus, Dr. average rate must be increased. That 109 Dr. Shapiro acknowledged that the Hicks/ McFadden’s attempt to confirm Dr. Marshall relationship would serve to import the increase made the downward Rubinfeld’s 2.0 interactivity adjustment interactivity adjustment smaller, when downstream elasticities into the upstream market is unnecessary.111 Consequently, the (the ‘‘derived demand’’ effect), unless the price it was combined with the fact that—on effects of those downstream elasticities were the other side of the coin—the 110 Also by way of repetition (and emphasis), the swamped by other factors. See 5/20/15 Tr. 5044–45 noninteractive services were required to (Shapiro). The principal ‘‘swamping factor’’ is the existence of a sharp dichotomy between listeners unwillingness of a substantial segment of streaming with a positive WTP for streamed music and those listeners to pay a positive price to listen to who have essentially a zero WTP for streamed listeners’ WTP are relevant to this determination, noninteractive services. Since, by definition, music precludes an extension of this ‘‘ratio and have been considered in this determination. subscribers have a positive WTP, that ‘‘swamping equivalency’’ beyond the subscription market. 112 (90 royalty bearing songs × 0.3 cents) + (10 factor’’ does not come into play if the analysis is 111 Of course, Dr. McFadden’s conjoint survey pre-1972 songs × 0 cents) = (0.27 cents + 0 cents) limited to the market for subscription services. and his findings regarding the bimodal nature of = 0.27 cents.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00036 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26351

pay royalties for skips as though they The Judges reject the notion that there SoundExchange examined, only one has were plays, unlike the typical may be some quantum of adjustments to an escalating rate provision. Id. at 2227– interactive service. proposed benchmarks that disqualifies 28. That lone agreement with an them from consideration. Some variant escalating rate provision—the iHeart/ i. Incorrectly Weighting Average of a ‘‘three strikes and you’re out’’ Warner Agreement—was the subject of Royalties by Revenue Instead of by Play approach seems decidedly devoid of substantial criticism and ultimate Another defect in Dr. Rubinfeld’s legal or economic reasoning. The Judges rejection by Dr. Rubinfeld, as approach, according to Dr. Katz, was Dr. are more concerned with the inappropriate for use as a benchmark in Rubinfeld’s decision to compute his importance, or weight, of any given the current proceeding. Id. at 2229. average per-performance royalty by criticism of a benchmark than they are Fourth, the record evidence indicates weighting that average according to the with the number of potential that rates in SoundExchange’s own revenue per play earned by a service. adjustments. Trivial or measurable proposed benchmark market, interactive See Rubinfeld CWDT ¶ 203; 5/5/15 Tr. adjustments may be relatively great in streaming services, have decreased in 1824 (Rubinfeld). According to Dr. Katz, number, yet pale in comparison to one recent years. Rubinfeld WDT, Ex. SX weighting the per-play average by or two critical assumptions that might 0017, ¶ 140; 5/8/15 Tr. 2736–37 service revenue, as done by Dr. necessitate the qualification or rejection (Shapiro); 5/15/15 Tr. 4142 (Lichtman); Rubinfeld, created an upward bias of a benchmark. 5/19/15 Tr. 4611 (Shapiro). Further, Dr. compared to the revenue actually This determination is evidence of that Rubinfeld testified that he ‘‘actually saw earned by on-demand services that point. Dr. Rubinfeld’s benchmark fails . . . decreases in the noninteractive comprised Dr. Rubinfeld’s benchmarks. to account for the fact that a large cohort rate’’ in the data he reviewed. 5/6/15 Tr. Katz AWRT ¶¶ 42–44, 162; 5/11/15 Tr. of the listening public simply will not 2231 (Rubinfeld). Thus, if there were to 2830–34; 2837–40 (Katz). pay for streamed music. Thus, his be annual rate changes, the Services Dr. Katz maintained that the more subscription benchmark fails to capture argue, the record supports a decrease in realistic approach would have been to the very market of listeners who flock to webcasting rates during the upcoming weight the individual on-demand ad-supported (free-to-the-listener) rate period. services in the benchmark market by the noninteractive services. That single The Services do note Dr. Rubinfeld’s number of plays per service, not by the qualification circumscribes the assertion that interactive and revenue per service. Applying actual usefulness of Dr. Rubinfeld’s noninteractive services are converging, data, Dr. Katz demonstrated that using benchmark. One other criticism of his id. at 2225–2226, but they respond by Dr. Rubinfeld’s revenue weighting benchmark, viz., its failure to capture an arguing that this purported (and approach would have implied that in ‘‘effectively competitive’’ market, dubious) convergence does not support the period considered by Dr. Rubinfeld, permits an adjustment within the the conclusion that the Judges should the on-demand services would have subscription market rate and does not impose on noninteractive webcasters received $112.2 million more (42% require the Judges to reject the use of Dr. what Dr. Rubinfeld himself more) in revenues than they actually Rubinfeld’s benchmark in the characterized as a ‘‘serious increase’’ received. Katz AWRT ¶ 162. noninteractive subscription market. during the rate period. Id. at 2223. The Judges find this criticism Moreover, Dr. Rubinfeld admitted that irrelevant as applied to the subscription k. SoundExchange’s Proposed Annual his proposed annual increases were not market. In the interactive sphere, record Rate Increases From 2016–2020 are Not due to past convergence, but to his company agreements with interactive Supported by the Evidence ‘‘anticipation that the technology will services are configured pursuant to the The Services object to annual create even more convergence going ‘‘freemium’’ model, designed to convert increases in the royalties as arbitrary forward.’’ 5/5/15 Tr. 1829 (Rubinfeld). ‘‘free’’ listeners into paying subscribers, and incompatible with the willing He admitted that this ‘‘anticipation’’ who generate user revenue. See 5/7/15 buyer-willing seller standard, for the was ‘‘not based on hard data,’’ and he Tr. 2401–02 (Wilcox); 5/13/15 Tr. 3509 following reasons. conceded that ‘‘I can’t prove to you for (Herring). In the subscription market First, the Services contend that there sure where we’re going to be because we where the positive WTP and functional is no basis to assume, without are talking about the future.’’ Id. 1829– convergence engenders strong supporting theory or evidence, that rates 30. competition for paying listeners, a would necessarily increase during the For the foregoing reasons, the Services willing seller in the subscription market next rate period. In that regard, the conclude that SoundExchange’s seeks to maximize subscriber revenue Services note that Professor Rubinfeld interactive benchmark does not provide and focuses on average revenue per user admitted that there is no ‘‘theoretical a basis to set the statutory rates for (ARPU), not revenue per play. See, e.g., reason why we would expect prices just commercial webcasters in this 4/28/15 Tr. 374 (Kooker); 4/30/15 Tr. to go up.’’ 5/5/15 Tr. 1761 (Rubinfeld). proceeding. 970 (A. Harrison); see also supra, Second, he acknowledged the absence The Judges find that SoundExchange section IV.B.2.c. of any basis for his self-described has failed to make a sufficient factual ‘‘‘empirical judgment’ where we think showing that would support the linear j. The Number of Adjustments Does Not rates are likely to be going for competing $0.00008 annual rate increase proposed Disqualify Dr. Rubinfeld’s Interactive products.’’ Id. Moreover, as Dr. by Dr. Rubinfeld. The Judges find it Benchmark Rubinfeld, testified, his proposed dispositive that Dr. Rubinfeld One of the economic experts for escalating rates are not based on acknowledged that his opinion in this iHeart, Dr. Lichtman, asserted that the anticipated inflation, anticipated regard was neither based on theory nor sheer number of adjustments, as increases in music industry inputs, or on empirical analysis. Further, the fact discussed supra, needed ‘‘to draw any the consumer price index. 5/6/15 Tr. that some agreements in the benchmark analogy’’ between the interactive and 2226 (Rubinfeld). markets have annual escalators and noninteractive markets is so Third, none of the benchmarks on some do not renders those agreements ‘‘overwhelming’’ that the result is a which SoundExchange relied contained unhelpful, absent some explanation as ‘‘mess’’ and not reliable. 5/15/15 Tr. annual rate escalators. Moreover, out of to the bases for the inclusion or 4053–54. all the potential benchmarks that exclusion of such escalators.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00037 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26352 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

Additionally, market forces in the doesn’t make any sense.’’ 5/19/15 Tr. at should be characterized, and the other future may cause rates to move in either 4526 (Shapiro). Economists for both party has a strong incentive to direction, or to stay constant, and the licensors and licensees agreed that the characterize the payments in a record does not suggest a basis for a statutory rate effectively sets a ceiling particular way to influence the course of credible prediction. So too is the record on rates for statutory services, since a a future rate proceeding. devoid of any sufficient predictive service can always fall back on the That additional evidence is lacking evidence as to whether there will be statutory rate if it is unable to negotiate here. The Services raise sufficient doubt further convergence and/or competition an equal or lower rate with the as to the characterization of the between interactive and noninteractive copyright owner. See, e.g., id.; 5/27/15 compensation flowing from Apple to services or, if so, what impact that might Tr. at 6025–26 (Talley). The fact that Dr. Warner and Sony to persuade the Judges have on the rates. That is, the record Rubinfeld concludes that the effective that they cannot rely on Dr. Rubinfeld’s does not indicate why convergence rates under the Apple agreements are analysis of the Apple agreements. There would not occur through a reduction in substantially higher than the statutory is insufficient evidence in the record to interactive rates, rather than through (in rates strongly suggests that something is support SoundExchange’s analysis and whole or in part) an increase in amiss in his analysis. use of the Apple agreements.113 noninteractive rates. In sum, the record One possible reason Dr. Rubinfeld’s The uncertainty resulting from a lack does not contain a sufficient basis to analysis finds effective rates under the of evidence cuts both ways. The Judges adopt any prediction about the future Apple agreements that exceed the will not consider the licensee services’ direction of noninteractive rates. statutory rates is that he attributes alternative analyses that seek to compensation to the iTunes Radio l. Dr. Rubinfeld’s Analysis of demonstrate that the Apple agreements service that should have been attributed support their rate proposals. See, e.g., Noninteractive Agreements Does Not to other services licensed by Apple. The Corroborate His Interactive Benchmark Pandora PFF ¶ 344; Shapiro SWRT at license agreements for the iTunes Radio 12–16 & Table 1. The Services oppose service between Apple, on one hand, SoundExchange’s use of agreements and Sony and Warner, respectively, on ii. Other Noninteractive Agreements with Apple and several interactive the other, are one part of a complex The Services urge the Judges to reject services for what Dr. Rubinfeld business relationship between Apple Dr. Rubinfeld’s analysis of four described as noninteractive offerings, and the record companies, covering a additional agreements for allegedly and argue that if the Judges consider the number of different services. At or near noninteractive services: Beats Music’s agreements, a proper analysis the time that Apple entered into its The Sentence; Spotify’s ‘‘Shuffle’’ corroborates their own rate proposals iTunes Radio agreements with Sony and service; Rhapsody’s ‘‘Unradio’’; and and not SoundExchange’s. See, e.g., Warner, the parties amended some of Nokia’s ‘‘MixRadio.’’ The Services argue Pandora PFF ¶ 344; Shapiro SWRT at their existing agreements for other that each service has features that 12–16 & Table 1. services, and specified that some exceed what a service operating under For the reasons set forth below, the compensation that Apple was to have the statutory license would be permitted Judges will not consider these paid out under other agreements would to offer. The Judges agree, and find that, agreements in establishing or be characterized as payments for the as with the Apple agreements, there is corroborating a willing-buyer, willing- iTunes Radio service. Shapiro SWRT at insufficient record evidence to support seller royalty rate. 4; SX Ex. 2072 ¶ 2 (Amendment a useful analysis of these four i. Apple Agreements [REDACTED] to Apple/Warner Sound agreements. Recording cloud Service Agreement); The Services contend that Dr. Ex. 2073 ¶ 2 ([REDACTED] Amendment (A) Extra-Statutory Functionality Rubinfeld’s analysis of the Apple to Amended and Restated Apple/Sony (1) Beats ‘‘The Sentence’’ agreements is deeply flawed and Digital Music and Video Download unreliable for several reasons. First, the Sales Agreement). The Sentence was a free (to the user) Services argue that Dr. Rubinfeld SoundExchange argues that the Judges feature offered by Beats Music (Beats) as improperly allocates [REDACTED] and are bound by the parties’ a means of encouraging users to pay for other compensation to the licenses for characterization of these payments as Beats’ subscription service.114 Rubinfeld the iTunes Radio service rather than to unambiguously expressed in their CWRT ¶ 179. It allowed users to other licensed services that Apple agreements. SoundExchange Reply PFF generate a playlist by providing provides. See, e.g., Fischel/Lichtman ¶ 487. If the Judges were resolving a contextual inputs such as location, SWRT ¶ 36. Second, the services argue contract dispute between the parties, mood, setting and genre. It was subject that Dr. Rubinfeld should have analyzed SoundExchange’s argument might have to limited functionality, such as limited the parties’ ex ante expectations, rather merit. However, the Judges’ task is to skips, no use of off-line or cached than ex post performance, in determine the economic significance of content, and no rewind feature. Id. determining what a willing buyer and the compensation that changed hands ¶ 179–180. Dr. Rubinfeld describes The seller would agree to. See, e.g., 5/19/15 between the parties, and the contracts Sentence as ‘‘effectively a Tr. at 4526 (Shapiro). Finally, the are but one (albeit vitally important) noninteractive service involving services critique other adjustments that piece of evidence of that economic functionality that is closely comparable Dr. Rubinfeld makes (or fails to make) significance. Where, as here, a to other statutory services.’’ Id. ¶ 180. to the headline rates in the Apple transaction is part of a complex, The Services contend the record agreements to account for non-statutory interlocking business relationship, it is demonstrates that The Sentence functionality in Apple’s service. appropriate—even necessary—for the includes extra-statutory functionality. The Judges credit Dr. Shapiro’s Judges to consider other evidence and observation that Dr. Rubinfeld’s analysis to determine the true economic 113 In light of this determination, the Judges need conclusion that Apple was willing to value of the transaction. See Fischel/ not reach the licensee services other arguments concerning the Apple agreements. pay substantially in excess of the Lichtman SWRT ¶ 31. This is 114 Beats was acquired by Apple and, as of statutory license rate for what is particularly true when one party is December 1, 2015, no longer exists as a separate essentially a statutory service ‘‘just agnostic as to how certain payments service.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00038 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26353

Specifically, the record company MixRadio thus has significant extra- The rate derived from this analysis is agreements with Beats [REDACTED]. statutory functionality. Dr. Rubinfeld set forth in the Rates Conclusion, infra. Fischel/Lichtman SWRT ¶ 11. This does not account for this in his analysis. C. GEO’s Rate Proposals additional functionality would be (B) Lack of Analysis of Business Context expected to push the royalty rates up. In this Web IV proceeding, the Judges See id. ([REDACTED] adjusted rates Like the Apple agreements, the record had the opportunity to hear directly upward expressly to account for companies’ agreements with Beats, from a singer-songwriter who produces additional functionality that Spotify, Rhapsody and Nokia, and markets his own music. Mr. George [REDACTED]’’) (quoting IHM Ex. 3543 respectively, are part of broader Johnson, dba GEO Music, filed a at 8 (1/1/2014 Email from [REDACTED] economic relationships that include Petition to Participate in the proceeding. to [REDACTED] and [REDACTED])). Dr. other services. Id. ¶ 30. Beats, Spotify He filed all the necessary papers and Rubinfeld does not account for extra- and Rhapsody each license content from testified on both direct and rebuttal, as statutory functionality in his analysis of the record companies for their well as delivering an opening statement Beats’ license agreements. respective subscription services. Nokia and closing argument. at one time licensed music that it Mr. Johnson eloquently stated the (2) Spotify ‘‘Shuffle’’ offered for unlimited download plight of the singer-songwriter-artist (bundled with its mobile phones). As who is self-published and self- Spotify’s Shuffle service is a free-to- produced. He also proposed an the-consumer streaming service that discussed in connection with Apple, the Judges must consider evidence and overarching reform to the way in which permits the user to select a certain rights owners of music—written, number of songs (a minimum of 20 analysis of context to determine the true songs or a single album) and hear only economic value of a transaction when that transaction is part of a complex adequately adjust for, the dominant ad-supported those songs in a random order. Fischel/ (free-to-the-listener) segment of the noninteractive Lichtman SWRT ¶ 14. The ability to business relationship. Dr. Rubinfeld market. See Web III Remand, 79 FR at 23118. This select specific songs and be assured that does not analyze that context. defect has even greater resonance in this proceeding, given the abundant evidence, discussed only those songs will be played (C) Conclusion Regarding Corroborative supra, that the vast majority of listeners do not have distinguishes Shuffle from Agreements a positive WTP for access to sound recordings on noninteractive services. The increased streaming services. However, the Judges have ‘‘ring- Because Dr. Rubinfeld failed to degree of interactivity would be taken fenced’’ this defect by limiting the applicability of account for extra-statutory functionality, Dr. Rubinfeld’s analysis to the noninteractive into account in setting royalty rates. Id. and failed to analyze the broader subscription market. Second, the Judges also Dr. Rubinfeld does not account for this criticized Dr. Pelcovits in the Web III Remand for context of these services within the functionality in his analysis of Spotify’s failing to analyze agreements between the business relationship between the agreements with the record companies. interactive services and independent labels. Id. As service providers and the record discussed supra, Dr. Rubinfeld looked at certain (3) Rhapsody ‘‘Unradio’’ companies, the Judges determine that independent deals, but only made an adjustment on the assumption that Indies’ royalties would be they cannot rely on the analyses of these lower by the absence of the value of [REDACTED] Rhapsody’s Unradio service offers agreements to corroborate found in some of the Majors’ agreements with users personalized playlists based on SoundExchange’s interactive benchmark interactive services. Third, the Judges also criticized the users’ favorite artists or songs. It is analysis. Dr. Pelcovits in the Web III Remand for failing to a paid subscription service, with a 14- adjust for the downward trend in rates in the 5. Conclusion Regarding interactive benchmark market. Id. Both Dr. day free (ad-supported) trial period. Pelcovits and Dr. Rubinfeld used periods ending Rubinfeld CWRT ¶ 196. Unlike statutory SoundExchange’s Interactive during the year in which the proceeding started services, Unradio permits unlimited Benchmark Per-Play Proposal (2009 and 2014 respectively). Dr. Pelcovits used an 18-month period, while Dr. Rubinfeld used a 12- skips and permits users to play up to 25 For these reasons, the Judges find that month period. Compare id. with Rubinfeld CWDT favorites and seed tracks on an on- Dr. Rubinfeld’s interactive benchmark is ¶ 32. However, Dr. Rubinfeld acknowledged—but demand basis. Fischel/Lichtman SWRT only applicable when: failed to account for—the continuing downward ¶ 9. Again, this is extra-statutory • Revenues in both markets are trend in his interactive benchmark rates. Instead, he merely assumed that the interactive and functionality that would be expected to derived from subscription revenues and noninteractive rates would converge through an affect the royalty rate, and that Dr. are thus reflective of buyers with a increase in noninteractive rates in the hypothetical Rubinfeld did not account for in his positive WTP for streamed music; market and a decrease in rates in the interactive analysis. • functional convergence and market. Again, such an assumption may be reasonable in the subscription market, where downstream competition for potential (4) Nokia ‘‘MixRadio’’ convergence in functionality appears to exist (as listeners indicate a sufficiently high nonetheless limited by the DMCA performance Mobile phone manufacturer Nokia cross-elasticity of demand as between complement). Again, the Judges’ decision to ‘‘ring- fence’’ a subscription rate eliminates any improper bundled MixRadio, a free-to-consumer interactive and noninteractive services, use of this assumed convergence in the ad- streaming service, with its handsets.115 provided the noninteractive supported (free-to-the listener) noninteractive MixRadio provides customized, ad-free subscription rate is reduced to reflect market. Finally, in the Web III Remand, the Judges the absence of the added value of also observed that the value of Dr. Pelcovits’ noninteractive streaming. Unlike benchmark analysis was ‘‘diminished by [the] lack statutory services, MixRadio permits interactivity; and • of sufficient data’’ relating to the number of users to play radio stations that are a steering adjustment is made to noninteractive performances per subscriber. Id. Dr. cached on their mobile phones. eliminate the complementary oligopoly Rubinfeld essentially avoided this problem by not Rubinfeld CWRT ¶ 199. In addition, effect and thereby provide for an accounting for differences in the number of 116 performances made by subscribers to interactive MixRadio permits users to share music effectively competitive market price. and noninteractive services, respectively. Again, the with non-subscribers. Fischel/Lichtman Judges find that because a willing seller in the SWRT ¶ 12. 116 The Judges find as well that Dr. Rubinfeld’s streaming subscription markets would seek to interactivity analysis failed to cure all of the defects equalize Average Revenue per User (ARPU) that the Judges found to exist in the similar (through Dr. Rubinfeld’s ratio equivalency 115 The service is now simply ‘‘MixRadio,’’ as a interactivity analysis proffered by Dr. Pelcovits and approach) this issue as well has been adequately result of Microsoft’s acquisition of Nokia, and rejected by the Judges in the Web III Remand. First, addressed by the Judges through their ‘‘ring- subsequent sale of the MixRadio service to Line and of greatest importance, Dr. Rubinfeld’s fencing’’ of Dr. Rubinfeld’s benchmark analysis to Corporation. interactivity model fails to take account of, or the subscription market only.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00039 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26354 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

published, performed, recorded, ‘‘monetize’’ the change of the music GEO proposes that royalty rates for broadcast—would be paid for their product paradigm from an ownership subscription webcast streams be the artistic creations. However, the current model (purchase of physical recordings) greater of a per-performance rate and a law thoroughly segments both the to an access model (log in to Internet percentage revenue rate: copyrights and the licensing services and use as much or as little mechanisms. The rights and their control as one wants to direct the music Per- Percentage treatment have evolved over time, programming). Year performance of revenue barely keeping pace with the technology GEO makes three separate rate rate that uses them. Further, part of the proposals. 2016 ...... $0.22 70 music royalty process, i.e., royalties for 1. GEO Proposal 1 2017 ...... 0.24 68 use of published ‘‘musical works’’ is 2018 ...... 0.26 66 managed by a U.S. District Court in New GEO proposes that royalty rates for 2019 ...... 0.28 64 York, with statutory admonition to the nonsubscription webcasting be the 2020 ...... 0.30 62 court not to consider the effect of the greater of a per-performance rate and a rates set by the Judges. See 17 U.S.C. percentage revenue rate: 114(i). The complete picture urged by Id. Mr. Johnson can only come into focus Per- Percentage 2. GEO Proposal 2 with a new copyright law. Year performance of revenue Nonetheless, by comparing an artist’s rate As an alternative, GEO proposes a revenues from physical phonorecords to 2016 ...... $0.10 70 combination of a one-time fee the current ten-thousandths of a cent 2017 ...... 0.12 68 (described as a ‘‘cloud locker’’ fee) and ‘‘per spin’’ calculations for digital 2018 ...... 0.14 66 a ‘‘usage’’ fee that is the greater of a per- performances, Mr. Johnson highlighted 2019 ...... 0.16 64 performance royalty and a percentage of very effectively one of the paramount 2020 ...... 0.18 62 revenue. As with Proposal 1, GEO factors complicating this proceeding. proposes separate rates for subscription The music makers, the music recorders, Introductory Memorandum to the and nonsubscription webcast streams. and the music ‘‘consumers’’—both Amended Testimony and Written Direct broadcasters and listeners—are Statement of George D. Johnson at 4 GEO’s proposed nonsubscription rates struggling with how to address and (Jan. 13, 2015). are:

Copyright Per- Year cloud locker— performance Percentage one time fee rate of revenue

2016 ...... $0.50 $0.01 70 2017 ...... 0.55 0.02 68 2018 ...... 0.60 0.03 66 2019 ...... 0.65 0.04 64 2020 ...... 0.70 0.05 62

Id. at 5. GEO’s proposed subscription rates are:

Copyright Per- Year cloud locker— performance Percentage one time fee rate of revenue

2016 ...... $0.50 $0.10 70 2017 ...... 0.55 0.12 68 2018 ...... 0.60 0.14 66 2019 ...... 0.65 0.16 64 2020 ...... 0.70 0.18 62

Id. Copyright Per- between.’’ Id. at 23.117 As discussed Year cloud locker— Performance above, the Judges conclude that the 3. GEO Proposal 3 one time fee rate evidence in the record before us does As a third alternative, GEO Proposal 2016 ...... $0.50 $0.01 not support a greater-of rate structure or 3 consists of a one-time ‘‘cloud locker’’ 2017 ...... 1.00 0.02 a percentage-of-revenue rate in the fee and a per-performance rate. Proposal 2018 ...... 1.50 0.03 current proceeding. GEO provided no 3, which GEO describes as being 2019 ...... 2.00 0.04 evidence to change that holding. derived from the inflation-adjusted cost 2020 ...... 2.50 0.05 of a record album in 1964, would apply Id. at 6. to both subscription and nonsubscription web streams. Id. at 6– 4. Judges’ Conclusions With Respect to 117 7. GEO’s Rate Proposals See also id. at 5 (‘‘the Per-Performance Rate and Copyright Cloud Locker One-Time Fee Rate are GEO requests that the Judges adopt what GEO is proposing’’). either Proposal 3 or Proposal 2, ‘‘or in

VerDate Sep<11>2014 21:36 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00040 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26355

Likewise, the Judges find no rate and he provided no evidence to D. Pandora Rate Proposal persuasive evidence to support a ‘‘cloud support such a rate, the Judges find that locker’’ fee of the type that GEO (and there is insufficient evidence in the 1. Proposed Royalties only GEO) proposes. Mr. Johnson record to support a cloud locker rate. Pandora is a noninteractive licensee, presented no expert testimony to Therefore, the Judges are left with Mr. and it represents itself as ‘‘the leading support a ‘‘cloud locker’’ rate, nor did Johnson’s proposed per-performance Internet Radio Service in the United he provide any evidence that such a rate rates. The per-performance rates he States.’’ PAN Ex. 5002 ¶ 5 (Fleming- structure even exists in the market. proposes range from a low of $0.01 per Wood WDT). Like SoundExchange, What he did provide is his statement: stream ((2016 in Proposal 2 Pandora proposes a greater-of rate ‘‘The streamer’s economic model leaves (nonsubscription) and Proposal 3) to structure. Commercial webcasters out one crucial element—the customer, $0.30 per stream (2020 Subscription). and the bundled copyright cloud locker As with the cloud locker proposal, Mr. would pay the greater of 25% of revenue or ‘streaming account’ forces payment Johnson provides no evidence, other from eligible transmissions and a range for all music copyrights up-front, one than his personal view, that such rates of per-performance royalty rates. time, like all other products.’’ Id. at 5– are reasonable, or reflect what a willing Pandora proposes separate ranges of 6. The rates the Judges adopt must be buyer and a willing seller would agree royalty rates for subscription and based on substantial evidence in the to.118 In the absence of such evidence, nonsubscription (advertisement record. As Mr. Johnson is the only the Judges cannot adopt Mr. Johnson’s supported) commercial webcasting as participant to propose a cloud locker proposed per-performance rates. follows:

LOW END OF PROPOSED RANGE 119

A royalty equal to the greater of (i) or (ii) below: Per-performance Per-performance Year (nonsubscription) (subscription)

(i) Per-Play Rate: 2016 ...... $0.00110 $0.00215 2017 ...... 0.00112 0.00218 2018 ...... 0.00114 0.00222 2019 ...... 0.00116 0.00226 2020 ...... 0.00118 0.00230

(ii) 25% of Revenue from Eligible Transmissions

HIGH END OF PROPOSED RANGE

A royalty equal to the greater of (i) or (ii) below: Per-performance Per-performance Year (nonsubscription) (subscription)

(i) Per-Play Rate: 2016 ...... $0.00120 $0.00224 2017 ...... 0.00123 0.00228 2018 ...... 0.00125 0.00232 2019 ...... 0.00127 0.00236 2020 ...... 0.00129 0.00240

(ii) 25% of Revenue from Eligible Transmissions

Pandora’s Second Amended Proposed Merlin Agreement, which established Agreement. PAN Ex. 5014; Shapiro Rates and Terms at 2–3. terms and conditions under which WDT at 23, 26; PAN Ex. 5007 ¶ 24 (Herring WDT). 2. Pandora’s Noninteractive Benchmark Merlin granted Pandora the right to perform of all the sound recordings in a. Merlin Pandora relies upon the Pandora/ the catalogs of those Merlin record Merlin Agreement to support its rate companies that would ultimately decide Merlin is a global rights agency that proposal. On June 16, 2014, Pandora to opt-in to the Pandora/Merlin represents and collectively negotiates on and Merlin entered into the Pandora/ behalf of thousands of independent

118 See, e.g., id. at 7 (‘‘[w]hoever says that songs which Mr. Johnson would then adjust for inflation Day’s Night,’’ both of which are still (or again) are too expensive in this rate hearing at $.00 are and round to 50 cents per song). Id. at 7–8. Mr. available, in vinyl, on Amazon.com for prices nothing more than con-men since they expect Johnson also estimates that a Beatles record generally ranging from $15 to $20. See American music creators to work literally for $.00 purchased for $5 in 1964 would have cost, after beatlesbible.com, referenced on Dec. 14, 2015; per-song when a song really costs $5 dollars [sic] adjusting for inflation, $38 in 2014. Id. at 6. Since Amazon.com, referenced Dec. 14, 2015. per song using government low-end inflation the Judges decline to adopt a cloud locker rate, they 119 The low and high ends of the proposed range calculations and a real world 1964 benchmark.’’). need not decide whether the mechanical rate in correspond to levels of overspinning (or ‘‘steering’’) To establish his proposed cloud locker rate, Mr. effect in 1909, adjusted for inflation, would be a of Merlin-member tracks under Pandora’s Johnson requests that the Judges adopt as a suitable benchmark for Section 114 and 112 rates benchmark agreement. The issue of steering and the benchmark a 2-cent mechanical (section 115) for 2016–2020. Interestingly, the Beatles released rate calculations derived from steering are license rate for musical works in effect in 1909, two albums in 1964, ‘‘Beatles for Sale’’ and ‘‘A Hard described elsewhere in this determination.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00041 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26356 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

record companies in the United States Pandora’s revenue, prorated based on FOR PANDORA’S SUBSCRIPTION and 38 other countries. Van Arman the share of performances on Pandora SERVICE WDT at 10; 6/1/15 Tr. 6865 (Lexton); see accounted for by the Merlin Labels. also 5/18/15 Tr. 4204 (Herring). Merlin’s • The 2014 ‘‘headline’’ per-play rates 2014 members include numerous prominent are $0.[REDACTED] for each ad- independent labels, which produce supported performance and Headline Rate ...... Steered Rate. commercially and critically successful $0.[REDACTED] for each subscription $0.[REDACTED] $0.[REDACTED]. music. See Pandora PFF ¶¶ 123–126 performance. The 2015 ‘‘headline’’ per- 2015 (and record citations therein). play rates are $0.[REDACTED] for each These independent record companies ad-supported performance and Headline Rate ...... Steered Rate. negotiate with digital services $0.[REDACTED] for each subscription $0.[REDACTED] $0.[REDACTED]. collectively through Merlin in order to performance. PAN Ex. 5014 ¶ 3(a); obtain more favorable terms and Herring WDT ¶ 26; Shapiro WDT at Thus, Pandora claims that steering transaction cost savings than they 26.121 reduced the headline rates for its ad- otherwise could achieve on an Steering Provisions supported, nonsubscription service by individual basis. Van Arman WDT at 10; Steering is the term Pandora uses to [REDACTED]% in 2014 and would 4/28/15 Tr. 626–7 (Van Arman); 6/1/15 describe a licensee’s ‘‘ability to control reduce those headline rates by Tr. 6856–7 (Lexton). Pandora notes that the mix of music that’s played on the [REDACTED]% in 2015. Moreover, one of the Majors has acknowledged service in response to differences in Pandora claims that steering reduced that Merlin is a ‘‘virtual [ ] major.’’ PAN royalty rates charged by different record the headline rates for its subscription Ex. 5349 at 9 (‘‘[REDACTED]); 5/5/15 Tr. companies.’’ 5/8/15 Tr. 2683–4 service by [REDACTED]% in 2014 and 1969:19–23, 1975:8–1977:4 (Rubinfeld). (Shapiro). Just as the ‘‘ratio equality’’ is would reduce that headline rate by Merlin established a procedure for its foundational to SoundExchange’s rate [REDACTED]% in 2015. PAN Ex. 5014 members to either opt-in or opt-out of proposal, the concept of ‘‘steering’’ is ¶ 4; Herring WDT ¶ 27; Herring AWRT the Pandora/Merlin Agreement (most foundational to Pandora’s rate proposal. ¶ 48; Shapiro WDT at 27. members could [REDACTED], whereas a Shapiro WDT at 27 (‘‘This reduced per- The calculation of these effective small number of members reserved the play rate in exchange for increased steered rates is explained in paragraph right to [REDACTED]). Members who plays is the central piece of the Merlin 4 of the Pandora/Merlin Agreement, were represented by independent Agreement.’’). which sets forth the following distributors (i.e., distributors According to Pandora, steering and provisions for calculating the rates unaffiliated with the Majors) delegated the concomitant discounting terms are resulting from steering, using the 2014 the decision as to whether to opt-in to feasible in the noninteractive market ad-supported headline rate of these distributors. In total, [REDACTED] because Pandora has now tested and $0.[REDACTED] as an example. Pandora promises to increase of approximately [REDACTED] proven its ability to modify its playlist- ‘‘quantity’’ (spins) by at least members, covering approximately selecting algorithms to rely more or less [REDACTED]% in the aggregate above [REDACTED] tracks—opted in to the heavily on the music of particular Merlin’s ‘‘Natural Performance Pandora/Merlin Agreement. 5/18/15 Tr. record companies so that it can steer its Rate.’’ 122 However, Pandora will not 4221, 4235 (Herring); 6/1/15 Tr. 6870 listeners toward or away from the music pay a ‘‘price’’ equal to the (Lexton). from any one record company, thereby $0.[REDACTED] headline rate for these Pandora notes that, by statute, the permitting ‘‘workable competition’’ to additional spins. Instead, in exchange opting-in Merlin members could have emerge in the relevant, noninteractive for its promise to play at least declined to enter into the Pandora/ webcasting market. 5/19/15 Tr. 4557 [REDACTED] % additional spins, Merlin Agreement and thus remained (Shapiro). By contrast, Pandora notes, Pandora will receive a ‘‘discount’’ on bound in 2014 and 2015 by the statutory no evidence of such a steering capability the price paid for [REDACTED]. rates that incorporated the Pureplay existed at the time of the Web II or Web That discount is calculated as settlement rates. See PAN Ex. 5014 III proceedings. Shapiro WDT at 16. ¶ 1(r); Herring WDT ¶ 25.120 [REDACTED]. PAN Ex. 5014 ¶ 4(a)–(c). Pursuant to the Pandora/Merlin In support of a statutory rate based on b. Key Provisions of the Pandora/Merlin Agreement, the ‘‘headline’’ per-play the steering aspects of the Pandora/ Agreement. rates can be reduced by steering as Merlin Agreement, Pandora advances According to Pandora, the key terms follows. several arguments. First, Pandora of the Pandora/Merlin Agreement are maintains that steering embodies ‘‘price those that set forth the rate structure, FOR PANDORA’S AD-SUPPORTED competition at work,’’ and therefore royalty payments, and steering NONSUBSCRIPTION SERVICE reflects an ‘‘effectively competitive’’ provisions: market. 5/19/15 Tr. 4561–64 (Shapiro). Rate Structure and Royalty Payments 2014 Effective competition results from the • power to steer because, according to Dr. The agreement employs a greater-of Headline Rate ...... Steered Rate. royalty structure, with Pandora paying $0.[REDACTED] $0.[REDACTED]. Shapiro, a streaming service that the greater of a per-play prong and a possesses an ability to ‘‘steer’’ towards percent-of-revenue prong. The percent- 2015 certain recordings, and away from of-revenue prong specifies 25% of others, will have ‘‘much more Headline Rate ...... Steered Rate. 120 The statutory Pureplay settlement rates for $ 0.[REDACTED] $0.[REDACTED]. 122 The Pandora/Merlin Agreement defines 2014 and 2015, respectively, are 13¢ and 14¢ per ‘‘Natural Performance Rate’’ as [REDACTED]. PAN 100 plays for advertising-supported services (or Ex. 5014 ¶ 1(k). More specifically, Pandora 25% of revenue, whichever is greater), and 23¢ and 121 There is no separate fee in the agreement for promised an aggregate increase of Merlin-label 25¢ per 100 plays, respectively, for subscription ephemeral copies of the recordings; such copies are spins of at least [REDACTED]%, while promising to services in 2014 and 2015. Notification of covered under and included within the [REDACTED] to increase the spins of individual Agreements Under the Webcaster Settlement Act of performance fees above. PAN Ex. 5014 ¶ 3(d); Merlin member labels by at least that amount. Id. 2009, 74 FR 34796, 34799 (July 17, 2009). Herring WDT ¶ 26. ¶ 4(a).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00042 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26357

bargaining power and be able to price competition that can obviate the 5. The Science Team is composed of 15 negotiate a lower royalty rate.’’ Shapiro need for any actual steering in the individuals, 13 of whom hold doctorate WRT at 19. hypothetical market. Shapiro WDT at 9 degrees in computer science, In theoretical terms, a service’s ability (‘‘The net result in a workably engineering, statistics, or economics to steer increases its price elasticity of competitive market may well be from leading academic institutions. Id. demand, reducing the extent to which a relatively little actual steering . . . .’’). ¶ 5. licensor can mark up its price over Pandora avers that the Pandora/ Pandora’s controlled experiments marginal cost. 5/19/15 Tr. 4561–64 Merlin Agreement’s steering provisions (including the steering experiments) (Shapiro); 5/8/15 Tr. 2725–27 (Shapiro); reflect these competitive forces, i.e., a consist of comparisons between Pandora PFF ¶¶ 147–148, 152–157 (and supplier offering a lower price in an randomly selected groups of listeners, record citations therein). The attempt to gain volume. Shapiro WDT at one group receiving a manipulated relationship among elasticity, price and 27 (‘‘This reduced per-play rate in experience (the ‘‘treated’’ group) and the costs as a basis to measure market exchange for increased plays is the other group receiving the standard power is described by the Lerner central piece of the Merlin Agreement. Pandora experience (the ‘‘control’’ Equation (or Lerner Index)—a This feature plainly demonstrates that group). Id. These experiments are fundamental economic pricing rule. the Merlin Agreement is embracing the randomized, controlled, and blind. Shapiro WDT at 5. In mathematical workings of a competitive market.’’); Id.124 terms, the Lerner Equation 123 can be Shapiro WRT at 19; see 5/19/15 Tr. Pandora initiated the steering expressed as: 4574–5 (Shapiro). experiments because: (1) It had the According to Pandora, from the general technological capability to ‘‘willing buyer’’ perspective, the ability perform more of one record company’s to steer provides Pandora with the sound recordings and/or fewer of ‘‘competitive incentive to play directly- another record company’s sound licensed tracks more heavily than [it] recordings; and (2) it recognized that, as Second, Pandora asserts that steering would otherwise.’’ Herring AWRT ¶ 48. a noninteractive service it has the is not only theoretical and a contractual On the other side of the transaction, economic incentive to ‘‘steer’’ its commitment, it is occurring under the according to Pandora, the record shows performances toward music owned by a Pandora/Merlin Agreement. that for a ‘‘willing seller,’’ i.e., a Merlin particular record company if that music Specifically, Pandora is actually steering member who opted-in, this steering- is available at a lower royalty rate. [REDACTED]% above Merlin’s ‘‘natural based agreement, constituted a ‘‘good Shapiro WRT at 22–25. Therefore, performance rate’’ of sound recordings, competitive move,’’ taken in the record Pandora decided to determine through greater than the [REDACTED]% it has company’s ‘‘self-interest.’’ 4/28/15 Tr. its steering experiments whether and to contractually committed to steer— 610–11 (Van Arman). what extent it could use this evidencing that Pandora’s steering Pandora further avers that its technological ability to steer behavior is motivated by ‘‘price ‘‘overspinning’’ of Merlin tracks by performances without negatively differences,’’ not merely by the [REDACTED]% has not resulted in any affecting listenership. Herring WDT contractual ‘‘steering commitment.’’ negative feedback from Pandora ¶¶ 22, 31–32; McBride WDT ¶¶ 12–22; Shapiro WRT at 41; see 5/18/15 Tr. 4229 listeners or any negative financial Shapiro WDT at 27; Shapiro WRT at 22– (Herring); Herring AWRT ¶ 50. impact. 5/18/15 Tr. 4229–33 (Herring) 25. Dr. Shapiro noted that when steering (explaining that Pandora increased Thus, from June 4, 2014, to September is possible, the mere threat (explicit or plays of Merlin tracks, on an aggregate 3, 2014 (13 weeks), Dr. McBride and his implicit) by the service to divert basis, by approximately [REDACTED]% colleagues at Pandora conducted a performances from one record company in 2014, but this change in the mix of series of steering experiments in order to another gives the service negotiating spins did not cause any increase in to answer two questions: (1) Whether leverage.’’ Shapiro WRT at 20 (emphasis ‘‘complaints about song quality from increases or decreases in performances added). In such a market, he opines, ‘‘[a] Pandora listeners). of sound recordings owned by a record company facing a webcaster with particular record company would have considerable ability to steer customers c. Pandora’s Steering Experiments a measurable impact on a key listener away from its music has a strong In support of its assertion that the metric (average hours listened per incentive to discount its music to effects of potential steering can be registered user; and (2) whether increase the number of performances of pervasive in the noninteractive market, Pandora’s engineers could precisely its music made by that webcaster.’’ Pandora relies in part on its own manipulate the share of music played Shapiro WDT at 9–10. Thus, according internal ‘‘steering experiments.’’ More according to the record company that to Pandora, the ability to steer creates particularly, in 2014, at Dr. Shapiro’s owns the recordings. McBride WDT direction, Pandora conducted a set of ¶¶ 7, 12, 15. 123 The Lerner Equation states that there is an inverse relationship between the firm’s margin (the steering experiments to test its ability to The Steering Experiments consisted of gap between price and marginal cost) and the firm’s overspin recordings owned by each of a group of 12 experiments. Each elasticity of demand. That is, the increase in a the Majors. experiment involved a combination of buyer’s (licensee’s) own elasticity of demand (n) The 2014 steering experiments were one of three target ownership groups reduces the price (P) paid by the licensee over the licensor’s marginal cost (MC) pursuant to the Lerner conducted by Pandora’s in-house Equation. Thus, an increase in own-elasticity n ‘‘Science Team’’ which has primary 124 ‘‘Randomized’’ means listeners are assigned (holding MC constant) reduces the value of each responsibility for designing and randomly to either the ‘‘treated’’ group or the side of the equation. See generally Edwin Mansfield analyzing ‘‘controlled experiments.’’ ‘‘control’’ group, to ensure valid causal inference. and Gary Yohe, Microeconomics 376 (11th ed. Id. at n.1. ‘‘Controlled’’ means the outcome is a 2004) (‘‘Economists often use the Lerner Index . . . PAN Ex. 5020 ¶ 7 (McBride WDT). comparison between those receiving the exposure to measure monopoly power or market power.’’). Pandora witness Dr. Stephen McBride is and those not receiving the exposure, to account for [NB: The formula for the Lerner Equation appeared a member of Pandora’s Science Team, the ‘‘placebo effect.’’ Id. ‘‘Blind’’ means in a footnote in the determination as issued to the which performs research and analyses experimental subjects are unaware of their parties and the public, but it appears here in the participation in an experiment and, therefore, are body of the publication because of Federal Register to measure the effectiveness of features also unaware of whether they have been assigned drafting requirements.] offered by Pandora. McBride WDT ¶¶ 1, to the treatment group or the control group. Id.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00043 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 ER02MY16.001 26358 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

(UMG, Sony or WMG) and a target • Ancillary Promotional Benefits: considered for potential additional ‘‘deflection’’ in share of spins (treatment Additional non-pecuniary promotional adjustments were: group) as compared to spins that would benefits for Merlin, including 1. Differences in the determination of occur according to the standard Pandora [REDACTED]. See PAN Ex. 5014 ¶¶ 6– which performances are compensable as music recommendation results (control 11. 125 compared to the statutory license (i.e., group). McBride WDT ¶ 15. The spin See Herring WDT ¶ 30; Shapiro WDT at consistent treatment of [REDACTED] share deflections (the ‘‘steering’’) were: 29. and [REDACTED]); ¥30%, ¥15%, +15%, and +30% for each of the three ownership groups e. Pandora’s Conclusion Regarding the 2. additional financial terms of the manipulated. Id. The experimental Benchmark Status of the Pandora/ Pandora/Merlin Agreement, including subjects of the Steering Experiments Merlin Agreement [REDACTED]; and were all Pandora listeners, each of Based on the foregoing, Pandora 3. non-pecuniary terms in the whom was randomly assigned to one of asserts that the Pandora/Merlin Pandora/Merlin Agreement. the 12 treatment groups, to the single Agreement is the best benchmark in this 5/19/15 Tr. 4592–93 (Shapiro); Shapiro control group, or were included in the proceeding because WDT Appendix D at D–1–D–9; see portion of listeners excluded from all • it constitutes a competitive and Shapiro WDT at 30. experiments. McBride WDT ¶ 16. arms-length direct license between a The experiments demonstrated that noninteractive webcaster and thousands ¥ i. Adjustment for Royalty Bearing Plays Pandora was able to steer 15% or of record companies; ([REDACTED]) +15% for all three Majors without • it concerns the same rights as are causing a statistically significant change covered by the statutory license; This adjustment is required, in listening behavior. McBride WDT • it covers the same type of products according to Dr. Shapiro, because, on ¶ 21. However, Pandora was unable to the one hand, the Pandora/Merlin ¥ at issue in this proceeding—public steer 30% or + 30% for Universal or performances of sound recordings on Agreement treats [REDACTED] as non- Sony without creating a statistically noninteractive Internet radio; and compensable and the performance of significant change in listening behavior. • it involves the same ‘‘willing [REDACTED] as compensable, but the Id. sellers’’ (record companies that own statutory licenses takes the opposite d. Additional Terms in the Pandora/ sound recording copyrights) and a tack on both issues—treating Merlin Agreement 126 ‘‘willing buyer’’ (Pandora) that exist in [REDACTED] as compensable and the performance of [REDACTED] as non- The Pandora/Merlin Agreement the hypothetical market. compensable. Id. To adjust for both of contains the following additional terms PAN Exs. 5014–5015; Shapiro WDT at that are specifically addressed by Dr. 24–25; see also 5/28/15 Tr. 6323–24 these factors Dr. Shapiro took the Shapiro in his benchmark analysis: (Rubinfeld) (agreeing that the Pandora/ following steps. • [REDACTED]: Pandora also agreed Merlin Agreement satisfied each such First, he calculated the total payment to provide the Merlin members who criterion). Pandora expected to make to the opting- opted in with a [REDACTED] in the 3. Pandora’s Calculation of Royalty in Merlin members for all sound event Pandora [REDACTED]. PAN Ex. Rates Implied by Its Proposed recordings under the Pandora/Merlin 5014 ¶ 3(e); Herring WDT ¶ 26; Shapiro Benchmark Agreement. WDT at 28–29. This provision has not Second, he divided that total payment been triggered, 6/1/15 Tr. 6897 (Lexton), Pandora and its economic expert, Dr. Shapiro, did not simply apply the by the number of performances of and Merlin’s negotiators understood it Merlin Label recordings that would be was unlikely ever to be triggered. Id. at steering-adjusted rates implied by the Pandora/Merlin Agreement, but rather compensable under the statutory license 6956–57; PAN Ex. 5110. (as currently defined). Shapiro WDT at • Compensable Performances: also considered potential further 30–31; Appendix D. Performances of [REDACTED] are non- adjustments that might be required for compensable. All other performances an ‘‘apples-to-apples’’ comparison of the Dr. Shapiro describes this calculation are subject to a fee. 5/18/15 Tr. 4227 terms in the Pandora/Merlin Agreement as yielding a per-play rate that the (Herring). Certain tracks designated as with the statutory terms applicable to Pandora/Merlin Agreement would [REDACTED] are compensable at only noninteractive licenses. See Shapiro establish if Pandora and Merlin had [REDACTED] the headline rates. 5/18/15 WDT at 20–21, 23–37, Appendix D negotiated an agreement with a fixed Tr. 4227 (Herring). (‘‘Analysis of Merlin Agreement’’). per-play rate that treated [REDACTED] • [REDACTED]: The Merlin members as compensable and performances of who opt-in are [REDACTED] to receive a. Potential Additional Adjustments [REDACTED] as non-compensable. Id. a specified [REDACTED]. PAN Ex. 5014 The three principal aspects of the To make the point more clearly, Dr. ¶ 5; Herring WDT ¶ 29. Merlin Agreement that Dr. Shapiro Shapiro offered the following example:

Calculation Value

Pandora Performances of Merlin Music ...... [a] ...... 1,000,000 Number of [REDACTED] ...... [b] ...... 200,000 Number of [REDACTED] ...... [c] ...... 100,000 Compensable Performances Under Merlin License ...... [d] = [a]¥[b] ...... 800,000 Payment Per Compensable Play Under Merlin License ...... [e] ...... $0.00125 Total Royalty Payment Under Merlin License ...... [f] = [d] × [e] ...... $1,000

125 The Steering Experiments operated through (the ‘‘B’’ group, or treated group) and then compares 126 Dr. Shapiro’s decision as to whether and to Pandora’s ‘‘A/B Framework,’’ by which the Science the effects to groups of listeners who did not what extent to adjust his benchmark to reflect such Team intentionally changes one aspect of the experience the change (the ‘‘A’’ group, or control additional terms is considered elsewhere in this Pandora experience for a sample group of listeners group). McBride WDT ¶¶ 7–8 and 16. determination.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00044 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26359

Calculation Value

Compensable Performances Under Statutory License ...... [g] = [a]¥[c] ...... 900,000 Effective Per-Play Rate Under Statutory License ...... [h] = [f] ÷ [g] ...... $0.00111

Shapiro WDT at 30–31; 5/19/15 Tr. iii. Potential Adjustments for Non- Shapiro took note that Pandora believed 4589–92 (Shapiro); see id.at 4594 Pecuniary Terms the presence of [REDACTED] might be (noting that $0.[REDACTED] rate was The Pandora/Merlin Agreement also ‘‘accretive to the listener experience’’ as ‘‘an illustrative example,’’ and ‘‘not a contains non-pecuniary financial terms well as a form of advertising, and that rate proposal’’).127 that are not permitted in the statutory Pandora was in fact planning controlled license. Dr. Shapiro attempted to tests to measure listener responses and ii. Potential Adjustments for Additional solicit listener feedback in order to determine whether it was appropriate to Financial Terms determine the appropriate nature and increase his proposed rate to reflect any frequency of [REDACTED] on values for these items. Shapiro WDT at The Pandora/Merlin Agreement [REDACTED] stations.’’ Id. In light of 29–31; Appendix D at D–10–19 (‘‘Non- contains additional financial terms not the mutually beneficial nature of Pecuniary Terms in the Merlin permitted in the statutory license. Dr. bumpers, Pandora personnel informed Agreement’’); see 5/19/15 Tr. 4595–98 Shapiro attempted to determine whether Dr. Shapiro that, even without a (Shapiro). it was appropriate to increase his contractual obligation to do so, Pandora proposed rate to reflect values for these (A) [REDACTED] on Pandora offered [REDACTED], gratis, along with items. Dr. Shapiro ultimately found no Dr. Shapiro did make an adjustment Pandora Premieres tracks. Shapiro WDT basis to increase his proposed rates to to increase his calculated ‘‘steered’’ rate at D–13 & n.26. reflect these items. Shapiro WDT at 28– In light of the foregoing, Dr. Shapiro by 0.0002¢ (i.e., $0.000002) per- 29 (Appendix D); see 5/19/15 Tr. 4592– could not conclude that the performance to reflect [REDACTED] [REDACTED] provision on balance 93 (Shapiro). Broadly, Dr. Shapiro found made available by Pandora to Merlin in no value in these additional terms created more value for Merlin than for [REDACTED] of the Pandora/Merlin Pandora, and therefore he made no because neither Pandora nor Merlin had Agreement. Shapiro WDT at 31; Shapiro calculated or even estimated any value adjustment to his proposed effective WDT at 31; Appendix D at D–11 to D– royalty rate on that basis. attributable to these items. More 12. particularly, Dr. Shapiro analyzed these (D) Access to Pandora Metrics additional financial terms in the (B) [REDACTED] Pursuant to the Pandora/Merlin following manner. Pursuant to the Pandora/Merlin Agreement, opting-in Merlin members Agreement, Pandora agreed to allow (A) The [REDACTED] Provision will receive [REDACTED] metrics each Merlin member that had opted-in regarding [REDACTED]. PAN Ex. 5014 Dr. Shapiro assigned no separate to [REDACTED]. PAN Ex. 5014 § 7. Dr. § 9 (Pandora/Merlin Agreement) see also value to Merlin’s contractual right to Shapiro did not make an adjustment to Shapiro WDT at D–14 & n.29); Herring receive [REDACTED]. According to Dr. increase the value his benchmark for WDT ¶ 30. Shapiro, he made no adjustment to his this non-statutory benefit, because However, Dr. Shapiro noted that, at proposed rate to reflect this term Pandora personnel told him that the time he prepared his testimony, because Pandora’s financial projections ‘‘[REDACTED] are mutually beneficial Pandora was also developing a service did not show that Pandora would to the Merlin Labels and to Pandora.’’ called the Artist Marketing Platform [REDACTED] in 2014 or 2015. Id. at Shapiro WDT at D–12. With regard to (‘‘AMP’’), expected to launch in October 4689–90. the benefit to Pandora, Dr. Shapiro was 2014, through which Pandora proposed informed by Pandora personnel that to provide these same metrics to all (B) The [REDACTED] ‘‘Pandora considers that [REDACTED] artists, not only to artists on the labels strengthen artist engagement with of Merlin members. Pandora did not Dr. Shapiro also assigned no separate Pandora and thereby drive incremental plan to charge for AMP. Shapiro WDT value to the [REDACTED] that provided listening and listeners to the service, at D–14 & n.30; see Herring WDT ¶ 30. Merlin with [REDACTED]. He testified build brand loyalty, and enhance Since Pandora stated that it intended that he declined to add a separate value listener retention.’’ Id.; see Westergren to make its AMP available to all artists for [REDACTED] because: WDT ¶ 38. Accordingly, Dr. Shapiro at no charge, Dr. Shapiro concluded that [The] rate proposal is based on payments could not determine that the value of no adjustment to the effective royalty that Pandora is making and will be making such [REDACTED] was greater to the rate was necessary to account for the to Merlin where the guarantee is binding. So Merlin members than to Pandora, and, Pandora Metrics to which Merlin Labels the insurance is coming in. And those consequently, he concluded that no would have access. Shapiro WDT at D– payments are included and, of course, raise adjustment to the effective royalty rate 14. the amounts of money that Pandora is paying was necessary. Shapiro WDT at D–13. (E) [REDACTED] and, therefore, they raise the rate that’s in my (C) [REDACTED] proposal, so it includes that. Under the Agreement, Pandora, Each Merlin member that opted-in to [REDACTED], may create a Id. at 4696. the agreement could elect to [REDACTED]. PAN Ex. 5014 § 10 [REDACTED]. PAN Ex. 5014 (Pandora/ (Pandora/Merlin Agreement); see also 127 Dr. Shapiro also made a small adjustment in Merlin Agreement § 8). Shapiro WDT at D–14, D–15 & n.31. his effective royalty rate calculation to reflect that According to Dr. Shapiro, Pandora personnel explained to Dr. certain tracks [REDACTED]. PAN Ex. 5014 (1)(c) and 3(c) . Dr. Shapiro assumed that [REDACTED] [REDACTED] are mutually beneficial to Shapiro that such [REDACTED] were would represent [REDACTED]% of Merlin tracks the opting-in Merlin members and to potentially mutually beneficial to the overall. Shapiro WDT at App. D–7. Pandora. Shapiro WDT at D–13. Dr. Merlin members and to Pandora. Id. at

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00045 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26360 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

n.32. The Merlin members benefit from labels. Id. Pandora estimates that Merlin percent of these albums are by artists [REDACTED], generating benefits to the member artists [REDACTED]. Id. whose labels are Merlin members and Merlin members in the form of Pandora acknowledges that Pandora Pandora estimates that participation by enhanced royalties and discovery of Presents generates promotional benefits artists whose labels are Merlin members their other artists. Id. For Pandora, these for the featured artists. However, will [REDACTED] to [REDACTED] [REDACTED] offer another context for Pandora also understands that Pandora percent. Shapiro WDT at D–18 nn.51, engaging listeners and, by increasing the Presents also generates marketing 52. Pandora also estimates that the number of Merlin member plays on benefits for Pandora with respect to number of Merlin label albums featured Pandora, these [REDACTED] work in advertisers, listeners, artists, and labels. on Pandora Premieres will [REDACTED] tandem with the steering provisions in Id. More particularly, Pandora not only from around [REDACTED] per year to the Pandora/Merlin Agreement. views the program as a marketing around [REDACTED] per year. Id. at By way of comparison, Dr. Shapiro platform that adds value for Pandora’s n.53. service, but Pandora has also required noted that Pandora is working with Dr. Shapiro acknowledges that that Pandora Presents events another entity to [REDACTED] that will Pandora Premieres generates [REDACTED]. Fleming-Wood WDT ¶ 29 feature specific artists. Id. at n.34; see promotional benefits for the featured & n.5; see Westergren WDT ¶ 38. Herring WDT ¶ 30 n.11. Pandora artists and their labels, but that benefit Pandora Presents events thus generate personnel informed Dr. Shapiro that is offset by (and evident from) the fact additional advertising revenue for neither Pandora nor the entity that labels waive royalties for the one- Pandora as well as promotion of the [REDACTED] is [REDACTED], which week period that an album is on Pandora brand with Pandora listeners. suggested to Dr. Shapiro that such Pandora Premieres. Shapiro WDT at D– Over the long run, Pandora considers [REDACTED] create ‘‘mutual and 18. Pandora also receives significant roughly equalized benefits for both that Pandora Presents events lead to increased listener satisfaction and benefits from Pandora Premieres, Pandora and the [REDACTED] creator.’’ because it offers a benefit to Pandora Shapiro WDT at D–15. retention, and thus to greater advertising and subscription revenue. Id. listeners, who receive an early For these reasons, Dr. Shapiro Because of the foregoing, Dr. Shapiro opportunity to listen to entire new concluded that no adjustment to the likened Pandora’s role in coordinating albums from artists they like and to buy effective royalty rate was necessary to Pandora Presents events to that of an the music. Fleming-Wood WDT ¶ 30. account for the [REDACTED] provision independent concert producer and On balance, therefore, Dr. Shapiro in the Merlin Agreement. Id. at D–15 to promoter. Therefore, Dr. Shapiro concluded that Pandora Premieres D–16. concluded that the [REDACTED] generates significant benefits both to the (F) Pandora Presents and Pandora Pandora Presents events, on balance, artists and label, on the one hand, and Premieres Events did not call for any adjustment to the to Pandora as well. Because the program effective royalty rate he had calculated. is mutually beneficial, and because Pursuant to the Pandora/Merlin Shapiro WDT at D–17. Pandora [REDACTED], Dr. Shapiro Agreement, opting-in Merlin members concluded that the [REDACTED] in (2) Pandora Premieres receive [REDACTED] in ‘‘Pandora Pandora Premieres does not call for an Presents’’ and ‘‘Pandora Premieres’’ Pandora Premieres is a program adjustment to the effective royalty rate events. PAN Ex. 5014 § 11 (Pandora/ through which Pandora promotes he had calculated. Shapiro WDT at D– Merlin Agreement). Dr. Shapiro albums in the week prior to their 19.128 considered these two types of events release. Fleming-Wood WDT ¶ 30. separately. Pandora sends an email inviting certain iv. Adjustments Over the 2016–2020 Period (1) Pandora Presents listeners (selected based on their listening tastes and profiles) to listen to Dr. Shapiro adjusted his proposed Pandora Presents is a program a new album during the week prior to launched in December 2011, through rates higher to reflect anticipated its release date. Id.; see also Shapiro inflation over the 2016–2020 statutory which artists perform live before an WDT at D–17 n.45. When selecting audience of fans that Pandora identifies period. Shapiro WDT at 35. However, at albums to feature on Pandora Premieres, the hearing, Dr. Shapiro testified that he and invites without charge. Fleming- Pandora reviews albums and artists Wood WDT ¶ 29. Each of these events would have preferred not to predict proposed by the record companies to future inflation, but rather to include a is designed for and sponsored by an ensure ‘‘a good fit with the program’’ advertiser. Pandora essentially plays the statutory term requiring the rates to be and to ‘‘generate a high volume of adjusted annually to reflect actual role of a concert producer and promoter, listening.’’ Fleming-Wood WDT ¶ 30. choosing artists to feature in Pandora inflation. 5/19/15 Tr. 4608–10 (Shapiro). Pandora provides these selected Dr. Shapiro did not make any other Presents events that will best speak to Pandora Premieres listeners with ‘‘click- the target audience of the sponsoring adjustments to reflect anticipated or to-buy functionality.’’ Id. at n.46. predicted changes over the statutory advertiser. Id. Pandora identifies and Pandora requires the labels to waive matches advertisers and artists that royalties for the one-week period that an appeal to a particular demographic, then 128 Dr. Shapiro also considered two factors album is on Pandora Premieres. Shapiro enumerated in the statutory willing buyer/willing books a location for the event and WDT at D–18. Pandora personnel seller formulation—Pandora’s potential role in markets the event to Pandora listeners informed Dr. Shapiro that Pandora has promoting or substituting for other Merlin label with a demonstrated interest in the never charged labels for their revenue streams, and Pandora and Merlin’s featured artist. Pandora [REDACTED]. ‘‘relative contribution.’’ He concluded that, as participation in Pandora Premieres and rational economic actors with access to information Pandora [REDACTED]; sometimes has no plans to do so. Id. at D–18 n.49. regarding such factors, the parties would attempt to Pandora [REDACTED]. Shapiro WDT D– Pandora Premieres features two to five make sure that such elements were ‘‘fully baked in’’ 17 n.43. albums per week, or about 150 albums and ‘‘automatically included’’ in the negotiated annually. Fleming-Wood WDT ¶ 30. rates. 5/19/15 Tr. 4605–06 (Shapiro). Given this There have been between fact, Dr. Shapiro made no further adjustments to the [REDACTED] Pandora Presents events Pandora personnel informed Dr. Shapiro rates he derived from the Pandora/Merlin per year featuring artists on Merlin that approximately [REDACTED] Agreement.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00046 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26361

period. His adjusted rates are set forth in the table below: 129

EFFECTIVE PER-PLAY ROYALTY RATES AFTER ADJUSTMENTS [2016 Through 2020 (¢)]

Inflation Advertising- rate * Subscription Blended 130 (%) supported

30% Steering: 2016 ...... 2.20 0.1105 0.2146 0.1225 2017 ...... 1.73 0.1124 0.2183 0.1246 2018 ...... 1.74 0.1144 0.2221 0.1268 2019 ...... 1.76 0.1164 0.2260 0.1290 2020 ...... 1.78 0.1185 0.2300 0.1313 12.5% Steering: 2016 ...... 2.20 0.1205 0.2238 0.1324 2017 ...... 1.73 0.1226 0.2276 0.1347 2018 ...... 1.74 0.1247 0.2316 0.1370 2019 ...... 1.76 0.1269 0.2357 0.1394 2020 ...... 1.78 0.1291 0.2399 0.1419 * The inflation rate reported for 2016 accounts for expected inflation from the mid-point of the period Q4 2014 through 2015 (May 2015) to the midpoint of 2016 (August 2016). The other inflation rates account for annual expected inflation to the mid-point (August) of each calendar year listed.

Dr. Shapiro explained why he recorded music can indeed be based on rate structure or a percent-of-revenue proposed two alternative rates: ‘‘[The webcaster revenues, at least in the case of structure with payments prorated rate selected] depends on how much Pandora. Furthermore, webcasters and many according to label share.’’ Shapiro WDT steering Pandora is doing. If they do other types of music users pay royalties to at 37–38. The relevant question for music publishers and composers, through more steering, that lowers the rate ASCAP and BMI that are set as a percentage purposes of rate-setting, therefore, they’re going to be paying, in fact, and of revenue. For example, the ASCAP rate according to Dr. Shapiro, ‘‘is whether so then that lowers the corresponding court recently established a royalty rate for the repertoires of the [Majors] would statutory rate derived from the Merlin Pandora of 1.85 percent of revenue for the command a higher rate per play or a Agreement.’’ 5/19/15 Tr. 4603–04 period 2011–2015 for its performance of higher percent-of-revenue than the (Shapiro). musical compositions in the ASCAP Merlin Labels in a workably competitive repertoire. This indicates to me that market.’’ Id. b. Pandora’s Proposed Greater-of Rate webcasting revenues can serve as a practical Pandora answers this question in the Structure Including a 25% of Revenue basis for royalty payments. negative, for two reasons. First, Prong Shapiro WDT at 23.131 according to Dr. Shapiro, the empirical In addition to the proposed per-play evidence demonstrates that there is no c. Pandora’s Proposed Application of rates, Dr. Shapiro’s rate proposal greater promotional effect on the sale of the Pandora/Merlin Rates to the Majors employs a greater-of structure, with the songs from the Majors (as compared to second prong set at ‘‘25 percent of the Pandora avers that the effective rates the Indies) from performances on revenue attributable to the licensed established by the Pandora/Merlin Pandora to support an upward music,’’ as such revenue is defined in Agreement are not only representative adjustment to the Merlin benchmark. 5/ the regulations proposed by Pandora. of the rates that Indies would receive as 19/15 Tr. 4623–64 (Shapiro). Second, Shapiro WDT at 20 & n.30; 5/19/15 Tr. willing sellers in the hypothetical Pandora has the same ability to steer 4608:16–23 (Shapiro). This is the same marketplace, but are also representative toward and away from the repertoires of greater-of rate structure adopted by the of the rates that the Majors would each of the Majors, just as it has done parties to the Pandora/Merlin receive in the hypothetical marketplace. with the Merlin Labels. See 5/19/15 Tr. Agreement. PAN Ex. 5014 ¶ 3(a). Pandora’s explanation as to why this 4624–30 (Shapiro); Shapiro WDT According to Dr. Shapiro, a greater-of extrapolation is warranted is based on Appendix F at F–6.132 formula with a ‘‘percent-of-revenue’’ its distinction between greater revenue To bolster this argument, Pandora prong is proper for the following derived from a higher number of plays notes that Dr. Rubinfeld’s analysis vis- reasons. as opposed to greater revenue from a a`-vis his own interactive benchmark higher per-play rate. As Dr. Shapiro reveals that Merlin receives essentially [T]he Merlin Agreement . . . specifies that Pandora’s royalty payments to the opined, Majors have a higher share of the same level of monetary participating Merlin Labels . . . will be at the overall plays on Pandora than the consideration as the Majors in the least 25 percent of its revenue attributable to Merlin Labels do, and thus they receive interactive market. Pandora concluded the music of those labels. These agreements more in royalty income because that therefore that the effective rates derived show that, as a practical matter, royalties for ‘‘occurs automatically under a per-play from the Pandora/Merlin Agreement

129 The rates in the table differ from the rates 131 Dr. Shapiro assigned no separate value to the incorporating the 25% of revenue royalty payment, proposed by Pandora because the proposed rates are 25% of revenue prong for adjustment of the per- he concluded that it would be ‘‘double counting or rounded. play prong, because he understood that the per-play just nonsensical’’ to add the value of that prong into 130 Dr. Shapiro blended the ad-supported and prong would result in a payment by Pandora to the per-play prong. Id. at 4686. 132 subscription rates to create his ‘‘blended’’ rate. Merlin of approximately [REDACTED]% of revenue Dr. Shapiro’s conclusion that noninteractive attributable to Merlin, thus not triggering the lower However, Pandora does not propose that the Judges services can steer away from the Majors as well as 25% prong. 5/19/15 Tr. 4683–4 (Shapiro). Further, the Indies is based upon Pandora’s ‘‘steering adopt such a ‘‘blended’’ rate. because Dr. Shapiro included a second prong experiments.’’

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00047 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26362 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

indeed can serve as benchmarks for the acknowledged on cross-examination that Pandora offered Merlin the rates to be paid by the Majors. See that he thought it would be important to [REDACTED] the Pandora/Merlin Pandora PFF ¶¶ 158–163 (and citations know Merlin’s expectations as to value Agreement as a counterproposal to to the record therein). in order to do a ‘‘proper analysis’’ of the Merlin’s proposal to [REDACTED]. SX value of the Pandora/Merlin Ex. 310 at 1; 6/1/15 Tr. 6986 (Lexton). 4. SoundExchange’s Criticisms of the Agreement.’’ Id. at 467–71. Moreover, In the same vein, Mr. Van Arman, co- Pandora Rate Proposal SoundExchange notes that the value founder and co-owner of the Indie SoundExchange opposes the use of analysis undertaken by Dr. Shapiro is record company (and Merlin member) the Pandora/Merlin Agreement as a not based on Pandora’s expectations Secretly Group, testified that the benchmark in this proceeding. Its that existed before the execution of the presence of the [REDACTED] provision opposition is based upon several Pandora/Merlin Agreement, but rather was one of the reasons his labels opted- principal arguments. on the valuation evidence he obtained in to the Pandora/Merlin Agreement. 6/ from Pandora after the Pandora/Merlin 2/15 Tr. 7172 (Van Arman). a. The Pandora/Merlin Agreement Agreement had been executed. Id. at Creates New Rights and New ii. The [REDACTED] Provision 4669. Obligations That Are Unavailable Under SoundExchange asserts that, had Dr. The Pandora/Merlin Agreement the Statutory License Shapiro considered the value placed on obliges Pandora to [REDACTED] to the SoundExchange asserts that the these extra-statutory elements of opting-in Merlin members. PAN Ex. Pandora/Merlin Agreement does not consideration by Merlin and its 5014 § 5. These [REDACTED] are not cover the same rights that are available members, the total value of the available under the statutory license and under the statutory license and also consideration would have at least are not replicated in Pandora’s rate creates new obligations that are equaled the existing Pureplay statutory proposal. SoundExchange notes that Mr. unavailable under the statutory license. settlement rates for 2014 and 2015. In Lexton testified that Merlin would not Specifically, SoundExchange avers that support of this point, SoundExchange have entered into the Pandora/Merlin the Pandora/Merlin Agreement contains relies in substantial measure on the Agreement if it had not contained these the following extra-statutory rights and testimony of one of Merlin’s two chief [REDACTED] commitments. 6/1/15 Tr. duties: negotiators of the Pandora/Merlin 6906 (Lexton). SoundExchange also • [REDACTED]; Agreement, Charlie Lexton, Merlin’s notes that Pandora itself viewed the • [REDACTED]; Head of Business Affairs and General [REDACTED] as a valuable [REDACTED] • [REDACTED]; Counsel. SX Ex. 13 ¶ 1 (Lexton WRT). provision. See SX Ex. 310 at 2 (a • [REDACTED]; Mr. Lexton testified that, in Merlin’s contemporaneous Pandora negotiating • [REDACTED]; view, the consideration provided to document, in which Mr. Herring wrote: • [REDACTED]; Merlin members by the Pandora/Merlin ‘‘[REDACTED]’’). • [REDACTED]; and • Agreement was, ‘‘at worst, no lower iii. Advertising/Promotional Benefits [REDACTED]. than the compensation under the See PAN Ex. 5014, §§ 1(c)(v), § 2(c) and existing statutory rate paid by Pandora.’’ Mr. Lexton testified that Merlin 13; see generally SX PFF ¶¶ 559–562 Id. at 17. would not have entered into the (and record citations therein). More particularly, SoundExchange Pandora/Merlin Agreement if it had not Given these differences between the relies on the following evidence and included the advertising and promotion Pandora/Merlin Agreement and the testimony with regard to items of extra- benefits ultimately embodied in the statutory license, SoundExchange statutory consideration. agreement. 6/1/15 Tr. 6909 (Lexton). concludes that the former at best is but According to Mr. Lexton, these benefits a weak benchmark for the latter. See SX i. The [REDACTED] Provision and clearly were of value to Merlin’s PFF ¶ 558 (quoting SDARS II, 78 FR at Merlin’s [REDACTED] members. Id. at 6880. He explained that 23064 (Apr. 17, 2013)) (Additional According to SoundExchange, the these advertising and promotion considerations and rights granted in [a evidence shows that Merlin and its provisions ‘‘provided considerable proposed benchmark] that are beyond members placed a value on the value that could not be replicated by the those contained in the Section 114 [REDACTED] provision, because Merlin statutory license.’’ SX Ex. 13 ¶ 43 license weaken the [benchmark’s] obtained this provision through its (Lexton WRT). ‘‘comparability as a benchmark.’’). negotiations with Pandora. 6/1/15 Tr. In like fashion, Simon Wheeler, 6894–95 (Lexton). Specifically, Merlin Director of Digital for another Merlin b. Dr. Shapiro Failed Adequately To had initially asked for [REDACTED], member, Beggar’s Group, testified that Value the Non-Statutory Consideration which Pandora refused to provide, one of his company’s motivations for and Thus Wrongly Failed To Increase leading to this [REDACTED] provision opting-in to the Pandora/Merlin His Benchmark as an alternative to [REDACTED]. Id. Agreement was that it afforded Beggar’s According to SoundExchange, not Further, Mr. Lexton testified that Merlin Group the ability to ‘‘tap into’’ these only is the Pandora/Merlin Agreement a ‘‘definitely’’ would not have entered promotional opportunities that were deficient benchmark, Dr. Shapiro also into the Pandora/Merlin Agreement if unavailable under the statutory license. wrongly failed to increase the value of the [REDACTED] provision had not SX Ex. 31 ¶ 23 (Wheeler WRT). that benchmark to reflect the value of been part of the agreement. Id. at 6898– SoundExchange also notes that Mr. the non-statutory consideration in the 99. Herring, one of Pandora’s negotiators, Pandora/Merlin Agreement. Mr. Lexton said that this provision likewise recognized that these SoundExchange asserts that Dr. Shapiro was important because Merlin believed, promotional tools had potential value to instead focused only on the lack of after considering [REDACTED], that Merlin, and, indeed, he acknowledged value attributed by Pandora to these there was a reasonable chance that his awareness that ‘‘Merlin believed that other forms of consideration. See [REDACTED] provision would be [these provisions] added value.’’ 5/18/ Shapiro WDT App. D at 1; 5/19/15 Tr. triggered, particularly during Pandora’s 15 Tr. 4275–76 (Herring). He further 4670 (Shapiro). However, fourth quarter of 2014. 6/1/15 Tr. 6896– acknowledged his awareness that SoundExchange notes that Dr. Shapiro 98 (Lexton). Mr. Lexton further noted Merlin had ‘‘sold’’ the promotional

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00048 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26363

benefits of the Pandora/Merlin d. The Pandora/Merlin Agreement 4203 (Herring). In fact, Mr. Herring Agreement ‘‘pretty strongly’’ to its Applies Only to a Single Webcaster recognized that Pandora would have members. Id. at 4279; see SX Ex. 2237 With Substantial Market Power been unable to negotiate the same terms at 1. SoundExchange notes that the with the Majors and would have to offer the Majors better terms. 5/18/15 Tr. iv. Access to Data Pandora/Merlin Agreement applies to only one licensee, Pandora, and the 4253 (Herring) (acknowledging that he When Pandora first proposed a direct terms of that license were not replicated ‘‘expected [to] . . . have to give more license to Merlin, Pandora offered in any other contract with any other favorable economic terms to a major record company than you would have to Merlin and its members access to licensee. SoundExchange finds this give to an independent record Pandora’s internal data. SX Ex. 104 at 5. point relevant because of Pandora’s ‘‘significant competitive strengths’’ company.’’). The right to such access was embodied To drive home this point, among webcasters, including its 77.6% in the final Pandora/Merlin Agreement. SoundExchange contrasts the absence of share of internet radio listening. PAN PAN Ex. 5014 § 9. Mr. Lexton testified evidence of any agreement between a Ex. 5012 at 11. According to that licensors do not have access to this Major and Pandora with the record SoundExchange, this large market share type of data under the statutory license. evidence of the iHeart/Warner afforded Pandora with market power Lexton WRT ¶ 40. Agreement. SoundExchange notes that, that was a meaningful factor in the pursuant to the iHeart/Warner Both Pandora and Merlin negotiations of the license with Agreement, SX Ex.33, per-play rates acknowledged that such data are Pandora. See SX Ex.19 at 6, 24–27 (i.e., even before any potential inclusion valuable to record labels generally. (Talley WRT) (noting that Dr. Shapiro of the value of other consideration) Westergren WDT at 16–17; SX Ex. 1736 failed to perform any analysis of range from $0.[REDACTED] to at 5; 6/2/15 Tr. 7157 (Van Arman); see meaningful allocations of buyer-side $0.[REDACTED] over the [REDACTED] 6/1/15 Tr. 7099–7100, 7106–07 (Simon power, including, for instance, whether period, greater than the rates in the Wheeler) (Access to data is something Pandora’s unique position in the market Pandora/Merlin Agreement.133 From Beggar’s Group ‘‘expect[s] of [its] major affected the terms of the Merlin this evidentiary distinction, direct licenses’’ and is ‘‘a part of every license.). SoundExchange concludes that the negotiation.’’). e. The Pandora/Merlin Agreement Was Services have not demonstrated that the SoundExchange also criticizes the ‘‘Experimental’’ rates in licenses between noninteractive usefulness of the Pandora/Merlin SoundExchange asserts that the services and Majors would match the Agreement as a benchmark for more Pandora/Merlin Agreement was merely lower rates in the Pandora/Merlin general reasons: an ‘‘experimental’’ modification of the Agreement. SX PFF ¶ 654; see also id. ¶ 656 (asserting iHeart/Warner c. The Pandora/Merlin Agreement Is restrictions created by the sound Agreement ‘‘confirm[s] that major Unrepresentative of the Larger Market recording performance complement. SX PFF ¶¶ 576–580 (and record citations record companies receive more SoundExchange asserts that the therein). At the hearing, Merlin consideration than independent record Pandora/Merlin Agreement pertains characterized the Pandora/Merlin companies when negotiating directly for only to record companies that represent Agreement as ‘‘experimental.’’ SX Ex. 13 licenses covering noninteractive less than [REDACTED]% of Pandora’s ¶ 27 (Lexton WRT) (describing the services.’’). performances and therefore cannot license as ‘‘an exercise in experimenting g. The Steering Provisions in the represent what the record companies— with direct licensing derived from the Pandora/Merlin Agreement Are Not including all three Majors—comprising existing statutory rates’’); see id. ¶ 25 Useful in Setting the Statutory Rate (‘‘Due to the fact Pandora offered us so Pandora’s other [REDACTED]% of SoundExchange rejects Pandora’s many additional benefits and other performances, would negotiate for in the foundational assumption that the added value that is not required by their hypothetical marketplace. SX RPFF steering provisions of the Pandora/ statutory license, we understood this as ¶ 753; SX PFF ¶ 507 (both relying on Merlin Agreement can be used to an opportunity for experimentation Shapiro WDT at 76). SoundExchange determine the statutory rate. given and within the constraints also avers that the Pandora/Merlin SoundExchange’s rejection of steering as imposed by Pandora’s existing statutory Agreement is not sufficiently probative a relevant benchmarking tool is based rates.’’); Wheeler WDT ¶ 9 (‘‘We knew of the rates that Indies would agree to on several factors: from the start that this was a short-term voluntarily because the bulk of the experiment. . . .’’) (emphases added). i. Steering Allegedly Creates ‘‘First Indies who opted-in [REDACTED]. 6/1/ Mover’’ Advantages That Cannot Be f. No Major Has Accepted a Similar 15 Tr. 6860, 6865–66 (Lexton). Replicated for All Licensees SoundExchange also notes that roughly Direct License With Pandora SoundExchange argues that as a 30% of the Merlin labels that opted-in SoundExchange emphasizes the do not regularly operate in the United matter of simple arithmetic a webcaster absence of what might otherwise be an cannot commit to steer to every record States. 6/1/15 Tr. 6863–64 (Lexton). important piece of evidence: No major company or label, because there is only Additionally, Mr. Lexton estimates that record company has agreed to a direct a total of 100% subject to steering. As of the [REDACTED] or so Merlin license with Pandora or any other one of its economic experts noted: members that opted-in directly (rather webcaster on the same rates and terms than through distributors or of the Merlin license. SoundExchange 133 SoundExchange also notes that [REDACTED]’s aggregators), approximately notes that this is unsurprising, in that licenses with [REDACTED], [REDACTED], and [REDACTED] have been affirmatively Pandora’s C.F.O. Mr. Herring, independent record companies for its [REDACTED] rejected by Pandora for inclusion in the acknowledged that Pandora regularly service likewise demonstrate that the major record companies receive considerably more consideration Merlin license, based on Pandora’s had conversations with the Majors, but than independent record companies. SX PFF ¶ 655, [REDACTED]. Id. at 6871. did not replicate the terms of the and Section XI.A therein (and record citations Pandora/Merlin Agreement. 5/18/15 Tr. therein).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00049 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26364 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

[A]n affirmative obligation to steer just iv. Pandora Did Not Test Steering Under repertoire to Pandora. In support of this can’t be implemented on a market-wide ‘‘Real-World’’ Conditions position, SoundExchange quotes Dr. basis. It’s just not possible for a service to say SoundExchange argues that Pandora Shapiro, who acknowledged that ‘‘a I’m going to steer listenership towards each failed to test steering under real-world record company with market power’’ label that I contract with. conditions, because there is no evidence could use that power to disable a webcaster’s threat of steering. 5/19/15 5/27/15 Tr. 6070 (Talley). that listeners were ever aware that steering was occurring. More Tr. 4576–77 (Shapiro). Dr. Talley Similarly, SoundExchange notes that particularly, SoundExchange points out similarly noted that, ‘‘in the an iHeart executive, Mr. Cutler, that Pandora has yet to experience any hypothetical market where there is no recognized the impossibility of potential negative listener reaction that background statutory rate . . . a label promising steering to all record may arise if and when competitors might say, okay, if you’re going to [steer companies: ‘‘Certainly, the share has advertise that Pandora has modified its against us], we may just walk to—its math has to add up to—a algorithm in a manner that contradicts away. . . .’’ 5/27/15 Tr. 6074 (Talley); hundred, so if someone goes from 20 to its long-standing claim to play ‘‘only the see also 5/1/15 Tr. 1429 (Harleston) (‘‘If 30, the rest of the pool must—those ten music listeners want’’ 134 in order to a service were to say we’re just not points must come from somewhere save money on royalty rates. See 5/19/ going to play your records because it else.’’ 6/2/15 Tr. 7239 (Cutler). 15 Tr. 4775 (Shapiro) (admitting that costs too much, the reality is we can Thus, as Dr. Rubinfeld noted, the Pandora did not test how people would go—we have other choices. We could lean into other services.’’). steering provisions provided Merlin react to learning ‘‘that Pandora was SoundExchange finds support for this factoring in royalty rates [in] how they with ‘‘first mover’’ advantages. position because the Services’ economic constructed the playlist.’’). Indeed, Dr. Rubinfeld CWRT ¶ 70. SoundExchange experts declined to conclude that the Shapiro ‘‘worried about’’ the question concludes therefore that Pandora cannot Majors were not ‘‘must haves’’ for whether a competitor could use such an escape from this ‘‘quandary’’ by noninteractive service. See 5/11/15 Tr. advertisement to ‘‘magnify’’ a negative discarding the [steering commitment], 2989–90 (Katz) (‘‘Q. Is it fair to say that reaction to steering. Id. at 4635–36. yet retaining the [discounted rates] from you . . . believe that the [M]ajors are Because successful steering in the real the Pandora/Merlin Agreement. must-haves for customized services According to SoundExchange, world depends on consumer reactions, such as Pandora? A. I would say I discarding the [steering commitment] SoundExchange concludes that Pandora believe that’s a possibility, yes.’’); 5/19/ would separate the rate in the agreement has failed to demonstrate a credible 15 Tr. 4582 (Shapiro) (Dr. Shapiro from the specific bargained-for threat of steering. testified that he was ‘‘offering no Additionally, SoundExchange notes consideration that Merlin obtained in opinion whether the [M]ajors are must- that Pandora has been unable to exchange for that rate. SX RPFF ¶ 764. have for Pandora.’’). generate as much ‘‘real world’’ steering ii. Revenue From Steering Is a Valuable as it intended under the Pandora/Merlin vi. Record Companies Can Utilize Benefit Not Available Under the Agreement. Specifically, the evidence Contract Clauses To Thwart Steering Statutory License actually shows that Pandora has not SoundExchange asserts that it can achieved the [REDACTED]% steering contract around a noninteractive SoundExchange asserts that the target for most Merlin labels. 5/19/15 Tr. steering provision provides Merlin with service’s proposal or threat to steer by 4676–16 (Shapiro). Dr. Shapiro also insisting upon a specific anti-steering a financial advantage that cannot be admitted that, as of November 2014, duplicated under the statutory scheme. clause or a more general ‘‘Most Favored Pandora had been unable to achieve the Nation’’ (MFN) clause.135 See SX Ex. 25 Therefore, SoundExchange avers, [REDACTED]% target for ‘‘a good Pandora’s proposed benchmark must be ¶¶ 14–19 (A. Harrison WRT) (‘‘UMG has number’’ of record labels. Id. Moreover, long recognized in our negotiations with adjusted upward to reflect that this non- for [REDACTED]% of Merlin labels, interactive services that they have the statutory value, like all non-statutory Pandora’s steering has been negative. SX ability to steer users away from UMG consideration, permitted a reduction in Ex. 2310. music through the music they feature the benchmark royalty rate. See SX PFF From these facts, SoundExchange and recommend through the service ¶¶ 701–708 (and citations to the record concludes that Pandora has failed to thereby decreasing our plays on the therein). provide sufficient real world evidence service and the revenue that flows to iii. Pandora Has Not Provided Support regarding its ability to steer, UMG and its artists.... We therefore for Its Claim That a ‘‘Threat’’ of Steering demonstrating a disconnect between the have negotiated for protections against Will Lead to Lower Rates theoretical case it has presented and the such steering.... [I]f we did not have realities it faces in the marketplace. these commitments the interactive SoundExchange challenges Dr. v. A Record Company Could Rebuff a services could effectively steer users Shapiro’s assertion that, in the Steering Proposal by Withholding Its toward other record labels artists and hypothetical market, the ability of a Entire Repertoire sound recordings through the music noninteractive service to steer among they highlight.’’); accord, 4/28/15 Tr. record companies would necessarily SoundExchange argues that a record 455–56 (Kooker); 4/30/15 Tr. 1144–45 create a ‘‘threat’’ of steering that would company could respond to a steering (Harrison); 6/2/15 Tr. 7202–05 cause rates to decline to an effectively threat by refusing to license 100% of its (Harrison); 5/7/15 Tr. 2487–88, 2490–93 or workably competitive level. (Wilcox) (all acknowledging on behalf of 134 Timothy Westergren, Pandora’s founder, had major record companies that anti- SoundExchange asserts that the record publicly stated that Pandora’s recommendations is bereft of any benchmark agreement would ‘‘be based on the genome, they will never be steering provisions are commonly used that reflects a ‘‘threat of steering,’’ let based on somebody buying the space.’’ SX Ex. 2369 alone that a ‘‘threat of steering’’ had at 1. In fact, Mr. Westergren explained in 2013 that 135 ‘‘In general, an MFN clause is a contractual allowed a noninteractive service to ‘‘[t]he only thing that drives what song [Pandora] provision that requires one party to give the other play[s] next for a listener is trying to deliver the best the best terms that it makes available to any obtain a lower rate. See SX PFF ¶¶ 609, possible listening experience for that individual.’’ competitor.’’ U.S. v. Apple, Inc., 791 F.3d 290, 304 709. Id. at 3. (2d Cir. 2015).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00050 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26365

in their agreements with the on-demand [REDACTED] enters into an agreement i. Pandora Has Been Unable To Perform services). with another major record label that Its Contractual Obligations Several such anti-steering contract provides more favorable terms for that SoundExchange avers that, even clauses were in evidence in the label regarding specified key provisions assuming the Pandora/Merlin proceeding: (including [REDACTED]), then Agreement otherwise had merit as a • The agreement between [REDACTED] must notify [REDACTED] potential benchmark, Pandora has been [REDACTED] and [REDACTED] contains of those more favorable terms and give unable to perform its contractual an anti-steering clause that prevents [REDACTED] the option to avail itself of [REDACTED] from steering towards obligations. In this regard, those terms. SX Ex. 80 at 25542–43; SoundExchange notes the following lower-priced music, including on PAN Ex. 5091; see also 4/28/15 Tr. 447– playlists, if that steering would result in problems that have hindered Pandora’s 50 (Kooker); ability to perform its contractual duties. lowering [REDACTED]’s share of total • The agreement between • plays to a level that is less than Staffing and capacity constraints; [REDACTED] and [REDACTED] contains • lack of reporting and payments, [REDACTED]’s market share. SX Ex. 37; an MFN providing that if [REDACTED] • a low fraction of labels who are see also 6/2/15 Tr. 7202–06 (Harrison); grants another label more favorable • The agreement between receiving payments pursuant to deal; financial terms, then [REDACTED] must • a low participation in the [REDACTED] and [REDACTED] contains also offer those terms to [REDACTED]. [REDACTED] program; and an anti-steering provision to prevent SX Ex. 36; see also 4/30/15 Tr. 1142–44 • a low percentage of labels receiving [REDACTED] from steering listeners (Harrison) (‘‘[REDACTED]’’); steering at or above [REDACTED]%. away from [REDACTED] content and • The agreement between SX Ex.1748 at 2 ; SX Ex. 2310. towards that of another label. 4/30/15 [REDACTED] and [REDACTED] contains SoundExchange further notes that Mr. Tr. 1145 (Aaron Harrison); the equivalent of an MFN provision (an • Herring candidly acknowledged that Mr. Harrison testified that ‘‘equal treatment’’ clause) by which Pandora had waited until after it [REDACTED]; 6/2/15 Tr. 7206 (Aaron [REDACTED] warrants that it has not executed the Pandora/Merlin Agreement Harrison); see Harrison WRT ¶¶ 15–16; provided [REDACTED] to another label. to determine the actual cost to Pandora SX Ex. 36 ¶ 7; In the event that [REDACTED] has • The agreement between of performing its contractual duties. 5/ violated this warranty, the [REDACTED] [REDACTED] and [REDACTED] 18/15 Tr. 4280 (Herring). Afterward, clause permits [REDACTED] to receive prohibits [REDACTED] from promoting Pandora’s Chief Scientist estimated that an immediate [REDACTED] to match the another label’s repertoire if it would Pandora would incur an annual cost of superior terms. SX Ex. 343; see also 5/ then exceed its market share, unless $[REDACTED] for the ‘‘initial build’’ 7/15 Tr. 2474–79 (Wilcox). Spotify offers the same increase in and $[REDACTED] annually in market share to [REDACTED]. SX Ex. 80 vii. Record Companies Could Thwart ‘‘ongoing support maintenance.’’ Id. at at 25537–38; see 4/28/15 Tr. 455–56 Steering by Requiring Up-Front Lump 4282; SX Ex. 1706 at 1. Pandora (Kooker). The practical effect of the Sum Royalties calculated internally that, just to clause is to prohibit [REDACTED] from SoundExchange notes that, as Dr. Katz provide the opting-in Merlin members increasing another label’s promotional candidly acknowledged, a record with the contractually promised access opportunities above its market share if company could neutralize a steering to data, Pandora would incur that would lower [REDACTED]’s threat by seeking a lump sum payment $[REDACTED] in initial costs and promotional opportunities to below its instead of per-play rates. 5/11/15 Tr. $[REDACTED] in ongoing annual costs. market share. 4/28/15 Tr. 456 (Kooker); 3015–6, 3019–20 (Katz).136 Id. at 20. Similarly, Pandora would need • The agreement between to spend almost [REDACTED] dollars in [REDACTED] and [REDACTED] contains h. Merlin’s Economic Interests Were Not initial costs and $[REDACTED] in an anti-steering provision that Fully Aligned With Those of Its annual costs to [REDACTED], two of the guarantees [REDACTED] will get Members advertising benefits contained in the [REDACTED] equivalent to its market SoundExchange addresses what it Pandora/Merlin Agreement. Id. share [REDACTED]. The provision suggests may be conflicts of interest as SoundExchange notes that these further provides that if any other record between Merlin and its distributor/ implementation issues have ‘‘impacted company receives an ‘‘uplift’’ over its aggregator-members, on the one hand, negatively’’ the willingness of Merlin Soundscan market share, [REDACTED] and the Merlin label members, on the members who opted-in to consider will receive the same ‘‘uplift.’’ SX Ex. other. First, Merlin and the distributors/ entering into this license in any future 343 at 20; SX Ex. 1814 at 26; SX Ex. 346 aggregators typically receive period. For example, Mr. Van Arman at 5; see 5/7/15 Tr. 2490–93 (Wilcox). [REDACTED] from members only if that testified that, [REDACTED] 6/5/15 Tr. More broadly, as noted above, member has opted-in. Second, Pandora 7158 (Van Arman); see also 6/1/15 Tr. SoundExchange asserts that, as in the paid Merlin a license fee directly that 7104–10 (Simon Wheeler) (detailing interactive market, the Majors could would vary, up to $375,000 (but in any implementation issues and concluding insist upon a general MFN clause in event no less than $250,000), depending [REDACTED]. each contract with a service, which upon the Merlin members [REDACTED]. would ensure that each Major gets the 5. Judges’ Conclusions Regarding SX Ex. 13 ¶ 56 (Lexton WRT). Thus, Pandora’s Benchmark Evidence benefit of the rates and terms set forth SoundExchange avers that Merlin had For the reasons set forth below, the in the service’s contracts with the other economic incentives to complete the Judges find that the noninteractive Majors. See 4/28/15 Tr. 449–450, 542 Pandora/Merlin Agreement and to urge benchmark proposed by Pandora is (Kooker); 4/30/15 Tr. 1142 (Harrison); 5/ its members to opt-in—incentives that informative as to the rates they shall set 7/15 Tr. 2473 (Wilcox). Several such were not necessarily consistent with the in this proceeding for a particular MFN contract clauses were in evidence interests of its members. in the proceeding: segment of the noninteractive • The agreement between 136 The dynamic economic effect of an up-front marketplace. That is, the Pandora [REDACTED] and [REDACTED] contains lump-sum royalty payment is discussed elsewhere benchmark is probative of the two an MFN provision providing that if in this determination. distinct royalty rates that a

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00051 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26366 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

noninteractive service would pay to 30%. All four licensors, including M, This process does not result, as some Indies in the: (1) Ad-supported (free-to- would thus be promised 60% + 45% + record industry witnesses suggested, in the-listener) market; and (2) the 30% + 15% = 150%. a ‘‘race to the bottom.’’ 137 Rather, it subscription market, respectively. SoundExchange’s point is that, by typifies a ‘‘race’’ to a workably or Pandora’s proposed benchmark is definition, it is mathematically effectively competitive price. On the premised principally on the provisions impossible for a noninteractive licensor licensees’ side of the market (the buyers’ of the Pandora/Merlin Agreement. to allocate more than 100% of its plays. side), the limit on the demand for lower SoundExchange raises two principal Thus, SoundExchange concludes, rates through steering is reached when challenges to Pandora’s benchmark: (1) steering can only work in a non- the noninteractive service is no longer The ability, vel non, of a noninteractive statutory setting and, even then, never in a position to make further service to ‘‘steer’’ or credibly ‘‘threaten’’ for all licensors. See 5/28/15 Tr. 6301 substitutions of one record company’s to steer in the hypothetical market; and (Rubinfeld); see also 5/27/15 Tr. 6070 sound recordings for another’s because (2) the potential value of other (non- (Talley) (‘‘[I]t’s almost like a Lake the potential for lost revenues exceeds steering) elements of consideration Wobegon effect, that not everyone can the cost savings.138 On the licensors’ Pandora provided to Merlin that might be above average, not everyone can side of the market (the sellers’ side), the offset the lower stated rates, thus receive steering.’’). limit on the willingness to supply leaving the effective rate unchanged This argument of course, in the static recordings at reduced rates is reached from the nonprecedential statutory sense, is mathematically correct. But, in when the licensor determines that any Pureplay Settlement rate. the dynamic sense, is it economically further reduction in the rate will not be In light of the importance of these two correct? Dr. Shapiro, for Pandora, sufficiently to cover all marginal and issues, the Judges first analyze these two responded to this argument in the recurring fixed costs (including contentious points, followed by a following colloquy with the Judges opportunity costs) for its particular discussion of SoundExchange’s other regarding the ‘‘threat’’ of steering: repertoire. (This is essentially stating in words the fundamentals of the Lerner objections to Pandora’s benchmark [THE JUDGES] proposal. Equation discussed at note 123 supra). Let’s . . . take . . . the market we’re Because the Judges are utilizing the dealing with here [and] address the first- a. ‘‘Steering’’ as a Mechanism for benchmark approach to rate setting—as Achieving Effective Competition in the mover criticism . . . that well, sure, you can steer to . . . record company A . . . but you both SoundExchange and Pandora Hypothetical Market can’t steer to all of them because you can’t endorse—the limits to steering (like the i. Could a noninteractive service steer play more than 100 percent of the music. Is value of promotion and substitution) are and credibly threaten to steer in the it . . . the threat of steering that pushes implicit in (‘‘baked-in’’) the terms of the hypothetical market? everybody . . . towards their original relevant benchmarks. That is, Pandora percentages to avoid being that odd man out and Merlin entered into their agreement SoundExchange argues that steering who was the holdout for the higher price? because each concluded that its steering creates merely a ‘‘first mover’’ advantage [DR. SHAPIRO] terms were advantageous.139 for those licensors who are able to enter That’s exactly—yes, absolutely. The SoundExchange argues that, even if into steering arrangements before their competitive outcome is when each of the the threat of steering could cause a competitors are able to obtain such record companies is at a rate where they’re reduction in rates in the hypothetical advantages. This argument is . . . not disadvantaged relative to the other noninteractive market, the Services have seductively simple: In its essence, it is guys .... This notion that you can’t steer, the 100% thing, it’s kind of offensive to an not provided any proof of an actual based on the elementary proposition threat of steering in the direct that no noninteractive service can steer antitrust economist . . . because it’s basically saying . . . price competition is some noninteractive licensing market, but more than 100% of its sound recordings. horrible thing. rather have presented only evidence of To take a simple example, assume there actual (not threatened) steering. See, are three Majors, U, S, and W, and one 5/19/15 Tr. 4561–63 (Shapiro); see e.g., 5/27/15 Tr. 6076 (Talley) (‘‘[N]ot Indie, M. Assume the ex ante steering Shapiro WDT at 9 (noting that the ‘‘net one of these transactions . . . is either allocation of plays was 40% for U, 30% result’’ of steering ‘‘in a workably negotiated in the shadow of a threat to for S, 20% for W and 10% for M, and competitive market may well be steer away or negotiated with an all plays were priced at $0.0020. Now, relatively little actual steering.’’). Dr. undertaking to steer away. It’s in the the noninteractive service strikes a deal Shapiro further notes that, in the opposite direction . . . a promise to with M to increase plays of M’s sound absence of steering, ‘‘[y]ou would be steer towards . . . as opposed to away recordings by 50% over the ex ante basically going to the rate that a cartel from. . . .’’). percentage, in exchange for, say, a 10% or monopolist would set.’’ 5/19/15 Tr. SoundExchange’s argument is reduction in per-play rates to only M. 4575 (Shapiro). unpersuasive, for two reasons. First, the Then, M’s noninteractive market share The Judges find that steering in the evidence shows that Merlin members increases by 50% from 10% to 15% hypothetical noninteractive market opted-in to the Pandora/Merlin (while its per-play rate declines by only would serve to mitigate the effect of Agreement specifically because they 10%, resulting in more revenue for M ex complementary oligopoly on the prices anticipated that Pandora might enter post steering). As a ‘‘first mover,’’ M paid by the noninteractive services and thus benefits. therefore move the market toward 137 See, e.g., Van Arman WDT at 14. However, the noninteractive licensee effective, or workable, competition. 138 The existence and identification of such a cannot promise all three other licensors, Steering is synonymous with price limit was the point of Pandora’s steering experiments. U, S, and W, the same 50% increase in competition in this market, and the 139 Likewise, iHeart and Warner entered into their plays via steering in the same contract nature of price competition is to cause steering-based agreement because it was mutually period. If it did, U would realize a prices to be lower than in the absence advantageous. By ‘‘advantageous,’’ the Judges are market share increase from 40% to 60%; of competition, through the ever-present noting the essence of the willing buyer/willing seller paradigm—that sophisticated commercial S would realize a market share increase ‘‘threat’’ that competing sellers will buyers and sellers are presumed to act rationally in from 30% to 45%; and W would realize undercut each other in order to sell their self-interest when entering into agreements a market share increase from 20% to more goods or services. that are not coercive.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00052 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26367

into steering agreements with other Shapiro—equating steering with price How can Licensor B avoid this loss? record companies, including the Majors. competition—to be correct. The ability By responding to this steering by In fact, SoundExchange’s’ own witness of noninteractive services to steer competing on price and lowering its testified that it was in his record toward lower priced recordings (and, by own price to $0.0018. company’s self-interest to act necessity therefore, away from higher How can Licensee obtain the lower ‘‘defensive[ly]’’ to enter the Pandora/ priced recordings) is the essence of price of $0.0018 without any actual Merlin Agreement, in light of the fact price competition. With Pandora (and steering? By threatening to steer and that Pandora might enter into ‘‘similarly iHeart) having demonstrated the thereby compelling Licensors A and B structured deals’’ with other record capacity and willingness to steer in this to compete for Licensee’s business by companies. 4/28/15 Tr. 610–11 (Van manner, it would be economically offering to accept a price of $0.0018. Arman); see 6/1/15 Tr. 6963 (Lexton). irrational for the other record companies Moreover, if Licensor B incurs the loss These facts reflect the general power of (that had not agreed to steering) to described above in one contracting steering as a threat in the marketplace. maintain their position and incur losses. period, that loss serves as the ‘‘threat’’ The Judges also find unpersuasive the To assume that record companies would necessary to avoid such losses in the criticism by SoundExchange that there ignore the ‘‘opportunity cost’’ of steering subsequent contracting periods by also is no record evidence of direct away from their repertoires would be a entering into an appropriate steering noninteractive agreements that were fundamental economic mistake. See 5/ arrangement. forged solely through a threat of 4/15 Tr. 1516–17(Lys) (emphasizing that Will there be a ‘‘race to the bottom?’’ steering. The point of the steering ‘‘opportunity costs are real costs’’). No. The so-called ‘‘bottom’’ will be argument is to demonstrate what would Dr. Shapiro’s point regarding the marked by the rate that equates: (1) An transpire in the hypothetical effectively economic ‘‘threat’’ posed, now that acceptable return to the Licensors given competitive market in which no steering is technologically possible, can their costs (including opportunity costs) statutory rate existed—not to be made clear through a hypothetical and the differentiated values of their demonstrate that a particular form of example: repertoires; and (2) an acceptable return • agreement is pervasive in the market Assume a Licensee was paying a to the Licensee by steering as far as with the extant statutory rate.140 It is market price of $0.0020 and historically possible (but no further), as limited by imperative not to confuse the (‘‘naturally’’) played 1,000,000 of its the potential loss of revenue if steering hypothetical market with the actual total number of songs from Licensor A, interferes with revenue as a regulated market.141 thus paying $2,000 to Licensor A. consequence of an inferior mix of sound • Moreover, the Judges find the Now, assume the Licensee and recordings. economic opinion expressed by Dr. Licensor A enter into a ‘‘steering’’ deal, ii. Is steering in the hypothetical market whereby Licensee promises to play an sufficient to establish an ‘‘effectively 140 One reason why steering is not yet more additional 200,000 songs whose competitive’’ rate? widespread in the market, as Dr. Shapiro noted, is copyrights are owned by Licensor A, that noninteractive services have developed the representing a 20% increase over the The Judges conclude, based on the steering technology only in the past few years since historical (‘‘natural’’) quantity of record evidence and expert testimony, the Web III proceeding. Shapiro WDT at 15 that the injection of steering into the (‘‘Pandora has now tested and proven its ability to 1,000,000 noted above. modify its playlist-selecting algorithms to rely more • In exchange, Licensee demands, hypothetical market provides for the or less heavily on the music of particular record and Licensor agrees, that Licensor A ‘‘effective competition’’ that the law companies.’’) (emphasis added). Now that this will receive less than $0.0020 per play, requires. Both Dr. Shapiro and Dr. Katz technological genie is out of the bottle, the Judges specifically, 10% less, i.e., only $0.0018. opined, and the Judges agree, that cannot minimize its impact in the hypothetical effective or workable competition arises market. Compare the two scenarios: 141 By way of comparison, Dr. Rubinfeld’s ‘‘ratio • Before steering, the money when licensees have the reasonable equality’’ benchmark royalty rate likewise does not exchanged equaled $2,000. (albeit still constrained) ability to select ‘‘exist’’ in the actual market. Rather, he derived that • After steering, the money sound recording inputs based upon benchmark rate by: (1) Looking at market data from exchanged is more, $2,160 (1,200,000 price. direct licenses; and (2) applying his economic × expertise to express certain economic opinions units $0.0018). The injection of steering into the regarding the necessary equality of the revenue-to- That is clearly a benefit to Licensor A, hypothetical market can occur in two royalty ratio in the interactive and noninteractive who has made an additional $160 ways, as it has in this determination. markets. (As noted infra, Dr. Rubinfeld’s ($2160¥$2000). First, as in the case of the Pandora/ ‘‘assumption’’ was revealed at the hearing to be premised on a model that serves to limit its The corresponding benefit to Licensee Merlin Agreement (and the iHeart/ applicability.). So too the steering-based proposed arises from the fact that it can now—ex Warner Agreement discussed infra), royalty rate is based on a benchmark analysis that post steering—play 1,200,000 songs at steering is incorporated by adopting a is tied to certain expert economic opinions $0.0018 per song for a total cost of benchmark that explicitly includes regarding market behavior. The Judges must weigh steering. Second, a steering adjustment and apply ‘‘economic . . . information presented by $2160. Ex ante steering, Licensee would the parties’’ as the bases for their rate have been required to pay the old can be made to a benchmark rate that is determinations, 17 U.S.C. 114(f)(2)(B), and therefore market price of $0.0020 per song to not otherwise effectively competitive. the expert opinions set forth by the parties’ another Licensor (call it Licensor B) for Such is the case with SoundExchange’s economists as to how the hypothetical market will interactive benchmark, which needs a perform are vital aspects of the record to be those 200,000 songs (which equals considered by the Judges. More broadly, the Judges $400), plus the $0.0020 Licensee also steering adjustment in order to note that the benchmarking approach, while highly paid to Licensor A ex ante steering for eliminate the ‘‘complementary instructive, is not the sole method for ascertaining 1,000,000 songs (which equals $2,000), oligopoly’’ effect discussed supra. The the statutory rate—indeed, the statute does not Judges note that adjustments to require the Judges to utilize the benchmark for a sum of $2,400 for 1,200,000 songs. approach. Here, the threat of steering has been Thus, Licensee has saved $240 in costs benchmark rates have regularly been demonstrated by a combination of benchmarks, ($2,400¥;$2,160). Since there is no made in § 114 proceedings—and indeed experiments and expert economic theorizing using ‘‘free lunch,’’ who loses? The loser is are required to be made—in order to fundamental principles of profit maximization and allow the benchmark to correspond to opportunity cost. This combination of proofs and Licensor B, who has lost the revenue arguments is actually more persuasive to the Judges from the foregone licensing of 200,000 the hypothetical market required by the than a mere benchmark standing alone. songs. statute. Here, as concluded supra, the

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00053 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26368 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

Judges have found as a matter of law ‘‘effectively competitive’’ market.142 Pandora promotes its service as playing that § 114 requires that they set a rate Indeed, even economists quite only the music the listener wants to which is effectively competitive. Thus, unwilling to assume that a given hear, the proof of the pudding, so to the steering adjustment is of a class with monopoly or oligopoly structure is speak, is in the listening, not in the any other adjustments necessary to inefficient and anticompetitive bristle at puffery used in advertising. harmonize the benchmark rate with the the idea that supranormal pricing Second, it is clear that Pandora has statutory requisites. See Web II, 72 FR arising from a complementary oligopoly not taken any steps to conceal that it has at 24092 (noting the Judges’ duty ‘‘to is reflective of a well-functioning engaged in such steering or that it determine if the benchmark agreements competitive market. See, e.g., Francesco intends to do so going forward. In the require any further adjustments based Parisi and Ben DePoorter, The Market present proceeding, the parties had the on any evidence of differences between for Intellectual Property: The Case of ability, which they exercised with the benchmark market and the target Complementary Oligopoly in The regularity, to enter into closed session to Economics of Copyrights: Developments hypothetical market.’’). avoid public disclosure of commercial in Research and Analysis (W. Gordon information they intended to maintain It is important to emphasize the and R. Watt eds. 2003) (noting the as confidential. However, at no time did limited nature of this sort of effective economic benefits of blanket licenses in Pandora attempt to close the competition. Price competition through reducing the greater-than-monopoly proceedings to prevent the public from steering does not diminish the stand- pricing of complementary oligopolists); learning of the introduction of steering alone monopoly value of any one sound Mark Lemley and Philip Weiser, Should into its music delivery model. The recording. Further, effective competition Property or Liability Rules Govern Judges note that no competing service through steering does not diminish the Information? 85 Tex. L. Rev. 784, 786– has advertised against Pandora or firm-specific monopoly value of each 87, 824 (2007) (comparing the ‘‘hold iHeart, attacking its use of steering. 5/ Major’s repertoire taken as a whole. up’’ (‘‘rent seeking’’) strategies of 19/15 Tr. 4775–76 (Shapiro). Thus, the Although Dr. Katz urged the Judges to copyright owners seeking supranormal evidence is not sufficient to indicate reduce the statutory rate to eliminate complementary compensation and of that Pandora would suffer an economic that market power as well, Katz WDT the owner of a parcel of real property loss merely from listener awareness that ¶ 43, the Judges decline to do so. There that is complementary to multiple other Pandora engages in steering. is absolutely no record evidence to parcels required for a large scale Third, although the extent of the suggest that the market power that a development, and noting that a steering may be economically Major enjoys individually by ownership compulsory license with a royalty rate significant to the licensors and of its collective repertoire is in any set by a regulatory authority (noting the licensees, the extent of steering at issue sense the consequence of improper CRB by name) can ‘‘minimize the in this proceeding may have little activity or that it is being used opportunity for rent-seeking behavior’’). noticeable impact on listeners. For individually by a Major to diminish iii. Did Pandora test steering under example, consider the result if, competition. That is, the Judges have no ‘‘Real World’’ conditions? hypothetically, a noninteractive service evidence before them to demonstrate were to steer away from Major A (which The Judges do not agree with had a pre-steering natural (historic) play that the Majors’ size and individual SoundExchange’s criticism that the market power is not the result of the rate of 40% on that service) by 12.5%. impact of steering is uncertain because Ex ante steering, the copyright on 4 in efficiencies and economies of scale and/ listeners were unaware that such or their superior operations. See every 10 songs played on that steering was being undertaken. The noninteractive service was owned by generally, Harold Demsetz, Industry Judges reach this conclusion for three Structure, Market Rivalry, and Public Major A. Steering away from Major A by reasons. 12.5% would reduce Major A’s play rate Policy, 16 J.L. Econ. 1, 3 (1973) (noting First, there is no evidence that by 5 percentage points (12.5% of 40% that ‘‘scale economies,’’ ‘‘[n]ew Pandora, or any noninteractive service, is 5 percentage points). Thus, ex post efficiencies’’ and ‘‘superior ability’’ can obtains and retains listeners by steering, Major A’s songs would form a ‘‘competitive basis acquiring a describing in any detail the technical constitute 35% of the plays on this measure of monopoly power’’). In the methodology it uses to select songs. The noninteractive service instead of 40% of absence of evidence that the Majors’ purpose of a streaming service is to the plays. market shares preclude effective provide songs to listeners—if they enjoy Consider a consumer who listened to competition, the Judges have no basis the music they will be satisfied, if they this noninteractive service for a period on this record to adjust rates lower to do not enjoy the music they will be of time sufficient to hear 20 songs. reflect that market concentration. unsatisfied, to the commercial detriment Ex ante steering, the consumer would of the service. While it is true that This holding must not be confused have heard 8 songs from Major A’s with the Judges’ holding regarding the repertoire (40% × 20 songs = 8 songs). anticompetitive effects of the 142 The Judges’ findings on this issue are not only consonant with the expert opinions of Drs. Shapiro Ex post steering, the consumer would complementary oligopoly that exists and Katz, but are also consistent with the expert have heard 7 songs from Major A’s among the Majors. Because the Majors economic testimony of SoundExchange’s own repertoire (35% × 20 songs = 7 songs). could utilize their combined market witness in Web III, Dr. Ordover. See Web III The one replacement song from Remand at 23114 (summarizing Dr. Ordover’s power to prevent price competition testimony as concluding that ‘‘if the repertoires of another record company’s repertoire among them by virtue of their all [Majors] were each required by webcasters (i.e., would not be a random song, but rather complementary oligopoly power—as if the repertoires were necessary complements) . . . would be the song the algorithm or proven by the evidence of the pro- each [Major] would have an incentive to charge a tastemaker selected after disqualifying monopoly price to maximize its profits . . . 143 competitive effects of steering and the constitut[ing] higher monopoly costs . . . paid by the eighth song from Major A. The admissions of Universal and its agents webcasters to each of the [Majors].’’) (emphasis discussed supra, section IV.B.3—the added). The Judges in this determination adopt this 143 In his oral testimony, Dr. Shapiro utilized Judges must establish rates that reflect economic reasoning and will not allow such another example, assuming a 15% steering ‘‘boost’’ complementary oligopoly power to be incorporated to a Major with a prior ‘‘natural’’ performance rate steering, in order to reflect an into the statutory rate. of 20%. According to Dr. Shapiro, such a steering

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00054 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26369

issue thus is whether such a change in v. Is the value of steering available Pandora/Merlin Agreement. Dr. Shapiro, song delivery would diminish under the statutory license? on behalf of Pandora’s direct case, went listenership to a noninteractive service SoundExchange argues that any through each item of additional to a point that would be economically benefits from steering must be treated consideration and explained why he harmful to the service, thus dissuading like any other consideration in a direct either adjusted his benchmark value the service from steering. In fact, license that is not authorized under the higher (as in the case of certain Pandora presented evidence regarding Act. That is, SoundExchange asserts that advertising consideration) or declined to this issue, to which the Judges now steering must be independently valued, adjust the benchmark for other elements turn. and the separate value must be added to of potential value. the statutory rate. The Judges The Judges do not find that the mere iv. What is the impact of Pandora’s disagree.146 Steering under the Pandora/Merlin presence of other items of potential Steering, as Dr. Shapiro emphasized, value serves to disqualify the Pandora/ agreement and in Pandora’s Steering is simply an example of price Merlin Agreement as a suitable experiments? competition at work. Further, benchmark. Benchmarks may be § 114(f)(2)(B) of the Act and prior imperfect in the sense that they include Pandora’s steering under the Pandora/ decisional law require that the Merlin Agreement, which guarantees a features that are ill-suited for adoption commercial rate reflect an ‘‘effectively in the statutory rate. To reject a [REDACTED]% level of steering, has not competitive’’ market. Therefore, the proposed benchmark for that reason resulted in any negative feedback or value of steering is a component of the alone would be—to put it colloquially— other deleterious consequence for statutory license—not extraneous to it— throwing out the baby with the Pandora. Likewise, the series of steering and should not be excluded through an bathwater. Because there is no single experiments conducted by Pandora adjustment process or otherwise from undifferentiated market for the statutory indicated that Pandora could steer away the rate ultimately set by the Judges.147 from or toward a Major’s repertoire by service, benchmarks must be borrowed b. Does the Pandora/Merlin Agreement a change of ± 15% without causing a from other markets or sub-markets and contain non-statutory value that either statistically significant change in will always be imperfect to some degree (i) disqualifies the Pandora/Merlin listening behavior. McBride WDT ¶ 21. and either in need of adjustment or Agreement as a benchmark; or (ii) limited in their applicability. But to Importantly, SoundExchange levels diminishes the value of steering in the ignore a benchmark for that reason no criticisms at Pandora’s steering Pandora/Merlin Agreement? alone would be an inappropriate experiments, save to make the point, i. The Potential Presence of Non- indictment of the benchmarking process rejected above, that the experiments did Statutory Value Does not Disqualify the itself. not reflect ‘‘real world’’ conditions. See Pandora/Merlin Agreement as a Further, Dr. Shapiro testified that he SX RPFF ¶¶ 780–784 (and record Benchmark found these elements of additional citations therein).144 The Judges consideration to either: (1) Provide joint likewise fail to identify any problems SoundExchange and Pandora both note that several additional elements of value to Pandora as well as Merlin with regard to Pandora’s steering potential value are present in the members; (2) be unlikely to be achieved; experiments. Thus, the evidence is or (3) be already incorporated into his undisputed that Pandora can steer at 146 ± The Pandora/Merlin Agreement allows for a valuation. There was no sufficient least 15% of its music toward or away very limited and conditional [REDACTED]. See rebuttal by SoundExchange witnesses to from the Majors without a negative PAN Ex. 50141(c)(v) and (2)(c). However, there is these points. As the Judges explain infra impact on listenership.145 no evidence in the record to suggest that such a limited and conditional [REDACTED] would be in their discussion of the same issue in exercised and, if so, how often. There is also no connection with the iHeart/Warner change would have ‘‘almost no perceptible impact evidence in the record to demonstrate the extent Agreement, an important general on the listening experience, as it would entail a this [REDACTED] would impact the effective rate change in ‘‘one [song] out of 30’’ or ‘‘one song every under the Pandora/Merlin Agreement. Therefore, consideration relating to this issue is the couple hours.’’ 5/19/15 Tr. 4630–35 (Shapiro) (and this contractual safeguard does not constitute a absence of evidence of value from a also explaining that steering need not result in a basis to adjust the Pandora/Merlin benchmark. party with regard to such additional change with regard to the seeded song or artist, but 147 SoundExchange attempts to impeach Dr. terms, when that party has the incentive rather would affect only subsequent songs played Shapiro on this point by seeking to use his rebuttal on the listener’s station). testimony against him. See SX PFF ¶ 705 (‘‘[Dr.] (as well as the means) to provide the 144 This is a curious criticism of an economic Shapiro also acknowledged that steering Judges with such evidence. experiment. By its very nature, an economic commitments have value. In response to [Dr.] Rubinfeld’s statement that ‘‘a direct license Additionally, SoundExchange’s experiment, or an economic model, is intentionally assertion that the additional items not designed to replicate real world conditions, but containing a binding steering commitment is unsuitable as a benchmark unless some adjustment rather to isolate certain conditions of the real world created sufficient value to offset the is made to reflect the value of the commitment to for testing and to hold the other conditions lower rate in the Pandora/Merlin the record company,’’ [Dr.] Shapiro agreed with constant. The particular condition that [Dr.] Rubinfeld that ‘‘some adjustment is Agreement strikes the Judges as SoundExchange claims the steering experiments appropriate.’’ Shapiro WRT at 41. However, economically irrational. If the supposed held constant—listener knowledge of steering in the SoundExchange omitted the remainder of Dr. additional value of the non-steering algorithm—seems wholly beside the point to the Shapiro’s testimony, which omission seriously Judges. To state the obvious, consumers listen to distorts his opinion: Without the omission, Dr. items in the Pandora/Merlin Agreement noninteractive services because of the quality of the Shapiro’s full testimony on this point states: ‘‘[Dr.] equals the difference between the non- music, not because of their interest in what goes Rubinfeld takes the position that a direct license steered rates and the lower steered rates, into the algorithmic ‘‘black box.’’ If the music is of containing a binding steering commitment is then what is the point of the parties poor quality, then listeners will vote with their unsuitable as a benchmark unless some adjustment feet—or, more correctly,—with their ears. is made to reflect the value of the commitment to incurring the transaction costs 145 iHeart did not run experiments regarding its the record company. I agree that some adjustment associated with negotiating such a deal? steering of sound recordings [REDACTED]. is appropriate, but only to the extent that the Why would Pandora commit to incur However, iHeart [REDACTED] and received steering commitment exceeds the amount of significant expenses to begin to set up complaints from noninteractive custom listeners steering that the webcaster would engage in just that [REDACTED]. See 6/2/15 Tr. 738–51 (Cutler); based on price differences. Id. (emphasis in an infrastructure necessary to perform SX Ex. 1037 [REDACTED]’’). original). the steering function?

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00055 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26370 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

ii. The Evidence Does not Support a Although Mr. Barros represents only increase the Pandora/Merlin benchmark Lessening in the Usefulness of the one Indie, SoundExchange selected him to reflect any extra-statutory Pandora/Merlin Agreement as a as a representative of the Indies’ consideration that was not already Benchmark for the Rates Indies Would position regarding the value of the accounted for by Dr. Shapiro. Pay in the Hypothetical Market Beyond Pandora/Merlin Agreement. Clearly, c. Is Merlin sufficiently representative of the Adjustments Made by Dr. Shapiro SoundExchange could not present the a segment of the sound recording In rebuttal to Dr. Shapiro’s item-by- testimony of more than [REDACTED] market? item consideration of the potential opting-in Merlin members, and the Judges therefore find the testimony The Judges reject SoundExchange’s additional items of value in the argument that Merlin is not sufficiently Pandora/Merlin Agreement, against interest by this Merlin member selected by SoundExchange to be representative of the independent sector SoundExchange did not introduce of the sound recording industry. The expert testimony to establish alternative particularly probative. Additionally, a May 15, 2014 internal Judges rely on several facts in reaching values. Rather, SoundExchange relied email written by Mr. Lexton appeared to this conclusion. on the narrative testimony of industry the Judges to reference Merlin’s strategy First, the Judges note that between witnesses Glen Barros, Darius van to attempt to obfuscate the usefulness of [REDACTED] and [REDACTED] Merlin Arman and Simon Wheeler to support the Pandora/Merlin Agreement as a members, out of approximately the position that these other items had benchmark in this proceeding: [REDACTED] total members opted-in to some unquantified value to the Merlin the Merlin Agreement. Thus, it is members. Although such after-the-fact [REDACTED] accurate to state that the evidence assertions can carry some weight, the SX Ex. 102. Thus, it appears to the Judges regarding the Pandora/Merlin Judges find such testimony to be that Merlin’s negotiation of additional terms was intended (at least in part) ‘‘to facilitate’’ Agreement relates—to use Dr. Talley’s inconsistent with Merlin’s conduct the very argument SoundExchange now term—to [REDACTED] to [REDACTED] during the negotiations. asserts through Mr. Lexton’s testimony ‘‘dyads’’ between licensors and a More particularly, although Merlin regarding the purported significance of the licensee. The Judges find this quantity has the ability to negotiate and evaluate unvalued additional terms. of contracts to be significant and agreements in a sophisticated manner, it In a subsequent email to Pandora dated probative with regard to: (1) Steering failed to value these additional elements June 3, 2014, Mr. Lexton made Merlin’s rates that Indies would accept; and (2) of consideration. See, e.g., 5/1/15 Tr. position in this regard even more explicit, by the principle that steering can be 125–52 (Simon Wheeler) (Merlin, is asking Pandora to include the following utilized as means of price competition ‘‘just as capable of understanding the proposed language in the final agreement: in the noninteractive market. [REDACTED] complexity of the rights and licenses at In addition, the Judges do not find issue in digital streaming as major PAN Ex. 5116 at SNDEX0315243. That persuasive SoundExchange’s argument record labels.’’); 5/28/15 Tr. 6513 request was rejected by Pandora and the that a majority of Merlin members who (Barros) (agreeing that independent requested language was never included opted-in to the Pandora/Merlin label ‘‘Concord’s assessment of the in the final Pandora/Merlin Agreement. Agreement did so through their value it receives from licensing its Id. Nonetheless, Merlin proceeded to agreements with aggregators and/or repertoire is just as sophisticated as any enter into the Pandora/Merlin distributors. These opting-in members other label.’’); 6/1/15 Tr. 6924–25 Agreement, anticipating that it would be delegated the decision whether to opt- (Lexton) (‘‘Merlin brings expertise to used by Pandora as evidence in this in to these distributors and aggregators bear on its negotiations with digital proceeding. See, e.g., 6/1/15 Tr. 6962, and there was certainly no evidence or music services.’’). If the extra-statutory 6966 (Lexton); id. at 7095 (Wheeler); SX testimony to suggest that these items were of particular and essential Ex. 102 at 3 (5/14/15/14 email among arrangements were coerced or that any value to Merlin, the Judges would have Merlin executives); PAN Ex. 5117 at Merlin members who opted-in through expected to be presented with evidence SNDEX0437582 (6/9/14 internal email this process disagreed with the decision. as to how Merlin valued these several from Mr. Lexton). Thus, the decision by Merlin members items. However, as noted, no such The foregoing emails and testimony, to delegate the decision whether to opt- evidence was presented.148 combined with Merlin’s and in to its agents is a component of the Additionally, one Merlin member SoundExchange’s failure to separately business model these Merlin members presented as a witness by value the other elements of chose to follow. The Judges cannot SoundExchange, Glen Barros, President consideration either during negotiation criticize the decision of these Merlin and C.E.O. of Concord Record Group, or during the proceeding, strongly members, and by extension, call into testified that ‘‘in all likelihood’’ he indicate to the Judges that Merlin found question their intention to be bound by would have opted-in to the Pandora/ the value in the Pandora/Merlin the Pandora/Merlin Agreement, merely Merlin Agreement even if these other Agreement to lie in the steering—that is, because they have arranged their elements of value had not been included the trade-off of more plays at a lower licensing affairs in this manner. By way in that agreement. 5/28/15 Tr. 6537–39 rate for more total revenue. of analogy, just as SoundExchange’s (Barros) (emphasis added).149 In sum, if there was any additional criticism of Pandora’s business model is value to Merlin from the other items not relevant to the setting of rates in this 148 In fact, with regard to one of the unquantified sufficient to reduce the overall value of proceeding, the Judges do not find items of alleged value—the [REDACTED] steering as adopted for a statutory relevant the business judgments of provision—contemporaneous correspondence license, the record evidence fails to Merlin members to utilize aggregators among Merlin members and personnel discounted provide a basis for such an adjustment. any value in the [REDACTED] provision in the and/or distributors as their agents in Pandora/Merlin Agreement. PAN Ex. 5110 at For these reasons, the Judges decline to this regard. SNDEX0374284 (Correspondence from Relatedly, the Judges find that the fact [REDACTED] stating that ‘‘[REDACTED]’’). testimony undercut the testimony quoted in the that Merlin negotiated collectively on 149 SoundExchange asserts that Mr. Barros’ text, supra. The Judges find Mr. Barros’ testimony subsequent testimony that he found the ability for as cited in the text, supra, to be credible, and they behalf of its members does not diminish his record company to receive royalties on pre-1972 find that his subsequent attempt to qualify that the value of Merlin as a party capable royalties to be a ‘‘gating’’ issue and that such testimony to be lacking in credibility. of entering into an agreement that is

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00056 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26371

otherwise an appropriate benchmark. members found the terms of the came from any single music user, then that Merlin members utilize the collective agreement beneficial to them, music user could well have monopsony capacities of Merlin in order to transact notwithstanding any alleged separate power over Merlin. But this is demonstrably licensing business in a more efficient benefits to Merlin as a collective not the case for Pandora. The Merlin Labels generate revenues from many different users manner, as described by a Merlin’s organization. of their sound recordings, including other testifying executive, Mr. Lexton: The Judges also reject the criticism noninteractive webcasters, interactive Merlin’s purpose is to allow independent that Merlin has not uniformly services, and from the sale of physical record companies to benefit from direct deals represented its members because albums and digital downloads. In fact, I negotiated by Merlin on a collective basis. As Pandora has used its editorial discretion estimate, based on data for the recorded such Merlin is a one stop shop for recorded to exclude (as of the time of the hearing) music industry overall, that Pandora music rights licensing. It represents recorded from its playlist sound recordings accounted for roughly 5 percent of the music rights owned and/or controlled by owned by some of the opting-in Merlin revenues received by the Merlin Labels in independent record labels and distributors 2013 for the licensing of their music in the members. There is no allegation that United States. Thus, Pandora’s share of the who are eligible and choose to join Merlin. Pandora promised to make all sound . . . Merlin’s core remit is to represent its Merlin Labels’ revenues is far short of the members in negotiating licenses with digital recordings available on its service, and level that would be necessary for Pandora to music services in the hope of overcoming therefore each Merlin member accepted have undue market power in its negotiations market fragmentation issues that have the risk that Pandora, in its editorial with Merlin. historically challenged the independent judgment, might not include some or all Shapiro WDT at 24–25 (emphasis music sector particularly in the digital of its sound recordings. added). The Judges find this explanation domain. Finally, the Judges do not find merit sufficient to contradict the assertion that Lexton WRT ¶¶ 11–12. Indeed, Merlin in SoundExchange’s argument that Pandora exercised undue market power apparently is sufficiently successful in Merlin is not a sufficient representative in negotiating the terms of the Pandora/ this endeavor that one of the Majors, of Indies in the marketplace. Merlin Agreement. [REDACTED], has characterized Merlin SoundExchange did not produce any There is an additional and separately as the ‘‘fifth Major.’’ PAN Ex. 5349 at 9 witnesses from Indies who were not sufficient reason why SoundExchange’s ([REDACTED] approvingly noting to members of Merlin to testify to this claim of Pandora’s monopsony power [REDACTED] that Merlin publicly effect. Rather, SoundExchange produced cannot be adopted. The assertion that presents itself as a ‘‘fifth major’’).150 witnesses whose Indie record Pandora exercised market power in Further, the Judges reject companies did opt-in to the Pandora/ these negotiations ignores the fact that SoundExchange’s assertion that Merlin Merlin Agreement. Given Merlin’s Merlin did not have to accept any of as a collective had different incentives capacity to negotiate and its well- Pandora’s terms—Merlin and its than its members that somehow regarded industry status, the fact that members could have fallen back on the diminish the value of the Pandora/ non-Merlin Indies are not covered by Pureplay statutory settlement rates Merlin Agreement as a benchmark. the Pandora/Merlin Agreement, in the rather than accede to any demand by These incentives included financial and absence of other evidence, is not Pandora. That is, by this particular status benefits to Merlin if its members sufficient to call into question the assertion, SoundExchange is assuming opted-in, which were distinct from usefulness of this benchmark. arguendo that the effective Pandora/ whatever benefits individual members d. Did Pandora have substantial market Merlin rates are below an appropriate might obtain by opting in to the market rate because of Pandora’s market Pandora/Merlin Agreement. The Judges power that is reflected in lower effective rates in the Pandora/Merlin Agreement? power.151 But why would Merlin and its understand this criticism to be based members voluntarily enter into an upon the classic principal-agency The Judges reject SoundExchange’s agreement to accept rates lower than the problem, in which the interests of the assertion that Pandora had significant statutory alternative and lower than principals (Merlin members) may not be market power that caused the effective what would exist in a competitive fully aligned with the interests of the rates in the Pandora/Merlin Agreement market? agent (Merlin). However, this is a to be lower than effectively competitive Therefore, the Judges reject the common problem when principals rates. Initially, the Judges note that this assertion that Pandora exercised undue delegate functions to agents. Unless the assertion is not supported by any market power in negotiating the evidence demonstrates that the agent empirical market data, analysis, or effective rates contained in the Pandora/ (Merlin) has engaged in a breach of duty comparison with other negotiated Merlin Agreement. toward its principals (Merlin members), comparable interactive rates. the lack of a complete alignment of More importantly, the issue of e. Was the Pandora/Merlin Agreement interests does not invalidate the Pandora’s ‘‘market power,’’ vel non, was merely ‘‘experimental?’’ benchmark status of the agreement anticipated and addressed by Pandora’s Two of SoundExchange’s witnesses entered into by the principal. Indeed, economic expert, Dr. Shapiro, who characterized the Pandora/Merlin because this is the principal-agent explained: Agreement as an ‘‘experiment,’’ as arrangement that the Merlin members Pandora is the largest noninteractive distinguished from an actual voluntarily created—including whatever webcaster. I have considered specifically marketplace agreement. The Judges misalignments in incentives might whether Pandora had undue market power in reject this attempt to characterize this theoretically exist—it is especially its negotiations with Merlin. In the language real agreement, involving the exchange representative of a marketplace of antitrust economists, I have considered of actual consideration, as an whether Pandora has monopsony power over transaction. The fact that approximately ‘‘experiment.’’ [REDACTED]-[REDACTED]% of Merlin. Pandora’s share of listening among noninteractive webcasters is not the key An economic experiment is Merlin’s [REDACTED] members opted- variable for determining whether or not undertaken under controlled laboratory in to the Pandora/Merlin Agreement is Pandora has monopsony power over Merlin. conditions, as distinguished from compelling evidence that the Merlin Rather, the correct variable upon which to focus is the share of the Merlin Labels’ 151 SoundExchange is thus assuming here that, 150 At the time, there were four Majors, Universal, revenues that comes from Pandora. If a very under section 114(f)(2)(B), a benchmark rate must Sony, Warner, and EMI. large share of the Merlin Labels’ revenues reflect an adequate level of competition.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00057 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26372 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

market transactions that take place in licensing of rights under the Pandora/ individual Merlin Labels, Pandora the real world. See Guillaume R. Merlin Agreement. represented in the Pandora/Merlin Frechette and Andrew Schotter, Pandora does not dispute that it had Agreement only to [REDACTED] to Handbook of Experimental Economic not (as of the hearing date) been able to increase spins by at least Methodology 21 (2015) (‘‘[T]o run an implement all the benefits promised in [REDACTED]% above the natural rate. experiment . . . experimenters are of the Pandora/Merlin Agreement. Thus, the individual members necessity engaged in market design in However, the Judges note that objectively cannot complain about the the laboratory.’’) (emphasis added). SoundExchange did not produce any level of overspinning at any point in Quite clearly, the Pandora/Merlin correspondence from Merlin or its time, unless they can also claim that Agreement was not and is not an members complaining about the failure Pandora had not been [REDACTED]. As of Pandora to perform, or any threat to ‘‘economic experiment.’’ noted above, SoundExchange did not SoundExchange’s witnesses may have terminate the agreement or sue Pandora produce any evidence suggesting that used the word ‘‘experiment’’ to suggest for nonperformance. Rather, the any individual members had lodged a tentative or impermanent relationship evidence suggests that Merlin such a complaint. between Pandora and Merlin. If so, that recognized that the structuring of criticism proves too much, as all performance needed to be an ongoing With regard to SoundExchange’s benchmark agreements—indeed and collaborative effort. As Pandora’s claim that Pandora has incurred virtually all agreements—could be Chief Financial Officer, Mr. Herring, substantial unexpected capital costs in characterized as ‘‘experiments,’’ in that testified: implementing a steering system, Mr. they have stated durations, and the [REDACTED] Herring testified that these investments, parties are free to vary the terms of their although motivated in the short-term 5/18/15 Tr. 4318 (Herring); see also economic relationship after the so- and in part by the Merlin Agreement, in PAN Ex. 5014 (Pandora/Merlin called ‘‘experiment’’ has expired. In this fact laid the groundwork for Pandora to Agreement, ‘‘Feature Implementation sense, the word ‘‘experiment’’ is Timeline’’), Exhibit C thereto implement steering more broadly across misused to cast a wide disqualifying net ([REDACTED]’’ (emphasis added). the non-interactive webcasting market. on all benchmark agreements. SoundExchange did not produce 5/18/15 Tr. 4313–17 (Herring) (‘‘some of f. Has Pandora’s performance under the evidence to call into question Pandora’s these costs are fixed costs to be Pandora/Merlin Agreement performance under this [REDACTED] amortized over time with the compromised the usefulness of that clause. anticipation of being applied to other benchmark? 152 direct licenses with other record More importantly, the evidence companies, and expensed at the time indicates that Pandora has performed its Even assuming that the Pandora/ that the costs are incurred, and therefore core obligation under the Pandora/ Merlin Agreement is, in principle, a ‘‘spread over those deals.’’). Thus, the Merlin Agreement: The increase in useful benchmark, SoundExchange asks existence of these costs does not the Judges to look to Pandora’s alleged spins of Merlin recordings, in the establish any fact to contradict the poor performance of its obligations aggregate, by at least [REDACTED]%, Judges’ finding that the Pandora/Merlin under the Pandora/Merlin Agreement. above their collective ‘‘natural’’ rate. In Agreement is a useful benchmark. In As detailed supra, SoundExchange fact the evidence shows that Pandora is fact, Pandora’s commitment to incur alleges that Pandora has failed to overspinning Merlin member recordings substantial build-out costs to create the perform certain contract obligations collectively by [REDACTED]%. On the (such as, e.g., [REDACTED]) and that the individual Merlin label level, the results steering architecture underscores that cost of performance is daunting for have been uneven—some Merlin labels this agreement (and the iHeart/Warner Pandora, which combine to create what have been overspun by [REDACTED]- Agreement) represents the cutting-edge one might call ‘‘seller’s remorse’’ among [REDACTED]% of their natural rate, see of a technological advance that can Merlin participants with regard to the 5/18/15 Tr. 4229–30, 4291–4293 ameliorate the anticompetitive effects of (Herring); SX Ex. 2310 (showing a complementary oligopoly. 152 hundreds of Merlin Labels with rates of A general issue of proof arose in this g. Do the steering experiments and the proceeding as to whether a benchmark’s value can overspinning exceeding Pandora/Merlin Agreement demonstrate be measured by the parties’ performance under a [REDACTED]%)—but other Merlin proposed benchmark agreement, in addition to the the rate to which a major would agree? parties’ expectations of value when the benchmark Labels are spinning at less than a was created. This issue arose in a different context, [REDACTED]% increase their above The Judges find this SoundExchange regarding whether iHeart’s ‘‘incremental’’ rate their prior levels. SX Ex. 1748 at 2; SX criticism to be meritorious. These analysis of its iHeart/Warner Agreement benchmark Ex. 2310.153 should be analyzed by reference only to the parties’ steering experiments reflect only a However, the only specific promise by quantity adjustment that could be expectations at the time of contracting, or whether Pandora of increased spins in the the Judges should also consider the parties’ attempted with regard to the Majors, not Pandora/Merlin Agreement was its performance under the iHeart/Warner Agreement. a rate adjustment arising from steering As discussed in detail infra, the Judges have promise [REDACTED] to increase to or from a Major. By contrast, the rejected iHeart’s ‘‘incremental’’ rate analysis, Merlin spins collectively by thereby mooting the issue of whether the parties’ Pandora/Merlin Agreement does reflect performance under that agreement affected the so- [REDACTED]%, and it appears undisputed that Pandora has performed the impact of steering on negotiated called ‘‘incremental’’ rate. With regard to the rates (as does the iHeart/Warner Pandora/Merlin Agreement, SoundExchange argues this obligation and, in fact, has far that Pandora’s performance under the Pandora/ exceeded the [REDACTED]% minimum. Agreement). Thus, while the Judges find Merlin Agreement indicates that the agreement is With regard to the underspinning of the steering experiments to be probative not usable as a benchmark. Because—as explained of the general principle that steering can in the text, infra—the Judges find that Pandora’s performance does not cause them to reject the 153 Labels owned by Beggars Group (whose be effected to some extent without a Pandora/Merlin Agreement as a usable benchmark, officer, Simon Wheeler claimed the Pandora/Merlin negative impact on listenership, the the question of whether evidence of performance is Agreement was a failure)—including XL Judges do not accept that this generally appropriate to consider when setting rates Recordings, Matador and Nation Records—are being constitutes direct evidence sufficiently need not be decided by the Judges in this overspun on Pandora by as much as determination. [REDACTED]%. SX Ex. 2310. probative of the rates that would result

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00058 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26373

from steering writ large in the Therefore, the Judges consider the rate clinic, 65 F.3d 1406, 1415 (7th Cir. marketplace.154 established by the Pandora/Merlin 1995) (Posner, J.) (‘‘It would be a strange Moreover, Pandora’s own witness Agreement to establish only one interpretation of antitrust law that testified in a manner that contradicts guidepost (i.e., a relevant financial point forbade competitors to agree on what Pandora’s attempt to bootstrap the of reference) to a statutory rate. The price to charge, thus eliminating price Pandora/Merlin rates onto the Majors. Judges are informed as to the limited competition among them, but allowed Mr. Herring, Pandora’s C.F.O., testified weight of this rate in the ultimate them to divide markets, thus that Pandora would have to offer a statutory rate they shall set, by the fact eliminating all competition among higher steering-based rate to a Major that Indie sound recordings reflect them.’’).156 than Pandora obtained in the Pandora/ approximately [REDACTED]% of the While the Majors’ individual market Merlin Agreement. 5/18/15 Tr. 4253 sound recordings played on Pandora. power is not in itself necessarily (Herring). The Judges have noted SX Ex. 269 at 73. improper, the hypothetical exercise of previously that the Majors’ repertoires that power in this manner in the h. Can the Majors avoid steering in the must be distinguished from those of the noninteractive market would be hypothetical market? Indies. See SDARS II, 78 FR at 23063 antithetical to the ‘‘effective (the Majors are distinguishable from the SoundExchange argues that any competition’’ requirement inherent in Indies ‘‘by virtue of the depth and attempt by a noninteractive service to the § 114(f)(2)(B) standard. That is, each breadth of their music catalogues impose steering on the record Major may well be entitled by its firm- [which] make up a critical portion of the companies would be rebuffed by the specific market power to higher rates sound recording market.’’).155 Majors. In particular, SoundExchange than the Indies, but the Majors cannot argues that the record companies would bootstrap that power into a further 154 The use of benchmarking serves to tie the respond to a steering threat by: (1) capacity to reap the benefits of a quantity aspect of steering to its impact on rates, Withholding their entire repertoires; (2) and the absence of a relevant Majors’ benchmark in complementary oligopolist by Pandora’s evidence prevents the Judges from imposing Anti-Steering or ‘‘Most brandishing such power as a sword determining a steered price for Majors from that Favored Nation’’ contract clauses; and/ against steering. evidence. Although Dr. Shapiro asserts that the or (3) requiring up-front lump sum Thus, in the present case, the steering experiments demonstrate that the Majors royalty payments from the hypothetical use by one or more of the should receive the same rate as the Indies in a market with steering, that opinion is contradicted noninteractive services. Majors of its power to boycott a by the higher rate set forth in the [REDACTED] i. Withholding the Entire Repertoire noninteractive service—one that had Agreement which also contains a significant sought to inject some price competition steering component. Dr. Shapiro attempts to explain A Major could respond to a threat of into the market via steering—would the higher [REDACTED] rate as a function of a so- steering by threatening to withhold its called ‘‘focal point,’’ ‘‘anchor’’ or ‘‘magnet’’ effect undermine the ‘‘effective competition’’ created by the extant applicable statutory rate, that entire repertoire from that standard that the D.C. Circuit, the allegedly raises the negotiated rate toward (yet still noninteractive service. There appears to Librarian of Congress and the Copyright below) the statutory rate. However, although this be a consensus that the repertoire of Royalty Judges have declared to be an theoretical effect is discussed in the economic each of the three Majors is a ‘‘must literature, Dr. Shapiro acknowledged that it is not essential element of the § 114(f)(2)(B) an ‘‘ironclad’’ economic law, and there is scant have’’ in order for a noninteractive standard. evidence in this proceeding why such a potential service to be viable. See 5/18/15 Tr. ‘‘focal point’’ or ‘‘magnet’’ effect would cause 4254 (Herring) (admitting that without ii. Anti-Steering or MFN Clauses unconstrained licensors to eschew a lower market the repertoire of a Major, it would be a In the interactive market, the Majors rate that would produce greater revenue. much different service); 5/18/15 Tr. 155 Dr. Shapiro opines that the Majors’ advantage commonly include anti-steering or MFN in the hypothetical market would be reflected 4472 (Shapiro) (declining to state the clauses in their agreements with the economically solely through the greater number of majors are not ‘‘must haves’’ for services. The Judges find that such noninteractive plays, rather than also in a higher noninteractive services); see also SX Ex. clauses have no purchase vis-a`-vis per-play rate. See, e.g., 269 at 74 (noting disproportionate share 5/20/15 Tr. 5058 (Shapiro) (testifying that the larger steering in exchange for lower rates in repertoires of the Majors ‘‘does not mean’’ that the of top spins from Majors’ repertoires). the noninteractive market. In the Majors deserve a ‘‘greater value per-performance.’’); However, the ability of the Majors to noninteractive market, an insistence by 5/19/15 Tr. 4730 (Shapiro) (rejecting use of market utilize such a boycott to defeat steering a Major that a noninteractive service share alone in determining ‘‘value per spin’’). would be a function of their However, Dr. Shapiro ignores the fact that there is abide by an anti-steering clause, or a apparently a greater per-song value overall for songs complementary market power. Simply MFN clause that has the same effect, is in the Majors’ repertoire, as evidenced by Pandora’s put, demands by the Majors to prevent tantamount to importing the own data—showing that the Majors account for steering by insisting that a anticompetitive complementary [REDACTED]% of ‘‘top 5% weekly spins,’’ noninteractive service not deviate from [REDACTED]% of the ‘‘top 10% weekly spins,’’ and oligopoly power of the Majors from the [REDACTED]% of the ‘‘top 20% weekly spins’’— an historical (‘‘natural’’) division of interactive market into the despite the fact that the Majors account for only market shares would be a classic noninteractive market. Dr. Rubinfeld’s [REDACTED]% of the total spins on Pandora. example of anticompetitive conduct. rebuttal testimony at the hearing is Compare SX Ex. 269 at 74 with SX Ex. 269 at 73. See, e.g., Blue Cross & Blue Shield These ‘‘top spin’’ figures are indicative of the ‘‘must telling: have’’ aspect of the Majors’ repertoire (leaving aside United of Wisconsin v. Marshfield the anticompetitive complementary nature of their 156 The Judges emphasize that their analysis in combined repertoires). Indeed, the record suggests chicken can result in both: (1) More revenue to the the text, supra, is not intended to suggest any to the Judges that the popularity of the Majors’ restaurant for each steak dinner compared with antitrust violations by any actor in the interactive spins is the reason why steering away from their each chicken dinner; and (2) more total revenue or noninteractive market. The Judges’ concern repertoires cannot be pursued beyond a certain attributable to the greater number of steak dinners under section 114(f)(2)(B) is to set rates that reflect level, and why Dr. Shapiro candidly declined to arising from the patron’s more frequent visits to the a hypothetical market that is effectively reject the idea that the Majors’ repertoires were restaurant to eat steak. In more formal economic competitive. If the hypothetical market posited by ‘‘must haves’’ even though noninteractive services terms, the typical listener (or the restaurant patron) one of the parties to this action would result in rates could steer away from them to an extent. To use an gets more ‘‘utility’’ from the Majors’ songs (or from that were not effectively competitive, then such a imperfect yet helpful analogy: A regular restaurant the steak) each time one is ‘‘consumed,’’ and also hypothetical market must be rejected—even if it diner might prefer steak to chicken, to the extent consumes those songs (and steaks) more often. The would be the result of tacit or other conduct that that she orders steak 7 out of every 10 meals at the seller can benefit from both the greater ‘‘utility’’ and might not rise to the level of a violation of the restaurant. This greater demand for steak versus the frequency of purchases. antitrust laws.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00059 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26374 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

Q: Now [Dr.] Shapiro has testified that the invalidation. See U.S. v. American This up-front lump sum strategy in threat of steering, alone, would lead to lower Express Co., 88 F. Supp. 3d 143, 189, actuality is merely another way in rates from record companies. What’s your 194 (E.D.N.Y. 2015) (‘‘anti-steering which a Major could bootstrap its view of that opinion? rules’’ can ‘‘block pro-competitive otherwise unobjectionable market [DR. RUBINFELD] I don’t think it’s likely to happen because I efforts’’ to the extent that ‘‘the market is power to preserve complementary don’t think the threat . . . is a credible broken,’’ when such rules prevent oligopoly power in the noninteractive threat—that would be the term we use in ‘‘price competition,’’ by not permitting market. The Judges note that economics—and the reason is . . . that, first buyers ‘‘to use their lowest cost SoundExchange’s expert economic of all, the record companies, as I have said supplier, as they can in other aspects of witness, Dr. Rubinfeld, has written that a number of times before, do have substantial their businesses.’’); United States v. ‘‘[i]n dynamically competitive bargaining power and they have responses to Apple, 791 F.3d at 320 (‘‘we are industries, where new product and the threat that takes away its credibility. In breaking no new ground in concluding features are an important part of the rather strong version, they could . . . that MFNs, though surely proper in competition, even licenses that include look to other sources of listeners and say many contexts, can be ‘‘misused to we’re going to consider not using your only fixed, or lump-sum payments, can service, but . . . they could say we’re not anticompetitive ends in some cases.’’). result in an anticompetitive lessening of going to feature all of the same artists, maybe The Judges likewise find the competition.’’ Daniel L. Rubinfeld and we’ll take some of our top artists off our hypothetical use by the majors of anti- Robert Maness, ‘‘The Strategic Use of offerings .... steering clauses in response to the threat Patents: Implications for Antitrust,’’ *** of price competition-via-steering would reprinted in Francois Leveque and [THE JUDGES] thwart ‘‘effective competition.’’ 157 Howard Shelanski, Antitrust, Patents Professor, do you think that the smaller iii. Up-Front Royalty Payments and Copyright 85, 91–92 (2005). In the independents have that same bargaining present context, the noninteractive power . . . to respond to the threat of SoundExchange asserts that a record steering . . . ? service that would be compelled to pay company could frustrate an attempt at to a Major an up-front lump-sum license [DR. RUBINFELD] steering by requiring noninteractive No. They wouldn’t have . . . quite the same based on the old per-play rate (or lose bargaining power. services to pay their royalties up-front access to 100% of the Major’s repertoire) *** in a lump sum, instead of on a per- would need to recover those fixed and [THE JUDGES] performance basis. Such a lump-sum sunk costs and thus forego price What do the independents lack that the requirement would frustrate steering in competition-via steering.158 [M]ajors have that makes the independents the following manner: If a licensee has In sum, each of the three contract unable to exercise that threat? already paid Record Company A a devices relied upon by SoundExchange [DR. RUBINFELD] required, large up-front fee (equal to its [T]ypically, they’re only going to have a few to defeat steering are dependent upon natural/historic play level multiplied by the exercise of market power to preserve artists that have really the name recognition the old, higher per-play rate) then the and the power to make a difference. the power of complementary oligopoly, [THE JUDGES] marginal cost going forward to the which would thwart effective So if the record company industry was more noninteractive service of playing a competition in the noninteractive atomistic, the threat of steering would be sound recording from Record Company market. Thus, all three contracting more credible, but because it’s not that A would be zero. By contrast, Record devices would be inconsistent with the atomistic . . . it makes the ability of the Company B—even if it offered a reduced statutory direction to set rates, based on [M]ajors to rebut the threat . . . more likely steering rate—would still be insisting on competitive information, that would be to be successful? a rate greater than the marginal rate of [DR. RUBINFELD] set between willing buyers and willing zero the licensee would be paying to sellers in an effectively competitive I think that’s true. . . . [T]hat’s a harder Record Company A. The noninteractive world for me to imagine because I have been marketplace in the absence of a in the world of seeing three or four major service would thus be compelled to statutory license. companies having a pretty big impact. either pay the up-front lump sum and lose the benefits of price competition, or i. Conclusion Regarding the Pandora 5/28/15 Tr. 6302–05 (Rubinfeld) refuse to pay the lump sum and lose Benchmark (emphasis added). access to 100% of the repertoire of For the foregoing reasons, the Judges This testimony underscores the point Record Company A. that the Majors’ capacity to undermine will utilize Pandora’s steering-based ‘‘price competition-via steering’’ is a benchmark as a guidepost to establish 157 Dr. Rubinfeld also speculated that in the the zone of reasonableness for the function of their complementary hypothetical market the Majors could ‘‘take some of oligopoly power. Once again, the Judges our top artists off our offerings’’ in response to an noninteractive royalty rates that would do not find that the mere size of the attempt at price competition-via steering. 5/28/15 be paid by Indies in the ad supported Majors or their share of the Tr. 6302 (Rubinfeld). But in that hypothetical (free-to-the listener) and subscription market, such an attempt by an entity with rights to markets. Pandora has proposed two sets noninteractive market is in itself collectively license a substantial market share anticompetitive (especially on this would invite scrutiny as anticompetitive. See of such benchmarks, depending upon record), but the Judges find that the ‘‘Dept. of Justice Sends Doc Requests, Investigating the level of steering the Judges find to ability of the Majors to leverage that UMPG, Sony/ATV, BMI and ASCAP Over Possible be appropriate for rate-setting purposes. ‘Coordination,’ ’’ Billboard.com (July 13, 2014). The Judges find that this guidepost market power to create the (‘‘The Department of Justice has sent out CIDs (Civil complementary oligopoly pricing Investigative Demand for Documents) to ASCAP, should be established by applying a rate problem can neither be imported into BMI, Sony/ATV Music Publishing and Universal Music Publishing Group in connection with their 158 The Judges are not stating that a requirement the noninteractive market nor assumed review of . . . whether partial withdrawals of of an up-front payment lump-sum royalty type to be part of the hypothetical effectively digital rights should be allowed.’’). Thus, such provision is per se inconsistent with effective competitive noninteractive market. behavior would not necessarily be consonant with competition. For example, in the [REDACTED] Indeed, in the hypothetical market ‘‘effective competition,’’ but rather an Agreement, discussed infra, [REDACTED] is anticompetitive leveraging of market power. The obligated to pay [REDACTED] to [REDACTED] even without a statutory rate, such anti- Judges thus decline to incorporate such licensor if [REDACTED]. SX Ex. 33 at 14–17, ¶¶ 3(a) and (d). steering clauses (and other anti-steering responses in the hypothetical effectively However, there is no evidence that this provision tools) would be ripe for judicial competitive market. would frustrate effective competition.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00060 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26375

premised upon the lower of the two 2. The Fischel/Lichtman Proposed Were not independently valued by the steering alternatives presented by Benchmark parties on a monetary basis; (2) Pandora: the [REDACTED]% steering a. The iHeart/Warner Agreement benefited both parties; and (3) therefore figure, rather than the higher 30% had an uncertain net value: figure.159 The lower [REDACTED]% Effective October 1, 2013, iHeart and • Warner’s grant to iHeart of sound level is appropriate because it is the Warner entered into an agreement (the recording rights [REDACTED]; level to which Pandora was willing to iHeart/Warner Agreement) that • iHeart’s commitment to provide commit [REDACTED]. PAN Ex. 5014 addressed, inter alia, the rates that Warner with no less than [REDACTED] ¶ 4(a). The Judges recognize the iHeart would pay to Warner for iHeart’s percent of total airplay devoted to a relatively nascent nature of steering. plays of Warner sound recordings on music advertising campaign that iHeart Although these factors certainly do not iHeart’s custom noninteractive service. provides on its webcast stations, known invalidate the Pandora/Merlin SX Ex. 33 (iHeart/Warner Agreement). as the Artist Integration Program Agreement as a usable benchmark, they As it pertained to these noninteractive (‘‘AIP’’); 161 do suggest to the Judges that the more plays, the iHeart/Warner Agreement • Warner’s [REDACTED] right to prudent course is to incorporate only provided that iHeart would pay the [REDACTED] and iHeart’s [REDACTED] the guaranteed 12.5% level of steering, greater of: (1) A per-performance fee on right to [REDACTED]); and and use the resultant rates as the custom performances; and (2) Warner’s • iHeart’s ‘‘most favored nation’’ appropriate guideposts for the rates pro rata share of a specified percentage protection vis-a` -vis [REDACTED], such attributable to the Indies portion of the of iHeart’s non-simulcast noninteractive that, if Warner were to enters into an statutory market.160 revenue. Specifically, the iHeart/Warner agreement to license sound recording Agreement calls for the following rates: rights for [REDACTED]’s [REDACTED] E. iHeart Rate Proposal and provide [REDACTED] with terms 1. Introduction IHEART/WARNER PER-PERFORMANCE that are more favorable than those iHeart proposes a per-play rate of ROYALTY RATES offered to iHeart, then iHeart would be $0.0005 for the § 114 license. In support afforded the option to adopt those Per-performance [REDACTED] terms. of this proposal, iHeart relies on the Calendar year rate analysis undertaken by its expert Fischel/Lichtman AWDT ¶ 38. Drs. Fischel and Lichtman described witnesses, Drs. Daniel Fischel and 2013 ...... $0.[REDACTED]. the [REDACTED] as an ‘‘insurance Douglas Lichtman, of rates set forth in 2014 ...... $0.[REDACTED]. policy’’ that benefited iHeart in the certain agreements entered into by 2015 ...... $0.[REDACTED]. event it would [REDACTED]. Likewise, iHeart in the market for noninteractive 2016 ...... $0.[REDACTED]. they described the AIP provision as an services. Each calendar year dur- $0.[REDACTED]. ing the Renewal Term ‘‘insurance policy’’ that benefited 159 The lower steering level results in a higher if any. Warner, because iHeart’s commitment to per-play rate. continue to provide the AIP benefit 160 Pandora attempted to corroborate its Pandora/ meant that Warner did not have to Merlin benchmark by introducing, in rebuttal, its IHEART/WARNER PERCENTAGE assume the risk that iHeart might charge agreement with a classical music record company, REVENUE ROYALTY RATES Warner for the right to access the Naxos of America, Inc. (Naxos), that had been benefits of AIP. See iHeart PFF ¶¶ 179– entered into as of January 1, 2015. PAN Ex. 5018 Period Percentage (the Pandora/Naxos Agreement). However, the 180 (and record citations therein). Judges reject the Pandora/Naxos Agreement as a First [REDACTED] [REDACTED]%. Drs. Fischel and Lichtman recognized corroborating benchmark for several reasons. First, the difficulty in quantifying the values Naxos, as a classical music label, is at best months after Effective representative of a narrow genre and therefore its Date. of what they described as these agreement cannot serve to be representative of a Months [REDACTED] [REDACTED]%. ‘‘insurance policy’’ equivalents. wider variety of sound recordings. 5/13/15 Tr. at after Effective Date. However, they aver that neither party 3512 (Herring). Second, the Pandora/Naxos Each month during the [REDACTED]%. assigned any values to these (and the Agreement does not contain any steering terms, but Renewal Term if any. rather sets a statutory per-play rate other) non-rate terms and that the net ($0.[REDACTED]), lower than the default rate value of these items therefore can only ($0.0014) established by the Pureplay settlement. SX Ex. 33 at 15–16 (iHeart/Warner be set at zero. Fischel/Lichtman AWDT PAN Ex. 5018. Although this difference, ceteris Agreement). ¶ 39. As Dr. Fischel further testified: paribus, would create an incentive for Pandora to The iHeart/Warner Agreement play more classical music owned by Naxos, there incorporates the same economic steering We followed the . . . real-world example was evidence, acknowledged by Dr. Shapiro, that of the parties . . . who did not price any of Pandora was constrained in any potential steering logic as the Pandora/Merlin Agreement. these terms. . . . [T]here was no separate toward Naxos by the fact that there was only one Specifically, at the time of the execution pricing in the agreement or separate other classical label, Decca, which would make it of the iHeart/Warner Agreement, hard for Pandora to steer away from the latter given valuation in the agreement in terms of the its share of the market. 5/17 Tr. 4706–07 (Shapiro) Warner’s actual share of iHeart’s custom spreadsheets . . . that I reviewed as (considering Naxos’s and Decca’s presence in noninteractive webcasts was background for the contract. . . . For that classical music market and acknowledging ‘‘there approximately [REDACTED]%. reason . . . the best answer, given the real- are issues with some specialized areas of music However, under the iHeart/Warner world data that we have, is to place a net where it might be harder to steer.’’) Further, Agreement, iHeart is obligated to value of zero on them because that’s what the Pandora did not conduct any steering experiments parties themselves did. with regard to steering away from Decca, as it did [REDACTED]. Drs. Fischel and with regard to steering away from the Majors. Third, Lichtman concluded that this provision 5/21/15 Tr. at 5336–40 (Fischel). Dr. Shapiro opined that, if steering did occur at the created an incentive for iHeart to Moreover, according to iHeart, even 30% level, Naxos would pay two different rates for plays on Pandora’s ad-supported and subscription increase Warner’s share of performances SoundExchange’s economic expert, Dr. services, respectively. Shapiro WRT, at 37–38. substantially [REDACTED]. Fischel/ However, the Pandora/Naxos Agreement does not Lichtman AWDT ¶ 36. 161 According to Drs. Lichtman and Fischel, bifurcate rates in this manner, but rather sets a The iHeart/Warner Agreement also under the AIP program, iHeart dedicates airtime to single per-play rate of $0.[REDACTED] that would promoting particular artists or songs, typically new apply to Pandora’s ad-supported and subscription contains the following additional artists or recently-released songs. These promotions services. PAN Ex. 5018. elements that, according to iHeart: (1) may include [REDACTED]. SX Ex. 33 at 19.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00061 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26376 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

Rubinfeld, admitted that none of the wouldn’t play them and Warner wouldn’t Lichtman AWDT ¶ 42; IHM Ex. 3034 at experts in this proceeding likewise receive any compensation for them. The 172 (‘‘Projected iHeartMedia-Warner ‘‘actually put[ ] a numerical value on royalty rate negotiated for this second part of Royalty Rates’’). the bundle, therefore, is a more appropriate Under scenario (1), without the these additional items.’’ 5/28/15 Tr. measure of what a willing buyer and a 6289 (Rubinfeld). In addition, iHeart willing seller would negotiate if steering of additional plays at lower notes, Dr. Rubinfeld acknowledged that unconstrained by government regulation. average rates, iHeart expected to pay several of these items were ‘‘terms that Warner licensed the rights to those Warner a total of $[REDACTED] in favor iHeart,’’ and yet were not performances to iHeart, and iHeart royalties. Under scenario (2), with the separately valued and priced by the compensated Warner for that license, at rates steering of additional plays at lower parties. Id. at 6435. that were acceptably profitable for both average rates, iHeart expected to pay However, iHeart does not conclude parties. The rate here was not determined by Warner a total of $[REDACTED]. regulation; it was determined by the give- Fischel/Lichtman AWDT ¶¶ 43, 51. from the foregoing that the iHeart/ and-take of a true negotiation. Warner Agreement sets forth a usable Dr. Fischel then divided the total benchmark rate that mirrors the stated Id. expected compensation under the rates of $0.[REDACTED] to Thus, Dr. Fischel needed to Today’s Growth Model ($[REDACTED]) $0.[REDACTED], or even the purported distinguish between the two bundles by the total number of performances lower rates of $0.[REDACTED] to that he had conceptualized, which projected in that model ([REDACTED]). $0.[REDACTED] resulting from the required him to consider the projected This calculation projected an average [REDACTED] adjustment applied by number of Warner plays in each bundle. per-play rate of $0.[REDACTED], Drs. Fischel and Lichtman (as discussed To perform this analysis, he relied upon rounded to $0.[REDACTED]. Fischel/ infra). Rather, according to Dr. Fischel, a set of projections that iHeart’s Board Lichtman AWDT ¶43; IHM Ex. 3034 at the foregoing rates reflect only the of Directors used when evaluating and 172 (‘‘Projected iHeart Media/Royalty average rates in or derived from the approving the iHeart/Warner Rates’’). iHeart/Warner Agreement. Dr. Fischel Agreement. Fischel/Lichtman AWDT Even before Dr. Fischel attempted to ¶ 40 (projections also served as basis for asserts that such an average rate ‘‘does determine his ‘‘incremental rate’’ under iHeart Board’s approval of stated rates not necessarily reflect the rate . . . that the iHeart/Warner Agreement, he in iHeart/Warner Agreement). a willing buyer and willing seller would emphasized that this average rate itself According to iHeart’s Head of Business have reached in a marketplace’’ was [REDACTED]% lower than the Development and Corporate Strategy, unconstrained by government regulation statutory rate of $0.0025 that iHeart Steven Cutler, this set of projections, or interference.’’ Fischel/Lichtman would otherwise pay under the referred to by iHeart as the ‘‘Today’s applicable NAB/SoundExchange AWDT ¶ 44. Growth’’ model, was [REDACTED], In an attempt to correct for this settlement. Fischel/Lichtman ¶ 43. representing the parties’ ‘‘best Additionally, Drs. Fischel and alleged defect, Dr. Fischel estimates’’ of performance under the Lichtman opined that this conceptualizes the Warner plays on iHeart/Warner Agreement. 6/2/15 Tr. $0.[REDACTED] rate needed to be iHeart as comprising two distinct 7247–48 (Cutler); see Fischel/Lichtman adjusted downward for a [REDACTED] economic bundles. Dr. Fischel states: AWDT ¶ 40; 5/21/15 Tr. 5365 (Fischel). adjustment, to reflect the fact that, As an economic matter, the [iHeart]- The Today’s Growth model projected under the iHeart/Warner Agreement, Warner agreement reflects a bundle of two that iHeart would play [REDACTED] [REDACTED] are not subject to a royalty distinct sets of rights. The first set provides total performances of all labels’ sound payment by iHeart to Warner. Id. ¶ 35. a license for iHeartMedia to play the same recordings over the [REDACTED] term number of Warner performances as it would They then noted that iHeart, had have played absent the agreement. The of the agreement. Fischel/Lichtman projected that an adjustment for second set of rights provides a license for AWDT ¶ 41 and Ex. A thereto [REDACTED] would reduce the effective iHeartMedia to play additional Warner (‘‘Projected Performances During Initial average per-play rate under the iHeart/ performances, above and beyond those it Term of iHeartMedia Agreement with Warner Agreement ‘‘to between would have played absent the agreement. Warner’’); IHM Ex. 3034 at 170. iHeart $0.[REDACTED] and $0.[REDACTED].’’ Id. ¶ 45. estimated Warner’s share of those Id. Accordingly, Dr. Fischel opines that performances under two key scenarios: Dr. Fischel then turned his analysis compensation for the first ‘‘bundle’’ of (1) The [REDACTED] scenario, which toward the calculation of his so-called rights is directly affected by the existing reflected iHeart’s expectations if no ‘‘incremental rate.’’ He noted the simple statutory rate, and therefore ‘‘provides agreement with Warner was reached; math demonstrating that, according to essentially no information about the rate and (2) the ‘‘Warner Direct License the Today’s Growth Model, the willing buyers and sellers would Terms’’ scenario, which reflected its difference in the number of Warner negotiate in the absence of government projections under the terms and plays on iHeart’s custom noninteractive regulation.’’ Id. ¶ 48. conditions of the Warner agreement as service between Scenario (2) However, Dr. Fischel opines that the signed. Fischel/Lichtman AWDT ¶ 42 ([REDACTED] plays) and Scenario (1) second ‘‘bundle’’ he conceptualizes is and Ex. B thereto (‘‘Projected ([REDACTED] plays) equaled ‘‘highly relevant to what willing buyers iHeartMedia/Warner Royalty Rates’’); [REDACTED] plays. He further noted and willing sellers would negotiate if IHM Ex. 3034 at 172. that the difference in royalties—again unconstrained by government Under scenario (1), iHeartMedia according to the Today’s Growth regulation.’’ Id. ¶ 49. In support of this expected Warner music to constitute Model—between Scenario (2) opinion, Dr. Fischel testified: [REDACTED]% of total performances, or ($[REDACTED]) and Scenario (1) [REDACTED] performances, on the ($[REDACTED]) equaled $[REDACTED]. This part of the bundle involves a license for iHeart custom service. Under scenario Fischel/Lichtman AWDT ¶¶ 50–51; IHM iHeart to play additional Warner performances, above and beyond those it (2), iHeart expected to increase Warner’s Ex. 3034 at 172 (‘‘projected iHeart would have played absent the agreement. share of performances to [REDACTED] Media/Warner royalty rates. Those additional performances are not percent, and thus expected to play Dr. Fischel then divided the directly influenced by the existing statutory [REDACTED] Warner performances over $[REDACTED] additional revenue by rate, because absent the agreement, iHeart the duration of the agreement. Fischel/ the additional [REDACTED] plays to

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00062 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26377

derive his ‘‘incremental rate’’ of [REDACTED] ‘‘incremental rate’’ by applying his ‘‘two $0.0005. Id. As noted supra, Dr. Fischel See generally IHM Exs. 3340, 3342, bundles’’ approach. He noted that iHeart opined that his so-called ‘‘incremental 3343, 3345, 3347, 3349, 3351–3370, expected to play an additional rate of $0.0005 was a better benchmark 3642. However, iHeart states that [REDACTED] performances and than the average rate of $0.[REDACTED] [REDACTED] of these 27 webcasters has expected to pay $[REDACTED] more in implied by the Today’s Growth Model paid royalties under the percentage of royalties. This incremental difference or the rates actually set forth in the revenue prong, because the per-play rate yielded the so-called ‘‘incremental rate’’ iHeart/Warner Agreement, because the has generated the higher royalty. of $0.[REDACTED] ($[REDACTED]/ so-called ‘‘incremental rate’’ was not Fischel/Lichtman AWDT ¶ 61. [REDACTED] plays). Fischel/Lichtman tainted by the upward influence of the Each of these 27 iHeart/Indies AWDT ¶ 68; IHM Ex. 3034 (Fischel/ statutory rate. Accordingly, Dr. Fischel Agreements contains a [REDACTED]- Lichtman AWDT, Ex. D thereto). opined, ‘‘this $0.0005 per-performance year term. Id. These iHeart/Indies Unlike the iHeart/Warner Agreement, these 27 Warner/Indies Agreements rate is the best available evidence on the Agreements also contain other rates that were not supported by an internal question at issue in this proceeding.’’ are not applicable to custom 162 projection of expected increased plays, Fischel/Lichtman AWDT ¶ 52. noninteractive webcasting. Id.; see As noted at the outset of this section, such as the ‘‘Today’s Growth’’ model Fischel/Lichtman AWDT ¶ 58. the iHeart/Warner Agreement contains a upon which Dr. Fischel relied for his As in the iHeart/Warner Agreement, greater-of rate structure. However, Drs. iHeart/Warner ‘‘incremental’’ analysis. the iHeart/Indies Agreements contain Fischel and Lichtman declined to Rather, Dr. Fischel testified that he and various additional items, some of which incorporate any greater-of formula into Dr. Lichtman ‘‘assumed (consistent with iHeart claims inure to its benefit, and their rate structure and they did not our understanding) that iHeart believed some of which benefit the labels. iHeart include any percentage-of-revenue that, after signing each of these deals, it points, by way of example, to the alternative rate in their proposed would increase each label’s share of all provision in all 27 agreements that benchmark. Dr. Lichtman explained this webcasts ([REDACTED]) by iHeart received a license for deviation from the iHeart/Warner [REDACTED] percent.’’ Fischel/ [REDACTED] and thereby avoided the Agreement: ‘‘[N]o one thought that Lichtman AWDT ¶ 66. Apparently, Dr. risk of [REDACTED] Additionally, in provision would be binding. So they Fischel did not use iHeart’s or his own many of those agreements, the Indies have a number that both parties looked ‘‘projections’’ of increased agreed [REDACTED]. Fischel/Lichtman at and said that number would never performances, as he did for his iHeart/ AWDT ¶ 62. actually be used in the real world, so Warner analysis, but rather ‘‘assume[d] As they analyzed the iHeart/Warner who cares what the number is . . ..’’ 5/ iHeart approximately met its projections Agreement, Drs. Fischel and Lichtman 15/15 Tr. 4016–17 (Lichtman); see also for . . . custom performances,’’ and concluded that the value of these terms 5/21/15 Tr. 5334 (Fischel) (same).163 therefore ‘‘the projections in [this] cannot be determined in isolation, and category[y] [are] equal to the actual b. The 27 iHeart/Indies Agreements found that there was no evidence number of performances.’’ Fischel/ indicating that the parties had explicitly iHeart also relies upon its separate Lichtman AWDT ¶ 66 (emphasis added). agreements with 27 Indies that, as of assigned value to them when analyzing Drs. Fischel and Lichtman concluded July 2014, accounted for approximately whether to enter into these 27 from the foregoing that the [REDACTED] percent of performances agreements. Accordingly, they $0.[REDACTED] ‘‘incremental rate’’ that on its custom service. Fischel/Lichtman concluded that it is appropriate to they estimated for the 27 iHeart/Indies AWDT ¶ 57 and Ex. C thereto; IHM Exs. assign a zero net value to the non- Agreements ‘‘demonstrates our main 3340, 3342, 3343, 3345, 3347, 3349, pecuniary terms. Id. conclusion, regarding the $0.0005 per- 3351–3370, 3642. Despite this relatively Therefore, Dr. Fischel proceeded to performance rate.’’ Fischel/Lichtman derive a so-called ‘‘incremental rate’’ for small percentage of plays (compared to ¶ 69.164 Warner), Drs. Fischel and Lichtman the 27 iHeart/Indies Agreements. He opine that ‘‘these 27 deals provide determined that, between 2012 and 3. SoundExchange’s Criticisms of the important additional evidence as to the 2014, and prior to the execution of these iHeart Rate Proposal rates negotiated by willing buyers and 27 agreements, iHeart expected to pay to a. Introduction willing sellers.’’ Fischel/Lichtman all these Indies $[REDACTED] (of which AWDT ¶ 57. $[REDACTED] was for custom webcasts) SoundExchange attacks the iHeart rate The principal custom noninteractive covering [REDACTED] performances (of proposal on six separate fronts. First, rate in these 27 agreements is which [REDACTED] were custom SoundExchange sets forth an overview [REDACTED]. Indeed, the 27 Warner/ webcasts), resulting in an average that purports to provide a different and Indies Agreements contain the following royalty rate of $0.[REDACTED] (iHeart more accurate understanding of the provision: was subject to the SoundExchange/NAB terms of the iHeart/Warner Agreement, settlement rates). IHM Ex. 3034 compared with the presentation put 162 Dr. Fischel then speculates as to whether even (Fischel/Lichtman AWDT, Ex. D). forth by iHeart. Second, SoundExchange the non-incremental plays would be priced higher Dr. Fischel then determined that, after or lower than $0.0005, but he comes to no 164 Drs. Fischel and Lichtman acknowledged the conclusion in that regard. Fischel/Lichtman AWDT the execution of these 27 iHeart/Indies obvious—that the $0.[REDACTED] ‘‘incremental’’ ¶ 53. Agreements, total performances would rate derived from the iHeart/Indies Agreements was 163 iHeart speculates that the percentage-of- increase to [REDACTED] (of which lower than the $0.[REDACTED] ‘‘incremental’’ rate revenue prong was added to the iHeart/Warner [REDACTED] were custom webcasts) derived from the iHeart/Warner Agreement. See Agreement by Warner to set a precedent for future 5/21/15 Tr. 5383 (Fischel). They opined that the rate-setting proceedings for sound recordings and and total royalties would increase to Indies might receive a lower rate because the Indies points to a document pertaining to Warner’s $[REDACTED] (of which $[REDACTED] artists may be ‘‘less well-known,’’ and because negotiations with [REDACTED] for support. See was for custom webcasts), resulting in Indies may have repertoires that are not ‘‘already IHM Ex. 3435 at 5; 5/15/15 Tr. 4024–25 (Lichtman). an average royalty rate of familiar to listeners.’’ Fischel/Lichtman AWDT However, iHeart does not identify any sufficiently ¶ 69. This testimony is generally consistent with the similar evidence that suggests the percentage-of- $0.[REDACTED]. Id. Judges’ finding, supra, with regard to the Pandora/ revenue prong in the iHeart/Warner Agreement was As with the iHeart/Warner analysis, Merlin Agreement, that Indies in fact receive lower included for this reason. Dr. Fischel then calculated his so-called royalty rates than the Majors.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00063 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26378 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

seeks to demonstrate the invalidity of at $[REDACTED], yielding a combined SoundExchange also notes in this Dr. Fischel’s ‘‘incremental rate’’ value for [REDACTED] such campaigns context, as it did in its opposition to approach. Third, SoundExchange avers of close to $[REDACTED] over the initial Pandora’s rate proposal, that the that iHeart’s analysis is also flawed term of the agreement. SX Ex. 32 at 14 steering elements of the iHeart/Warner because it fails properly to consider and n.9 (Wilcox WRT); 6/3/15 Tr. 7403 Agreement provide only ‘‘first mover’’ give value to other elements of (Wilcox). advantages’’ that would be consideration in the iHeart/Warner • iHeart also agreed to pay royalties ‘‘mathematically impossible’’ to Agreement, which would result in a to Warner for [REDACTED]. SX Ex. 33 replicate across the industry. 5/7/15 Tr. significantly higher benchmark per-play at 10 § 1(pp); SX Ex. 32 at 14 (Wilcox 2374 (Wilcox); Rubinfeld CWDT at 46 rate. Fourth, SoundExchange takes issue WRT). ¶ 183; 6/2/15 Tr. 7239 (Cutler). • with iHeart’s failure to account for the iHeart agreed to pay Warner a Moreover, SoundExchange noted that parties’ actual performance under the $[REDACTED] fee for a [REDACTED] iHeart found its ability to steer toward iHeart/Warner Agreement. Fifth, provision, the [REDACTED] agreement, any particular record company to be SoundExchange takes issue with which iHeart requested be in a separate limited. As noted in the Judges’ iHeart’s reliance on a single projection agreement but ultimately was included discussion of the Pandora rate proposal, made by iHeart during negotiations (the in the iHeart/Warner Agreement. 6/3/15 SoundExchange asserts that, when 166 ‘‘Today’s Growth’’ model) to establish a Tr. 7387 (Wilcox). iHeart tried to [REDACTED] it created benchmark in this proceeding, and its Through testimony at the hearing, ‘‘challenging listening experiences.’’ For failure to consider other SoundExchange and Warner asserted example, a listener’s seeded contemporaneous alternative that Warner perceived the additional ‘‘[REDACTED] Radio Station’’ projections. Sixth, SoundExchange items it received, combined with the [REDACTED] turned into a de facto seeks to discredit the 27 Warner/Indies rate and steering terms, as greater than ‘‘[REDACTED] Radio Station,’’ Agreements as proper benchmarks. what it would have received under the [REDACTED] and a listener’s seeded statutory license. 5/7/15 Tr. 2370 b. SoundExchange’s Overview of the ‘‘[REDACTED] Radio Station’’ (Wilcox) (Warner received ‘‘a package of [REDACTED] turned into a de facto iHeart/Warner Agreement consideration that is material and ‘‘[REDACTED] Radio Station greater and different in positive ways SoundExchange begins its critique by [REDACTED]. Thus, iHeart concluded than what we would be obtaining just referring to the negotiation period before that too much steering (to through a compulsory statutory deal.’’). the iHeart/Warner Agreement was [REDACTED]%) was ‘‘[REDACTED] all executed. It notes that iHeart originally Further, Mr. Wilcox testified that he did to the detriment of our custom offered Warner [REDACTED]. IHM Ex. not think this ‘‘deal’’ would ‘‘go forward product.’’ SX Ex. 1037. 3114 at 10. Warner rejected that on the existing terms if one of these proposal and according to Dr. Fischel, were missing.’’ 6/3/15 Tr. 7416 c. SoundExchange’s Criticism of the Warner ultimately achieved a ‘‘better (Wilcox). However, SoundExchange did ‘‘Incremental Rate’’ Approach of Drs. deal than [REDACTED]. 5/22/15 Tr. not proffer evidence or testimony that Fischel and Lichtman 165 5542, 5551 (Fischel). was contemporaneous with the SoundExchange begins its critique • When SoundExchange turns its negotiation of the iHeart/Warner with these undisputed assertions: attention to the several non-rate and Agreement that was probative as to non-steering aspects of the iHeart/ whether Warner required the other • None of these agreements—or any Warner Agreement, it notes the contract terms in order to avail itself of other agreement submitted by any other following provisions that were the rate and steering terms. party—has $0.[REDACTED] as the stated essentially ignored by iHeart. iHeart SoundExchange notes, however, per-performance rate or within any agreed to provide to Warner the greater (regarding the additional contract items range of stated rates. of [REDACTED]% of all AIP inventory of potential value to Warner) that iHeart • There is not a single document in that iHeart offers in the marketplace and did not produce a fact witness who evidence showing that any parties—not AIP having a ‘‘fair market value,’’ as testified regarding the actual value of just Warner and iHeart—ever had a stated in the iHeart/Warner Agreement, these terms to iHeart. ‘‘meeting of the minds’’ as to a rate of of at least $[REDACTED] per agreement SoundExchange also notes, as did $0.[REDACTED] per-performance. year. SX Ex.33 at 19–20 § 5(a). iHeart, that the latter also received • There is not a single • In addition to this ‘‘[REDACTED] additional contractual consideration communication between iHeart and AIP,’’ iHeart agreed to provide Warner beyond the right to perform Warner’s Warner citing a rate of $0.[REDACTED] with another advertising opportunity, to sound recordings under the agreement. under the iHeart-Warner agreement. participate in two ‘‘[REDACTED]’’ See Fischel/Lichtman AWDT at 20 • No internal iHeart document shows campaigns each year. This (‘‘insurance policy’’ allowing iHeart to such a rate for the iHeart-Warner ‘‘[REDACTED]’’ guarantees at least avoid [REDACTED] if [REDACTED] and agreement. [REDACTED] insertions of ads in [REDACTED] protection if [REDACTED] • There is no evidence in the record duration up to [REDACTED] seconds granted better terms to [REDACTED] for showing that a willing copyright owner each on iHeart’s terrestrial stations for [REDACTED] service); SX Ex. 33 at 31. would agree to license the performance artists selected at Warner’s discretion. However, despite the absence of any of its sound recordings at a rate of Each advertisement also must include a actual values being placed by the parties $0.[REDACTED]. [REDACTED]. SX Ex. 33 at 19–20 § 5(a); on these additional items, Mr. Wilcox • None of the other economic experts 81, Exhibit F. Warner calculated the concluded that the net value of all the who testified used such an approach in value of a single [REDACTED] campaign other consideration provisions is his written testimony. ‘‘heavily weighted to the Warner Music SX PFF ¶¶ 768–69 (citing 5/22/15 Tr. 165 SoundExchange also notes that Sony and Group.’’ 6/3/15 Tr. 7385 (Wilcox). 5489–90 (Fischel); Rubinfeld CWRT Universal turned down a similar offer from iHeart ¶ 23); Id. ¶¶ 784–88 (and additional because ‘‘[REDACTED].’’ SX Ex.1139; SX Ex. 25 at 166 In pertinent part, the [REDACTED] Agreement citations to the record therein). 12, ¶ 35 (Harrison WRT); 4/28/15 Tr. 509–510 (A. provided that, in exchange for a $[REDACTED] to Harrison) (describing iHeart’s proposal as Warner by iHeart, Warner granted to iHeart Next, SoundExchange takes ‘‘[REDACTED].’’) [REDACTED] SX EX. 1339. substantive aim at the ‘‘two bundles’’ of

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00064 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26379

rights approach. SoundExchange appeal of such a range of rates to ([REDACTED]). SX PFF ¶¶ 810–814, 827 (accurately) summarizes this opinion as Pandora, its own expert, Dr. Shapiro, (and citations to the record therein). stating that, according to Drs. Fischel did not adopt such an incremental rate, SoundExchange also takes issue with and Lichtman, the only relevant but rather recommended a rate that was iHeart’s claim, as asserted by Dr. information regarding the rate to which multiple times greater. Rubinfeld CWRT Fischel, that the absence of any willing buyers and willing sellers would at 22, ¶ 79. projections or spreadsheets detailing the agree, absent a statutory license, can be For these reasons, SoundExchange value of these additional items is found in the number of performances asserts that the so-called incremental evidence that the parties did not assign and revenue in the second bundle.167 As per-play approach of Drs. Fischel and values to them. However, SoundExchange continues to correctly Lichtman must be rejected, in favor of SoundExchange acknowledges that note, they then claim that dividing the an approach that determines per-play ‘‘when the Judges asked Mr. Wilcox so-called ‘‘incremental’’ revenue by the rates on an average royalty basis. whether Warner had assigned a number ‘‘incremental’’ number of performances value to . . . many of these provisions,’’ yields the precise per-play royalty rate d. The Alleged Importance of the Value his ‘‘consistent’’ response was that he to which the parties would have agreed of Non-Rate/Steering Items in the ‘‘could not be certain’’ of the number for 100% of the performances expected iHeart/Warner Agreement value. SX PFF ¶ 827. under their agreement in a world SoundExchange criticizes Drs. Fischel i. AIP and [REDACTED] without the statutory license. See SX and Lichtman for failing to make a Among the non-royalty and non- PFF ¶ 771 (and record citations therein). sufficient attempt to attach monetary steering elements within the iHeart/ The fundamental problem with this values to provisions in the iHeart/ Warner Agreement, SoundExchange ‘‘incremental’’ approach, according to Warner Agreement. See Fischel/ emphasizes iHeart’s failure to adjust its SoundExchange, is that it artificially Lichtman AWDT ¶ 39. More benchmark to reflect the value of two and erroneously divides the royalty particularly, SoundExchange rejects items referred to supra, AIP and payments by breaking the single actual their assumption that the non-royalty [REDACTED]. bundle of performances under the rate term provisions benefiting Warner, agreement into two hypothetical and those benefiting Heart, have a net (A) AIP bundles. According to SoundExchange, value of zero. See 5/21/15 Tr. 5/21/15 that approach artificially and SoundExchange notes that the iHeart/ Tr. 5340 (Fischel); (Fischel/Lichtman Warner Agreement itself states that AIP erroneously divides consideration into AWDT at 20–21). separate bundles that the parties did not has a ‘‘fair market value’’ of at least Rather, SoundExchange asserts the negotiate. To make the point, Dr. $[REDACTED] over [REDACTED] years. record reflects that this ‘‘net zero value’’ Rubinfeld, on behalf of SoundExchange, SX PFF ¶¶ 807–808 (and citations to the conclusion is inaccurate. The ‘‘record’’ applied an analogy: In a ‘‘buy one, get record therein). Thus, according to to which SoundExchange cites to one free’’ transaction, the price of the SoundExchange, it is irrelevant whether support this position is a conclusory second product is not zero; the second the parties had internal projections or statement made by Warner’s testifying product could not be obtained without spreadsheets establishing the value of executive, Mr. Wilcox, who stated that paying the full price for the first. AIP. See SX Ex 33 at 19, ¶ 5(a)(ii) the net value of the non-royalty rate Accordingly, the appropriate price for (declaring that AIP has a ‘‘fair market provisions is ‘‘heavily weighted to the each of the two products is not the value of at least [REDACTED] Dollars Warner Music Group.’’ 6/3/15 Tr. 7385 USD $[REDACTED] per Agreement ‘‘incremental price’’ of the second item, 168 but rather the average price of the two (Wilcox). SoundExchange further Year’’). items. Rubinfeld CWRT at 6, ¶ 24. seeks to buttress its argument that the Additionally, SoundExchange points SoundExchange also notes that Drs. iHeart benchmark fails to adjust for the to internal iHeart documents in which Fischel and Lichtman analyzed the value of items that favored Warner by Bob Pittman, iHeart’s C.E.O., asked of Pandora/Merlin Agreement through the reciting the list of such items and noting his employees, with regard to AIP, lens of their so-called incremental that Mr. Wilcox, in his oral and written [REDACTED]’’ SX Ex. 207. approach and concluded that the proper testimony, characterized such items as SoundExchange further notes that, in an rate derived from that agreement—for ‘‘incredibly important’’ ([REDACTED]); attempt to bridge differences in the use as the statutory benchmark—is ‘‘so important’’ ([REDACTED]); a ‘‘floor ongoing negotiations, Mr. Pittman between $0.0002 and negative $0.0002 valuation’’ ([REDACTED]); an suggested that iHeart asked Warner if (i.e., a rate at which the record ‘‘immediate uptick’’ in value AIP has value to Warner, because it has companies would pay the value to iHeart. SX Ex. 1372. noninteractive services rather than 168 Actually, Mr. Wilcox made this statement with Additionally, SoundExchange points to regard to a list of contractual items that would Mr. Wilcox’s written and oral testimony, receive royalties from these services). provide value only to Warner, not the entirety of See Fischel/Lichtman AWDT at 40–41. other non-royalty/steering items that Drs. Fischel in which he claims to recall that In attempting to highlight the purported and Lichtman asserted had value to both parties [REDACTED] indicated that iHeart absurdity of this result, SoundExchange and should be weighed and deemed for rate intended to [REDACTED], but he cannot purposes to have a net value of zero. See id. at identify a document confirming that notes that, despite the clear economic 7384–85 (Mr. Wilcox responding to a question regarding a demonstrative list of contractual items alleged representation by [REDACTED]. 167 SoundExchange also accurately summarizes and testifying that ‘‘they’re heavily weighted to the Wilcox WRT ¶ 23, 6/3/15 Tr. 7460–61 the contents of the two bundles: ‘‘The first is a Warner Music Group. These were, every one of (Wilcox) ‘bundle’ for the purported right to perform sound them, things that were important wins for us, if you SoundExchange also points to recordings up to the number of performances [Drs.] will, in the negotiation and were key to getting to Fischel [and]Lichtman say the parties expected to yes.’’). Drs. Fischel and Lichtman did not dispute numerous documents in which iHeart occur under the statutory license in the absence of that some contractual items had value to Warner, confirms the substantial value to record a direct license,’’ and ‘‘[t]he second is a ‘bundle’ for but rather concluded that the absence of valuations companies of AIP participation. See, the purported right to make all the additional by the parties required an expert to net the e.g., IHM Exs. 3114 at 5, 10; 3121 at 4; performances over and above those in the first offsetting values at zero. Thus, the cited testimony bundle that [Drs.] Fischel [and]Lichtman say the does not support SoundExchange’s assertion in the 3225 at 2. Further, during negotiations, parties expected to occur because of the direct text, supra, that ‘‘the record’’ reflects a net value for iHeart emphasized to Warner that AIP license.’’ SX PFF ¶ 770. these other items tilted toward Warner. had substantial stand-alone value. See

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00065 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26380 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

SX Ex. 93 at 1. Additionally, at the non-expert, Mr. Wilcox, the Warner value they claim was missing from the hearing, witnesses for both iHeart and executive, calculated his understanding valuation undertaken by Drs. Fischel Warner acknowledged the significant of the value of a [REDACTED] campaign and Lichtman, they conclude that under value of AIP to a record company. 5/21/ at $[REDACTED] per year, or iHeart’s ‘‘Today’s Growth’’ model, the 15 Tr. 5194–95 (Poleman) (iHeart approximately $[REDACTED] for the benchmark per-play rate would equal or executive describing AIP as [REDACTED] campaigns to which exceed $0.[REDACTED]. See SX PFF ‘‘invaluable’’); 6/3/15 Tr. 7392 (Wilcox); Warner was entitled over the initial ¶¶ 846–853 (and record citations Wilcox WDT at 12–13; (Warner term of the agreement. Wilcox WRT at therein). executive describing AIP as 14 n.9; 6/3/15 Tr. 7403 (Wilcox). e. Performance Under the iHeart/Warner ‘‘[REDACTED]’’). SoundExchange notes that no iHeart fact Agreement Has Not Matched the Based on such reasoning, iHeart witness disputed this attempted Projections in iHeart’s ‘‘Today’s estimated the quantity of AIP to be valuation. given to Warner not only [REDACTED], For these reasons, SoundExchange Growth’’ Model but also [REDACTED], as set forth on disputes the decision by Drs. Fischel In this proceeding, SoundExchange iHeart’s rate card.’’ See 5/20/15 Tr. and Lichtman to assign no independent did not rely in its direct case upon any 4885–86 (Pittman). As SoundExchange value to the [REDACTED] benefits of Warner’s projections reflecting its further points out, Mr. Poleman also contained in the iHeart/Warner expectations at the time the iHeart/ noted that access to AIP slots could in Agreement. Warner Agreement was negotiated and the future be [REDACTED], and, if so, executed. Rather, SoundExchange relies ii. [REDACTED] Agreement Warner would [REDACTED]. 5/21/15 upon an analysis by Dr. Rubinfeld of Tr. 5189–90 (Poleman). See also SX Ex. Another non-royalty/steering available data regarding performances 1139 ([REDACTED]. provision identified in the iHeart/ and royalties paid during the first eight For these reasons, SoundExchange Warner Agreement is a reference to a months of the iHeart-Warner avers that iHeart erred in declining to separate agreement—the ‘‘[REDACTED] agreement—from October 2013 to May attribute value to AIP in its iHeart/ Agreement’’ between the parties. 2014. Dr. Rubinfeld relied upon this Warner benchmark.169 SoundExchange avers that Drs. Fischel slice of performance data, rather than and Lichtman wrongly omitted the the expectations of the contracting (B) [REDACTED] value of this $[REDACTED] payment parties, because he found that According to SoundExchange, the from their calculation. According to ‘‘performance data reflect actual value of [REDACTED] is different from SoundExchange, this omission was experiences in the marketplace [and] [REDACTED] AIP in a way that improper because Mr. Wilcox testified [t]he most recent performance data is enhances record company promotional that ‘‘it was ‘‘worth . . . $[REDACTED]’’ likely to be the best predictor of what programs on iHeart. First, unlike AIP, 6/3/15 Tr. 7385 (Wilcox). Mr. Wilcox will happen in the immediate future.’’ Warner was not [REDACTED], and further testified that iHeart had Rubinfeld CWRT ¶ 27. However, Dr. iHeart did not [REDACTED]. 6/3/15 Tr. requested that this ‘‘[REDACTED] Rubinfeld also cautioned that ‘‘review of 7405 (Wilcox). transaction be set forth in a separate a longer period of performance data may The iHeart/Warner Agreement’s agreement, but Warner preferred that it offer additional value if the review [REDACTED] provision guarantees be included—as it ultimately was—in reveals important trends in the Warner at least [REDACTED] of up to the iHeart/Warner Agreement. 6/3/15 industry.’’ Id. SoundExchange also [REDACTED] for [REDACTED] on all of Tr. 7387 (Wilcox). SoundExchange also points out that Dr. Katz (the NAB’s iHeart’s [REDACTED] of [REDACTED] notes that iHeart does not dispute that economic expert), Mr. Cutler (an iHeart chosen by Warner. SX Ex. 33 at 19–20 the $[REDACTED] was executed on the executive), and Aaron Harrison (a § 5(a); id. at 81, Exhibit F, §§ 1–2. same day. 6/2/15 Tr. 7304 (Cutler); 5/ Universal executive) all recognized the According to Warner, both the 22/15 Tr. 5505 (Fischel). Further, importance of using current [REDACTED] and the fact that SoundExchange points out that none of performance data to update prior [REDACTED] are unique to this iHeart’s fact witnesses testified that the projections or expectations. See SX PFF program, [REDACTED]. 6/3/15 Tr. 7401 $[REDACTED] was not consideration ¶¶ 800, 803–04 (and citations to the (Wilcox). Further, the [REDACTED] tied closely to the webcasting record contained therein). provisions require iHeart to include a agreement. From the 8-month slice of data that he [REDACTED] and give Warner the right SoundExchange acknowledges that reviewed and about which he opined, to [REDACTED], and to [REDACTED]. the ‘‘[REDACTED] Agreement’’ contains Dr. Rubinfeld calculated an alternative SX Ex. 33 at 82, Exhibit F, § 7. an [REDACTED]. See SX Ex. 1339 at 1– average per-play royalty rate. Rubinfeld Warner did not attempt to value 2. However, SoundExchange argues that CWDT at 57–59, ¶¶ 229–236); SX Ex. 64 [REDACTED] contemporaneous with the iHeart is inconsistent by claiming that (Rubinfeld App. 1b, backup negotiations, and did not include a the Judges should apply that express calculations).170 For custom stated value for [REDACTED] in the clause, yet they should ignore the noninteractive performances, Dr. iHeart/Warner Agreement. express valuation of AIP at Rubinfeld calculated a per-play rate of SoundExchange did not utilize an $[REDACTED] in the iHeart/Warner $0.[REDACTED] ($0.[REDACTED] expert to value [REDACTED] in the Agreement. See SX PFF ¶ 830. rounded). When he attributed the value hearing. However, for this proceeding, a Additionally, SoundExchange avers that of AIP to the per-play rate, his eight- Warner would not have executed the month performance-based rate rose to 169 SoundExchange, noting one of iHeart’s webcasting agreement (all else equal) rebuttals on this issue, acknowledges that in the $0.[REDACTED] per play past, iHeart provided AIP [REDACTED]. Therefore, absent the $[REDACTED] payment. 6/3/ ($0.[REDACTED] rounded). SX Ex. 66. SoundExchange recognized that AIP provisions 15 Tr. 7388 (Wilcox) (‘‘It was a material Dr. Rubinfeld then attempted to could be construed as a form of ‘‘insurance’’ against amount of money and important to us equalize the iHeart/Warner and derived [REDACTED]. SoundExchange asserts that the as part of the total list of consideration threat that iHeart would [REDACTED] AIP was real, potential statutory rate to equalize so any ‘‘insurance’’ value would be quite high, we were getting . . .’’). albeit indeterminate. See SoundExchange PFF ¶ 823 In sum, when Dr. Rubinfeld and 170 Dr. Rubinfeld also updated his calculations to (and citations to the record therein). SoundExchange account for all of the include June to September 2014). SX Ex. 133.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00066 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26381

royalty-bearing performances by Warner Agreement, one of the models g. The Alleged Deficiencies in the 27 adjusting for skips and for the playing that was [REDACTED]—the iHeart/Indies Agreements and in The of [REDACTED]. To that end, he used ‘‘[REDACTED]’’ Model—proved to be a Analysis of Their Terms by iHeart’s the same adjustment factor, 1.1, as he more accurate estimate of [REDACTED]. Experts had used when performing his own See 5/22/15 Tr. 5494 (Fischel); 6/2/15 SoundExchange raises several interactive benchmarking analysis. Tr. 7264–65 (Cutler). This consistency challenges to iHeart’s attempt to use the Rubinfeld CWDT at 58 ¶ 234; SX Ex. 66. between the ‘‘[REDACTED]’’ model and 27 iHeart/Indies Agreements as SoundExchange avers that Dr. initial actual performance existed, benchmarks in this proceeding. First, Rubinfeld’s calculations as they relate to according to SoundExchange, because SoundExchange avers that the status of custom webcasting are conservative for iHeart had [REDACTED]. 5/22/15 Tr. these licensees as Indies renders them the following reasons: 5522 (Fischel); 5/20/15 Tr. 4839–40 • unrepresentative of the rates and terms He makes no adjustment upward for (Pittman) ([REDACTED]). that a noninteractive webcaster would the certainty of value that Warner SoundExchange surmises that such negotiate with a major recorded music receives as a result of getting [REDACTED] policies were put into company. SoundExchange notes that [REDACTED]. Rubinfeld CWDT at 57, effect, and thus contributed to the actual even Dr. Fischel acknowledged, ¶ 229. initial performance under the iHeart/ ‘‘Warner got a [[REDACTED]%] better • He does not account for any Warner Agreement that resembled the 171 deal than the Indies’’ from iHeart. 5/22/ additional value from [REDACTED]. ‘‘[REDACTED]’’ model rather than the 15 Tr. 5542 (May 22, 2015) (Fischel). f. iHeart Relies on Projections From ‘‘Today’s Growth’’ model. Whatever the Second, SoundExchange notes that Only One Model—the ‘‘Today’s reason, as Mr. Cutler of iHeart the greater-of rate structure in the Growth’’ Model acknowledged, iHeart’s growth in iHeart/Indies agreements for custom Warner plays over the initial contract noninteractive webcasting are SoundExchange avers that Drs. period has been [REDACTED]. 6/2/15 [REDACTED], and thus are unduly Fischel and Lichtman relied exclusively Tr. 7264–65 (Cutler). influenced by that statutory rate. See, on one specific projection that applied SoundExchange notes as well that Dr. e.g., IHM 3340, Tab 7/Ex. F (agreement certain ‘‘assumptions’’ regarding future Fischel admitted on cross-examination between Indie DashGo and iHeart at 4, performance under the iHeart/Warner that he had performed an analysis of the 8) Third, SoundExchange avers that Agreement. These expectations were effective incremental rates under the these Indies comprise in total no more contained in the ‘‘Today’s Growth’’ ‘‘[REDACTED]’’ model (but did not than [REDACTED]% of plays on the model presented to iHeart’s Board of submit evidence of that calculation or service in July 2014, and most account Directors in mid-2013. Fischel/ testify as to that calculation). On cross- for less than [REDACTED]% of plays Lichtman AWDT at 21 ¶ 40. examination, Dr. Fischel further 175 Although Drs. Fischel and Lichtman See SX PFF 863. acknowledged that the incremental rate SoundExchange notes that Drs. state that they chose the ‘‘Today’s he had calculated equaled Fischel and Lichtman determined both Growth’’ model because the iHeart $0.[REDACTED] per play under the average and incremental rates related to Board purportedly ‘‘relied on [it] as the ‘‘[REDACTED]’’ model. 5/22/15 Tr. 5523 these 27 iHeart/Indies Agreements. most realistic [case]’’ when approving (Fischel).173 iHeart calculated an average royalty rate the iHeart-Warner Agreement, 5/21/15 SoundExchange additionally points to of $0.[REDACTED] from these 27 Tr. 5322 (Fischel), SoundExchange an effective per-play rate that iHeart agreements, and an incremental rate of notes that iHeart actually [REDACTED]. supposedly wrongly ignored—the rate $0.[REDACTED] from these 27 IHM Ex. 3338 (Cutler WDT); see also 6/ derived from a model [REDACTED]. See 172 agreements. Fischel/Lichtman AWDT, 2/15 Tr.7263–64 (Cutler). SX Ex. 367 at 005; 6/3/15 Tr. 7552–53 Although there is no evidence that the Ex. D. (Wilcox); see also SX Ex. 92 at 15 However, with regard to the iHeart Board relied on the (alternative model comparisons). incremental rate, SoundExchange notes ‘‘[REDACTED]’’ or ‘‘[REDACTED]’’ Applying this model, according to that Drs. Fischel and Lichtman did not models, SoundExchange avers (albeit SoundExchange, yielded an average possess the same contemporaneous without supporting evidence) that performance rate above projections from iHeart (or the Indies) as because iHeart executives [REDACTED], $0.[REDACTED], and an incremental ‘‘it was wrong for Drs. Fischel and rate of approximately $0.[REDACTED]. assumptions we used throughout the course of Lichtman to ignore them completely.’’ Once again, these rates were negotiating this deal to stress-test the, you know, SX PFF ¶ 779. SoundExchange further mathematically derived by edge cases, you know, trying to figure out that this deal would perform positively for us in as many notes that, although Mr. Cutler testified SoundExchange, not its witnesses, that he viewed the Today’s Growth situations as we can throw at it. That’s, sort of, the based on ‘‘the simple math that Prof. point.’’ 6/3/15 Tr. 7421 (Wilcox). Thus, it is unclear model as the best estimate, neither he Fischel described’’ as applicable to as to exactly what model or models were nor any other iHeart witness testified calculating these rates. See SX PFF [REDACTED]. Moreover, Mr. Wilcox did not identify in his written testimony which model or that [REDACTED]. Id. Consequently, ¶ 794.174 SoundExchange asserts that the Fischel/ models were [REDACTED]. The Judges find Mr. Wilcox’s oral testimony on this subject to be neither Lichtman analysis is compromised 173 Although Dr. Fischel did not identify the credible nor informative. because they failed to test [REDACTED]. average rate derived from the ‘‘[REDACTED]’’ 175 SoundExchange does not provide a citation to See 5/22/15 Tr. 5496–97 (Fischel). model, the basic math derived from iHeart’s the record for these statistics, referring only to SoundExchange noted when it looked ‘‘[REDACTED]’’ model projections reveal an average ‘‘iHeart’s data.’’ SX PFF ¶ 863. By contrast, Drs. at actual performance under the iHeart/ royalty rate of $0.[REDACTED]. for the entirety of Fischel and Lichtman stated in their written performances under the iHeart/Warner Agreement testimony that ‘‘[a]s of July 2014, these 27 labels if the ‘‘[REDACTED]’’ model had been applied. SX accounted for approximately [REDACTED]% of 171 Dr. Rubinfeld claims his estimate is also Ex 207; See SX PFF ¶ 793. webcast performances on iHeart,’’ but it was conservative because he applies the conservative 174 Although Mr. Wilcox testified that this model unclear from their testimony whether that pre-deal market share of [REDACTED]% despite a indicating higher rates was [REDACTED], he did percentage combined custom and simulcast claim by Warner that its actual market share on not clearly identify a model upon which performances. See Fischel/Lichtman AWDT ¶ 57 & iHeartRadio was approximately [REDACTED]%. [REDACTED]. Indeed, Mr. Wilcox testified that that n.51. Thus, the record is unclear what percentage Rubinfeld CWDT at 59 n. 135. the model that he identified as having been of plays on iHeart’s custom noninteractive service 172 [REDACTED]. Cutler WDT, Ex. DD. [REDACTED] ‘‘was just one of many sets of is comprised of these 27 Indies’ recordings.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00067 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26382 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

they had relied upon to determine the iHeart expected to be played under the a common example, tire sellers will incremental rate under the iHeart/ higher existing statutory rate. The often advertise a special offer: A buyer Warner Agreement. 5/22/15 Tr. 5543 second conceptual value-bundle can pay for three tires and get the fourth (Fischel). Accordingly, the presumption (Scenario 2) consists of the number of tire free. This is economically (and by Drs. Fischel and Lichtman that iHeart performances (with steering, from mathematically) equivalent to a 25% would increase performances by [REDACTED]% to [REDACTED]% reduction in the price of four tires. No [REDACTED]% is not based on any market share) iHeart expected to be one could go to the automotive store iHeart projection, nor is it supported by played under the lower direct deal rate. and receive only the ‘‘free’’ fourth tire! any provision of the 27 contracts. 5/22/ Drs. Fischel and Lichtman then consider iHeart attempts to distinguish the ice 15 Tr. 5544 (Fischel). Moreover, the the expected difference between the cream cone example by noting that, in starting point, pre-agreement higher revenues arising from the direct the present case, Drs. Fischel and performance numbers were based upon deal. Finally, they divide the Lichtman are not eliminating a market- iHeart’s actual performances of Indie incremental revenue by the number of based price for the pre-incremental recordings. Id. at 5545.176 From this incremental plays to determine their bundle, but rather are eliminating a number, Drs. Fischel and Lichtman ‘‘incremental rate.’’ government-set rate that casts a extrapolated an ‘‘expectations’’-based This methodology intentionally ‘‘shadow’’ on the market. There are [REDACTED]% increase in the number attributes no market value to the rate several errors in this reasoning. First, of post-execution performances. Id. and revenue paid for the pre- the statutory rates were set after market Finally, SoundExchange notes the incremental performances. Although, as participants provided the Judges in the testimony of one Indie representative, noted above, Drs. Fischel and Lichtman prior proceeding with market evidence. Mr. Barros of Concord, who stated that engage in this process in order to There is no a priori reason to conclude Concord would not have entered into remove the alleged impact of the that the rates set in that earlier this agreement with iHeart to reduce ‘‘shadow’’ of the statutory rate, they proceeding failed to reflect or custom noninteractive webcasting rates merely replace one supposed problem approximate market forces, and iHeart to [REDACTED] if the agreement did not with a very real and more serious does not provide evidence as to why the also include the [REDACTED] and problem. That is, they replace the Judges should re-litigate prior rates and compensation for performances of statutory rate with an effective rate of reach such a conclusion.179 Second, to [REDACTED]. 5/28/15 Tr. 6506 zero for the pre-incremental use a zero rate in order to remove the (Barros).177 According to performances. There was no evidence alleged shadow of the Judges’ statutory SoundExchange, Drs. Fischel and presented in this proceeding, indeed no rate or a settlement rate would be, to put Lichtman erred by failing to adjust their logical evidence could be presented, to the matter colloquially, ‘‘throwing out proposed rates to account for this support an assertion that the bulk of the the baby with the bathwater.’’ A additional consideration. pre-incremental performances under functionally zero rate for the pre- iHeart’s ‘‘two bundle’’ concept would be incremental performances is no mere 4. The Judges’ Analyses and Findings priced at zero in an actual market. To Regarding iHeart’s Rate Proposal potential ‘‘shadow;’’ it is an ink blot that state the obvious, the creation of sound obliterates any economic value inherent a. The Judges Reject iHeart’s recordings is not costless, and prices are in the majority of the performances for ‘‘Incremental’’ Rate Analysis positive because costs must be 180 178 which the rates must be established. recovered. Accordingly, the Judges reject iHeart’s The Judges agree with Relatedly, although iHeart would like SoundExchange’s critique that the incremental approach and they reject the Judges to focus only on the the $0.0005 rate its experts derived by ‘‘incremental approach’’ advanced by incremental number of performances using the incremental approach. To be iHeart is an inappropriate method for and the incremental revenue, those clear, that incremental $0.0005 determining rates under § 114. There are incremental values cannot exist without proposed rate does not constitute a a number of reasons why the iHeart first paying for the pre- benchmark or a guidepost which the ‘‘incremental approach’’ is improper. incremental performances at pre- First, the basic premise of the Judges have relied for any purpose, and incremental rates. To put the point that incremental rate and the analysis approach is erroneous. In an effort to colloquially, ‘‘you cannot get there from from which it was derived has not avoid the so-called ‘‘shadow’’ of the here.’’ That tautological point is not influenced the Judges in their statutory rate, Drs. Fischel and avoided by arbitrarily attributing a zero determination of the statutory rate in Lichtman essentially substitute a rate of value to the pre-incremental this proceeding.181 zero for the number of sound recordings performances. played under the existing statutory rate. SoundExchange makes this point well 179 Then, they conceptually divide the by analogizing to a ‘‘buy one, get one Similarly, iHeart has not proffered evidence expected total of performances under sufficient to show why the rates set in settlements free’’ offer. If a vendor offered an ice between parties, that both parties agree may be the direct license (the iHeart/Warner cream cone (to adopt SoundExchange’s evidence of a market rate, fail to reflect, or at least Agreement) into two value-bundles. The demonstrative example at the hearing) approximate, market rates as of the time they were first conceptual value-bundle (Scenario for $1.00, but offered two ice cream set. 1) consists of the lower number of 180 On a less colloquial and more economic basis, cones for $1.06, it would be absurd to iHeart has confused an elasticity-type concept with performances (without steering) that conclude that the true market price of price. iHeart calculates the change in total revenue an ice cream cone is the incremental six divided by the change in quantity. Such a 176 SoundExchange also points out that Drs. cents. Rather, this offer indicates a proportionate change is not equivalent to a unit Fischel and Lichtman only had performance data market price of $0.53, the average price price. for [REDACTED] of the 27 Indies, so they 181 iHeart attempts to support its ‘‘incremental’’ extrapolated the data that they had. Id. at 5548; see for the two ice cream cones. Or, to take analysis with three arguments that it claims are also SX Ex. 2347. confirmatory of the $0.0005 rate. See iHeart PFF 177 As noted in the Judges’ analysis of the 178 It is also unsupported by the evidence that ¶¶ 236–260 (and citations to the record therein). Pandora/Merlin Agreement, Mr. Barros did not record companies would forego all royalties in the The Judges note that their rejection of this indicate that Concord, or anyone on its behalf, hypothetical market merely to obtain a promotional ‘‘incremental’’ analysis moots the relevance of any established a monetary value for these other value from the playing of their recordings on a attempt to confirm its purported contextual contractual items. noninteractive service. reasonableness. Further, the fact that iHeart did not

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00068 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26383

b. The Judges Find the Average per-Play Willing buyer and seller test: The rates license test.’’ The Judges disagree. As Rate Indicated by the iHeart/Warner are intended to be those that would with regard to the Pandora/Merlin Agreement to be a Useful Benchmark have been negotiated in a hypothetical Agreement, it is crucial to appreciate 184 Unlike the incremental rate derived marketplace between a willing buyer that the adjusted effective rate in the by iHeart’s experts, the ‘‘average rate,’’ and a willing seller. direct license is less than the default i.e., the stated per-play rate contained in There is no dispute that Warner was rate that would otherwise control (the the iHeart/Warner Agreement is a useful a willing seller in connection with the SoundExchange/NAB settlement rates benchmark that, after adjustment, is iHeart/Warner Agreement. As one of the for iHeart, and the Pureplay rates for probative of the rate that would be paid three Majors, Warner is a sophisticated Pandora). Accordingly, Warner was by a Major, as a willing seller/licensor, entity capable of negotiating direct under no compulsion to accept the to a noninteractive service, as a willing agreements in a manner that it lower rate (compared to the buyer/licensee.182 understands will advance its economic SoundExchange/NAB settlement rate) interests. Likewise, iHeart is a leading set forth in the iHeart/Warner i. The Benchmark Passes the ‘‘Four-Part noninteractive webcaster—not to Agreement; it could have rejected that Test’’ Derived From the Judges’ Prior mention one of the largest transmitters rate and defaulted to the higher Decisions of music across various platforms. SoundExchange/NAB settlement rate. The iHeart/Warner Agreement iHeart thus without dispute is also Instead, Warner agreed to the lower rate, satisfies the sub-tests implicit in the clearly capable of representing its in exchange for the anticipated steering Judges’ prior determinations, as economic interests in negotiating direct by iHeart of additional webcast outlined by Dr. Rubinfeld: agreements. performances of Warner sound In the present case, the record is recordings (from approximately propose these approaches as benchmarks or as other replete with voluminous submissions [REDACTED]% to [REDACTED]% of independent bases to set the rates makes them and substantial testimony indicating the total sound recordings). Accordingly, unhelpful and inappropriate as evidence to support diligence of both iHeart and Warner in iHeart’s rate proposal. However, in the interest of the Judges find that the ‘‘statutory completeness, the Judges note the following with negotiating this direct agreement. license test’’ has also been satisfied by regard to those arguments. First, Drs. Fischel and Clearly, each party was a willing the iHeart/Warner Agreement. Lichtman undertook what they called a ‘‘thought participant in the legal sense; that is, Further, and as discussed in experiment,’’ whereby they attempted to estimate a each party was under no compulsion to rate necessary for sound recording copyright connection with the Pandora/Merlin holders to maintain revenue at current levels if enter into the iHeart/Warner Agreement, Agreement, the steering aspects of the 100% of all listening to recorded music migrated to and each party had the opportunity to iHeart/Warner Agreement also satisfy a noninteractive webcasting. (They concluded that avail itself fully of all facts that it the rate would be $0.[REDACTED] per play.) They statutory ‘‘test’’ omitted from Dr. also did the same analysis on the assumption that deemed pertinent before executing that Rubinfeld’s four-part approach: The only 25% migrated to noninteractive services. agreement. See, e.g., Amerada Hess ‘‘effective competition’’ test. The (They concluded that the rate would be Corp. v. Comm’r, 517 F.2d 75, 83 (3d steering aspect of the iHeart/Warner $0.[REDACTED] per play.) However, Drs. Fischel Cir. 1975) (defining a ‘‘willing buyer’’ and Lichtman acknowledge that this ‘‘thought Agreement reflects price competition— experiment’’ is ‘‘not evidence of what a willing and a ‘‘willing seller’’ as parties not an increase in quantity (more buyer and willing seller would negotiate.’’ Fischel/ ‘‘being under any compulsion to buy or performances) in exchange for a lower Lichtman AWDT ¶ 128 (emphasis added). to sell and both having reasonable price (a lower rate). All of the reasons Therefore, such speculation is irrelevant to the knowledge of relevant facts.’ ’’). Judges. Second, Drs. Fischel and Lichtman set forth in this determination in the performed an ‘‘Economic Value Added (‘‘EVA’’) Same parties test: The buyers in this analysis of the Pandora/Merlin analysis of the costs, revenues and necessary ROI hypothetical marketplace are the Agreement regarding the pro- of a ‘‘hypothetical simulcaster’’ to determine the statutory webcasting services and the competitive aspects of such steering, rate necessary for it to remain in business in the sellers are record companies. long-run, which they determined to be between including the dynamic effect of a threat In the iHeart/Warner Agreement, the $0.[REDACTED] and $0.[REDACTED] per play. of steering, apply with equal force to the However, as the Judges have repeatedly held, rate buyer/licensee, iHeart, is a statutory iHeart/Warner Agreement.185 proceedings under section 114 are not public utility webcasting service. The seller/licensor, Same rights test: The products sold style proceedings whereby parties are guaranteed a Warner, is a record company. Clearly, rate of return. See, e.g., Web III Remand, 79 FR at consist of a blanket license for digital this aspect of the benchmark test is 23107. Further, their EVA model was based on a transmission of the record companies’ sample of terrestrial radio firms that is not satisfied. necessarily representative of simulcasters. Statutory license test: The complete repertoire of sound recordings, Additionally, their EVA analysis fails to consider hypothetical marketplace is one in in compliance with the DMCA the rates necessary for record companies to obtain requirements. a sufficient rate of return, so they have simply which there is no statutory license. focused on the demand side of the market and The iHeart/Warner Agreement is a 184 ignored the ‘‘willing sellers’’ on the supply side. direct agreement between the parties. The Judges’ determination of the adjusted Third, Drs. Fischel and Lichtman compare the effective rate under the iHeart/Warner Agreement is statutory rate for satellite digital audio radio The rates established in this agreement discussed infra. services (SDARS) and find that it suggests a per- are not statutory rates. More 185 iHeart notes that the threat of steering could play rate of $0.[REDACTED] to $0.[REDACTED]. particularly, at the time the iHeart/ cause steering to occur in a number of differentiated However, rates set by the Judges in other types of Warner Agreement was executed, iHeart ways, e.g., with one service making steering deals proceedings are not probative of rates that should with several licensors, several licensees making be set in this proceeding, especially when the was obligated to pay royalties to Warner similar deals with the same licensor(s), or a licensee standards in the two proceedings are different. The according to the schedule of rates set making different deals with different licensors over rate standard in SDARS proceedings is different forth in the SoundExchange/NAB time. See iHeart RPFF at 6 n.15. However, the from the standard in section 114(f)(2)(B) for settlement.183 Judges need not rely on such specific predictions. noninteractive services. See 17 U.S.C. In whatever ways in which the reality of steering § 801(b)(1)(A)–(D) (setting forth particular objectives SoundExchange asserts that, and the concomitant threat of steering-induced that the rates must achieve). nonetheless, the rates in the iHeart/ price competition develop, it is clear to the Judges 182 In discussing the reasons why this average rate Warner Agreement are too heavily that, as Dr. Shapiro explained, steering is the is a useful benchmark, the Judges find it helpful to influenced by the ‘‘shadow’’ of the mechanism by which the complementary oligopoly organize their finding by adopting Dr. Rubinfeld’s power of the Majors is offset, allowing the Majors characterization of the elements of the statutory test statutory rates to satisfy this ‘‘statutory to realize only their considerable (non- implicitly set forth in section 114. See Rubinfeld complementary) oligopolistic power generated by CWDT ¶ 122(a)–(d). 183 See note 28, supra. their repertoires and their organizational acumen.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00069 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26384 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

It is not disputed that the iHeart/ ‘‘two-bundle’’ theory, the parties’ a per-play basis. More particularly, the Warner Agreement provides in pertinent expectations and projections are baked Judges would expect that the record part for a license from Warner to iHeart into the single explicit annual rate companies’ internal valuations and to play Warner sound recordings on contained in the iHeart/Warner spreadsheets would set forth their iHeart’s noninteractive webcasting Agreement. Regardless of whether understanding of these monetary values service. See SX Ex. 33 at 8 ¶ 1(y) actual performance eventually (not merely the existence of some (defining ‘‘[REDACTED]’’); id. at 11, resembles the ‘‘Today’s Growth Model’’ unquantified value). Similarly, the ¶ 2(a)(1) (granting right to play relied upon by the iHeart Board, or Judges would anticipate receiving ‘‘[REDACTED]’’ on ‘‘[REDACTED]’’). some more pessimistic or optimistic expert testimony from SoundExchange’s Pursuant to the iHeart/Warner model of projections considered by economic witnesses, ascribing a Agreement, a ‘‘[REDACTED]’’ must iHeart or Warner, iHeart was monetary value to such additional ‘‘[REDACTED]. Id. at 8, ¶ 1(y). In turn, contractually bound to pay a fixed contractual consideration allegedly Exhibit A to the iHeart/Warner royalty per year, and Warner had the benefiting the record companies, Agreement permits [REDACTED]; duty to provide iHeart with access to especially if there were no requires iHeart to [REDACTED]; and Warner’s sound recordings if those fixed contemporaneous internal valuations allows a listener [REDACTED]. Id., Ex. per-play payments were made. made by the record companies A. Accordingly, the Judges look to the themselves. Accordingly, the Judges find that average rate agreed to by the parties in Reciprocally, the Judges would also iHeart/Warner Agreement satisfies the the iHeart/Warner Agreement for 2016, expect to receive evidence from the core of the ‘‘same rights test.’’ which coincides with the first year of webcasters/licensees with regard to ii. The Average Rate in the iHeart/ the statutory 2016–2020 period. That their contemporaneous calculation of Warner Agreement agreed-upon rate is $0.[REDACTED] per the monetary value of contractual play. consideration they allege to have The Judges agree with However, that average, stated per-play SoundExchange that any use of the received in addition to the basic right to rate is not necessarily applicable, play sound recordings. Also, and iHeart/Warner Agreement as a standing alone, as a benchmark, if it is benchmark must apply the effective especially if such evidence did not subject to necessary upward or exist, the Judges would expect to receive average rate contained in that downward adjustments to account for agreement.186 See SX RPFF ¶ 844 (‘‘The evidence from the economic experts other forms of consideration or to more testifying on behalf of the webcasters/ average effective rate approach . . . is accurately account for probative the proper analytical method. . . .’’) licensees regarding the monetary value evidence related to the rights available of such additional forms of (emphasis in original). The iHeart/ under the statutory license. The Judges Warner Agreement sets forth different consideration supposedly benefiting the turn to these issues in the next section webcasters/licensees. per-play rates for [REDACTED]. The of this determination. record does not reflect the reason(s) why The Judges’ expectation that such iHeart and Warner negotiated an iii. Potential Adjustments to the Rate evidence would be proffered is increase in the rates from a low of Derived From the iHeart/Warner heightened by the accurate accusations $0.[REDACTED] in [REDACTED] to a Agreement hurled by each side that the other side was manipulating the terms of the high of $0.[REDACTED] in [REDACTED] (A) General Considerations (and for any renewal term thereafter). In potential benchmark in order to any event, the parties’ inclusion of A potential benchmark can include influence the Judges in this proceeding. specific per-play rates paid to Warner in terms that provide a licensor with See, e.g., 4/30/15 Tr. 1141–42 (A. exchange for the right granted to iHeart additional compensation, whether in Harrison) ([REDACTED]); 4/28/15 Tr. to play Warner’s sound recordings cash or in kind, beyond the simple 508–09 (Kooker) [REDACTED]); 6/1/15 reflects the parties’ WTA and WTP for receipt of money in exchange for the Tr. 6962 (Lexton) (acknowledging that the particular years. In the absence of right to play sound recordings. In any deal Merlin concludes will be relevant evidence necessitating similar fashion, a potential benchmark available as evidence in CRB hearings); adjustments or legal conditions extrinsic can also provide a licensee with SX Ex.102 at 3 (5/14/14 email among to the parties’ agreement, the Judges additional compensation, beyond the Merlin executives); PAN Ex. 5117 cannot second-guess the rates to which basic right to play sound recordings in (same); 5/19/15 Tr. 4760 (Shapiro) (‘‘My the parties have agreed in a benchmark exchange for the payment of money. working assumption is that everybody is contract that otherwise satisfies the When the parties’ proposed benchmark aware of this proceeding and how . . . statutory test for a usable benchmark. agreement has bundled such other items deals they cut might affect it.’’) By applying the average rate explicitly with the simple payment-for-plays (emphasis added); IHM Ex. 3517 set forth in the iHeart/Warner obligation that mirrors the rate [REDACTED]). It would be surprising, to Agreement (subject to potential provisions of § 114, the issue arises as say the least, if parties who anticipated adjustments), the Judges have obviated to whether and how, if at all, to value that a direct deal would be used by an the protracted dispute between the these non-statutory items. adversary improperly in this proceeding parties regarding the probative value of As an initial matter, the Judges note did not develop evidence sufficient to different models and projections of that the parties have a strong self- rebut that attack, unless no such future growth of performances and interest to establish values for non- evidence—factual or expert—could royalties. That is, in the absence of a statutory items that would support their reasonably be presented. Thus, when a positions. Thus, the Judges would party fails to provide such important, 186 The stated per-play rate is the equivalent of anticipate that the record companies competent and probative factual or the ‘‘average’’ rate because it is the same rate paid and SoundExchange would present expert evidence, the Judges are left with for each performance. To use iHeart’s parlance, specific evidence of the monetary value no evidentiary basis to support the there is only one ‘‘bundle’’ of rights, with each assertion that the alleged additional performance priced at the same rate. The issue of for the non-statutory consideration they how to adjust, if at all, that ‘‘average’’ rate into the received under the contract that must be value of other contractual items is average ‘‘effective’’ rate is discussed infra. added to the stated (‘‘headline’’) rate on sufficient to alter the rates and terms of

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00070 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26385

the benchmark agreements in which parties agreed, inter se, that the value of The Judges also find that iHeart’s they are contained. the AIP terms is $[REDACTED]. Rather, willingness to provide AIP [REDACTED] With those general considerations in the iHeart/Warner Agreement sets forth to record companies was rational. As mind, the Judges now analyze particular a purported general economic fact Mr. Poleman testified, see supra, AIP issues disputed by the parties regarding regarding a ‘‘market,’’ i.e., that that there campaigns provided information about the valuation of certain items in the [REDACTED]. However, that assertion of sound recording artists that served to iHeart/Warner Agreement. supposed ‘‘fact’’ is belied by the record. build iHeart’s brand as a trusted It is undisputed that iHeart provided ‘‘curator’’ of music for its listeners. (B) AIP AIP [REDACTED] to Warner (and to Thus, AIP had value to both the record AIP, iHeart’s Artist Integration Sony and Universal) prior to the companies and iHeart, which would Program, allows Warner’s artists to formation of the iHeart/Warner explain why a sophisticated entity such benefit from particular advertising on Agreement, and that iHeart continued to as iHeart would [REDACTED] AIP time iHeart’s music-formatted radio stations provide AIP—[REDACTED]—to Sony [REDACTED] to record companies. and iHeart’s Web sites, in the form of and Universal after the execution of the Relatedly, the Judges note internal [REDACTED].’’ SX Ex 33 at 19 § 5(a)(i). iHeart/Warner Agreement. 5/21/15 Tr. iHeart communications indicating that Clearly, such advertising inures to 5343–44, 5348 (Fischel); 6/2/15 Tr. iHeart [REDACTED]. Warner’s benefit. 7312, 7335 (Cutler). It is also The Judges further find that the Additionally, the iHeart/Warner undisputed, and clear from the iHeart/ testimony by Warner’s executive, Mr. Agreement contains an express Warner Agreement, that [REDACTED], Wilcox, confirms that the provision stating that this further negating the existence of any ‘‘$[REDACTED]’’ figure was used as ‘‘[REDACTED] AIP Commitment’’ has market value. SX Ex. 33 at 34, ¶ 18(g). [REDACTED] rather than a statement of an annual ‘‘fair market value of As Mr. Poleman, an iHeart witness, value that the Judges could simply add [REDACTED] Dollars (USD testified: ‘‘these monetary figures serve to the effective rate under the iHeart/ $[REDACTED]).’’ Id. at § 5(a)(ii) no other purpose than [REDACTED]. Warner Agreement. The following (emphasis added). SoundExchange These monetary figures do not reflect testimony on direct examination is argues that there is no reason to require [REDACTED] Poleman WRT ¶ 22. telling: evidence of an internal valuation when The Judges find these undisputed Q: Did iHeart represent to you [AIP] had the parties have agreed to a ‘‘fair market facts to demonstrate that there was no value, monetary value? value’’ on the face of their contract. actual ‘‘market’’ in which Warner A: Yes. iHeart makes several arguments in an procured AIP from iHeart. If such a Q: What was that amount? attempt to disavow this agreed-upon market existed, with a fair market value A: Well, ultimately it was agreed on that valuation: of $[REDACTED] for the AIP provided we would say that it was [REDACTED]. They • AIP provides value to iHeart and to to Warner, it would have been irrational were contending it was worth more and that Warner because AIP content is valuable for iHeart simply to give away such was a conservative estimate. Ultimately, they gave us the $[REDACTED] CPM number as a to listeners and therefore also ‘‘helps substantial value (e.g., the equivalent of way to value the different impressions that build [iHeart’s] brand . . . as [a] trusted [REDACTED]% of Dr. Rubinfeld’s were available to us through AIP. So that was curator[] . . . .’’ 5/21/15 Tr. 5189–92 proposed rate for 2016 and of the NAB/ ultimately where we agreed to settle in terms (Poleman). SoundExchange settlement rate for of valuing it.188 • Warner received [REDACTED] AIP 2015). See 5/28/15 Tr. 6284 (Rubinfeld) 6/3/15 Tr. 7388–89 (Wilcox). This [REDACTED] and the $[REDACTED] (AIP at a value of $[REDACTED] per testimony reveals two points: First, the reference was intended to reflect year would raise the effective rate by valuation was negotiated to establish a [REDACTED]. 6/2/15 Tr. 7312 (Cutler). $0.[REDACTED] per play). quantity term for AIP. Second, this • Rather, the Judges find guidance for iHeart’s commitment to testimony does not indicate any the meaning and of this [REDACTED] AIP therefore was in the reference in the negotiations to a ‘‘fair ‘‘$[REDACTED]’’ figure as it relates to nature of ‘‘insurance,’’ rather than a market value’’ for AIP that the parties the setting of rates in this proceeding in granting of an additional right. See IHM later simply plugged into the iHeart/ the context of the contractual clause in RPFF ¶ 815 (and citations to the record Warner Agreement. See also 6/2/15 Tr. which the figure is contained. The therein). 7318 (Cutler) (‘‘This is a sort of a quick- • Neither iHeart, Warner, nor contract states: ‘‘[iHeart] shall provide and-dirty formula where we took a Universal treated AIP as a Warner AIP insertions in each hugely averaged rate and applied it to ‘‘[REDACTED],’’ and iHeart Agreement year . . . that (i) have a fair what we—you know, ultimately these [REDACTED]. Id. ¶ 817 (and citations to market value of at least . . . promotional spots in these AIP the record therein). $[REDACTED] per Agreement Year; and programs.’’). • represent at least . . . [REDACTED]% of The $[REDACTED] was derived The Judges also find credible and all AIP inventory in each daypart and from iHeart’s advertising ‘‘rate card’’ as important the undisputed fact that no market.’’ SX Ex. 33 at 19 ¶ 5(a)(ii). This a means to measure that Warner got party, and no record company, [REDACTED]. 5/21/15 Tr. 5190 provision is consonant with iHeart’s (Poleman). explanation that the $[REDACTED] • then ([REDACTED]) Warner may [REDACTED]. Id. In its own projections, Warner figure was used to establish Thus, as a remedy for breach [REDACTED]. This declined to value AIP because AIP [REDACTED], and therefore is not a remedial provision further indicates that Warner ‘‘[REDACTED].’’ 6/3/15 Tr. 7500 monetary value that the Judges may had obtained in the iHeart/Warner Agreement (Wilcox). simply pro-rate, and thereby grossly [REDACTED] which, upon an iHeart breach, [REDACTED]. Additionally, [REDACTED]. See id. 187 The Judges find that the AIP provision inflate the benchmark rate. (‘‘[REDACTED]’’). in the iHeart/Warner Agreement does 188 ‘‘CPM’’ is cost per thousand advertising not support an increase in the effective 187 The Judges find that the contractual remedial impressions. 4/28/15 Tr. 419 (Kooker). Thus, the average per-play rate derived from that provisions relating to AIP support their findings in $[REDACTED] per 1,000 impressions factor can be this regard. Performance of the AIP terms required used to determine the quantity of impressions if benchmark. As an initial matter, the AIP iHeart and Warner to [REDACTED]. Id. ¶ 5(a)(i). In $[REDACTED] is substituted for the $[REDACTED] language in the iHeart/Warner turn, the iHeart/Warner Agreement provides that, if figure. Impressions are viewed or heard ads. 6/3/15 Agreement does not state that the Warner and iHeart disagree regarding [REDACTED], Tr. 7403–04 (Wilcox).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00071 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26386 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

considered that AIP could be valued as party through a fact or expert witness Mr. Wilcox did not dispute that Warner a cash equivalent. That is consistent presented any basis to create a monetary had [REDACTED]. 6/3/15 Tr. 7452 with the finding that the AIP term in the value for this ‘‘insurance.’’ Therefore, (Wilcox). Thus, the Judges find that, iHeart/Warner Agreement was intended the Judges are presented in this context even to the extent that post-contract as an [REDACTED], rather than a with the conundrum of an item of performance might be helpful in valuing mechanism for dramatically ostensible (insurance) value that has not determining value, Mr. Wilcox’s inflating the effective per-play rate in been valued by the parties, but is testimony as to a value in excess of that agreement. tendered to the Judges without $[REDACTED] for [REDACTED] is The Judges’ decision on this issue is evidentiary guidance. The Judges return simply not credible. also informed by the negotiating to the point made in the General In this context as well, neither party’s position taken by Warner. In particular, Considerations section. SoundExchange, negotiators nor its economic experts set under cross-examination, Mr. Wilcox, through Dr. Rubinfeld, acknowledges forth a monetary value. The rebuttal the testifying Warner executive, when that there is some insurance value in the performance-based testimony that asked if ‘‘you told the iHeart conversion of AIP into a contractual SoundExchange relies upon from Mr. representatives during negotiations that commitment, yet SoundExchange did Wilcox to demonstrate that you thought AIP was worth zero,’’ not present a method for valuation. [REDACTED] had value is simply testified: ‘‘I don’t have a specific iHeart, through Dr. Fischel, avers that insufficient when considered against recollection right now, but . . . that this ‘‘insurance’’ value would be quite Warner’s failure to [REDACTED], and in would have been consistent with the small, and he too did not provide a light of the fact that the Judges did not negotiating posture that I might have monetary value. If a party had the find Mr. Wilcox to be a particularly taken.’’ 6/3/15 Tr. 7466 (Wilcox) understanding that an element within a credible witness. Accordingly, the (emphasis added). This testimony benchmark could be valued in a manner Judges do not find that the inclusion of undermines Warner’s assertion that the that would further support its position, [REDACTED] rights in the iHeart/ Judges should simply add the Judges would expect that party to Warner Agreement supports an increase $0.[REDACTED] to the per-play rate present evidence in that regard. Here, in the effective average per-play rate derived from this benchmark, when SoundExchange declined to do so with derived from that agreement. Warner’s own witness had claimed in regard to the ‘‘insurance’’ value of the (D) The [REDACTED] Agreement negotiations that AIP had no value. conversion of AIP into a contractual Moreover, even if Mr. Wilcox’s assertion commitment. The Judges therefore find The Judges decline to include in the represented only his ‘‘negotiating that such unquantified ‘‘insurance’’ average effective rate any value derived posture,’’ then the Judges find that value cannot be added to the effective from the $[REDACTED] payment by iHeart’s representation of a positive per-play rate under the iHeart/Warner iHeart to Warner for rights under the value, including the $[REDACTED] Agreement.190 [REDACTED] Agreement. As an initial matter, this agreement is not even part figure plugged into the agreement, was (C) [REDACTED] also the consequence of negotiation of the iHeart/Warner Agreement. rather a declaration of fact as to the [REDACTED] the [REDACTED], is a Second, the [REDACTED] Agreement existence of a ‘‘fair market value’’ of program by which Warner may contains an integration clause that, as $[REDACTED].189 Finally, the Judges do [REDACTED]. See SX Ex. 33, Ex. F iHeart correctly notes, by its plain not find credible Mr. Wilcox’s testimony thereto. SoundExchange asserts that it language declares that it is the entire that he was informed by iHeart that it has a quantifiable value to Warner that agreement between the parties and thus would [REDACTED] AIP, in light of the must be pro-rated across the number of excludes reference to any other absence of any document sufficient to performances and added to the per-play agreement, such as the iHeart/Warner corroborate that assertion, and in light rate. However, the record indicates that Agreement. SX Ex. 1339. The Judges of the fact that iHeart has not Warner did not engage in any valuation further note that the iHeart/Warner [REDACTED] AIP. Moreover, even if of [REDACTED] contemporaneous with Agreement [REDACTED]. SX Ex. 33 iHeart had taken such a negotiating the negotiation of the iHeart/Warner ¶ 18(c). Third, the [REDACTED] position, the Judges do not find, after Agreement and that Dr. Rubinfeld did Agreement provides for a payment of listening to Mr. Wilcox’s testimony, that not perform any such expert economic $[REDACTED] in exchange for a specific set of rights unrelated to iHeart’s right he genuinely believed such a change in valuation. 5/28/15 Tr. 6437 (Rubinfeld). Rather, SoundExchange’s entire to play Warner sound recordings on AIP policy was forthcoming. The Judges do recognize that, by argument in support of a valuation, in iHeart’s noninteractive service. Fourth, converting AIP from a discretionary, excess of $[REDACTED], for it is irrelevant that Warner was aware of, voluntary program to a contractually [REDACTED] is based upon the hearing and made reference to, the [REDACTED] binding commitment, iHeart provided testimony of Mr. Wilcox. He derived Agreement value when it considered the Warner with what Drs. Fischel and this value from a single [REDACTED] value of its forthcoming relationship Rubinfeld both considered to be campaign undertaken by Warner after with iHeart. Indeed, as iHeart points ‘‘insurance’’ value. However, neither the iHeart/Warner Agreement had been out, Warner’s internal models and other executed. Wilcox WRT at 14 n.9. documents identified the [REDACTED] 189 The irony surrounding this issue is not lost on However, as iHeart points out, Warner’s Agreement’s $[REDACTED] payment the Judges. In this proceeding, Warner claims AIP post-execution performance—or more obligation as a distinct payment for has significant value, in order to inflate the accurately, non-performance— [REDACTED]. See iHeart RPFF ¶ 828 benchmark, but claimed during negotiations that contradicts this attempt at a (and citations to the record therein). AIP had no value, in order to [REDACTED]. 6/3/15 Tr. 7466 (Wilcox). Likewise, during negotiations, performance-based valuation. That is, The Judges also agree with iHeart’s iHeart touted the benefits of AIP, but minimizes its argument that the $[REDACTED] significance during this proceeding, in an attempt 190 Also, the unquantified value of any payment obligation in the [REDACTED] to avoid an increase in the effective benchmark rate. ‘‘insurance’’ aspect of the contractual AIP Agreement presents the Judges with an Such switching of positions, combined with the commitment would have had to be offset against the other issues discussed in this section regarding AIP, value of other non-pecuniary items in the iHeart/ issue of allocation rather than valuation. underscore the indeterminacy of AIP’s impact, if Warner Agreement that favor iHeart, as discussed See iHeart RPFF ¶ 830. The fact that the any, on this benchmark. infra. [REDACTED] Agreement contains an

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00072 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26387

unambiguous integration clause otherwise proper and reasonable than [Dr.] Fischel’s proposed underscores the fact that the rights and benchmark. incremental rate.’’ Id. at 6439. payments under that contract must be (F) Offsetting Value to iHeart in the (G) Adjusting the iHeart/Warner allocated only to that contract. The iHeart/Warner Agreement Benchmark Rate to Account for Judges therefore find that the [REDACTED] and Thereby Equalizing $[REDACTED] payment to Warner by iHeart points out that the iHeart/ the Number of Royalty-Bearing Plays iHeart under the [REDACTED] Warner Agreement also provides value Between the Benchmark and the Statute Agreement is properly allocated to that to iHeart in the form of: (1) A agreement for the provision of [REDACTED] royalty ceiling that serves Drs. Fischel and Lichtman note that [REDACTED], and cannot be attributed as de facto insurance against an iHeart listener is entitled to to the valuation of the parties’ rights— [REDACTED] and (2) most-favored- [REDACTED] 192 per hour per station or and rates—under the iHeart/Warner nation status at least equalizing iHeart’s channel, for which iHeart is not Agreement. terms with Warner’s terms in any required to pay royalties. Fischel/ Lichtman AWDT ¶ 35; SX Ex 33 at 15 (E) Other Unvalued Contract Items agreement with [REDACTED] Fischel/ Lichtman AWDT ¶ 38. However, the ¶ 3(b)(i); id. at 38 Ex A therein. They As noted supra, SoundExchange chronic problem the Judges have note, after setting forth the number of asserts that the effective average rate referenced supra applies here as well: [REDACTED] and performances that, under the iHeart/Warner Agreement iHeart did not attempt to place a value ‘‘[i]n July 2014, [REDACTED] . . . must be increased to reflect the value of on such items. Id. ¶ 39 (‘‘It is difficult constituted approximately [REDACTED] additional contract items, including: percent of all iHeart custom • to precisely quantify the value of these The guarantee that iHeart would various non-pecuniary terms’’ and performances, so that the functional per- [REDACTED] even if such steering fell iHeart ‘‘made no explicit attempt to performance rate paid on these contracts short of that level.191 • value these terms.’’). is approximately [REDACTED]% lower The alternative percentage-of- than the statutory per-performance revenue rate in the greater-of However, Drs. Fischel and Lichtman point out that because both parties pureplay rate.’’ Fischel/Lichtman formulation. AWDT ¶ 61 & n.9. This [REDACTED] • The additional $[REDACTED] failed to value such terms, it is acceptable to ‘‘assume[] a net value of adjustment is very close to Dr. payment guarantee by iHeart even if it Rubinfeld’s skips adjustment factor of never played any Warner sound zero for these terms.’’ Id.; see 5/28/15 Tr. 6435–37 (Rubinfeld) (acknowledging [REDACTED], which also included an recordings. offset for increased plays by virtue of • The guarantee that Warner would that he failed to attribute numerical the royalty value of [REDACTED] under receive at least the same [REDACTED], dollar values to items in the iHeart/ his interactive benchmark agreements). as it did prior to the iHeart/Warner Warner Agreement that benefited each If Drs. Fischel and Lichtman had Agreement. party respectively). • Warner’s [REDACTED], which The Judges disregard these unvalued applied that [REDACTED]% reduction iHeart could [REDACTED]. items; not because, as Drs. Fischel and to the otherwise stated average rate of • Royalties paid for [REDACTED]. Lichtman assert, they should be $0.[REDACTED] for 2013 in the iHeart/ See SX RPFF ¶ 889 (and citations to presumed to have a net value of zero. Warner Agreement, they would have the record therein). Rather, as stated in the General equalized that rate to a statutory rate of With regard to all of these items, Considerations section, the Judges tie $0.[REDACTED]. However, Drs. Fischel notwithstanding any potential monetary the indeterminacy of the net value of and Lichtman adjust their 2013 stated value that might be associated with these offsetting items to a (perhaps average rate from $0.[REDACTED] to them, neither Warner nor tactical) failure of proof of value by $0.[REDACTED]. SoundExchange avers SoundExchange established values for sophisticated parties. As Dr. Rubinfeld that it appears from iHeart’s own these items. Indeed, SoundExchange acknowledged in a colloquy with the documents however that this acknowledges that, when the Judges Judges: $0.[REDACTED] rate reflects an asked Mr. Wilcox whether Warner had incorporation of the Pureplay rate rather assigned a number value to ‘‘these [JUDGES] than a calculation to adjust for [I]f iHeart is paying a . . . rate based on [REDACTED] See SX Ex. 221 at 1, 4 & provisions,’’ he admitted that Warner dollar denominated items and gets some ‘‘could not be certain.’’ 6/3/15 Tr. 7409 other non-dollar denominated value—net n.21. (Wilcox). As the Judges noted in the value to iHeart as if it was paying some lower In response to this criticism, iHeart General Considerations section of this rate because it got new items of value—. . . does not refer the Judges to any analysis of the iHeart proposal, if the we just can’t value them because nobody did evidence of calculations it did to party that seeks to increase (or decrease) and we don’t have the evidence to do so. support a [REDACTED] reduction from an otherwise effective benchmark rate to [DR. RUBINFELD] $0.[REDACTED] to $0.[REDACTED]. account for other items of potential Yeah, that’s possible. Rather, iHeart simply declares value cannot or has not provided 5/28/15 Tr. 6439. Continuing, the Judges SoundExchange’s reliance on SX Ex. evidence of such value, when it was in reiterated that for these other items of 221, iHeart’s own document, is its self-interest to do so, the Judges value, ‘‘the sign is moving plus and insufficient to call into question the cannot arbitrarily adjust or ignore that minus’’ but ‘‘without dollar values [REDACTED] adjustment proposed by attached by the experts or the parties in iHeart. See iHeart RPFF at 119–20. 191 The parties disputed whether the pre- their contracts or their negotiations,’’ The Judges find that SoundExchange’s agreement pro rata level was [REDACTED]% or and lamented that they ‘‘have no way of criticism is appropriate. In order to [REDACTED]%. That dispute related to a reflect not only the [REDACTED] measurement of the ‘‘two bundles’’ hypothesized by valuing them . . . .’’ Dr. Rubinfeld Drs. Fischel and Lichtman, but rejected by the responded by commiserating, adjustment, but also to make an Judges in this determination. Under an average rate acknowledging that he too did not, and approach with a steering-based [REDACTED]% pro 192 [REDACTED] custom performances are rata share, it is irrelevant whether the pre-contract instead he simply fell back to a non- defined in the iHeart/Warner Agreement as pro rata Warner share on iHeart was [REDACTED]% sequitur: that his proposed rate was performances ‘‘that are [REDACTED].’’ SX Ex. 33 at or [REDACTED]%. closer to the ‘‘actual NAB rates . . . p. 15, ¶ 3(b)(i).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00073 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26388 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

adjustment to reflect plays of iHeart/Indies Agreements apply the per- who participated in this proceeding. [REDACTED], the Judges adopt Dr. play rates that have a set (i.e., average) That lower bound, according to Sirius Rubinfeld’s [REDACTED] adjustment to per-play rate that controls for each XM, is $0.0011. See Sirius XM PFF equalize the number of plays as between year.196 Those per-play rates are all ¶¶ 65–68.198 this benchmark and the statutory rate. equal to the [REDACTED] rates and Sirius XM did not produce an expert Thus, the 2013 rate of $0.[REDACTED], therefore are less than $0.[REDACTED]. witness to testify in support of its rate as noted above, would equalize to See, e.g., IHM Ex. 3353 ¶ 1(w) (the proposal. Rather, as noted above, Sirius $0.[REDACTED]. iHeart/Next Plateau Entertainment XM relies upon the lowest rate within More importantly, for the first year of Agreement). Thus, iHeart apparently has its WSA with SoundExchange and the the statutory period at issue, 2016, the derived that $0.[REDACTED] rate by work of the other Services’ economic stated average rate is $0.[REDACTED]. adding to the stated custom rates its per- witnesses to support its range, Applying a [REDACTED] adjustment of play calculation of additions to the rate endpoints and proposed rate. Thus, the [REDACTED] results in an equalized arising from the [REDACTED] revenue probative value of the Sirius XM rate is rate of $0.[REDACTED]. (Even applying to which Indies are entitled under the dependent in large measure upon the iHeart’s proffered [REDACTED]% rate iHeart/Warner Indies Agreements. As Judges’ analysis and conclusions reduction for this factor would result in the Judges noted with regard to the regarding the models proffered by these an adjusted rate of $0.[REDACTED], [REDACTED] revenues in their analysis other experts. Indeed, Sirius XM does before any consideration of additional of the proposed rates for simulcasting, not attempt to independently support [REDACTED].).193 these revenues are simply too the work of those other experts. Instead, indeterminate to support a rate analysis Sirius XM devotes the bulk of its c. The Percentage of Revenue Provision by the Judges. The Judges incorporate independent argument to an analysis of in the iHeart/Warner and iHeart/Indies those findings here, and find that the 27 its WSA settlement agreement.199 Agreements iHeart/Indies Agreements are not usable 2. Sirius XM’s Arguments in Favor of Its The iHeart/Warner Agreement as benchmarks, guideposts or other Rate Proposal contains a greater-of rate formula that evidence to support the rates set in this includes a [REDACTED]%– proceeding.197 Sirius XM’s primary business is broadcasting on a subscription fee basis [REDACTED]% rate, depending upon F. Sirius XM Rate Proposal the year of the agreement. SX Ex. 33 at over its two proprietary satellite 15–16, ¶ 3(b)(ii).194 1. Proposed Royalties systems. However, it also provides a simulcast of its satellite broadcast over For the reasons set forth in the Judges’ Sirius XM proposes that the § 114 the Internet. SXM Ex. 6000 ¶ 20 (Frear comprehensive rejection of a greater-of digital sound recording public WDT). Thus, Sirius XM’s Internet radio structure with a percentage-of-revenue performance royalty rate applicable to service is primarily a simulcast of Sirius prong, the Judges do not include these commercial webcasters for the 2016– XM’s satellite service. Id. ¶ 27 (emphasis iHeart greater-of provisions in the 2020 rate period be $0.0016 per- benchmarks they derive from the iHeart/ added). performance. Introductory Sirius XM also offers as an Internet Warner Agreement and the iHeart/ Memorandum to Sirius XM WDS at 1 service a noninteractive feature, ‘‘My Indies Agreements. (October 7, 2014). In support of this rate, Sirius XM,’’ at no extra charge to its Sirius XM avers that a zone of d. The Judges Consideration of the 27 Internet radio subscribers. Id. ¶ 28. reasonableness can be established for iHeart/Indies Agreements (Sirius XM also offers an on-demand the statutory rate. The high end of the service, ‘‘Sirius XM On Demand,’’ that iHeart has calculated an average zone, according to Sirius XM, is the is not subject to the § 114(f)(2)(B) rates). royalty per play for Indies of $0.0016 per-performance rate, which The noninteractive, non-simulcast $0.[REDACTED]. Fischel/Lichtman represents the lowest rate contained in 195 service, My Sirius XM, allows AWDT Ex. D therein. However, the the 2009 WSA settlement agreement subscribers to slightly personalize a between SoundExchange and Sirius 193 SoundExchange also takes issue with iHeart’s XM. The low end of the zone, according select group of music and comedy alleged application of a [REDACTED] adjustment to channels from the satellite service, to [REDACTED] webcasts [REDACTED], which to Sirius XM, is represented by several ‘‘guideposts,’’ i.e., the low end of the adjust for characteristics like library SoundExchange avers cannot be adjusted for depth, familiarity, and music style. Id. [REDACTED] because these stations, [REDACTED], estimated range of proposed rates do not [REDACTED]. See SX PFF ¶¶ 849–850 (and proffered by the economic experts who ¶ 28. citations to the record therein). iHeart disputes that Although introduced as a response to testified on behalf of the other Services assertion. See IHM RPFF at 120 (and citations to the truly customized Internet radio like record therein). SoundExchange also combined its Pandora, My Sirius XM does not [REDACTED] criticism in this regard with a ‘‘incremental’’ rate they proffered under the iHeart/ separate criticism regarding the treatment of Warner Agreement. provide the same amount of ‘‘digital only’’ transmissions by iHeart, leading Dr. 196 The greater-of percentage of revenue customization. My Sirius XM begins Rubinfeld to make a $0.[REDACTED] upward alternative was never triggered. Fischel/Lichtman from the same playlist created by adjustment to account for both of these issues. See AWDT ¶ 61. human curators for a satellite radio SX PFF ¶ 851 (and citations to the record therein). 197 To be clear, the Pandora/Merlin effective rate channel, and narrows that playlist SoundExchange did not clearly and sufficiently is $0.[REDACTED]—below the Pureplay rate explain its position on these combined issues, and because of the steering provisions in that the Judges therefore decline to make the $0.0001 agreement. See supra. Pandora had been subject to 198 Although Sirius XM asks the Judges to rely on upward adjustment advocated by Dr. Rubinfeld. the Pureplay rates and utilized steering to induce the low end of these ‘‘guideposts,’’ it notes that the 194 The iHeart/Indies Agreements contain a the Merlin members to agree to a lower rate in high end of these ‘‘guidepost’’ ranges from the other greater-of structure that, as noted above, fixes the exchange for more plays. The same concept (albeit Service economic experts is $0.0017, higher than percentage-of-revenue prong at [REDACTED]%. with different rates) underlies the 27 iHeart/Indies the top of its proposed range and its proffered See, e.g., IHM Ex. 3353, at 7–8, ¶ 4(a)(iii)(A). Agreements. These 27 Indies agreed to reduce the benchmark of $0.0016. However, as stated in the text, supra, the Judges rate to $0.[REDACTED] in [REDACTED], from the 199 For this reason, the Judges need not discuss find these agreements not to be probative. $0.[REDACTED] settlement rate on which they the merits of Sirius XM’s proposed range or, in 195 Drs. Fischel and Lichtman also calculated an could have insisted, in exchange for a lower rate particular, the low end of that range. The relative ‘‘incremental’’ per-play rate for Indies of that incentivizes iHeart to steer more plays to them merits of the benchmarks on which Sirius XM relies $0.[REDACTED]. Id. The Judges reject that rate for plus some indeterminate amount of [REDACTED] are discussed in the sections of this determination the same reason they rejected the $0.0005 revenues. dealing directly with those other benchmarks.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00074 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26389

slightly by manipulating a few sliders, In the context of the severe financial could provide any effective relief from which emphasize or deemphasize broad stress affecting Sirius XM’s entire those rates; and (3) the exacerbation of characteristics common to the relevant business, and the Internet radio that imbalance in bargaining power genre. 5/22/15 Tr. 5419–21 (Frear). For services’ extremely low usage and caused by various unrelated example, listening to the ‘60s channel importance to its core business, Sirius circumstances affecting Sirius XM at the through My Sirius XM might allow the XM believed it had no sensible option time of the negotiations. Sirius XM Ex. subscriber to emphasize more late ‘60s other than to accept the deal offered by 6000 ¶ 52. Sirius XM further avers that, music, more early ‘60s music, more SoundExchange. If it had not taken the by contrast, neither SoundExchange nor electric music, or more acoustic music. deal, Sirius XM would have been its constituent record companies had Id. at 5419:19–25. My Sirius XM allows required to continuing paying the higher similar countervailing pressures that users to shrink the playlist by adjusting Webcasting II rates. At the same time, could have mitigated this extreme for these characteristics—but does not NAB simulcasters with which Sirius imbalance. Id. ¶ 57 (and citations to the permit users to expand the playlist from XM’s Internet radio service competes record therein). that of the satellite radio channel. Id. would be paying the lower WSA Nonetheless, Sirius XM proposes that The Sirius XM Internet radio service settlement rates, and Pandora would be the Judges rely on the WSA settlement is a minor part of Sirius XM’s overall paying a small fraction of the agreement between Sirius XM and business, with its self-pay subscription Webcasting II rates, putting Sirius XM at SoundExchange, by adopting its lowest revenue (i.e., excluding trial a significant competitive disadvantage. rate, $0.0016, not only as the ‘‘the outer subscriptions) accounting for only Although Sirius XM could have boundary of a range of reasonable [REDACTED]% of Sirius XM’s total refused to sign the WSA with rates,’’ but also as the rate to be set in revenue. Frear WDT ¶ 29. Usage of the SoundExchange and instead sought the present proceeding. See Sirius XM non-simulcast My Sirius XM is low lower rates in the then-forthcoming Web PFF ¶ 64. Additionally, Sirius XM does even in comparison to the usage of III proceeding, the low listenership to not propose any rate escalation or Internet radio simulcast channels. Id. the Internet radio service meant that the reduction over the 2016–2020 period, ¶ 28. cost of participation in that proceeding whether to reflect inflation, deflation, or Sirius XM points out the relatively could far exceed any possible future any other factor. Finally, Sirius XM does low importance of noninteractive savings in royalty payments. Although not propose a two-prong rate structure services to its overall business model in Sirius XM attempted repeatedly to embodying any other rate formula than order to explain why it entered into the negotiate a more significant reduction, the per-play structure. WSA with SoundExchange in 2009— SoundExchange consistently refused to 3. SoundExchange’s Opposition to the and why that settlement agreement was materially move off its opening offer of Sirius XM Rate Proposal and remains not probative of market essentially matching the NAB rates. 5/ value and lacked the persuasive value 22/15 Tr. 5435–36 (Frear). With no other SoundExchange opposes the Sirius attributed to it in the Web III Remand. option that would have a less costly net XM rate proposal on several grounds. In this regard, Sirius XM avers: result, Sirius XM entered into the WSA First, SoundExchange rejects Sirius • As a result of the Webcasting II settlement agreement with XM’s suggestion that its settlement rates, Sirius XM made the decision to SoundExchange. Id. at 5434–35. contained above-market rates, because drop all free streaming on both the Then, according to Sirius XM, two Sirius XM voluntarily agreed to those Sirius and XM platforms, a decision that days before the deadline on which rates, even though it was under no resulted in a [REDACTED]- Sirius XM and SoundExchange were compulsion to negotiate with [REDACTED]% drop in the Internet required to close negotiations—and after SoundExchange. See SX RPFF ¶ 1022. radio service’s reported listening hours the parties had already agreed on the Second, SoundExchange states that and a resulting decrease in royalty rate schedule and finalized their deal— Sirius XM is flatly wrong to suggest that payments to SoundExchange. Id. ¶ 35; Michael Huppe (the party negotiating on its negotiation with SoundExchange did 5/22/15 Tr. 5416–17 (Frear). behalf of SoundExchange) added an not ‘‘mov[e] the needle with respect to • By late 2008, Sirius XM had extra term into the Agreement, requiring royalty rates.’’ In fact, Sirius XM was insufficient cash to repay hundreds of that it be precedential under the WSA. not only able to negotiate rate lower millions of dollars of debt scheduled to 6/3/15 Tr. 7627–29 (Huppe); 5/22/15 Tr. than the then-prevailing statutory rates come due in February 2009, and was 5443–54 (Frear). Having already failed for 2009, 2010, and 2011, but it was also unable to access the capital markets to to advance its other interests in able to negotiate lower rates for 2013, refinance this, and other, debt. Frear negotiations, Sirius XM agreed to this 2014, and 2015 than were contained in WDT ¶ 40. new term requiring its WSA settlement the NAB settlement. SX PFF ¶ 1079; SX • The pre-merger predecessors to agreement to be precedential, RPFF ¶ 1027. Sirius XM, Sirius and XM, had recently concluding negotiations and Third, when SoundExchange, through spent over $150 million on merger costs consummating the agreement before the Mr. Huppe, informed Sirius XM that alone. Id. ¶ 46. statutory deadline. Id. at 5444. SoundExchange wanted the settlement • Sirius XM narrowly avoided filing For the foregoing reasons, Sirius XM agreement to be precedential under the for bankruptcy protection when a maintains that the rates in the Sirius XM WSA, Sirius XM voiced no objection potential lender agreed to provide a loan WSA settlement agreement do not whatsoever in its email response less that narrowly enabled Sirius XM to reflect any industry-wide fair market than an hour later. NAB Ex. 4235. avert a default on its debt and value for the license. Instead, it claims Fourth, SoundExchange argues that bankruptcy. Id.; 5/22/15 Tr. 5430 that the rates are a product of: (1) The basic economics suggests that any (Frear). Web II rates, which, in Sirius XM’s financial distress Sirius XM was • The Sirius XM stock price fell from view, Congress found to be so wildly experiencing at the time should have over $4.00 per share in January 2007 to supracompetitive as to warrant reduced, not increased, its willingness a low of $0.05 per share on February 11, Congressional intervention and which to pay royalties for webcasting. SX Ex. 2009. Frear WDT ¶ 45. On September would continue to apply in the absence 29 ¶ 228 (Rubinfeld Corr. WRT). 15, 2009, Sirius XM received a delisting of a settlement; (2) SoundExchange’s Fifth, Sirius XM had a number of notice from NASDAQ. Id. monopoly power as the only entity that alternative options in addition to

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00075 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26390 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

agreeing to the settlement with that agreement as its proposed rate in economic expert witness, Professor SoundExchange. Specifically, the present proceeding.200 Michael Katz. Dr. Katz did not perform SoundExchange notes that Sirius XM Second, the Judges are unpersuaded a benchmark analysis to arrive at a rate. instead had the option to: by the fact that Sirius XM apparently Rather, he selected guideposts that • litigate in the Web III proceeding can afford the $0.0016 rate it now define the lower and upper bounds of and seek lower rates from the Judges; proposes, in contrast to earlier years what he described as a range of • avoid the cost of litigating Web III when it was financially in extremis. As reasonable rates that a willing buyer and and simply awaited the Judges’ rate the Judges held in the Web III Remand, a willing seller would agree to in a determination (a ‘‘costless option’’ and have consistently held, workably competitive market. See Katz according to SoundExchange); or § 114(f)(2)(B) does not require the Judges WDT ¶80. The NAB’s proposed rate of • avoid the statutory license to set a rate that ensures the financial $0.0005 per-performance presumably completely and enter into direct viability of any entity. Thus, the fact falls somewhere within that range.202 licenses with the various record that Sirius XM may be able to afford the Dr. Katz determined the low end of companies. $0.0016 rate now, but might not be able his ‘‘zone of reasonableness’’ by SX PFF ¶ 1077 (and citations to the to afford any higher rate, is simply not reference to terrestrial radio. See Katz record therein). pertinent to the Judges’ determination. WDT ¶¶ 81–84. Radio broadcasters are Sixth, SoundExchange notes that Moreover, the fact that Sirius XM not required to pay royalties for Sirius XM—despite its asserted acknowledges that noninteractive terrestrial broadcasts of sound financial difficulties—continued and streaming is only an ‘‘ancillary’’ part of recordings, and typically do not do so. expanded its noninteractive services, its business (in contrast to its satellite See 17 U.S.C. 114(a); Katz WDT ¶ 82. even though it asserted that such service) indicates that the impact of the Nevertheless, Dr. Katz points out, record services were an insignificant portion of rates on its noninteractive service companies seek out radio airplay to Sirius XM’s total subscribership cannot be a driver of the statutory rate promote other income streams, such as revenue. Moreover, SoundExchange determination. The Judges note that sales of CDs and permanent downloads. notes, Sirius XM’s internet revenue Sirius XM was willing to accept rates in See Katz WDT ¶ 82. He argues that grew from $[REDACTED] in 2010 to its 2009 WSA settlement at least in part economic theory predicts that this $[REDACTED] in 2014 while Sirius XM because of the ancillary nature of its promotional effect would drive down was paying rates under its WSA noninteractive service. Because that royalty rates, possibly even resulting in settlement agreement with noninteractive service remains ancillary negative royalty rates if the law SoundExchange. SX PFF ¶ 1078 (and in nature to Sirius XM, the Judges permitted record companies to pay citations to the record therein). cannot conclude that impact of the rates broadcasters to play their music (i.e., Seventh, SoundExchange asserts that set in this proceeding have any greater payola). Id. ¶¶ 81–82. Sirius XM’s rate proposal has no sound particular importance to Sirius XM now. Dr. Katz then argues ‘‘available basis. According to SoundExchange, the evidence indicates that promotional proposal was simply plucked from the G. NAB Rate Proposal benefits also arise from web simulcasts first year of the Sirius XM WSA 1. Proposed Rates of terrestrial broadcasts.’’ Id. ¶ 83. In settlement. Id. ¶ 61. Moreover, effect, he equates simulcasting with The NAB proposes a two-tiered rate according to SoundExchange, Sirius terrestrial radio and concludes that the structure for webcasts by simulcasters. XM’s reliance on the low-end rate in an lower bound of the range of reasonable Broadcasters that transmit fewer than agreement that its principal witness, Mr. rates for simulcasting is ‘‘near zero.’’ Id. 876,000 ATH would pay only the Frear, now expressly disavows, is ¶ 84. minimum fee. NAB Proposed Rates and arbitrarily selective and internally To set an upper bound to his zone of Terms at 3 (October 7, 2014). All other inconsistent. SX PFF ¶ 1081. reasonableness, Professor Katz looked to broadcasters would pay a per- the Judges’ decision in SDARS II. Id. 4. The Judges’ Analysis of the Sirius XM performance royalty rate of $0.0005 to ¶ 85. According to Professor Katz, Rate Proposal simulcast for each year of the rate term. The Judges reject Sirius XM’s Id. at 3–4. In SDARS II, the judges found that 13 percent [of gross revenue] constitutes a argument for a number of reasons. First, NAB’s rate proposal is limited to simulcasts (retransmissions by sensible upper bound on the zone of the Judges decline to re-litigate the reasonableness before adjusting to account probative value of the 2009 WSA broadcasters of programming for Section 801(b) factors. The rate was then settlement agreement between Sirius transmitted over their AM or FM radio reduced by an additional two percent for the XM and SoundExchange. That stations), and does not cover other third 801(b) factor, which was specific to agreement was entered into more than commercial webcasts. Id. at 2 (definition Sirius XM and the SDARS II proceeding. six years ago, and therefore does not of Eligible Transmission). Having Id. (footnotes omitted). He adopted 13 represent the present state of the rejected the NAB’s proposal to apply a percent of gross revenue as ‘‘an initial 201 noninteractive market, absent separate rate to simulcasters, the guidepost’’ for determining his range of affirmative evidence to the contrary. Judges consider the NAB’s proposed reasonable rates for simulcasters, subject Whether Sirius XM was compelled by rate as a rate that would apply to all to two adjustments to account for its financial circumstances or not to commercial webcasters. For the reasons differences between SDARS (satellite enter into that settlement might have detailed below, the Judges reject the affected the relevance of that agreement NAB’s rate proposal. 202 As discussed below, the upper bound of the as a benchmark in Web III, but it has no NAB’s range of reasonable rates is expressed as a 2. Analysis of Economic Evidence percentage of revenue. The NAB’s proposed rate is significance to the Judges in the present expressed as a per-performance royalty, however, proceeding. Indeed, as SoundExchange The NAB presented its methodology for arriving at a rate proposal through its and there is insufficient data in the record to notes, it is inconsistent for Sirius XM, convert the per-performance rate to a percentage of on the one hand, to criticize the revenue (and vice versa). Since the Judges deem it 200 The Judges have also analyzed the impact, if highly unlikely that the NAB would propose a rate benchmark value of its 2009 WSA any, of the other 2009 WSA settlement agreement— that exceeds the upper bound of its own expert’s settlement agreement, and then to between the NAB and SoundExchange. See supra. zone of reasonableness, the Judges presume that the expressly adopt the lowest rate from 201 See discussion supra, section I.A.3 proposed rate falls below that upper bound.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00076 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26391

radio) and simulcasters. Id. ¶¶ 86–87. the upper and lower bounds of his zone SOUNDEXCHANGE PROPOSED PER- The first adjustment (the ‘‘music- of reasonableness are not market rates. PERFORMANCE RATES FOR PER- listening adjustment’’) accounted for the There is no market for licensing of FORMANCES ABOVE 159,140 ATH fact that music accounts for a lower sound recordings for transmission by percentage of listening on AM/FM radio terrestrial radio stations, since there is Per- than on satellite radio. The second no general public performance right for Year performance adjustment (the ‘‘music-revenue sound recordings. That would be rate adjustment’’) accounted for ‘‘the fact sufficient reason to reject Dr. Katz’s 2016 ...... $0.0025 that non-music-formatted stations proposed lower bound of ‘‘near zero’’ 2017 ...... 0.0026 generally will not be paying royalties.’’ that he derived from terrestrial radio. 2018 ...... 0.0027 Id. ¶ 89. Moreover, Dr. Katz relies on an 2019 ...... 0.0028 The net effect of the two adjustments assumption that the promotional effect 2020 ...... 0.0029 essentially offset each other, resulting in of simulcasting is essentially the same an adjustment factor of one. Id. ¶ 92. as the promotional effect of terrestrial Id. at 4–5. These are the same per- Consequently, Dr. Katz determined that broadcasting, because they carry the performance rates the SoundExchange the upper bound to his zone of same content. As discussed above, proposes for commercial webcasters. reasonableness is 13 percent of gross broadcasters’ use of technologies to 2. NRBNMLC simulcasting revenues. Nevertheless, he substitute songs in their simulcast argues ‘‘there are strong reasons to streams destroys the underlying premise The NRBNMLC proposes what it conclude that the actual upper bound of that the content of a simulcast stream is describes as a ‘‘tiered and capped flat the zone of reasonableness is the same as the terrestrial broadcast. fee structure.’’ NRBNMLC PFF ¶ 80. significantly lower than 13 Even if the content is the same, the Under the NRBNMLC proposal, each percent.’’ 203Id. ¶ 93. Judges do not find sufficient persuasive noncommercial webcaster would pay a Dr. Katz’s approach is similar in some evidence supporting the conclusion that $500 annual fee for all performances of respects to the approach that the Judges simulcasts have the same promotional sound recordings up to a threshold of took (and the Court of Appeals affirmed) effect as terrestrial broadcasts.204 400 average concurrent listeners (3,504,000 ATH) annually, and an in SDARS II. In that case, the zone of As for Dr. Katz’s use of the SDARS II additional $200 for each additional 100 reasonableness that the Judges rate to establish an upper bound to his average concurrent listeners (876,000 determined based on the parties’ zone of reasonableness, that too is not ATH) annually, up to an annual fee cap benchmarks was extremely broad. In a market rate. It is a rate established by of $1,500 per station or channel. See order to narrow down the possible rates the government by means of a CRB Introductory Memorandum to Written within that zone, the Judges referred to proceeding. Moreover, it is not even a Direct Statement of NRBNMLC at 3 several ‘‘guide posts,’’ including the 13 rate that is intended to replicate market (October 7, 2014) (NRBNMLC percent rate that had been the basis for conditions. It is a section 801(b) Introduction); The NRBNMLC’s the rate that the Judges set in SDARS I. reasonable rate, albeit one that was Proposed Noncommercial Webcaster SDARS II, however, is distinguishable informed by marketplace evidence Rates and Terms at 3 (October 7, 2014) from the present case. In SDARS II the (though from a somewhat different (NRBNMLC Proposed Rates and Terms). Judges had little confidence in the market). In short, neither end of Dr. The NRBNMLC would define ATH to benchmark analyses offered by the Katz’s zone of reasonableness is include only transmissions of recorded parties which, in any event, yielded a anchored in the noninteractive music. Id. at 1. range of possible rates that was too streaming market that the Judges are broad to provide useful guidance to the seeking to replicate in this proceeding. 3. IBS and Harvard Broadcasting/WHRB Judges. Thus the Judges found it The Judges find Dr. Katz’s zone of necessary to consider other available Section 351.4 of the Judges’ reasonableness unhelpful in setting a procedural rules sets forth the required evidence as guideposts. In the instant rate for commercial webcasters, and case, the Judges have sufficient contents of a participant’s WDS, reject the NAB’s proposed rate that it including the requirement that, in a rate confidence in the available benchmark derived from Dr. Katz’s analysis. analyses to proceed without reference to proceeding, ‘‘each party must state its other guideposts. V. Judges’ Determination of requested rate.’’ 37 CFR 351.4(b)(3) In SDARS II, the Judges were not Noncommercial Webcasting Rates (required contents of WDS). The rule goes on to permit participants to revise determining a market rate under the A. Parties’ Proposals willing-buyer, willing-seller standard. their rate proposals at any time up to the The Judges decided SDARS II under the 1. SoundExchange filing of proposed findings of fact and section 801(b) reasonable rate standard. conclusions of law. Id. SoundExchange proposes that IBS’s WDS does not contain a rate As the Court of Appeals emphasized, noncommercial webcasters pay a flat under that standard ‘‘[t]he Copyright proposal, or anything that the Judges annual fee of $500 per station or could reasonably interpret as a rate Act permits, but does not require, the channel for all performances up to a cap Judges to use market rates to help proposal. It consists solely of the three- of 159,140 ATH per month. page written testimony of Frederick determine reasonable rates.’’ Music SoundExchange Rate Proposal at 4 Choice v. Copyright Royalty Bd., 774 Kass. Captain Kass introduces himself (October 7, 2014) SoundExchange and IBS, and briefly discusses the F.3d 1000, 1010 (D.C. Cir. 2014). That is proposes that, in any month that a not the case under § 114(f)(2)(B). The nature of IBS members’ webcasting noncommercial webcaster exceeds activities: Judges must determine market rates, yet 159,140 ATH, the webcaster pay per- the rates used by Dr. Katz to determine performance royalties at the following [IBS member] stations operate as non-profit entities within the meaning of the statute, as rates for its transmissions in excess of amended. They use digitally recorded music 203 Professor Katz’s primary argument that the 13 159,140 ATH: percent figure is too high is that it was derived in as instructional media for announcers and SDARS I from analysis of a market that was not programmers. The instantaneous listenership effectively competitive. Id. 204 See discussion, supra section III.A.3.c.v. to such music on member stations is

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00077 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26392 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

typically on the order of five listeners, with propose a royalty rate.207 In short, the noncommercial counterparts.’’ Id. at the exception of course-related music and only arguable reference by IBS to a rate 24097–98. To avoid this convergence or other on-campus events. In contrast, proposal was made by counsel in his overlap, the Judges adopted a cap on the audiences for live sports broadcast live closing arguments. The Judges do not size (as measured by audience size) of musical performances, and lectures and other credit this statement by counsel as a rate noncommercial webcasting stations or live on-campus originations are typically much larger than the audience for digitally proposal by IBS for three reasons. First, channels that are eligible for the recorded music. introducing a rate proposal for the first noncommercial rate. See 37 CFR IBS Members provide significant science, time in closing arguments does not 380.3(a)(2) (applying flat $500 royalty technology, engineering, management, comply with the Judges’ rules and is rate up to 159,140 ATH per month).208 media, and communication skill set training. grossly unfair to the other parties. SoundExchange’s proposal to The stations typically act as learning Section 351.4(b)(3) is extremely liberal continue to impose of a limit on the size laboratories where students may learn and regarding revisions to a party’s rate of noncommercial webcasters that are perfect their skills. proposal, but it presupposes that the eligible for a separate noncommercial IBS Ex. 9000 at 3 (Kass WDT). party has made a proposal as part of its rate is supported by the testimony of WDS, thus giving the other parties an Professor Thomas Lys. Professor Lys Similarly, WHRB’s WDS does not opportunity to analyze it prior to noted that, as a matter of economic contain a rate proposal, or anything the presenting their rebuttal evidence. logic, ‘‘there is no real difference Judges could reasonably interpret as a Second, ‘‘around $20 a year’’ is not between a noncommercial and a rate proposal. WHRB’s WDS is sufficiently definite or specific to commercial broadcaster.’’ SX Ex. 28 comprised of the WDT of Michael constitute a rate proposal. For example, ¶ 256 (Lys WRT); 5/29/15 Tr. at 6738. Papish, one of the station’s board which noncommercial webcasters The Judges credit this testimony, but do members. In three pages of written would pay ‘‘around $20 a year’’? All of not reach precisely the same ultimate testimony, Mr. Papish merely them? Only ones that transmit below a conclusion as Professor Lys. While introduces himself and describes certain ATH threshold? What threshold? Professor Lys apparently argues that WHRB’s operations. See generally IBS does not say. there should be no distinction between WHRB Ex. 8000 (Papish WDT). Third, even if the Judges were to commercial and noncommercial rates, Neither Captain Kass nor Mr. Papish consider this to be a proposal, IBS has he did not consider (and was apparently presented a rate proposal in the course offered only statements of counsel to unaware of) the revealed preference in of their respective live testimony at the support it. The record is devoid of any the marketplace for a separate hearing. The only hint of a proposal evidence to support IBS’s ‘‘proposal’’ or noncommercial rate. The Judges resolve might be gleaned from a colloquy the analysis from which it was the tension between Professor Lys’s between the Judges and counsel 205 purportedly derived. Nothing will come observation concerning economic logic during closing arguments: of nothing. Neither IBS nor WHRB has and the revealed preference in the [THE JUDGES]: So what exactly is IBS offered a rate proposal that the Judges marketplace by limiting the differential proposing here? can consider in this proceeding. treatment of noncommercial webcasters MR. MALONE: All right. In our pleadings B. Analysis and Conclusions to smaller players that have a as early as the agreement between correspondingly smaller impact on the SoundExchange and CPB, NPR became 1. Upper Threshold for Noncommercial commercial market. The Judges thus public when you published it in the Federal Rate agree with SoundExchange that Register, we have computed to the best of our The Judges have recognized eligibility for a noncommercial rate ability that there is a rate per ATH of should be limited to those 0.0011940. And we think that this is a noncommercial webcasting as a separate marketplace agreement entered into submarket in prior decisions only ‘‘up noncommercial webcasters whose voluntarily by one of the big companies in to a point.’’ Web II Original audience size falls below a fixed the market, and we think that sets the Determination at 24097. The Judges threshold. appropriate rate. stressed that there must be limits to the While SoundExchange proposes a Then when you scale that down to show differential treatment for threshold above which a the number of ATH that these college noncommercials to avoid ‘‘the chance noncommercial webcaster ceases to be stations, high school stations, academy that small noncommercial stations will eligible for a noncommercial rate, the stations, and the like are operating, it works NRBNMLC does not. The NRBNMLC out to around $20 a year. cannibalize the webcasting market more generally and thereby adversely affect does, however, propose a threshold 7/21/15 Tr. at 7949 (Kass). the value of the digital performance above which a noncommercial webcaster must pay an additional flat In its proposed findings, IBS directed right in sound recordings.’’ Id. (internal royalty fee (this structure is described its efforts to arguing against adoption of quotes and citations omitted). The supra, section V.A.2). Under either the SoundExchange/CBI settlement Judges concluded that any separate rate proposal a flat fee of $500 pays for all agreement 206 and, once again, failed to for noncommercial webcasters must ‘‘include safeguards to assure that, as performances of sound recordings up to the threshold. 205 the submarket for noncommercial William Malone, Esq., jointly represented IBS SoundExchange proposes that the and WHRB in this proceeding. In closing arguments webcasters that can be distinguished Mr. Malone, on behalf of WHRB, briefly discussed from commercial webcasters evolves, it threshold remain the same as the a matter related to terms. 7/21/15 Tr. at 7946. The does not simply converge or overlap 208 remainder of his closing argument, including the with the submarket for commercial Although the Judges and the parties discuss colloquy quoted in the text, was apparently on the ATH threshold as a ‘‘cap’’ on eligibility for a behalf of IBS alone. webcasters and their indistinguishable reduced noncommercial rate, this is not entirely 206 Those efforts were both untimely and not in accurate. A noncommercial webcaster that exceeds accordance with the procedures established in the 207 IBS goes through a series of computations in the cap in any given month does not pay Act, the Judges’ rules for submitting comments on its PFF in an effort to show that the proposed commercial rates for all of its transmissions in that a proposed settlement, and the Judges’ Federal settlement rates ‘‘in no way meet the comparability month, but only those beyond the cap. This results Register notice. See 17 U.S.C. 801(b)(7)(A); 37 CFR test for noncommercial royalty rates.’’ IBS PFF, at in noncommercial webcasters paying a lower 351.2(b)(2); 79 FR 65609 (November 5, 2014) 10. In the course of those computations, IBS comes average per-play rate than a commercial webcaster (SoundExchange/CBI agreement); 80 FR 15958 up with a $20/year figure, but it is unclear what that (that pays at the commercial rate for every (March 26, 2015) (SoundExchange/NPR agreement). figure represents. Id. performance).

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00078 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26393

current threshold for noncommercial audience growth. See Emert WDT ¶ 29; The NRBNMLC’s proposal to increase webcasters: 159,140 ATH per month Ex. 7010; 5/21/15 Tr. at 5275–77 the threshold to 400 concurrent listeners (218 concurrent listeners, on average, (Henes). Their stations could achieve is unsupported by the record. By for a webcaster that transmits 24 hours significant audience growth under contrast, the evidence demonstrates that per day). 307 CFR 380.3(a)(2). That is SoundExchange’s proposed rate the current threshold of 159,140 ATH also the threshold in the structure without subjecting themselves per month that SoundExchange SoundExchange/CBI settlement to additional royalty costs. proposes to retain has resulted, for the agreement above which a To the contrary, there is ample record vast majority of noncommercial noncommercial educational webcaster evidence to demonstrate that the vast webcasters, in no additional liability for (NEW) ceases to be eligible for the majority of noncommercial webcasters royalties beyond the minimum fee. settlement rate. See Digital Performance do not exceed the existing threshold. Moreover, the willingness of Right in Sound Recordings and SoundExchange payment data show that SoundExchange and CBI to adopt that Ephemeral Recordings: Proposed Rule, between 2010 and 2014, noncommercial threshold in their current settlement 79 FR 65609, 65611 (November 5, 2014) webcasters 210 paid usage fees 112 times agreement, after years of experience (proposed 37 CFR 380.22). By contrast, out of 3917 noncommercial webcaster with the identical threshold under the the NRBNMLC proposes a much higher payments (2.86%). NAB Ex. 4141; NAB current rates, demonstrates that it is threshold of 400 average concurrent Ex. 4149; see also SX Ex. 2 at 14 (Bender reasonable and workable. The Judges listeners, or 3,504,000 ATH annually WDT) (‘‘approximately 97% of hereby adopt it. 209 noncommercial webcasters paid only (292,000 ATH per month on average). 2. Consequences of Exceeding the The NRBNMLC argues that the [the] minimum fee’’). The NRBNMLC Threshold existing threshold should be increased seeks to counter this evidence with because it was originally established in testimony from Mr. Emert and Mr. SoundExchange proposes that a 2006 (based on 2004 survey data). Henes that they were ‘‘aware of’’ some noncommercial webcaster’s NRBNMLC PFF ¶ 143. In addition, the noncommercial broadcasters that transmissions beyond the 159,140 ATH NRBNMLC argues that an increase is impose listener caps on their simulcast threshold should no longer enjoy a necessary to provide noncommercial streams to avoid exceeding the existing reduced royalty rate. The NRBNMLC webcasters with ‘‘breathing room.’’ See threshold. Emert WDT, ¶ 38; 5/21/15 Tr. proposes that a reduced royalty rate, Emert WDT ¶ 40. These arguments are at 5271 (Henes). The NRBNMLC’s structured in $200 increments for each unpersuasive. evidence is vague and anecdotal. It was 876,000 ATH annually, should apply to While it is correct that the current not derived from the witnesses’ own transmissions beyond the threshold. experiences, but rather from something 159,140 ATH threshold was adopted a. The NRBNMLC’s Proposal originally in Web II based on survey they heard elsewhere. Even if the Judges evidence presented in that proceeding, were to deem this testimony credible, The Judges explained in Web II that that is not the only source for that the most that it reveals is the existence the threshold on the noncommercial number. See Web II, 72 FR at 24099. of some isolated instances of webcasting rate serves as a ‘‘proxy that SoundExchange and CBI adopted noncommercial webcasters that are aims to capture the characteristics that 159,140 ATH as the threshold in their constrained by the existing threshold. delineate the noncommercial settlement agreement, which is The testimony emphatically does not submarket.’’ Web II Remand, 72 FR at contemporaneous with this proceeding demonstrate that a substantial number 24099. As discussed in section V.B.1, and covers the same rate period. See of noncommercial webcasters are the Judges do this to assure that the NRBNMLC Ex. 7034, Attachment at 2– operating near the threshold and taking submarket for noncommercial 211 3 (SoundExchange/CBI Joint Motion to steps to keep below it. webcasters does not converge or overlap Adopt Partial Settlement). By contrast, with the submarket for commercial 210 the NRBNMLC cannot point to any These are webcasters that are coded ‘‘NCW– webcasters. SoundExchange’s proposal CRB’’ (noncommercial webcaster paying statutory is consistent with this rationale; the marketplace agreement rates), ‘‘NCW–WSA’’ (noncommercial webcaster (contemporaneous or otherwise) that paying WSA settlement rates) and ‘‘NCEDW’’ NRBNMLC’s is not. Not only would the employs the threshold it proposes. (noncommercial education webcaster paying under NRBNMLC’s proposal grant As to the NRBNMLC’s argument that the SoundExchange/CBI settlement) in the substantially reduced rates to large SoundExchange data. For purposes of this analysis, noncommercial webcasters whose noncommercial webcasters need the the Judges have excluded noncommercial ‘‘breathing room’’ that an increased microcasters which, by definition, stream far below operations compete with commercial threshold would provide, there is no the threshold and pay no usage fees. See webcasters’, but the effective rate for persuasive record evidence to support Noncommercial Microcasters, available online at such large noncommercial webcasters http://www.soundexchange.com/service-provider/ would actually decline as they grow that proposition. Mr. Emert did testify to noncommercial-webcaster/noncommercial- this effect. Emert WDT ¶ 39; see also microcaster-wsa/ (visited September 8, 2015). The larger due to the effect of the proposed 5/21/15 Tr. at 5271–71 (Henes). Judges consider a webcaster to be paying usage fees $1,500 cap on royalties. The NRBNMLC if the fees collected by SoundExchange in a offers no economic rationale for this However, that testimony was an particular year (a) exceed the $500 flat fee, (b) do expression of opinion, unsupported by not equal $600 (which most likely represents the result. See Lys WRT, ¶¶ 256–257. any factual evidence. Mr. Emert’s and $500 flat fee plus a $100 proxy fee in lieu of census The NRBNMLC does not address this Mr. Henes’ testimony that that the reporting) and (c) are not an even multiple of $500 issue directly. Instead, the NRBNMLC (most likely representing payment of the minimum dozen or so radio stations they operate argues that its proposed ‘‘tiered and fee for multiple channels). This is the approach that capped flat rate structure’’ is what a stream far below the existing threshold the NRBNMLC employed in interpreting these data. tends to contradict their statements See, e.g., NRBNMLC PFF ¶ 95. willing buyer and a willing seller would concerning the need to increase the 211 The NRBNMLC candidly admits that it does not know the extent to which noncommercial noncommercial webcaster community. While such threshold to accommodate future webcasters impose listener caps, noting that a survey may not have provided a definitive answer ‘‘[t]here is no way of knowing exactly how many for the entire population of noncommercial 209 This threshold effectively would be higher Noncommercial entities have done this . . . .’’ webcasters, it would have revealed far more about still as a result of the NRBNMLC’s proposal to NRBNMLC PFF ¶ 23. This statement is only the current state of affairs across the noncommercial exclude certain non-music intensive programming partially correct: The NRBNMLC could have webcasting market than the hearsay testimony of from the definition of ATH. surveyed its members or the broader these two witnesses.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00079 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26394 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

agree to in an effectively competitive SoundExchange’s rate proposal (filed not currently enjoy any legal right with market (i.e., a market rate). See with the Judges on the same day as the respect to the transmission of its sound NRBNMLC PFF ¶80. The NRBNMLC proposed settlement) would also recordings over terrestrial radio. There cited the testimony of its two witnesses contain usage fees. At most, the is no basis for the Judges to conclude as establishing the need of omission of usage fees from the that what the industry may be willing to noncommercial webcasters for rates that agreement reflected the parties’ decision accept in legislation that establishes are affordable and predictable. not to resolve the issue of what rates such a right is the same as what it would NRBNMLC Ex. 7011 ¶¶ 25–27, 30 would apply beyond the threshold, and bargain for in an arms-length transaction (Henes WDT); Emert WDT ¶¶ 31–32, to leave it for the Judges to determine against the backdrop of an existing 34–37, 41. The fatal flaw in this in the proceeding. statutory right of remuneration. argument is that it is unsupported by The NRBNMLC is correct in pointing b. SoundExchange’s Proposal any marketplace evidence and any out that the SoundExchange/NPR evidence of sellers who would be settlement agreement imposes a flat Although SoundExchange’s proposal willing to accept the NRBNMLC’s royalty rate with no additional usage to impose commercial rates above the proposed structure. Mr. Henes and Mr. fee. However, the SoundExchange/NPR 159,140 ATH threshold is consistent Emert may be willing, even eager to settlement differs so fundamentally in with the Judge’s rationale for limiting license music on this basis, but their so many ways from what the NRBNMLC the applicability of noncommercial testimony tells the Judges nothing about is proposing that it cannot serve as a rates, the NRBNMLC levels multiple the sellers’ side of the equation. As support for that proposal. The criticisms against it. These include: • discussed in greater detail in the SoundExchange/NPR settlement entails SoundExchange’s entire rate following paragraphs, none of the a single annual payment by a single proposal for noncommercial webcasters payer (CPB), in advance, to cover over lacks evidentiary support; marketplace evidence that the • NRBNMLC cites pertains to a rate 500 NPR member radio stations. 80 FR The specific usage rates that structure remotely similar to the one at 59590–91. The stations include a SoundExchange proposes are proposed by the NRBNMLC. range of formats, some of which entail ‘‘inappropriate for commercial As additional evidence to support very limited use of recorded music. webcasters and even more inappropriate Unlike the NRBNMLC’s rate proposal, for noncommercial webcasters’’; and their argument that a ‘‘tiered and • capped flat rate structure’’ is a market the settlement does not include tiered The fact that few noncommercial rate, the NRBNMLC cites the payments above the flat royalty rate, but webcasters have paid usage fees SoundExchange/CBI settlement does include a cap on the aggregate confirms that the proposed fees are agreement, the SoundExchange/NPR amount of recorded music that may be unreasonable. settlement agreement, the rates performed. NPR consolidates the reports NRBNMLC PFF ¶ 113. established for musical works under 17 of use for all of the stations covered by i. Evidentiary Support (or Lack Thereof) U.S.C. 118, and the position taken by the agreement. The NRBNMLC’s for SoundExchange’s Rate Proposal proposal does not provide for SoundExchange on legislation to create As Professor Rubinfeld readily a public performance right for sound consolidated reports of use. On the whole, the terms of the SoundExchange/ conceded, there are no current recordings that covers transmissions marketplace benchmarks from which to over terrestrial radio. Id. The Judges NPR agreement provide SoundExchange with significant benefits—reduced risk derive SoundExchange’s entire rate reach different conclusions based on proposal for noncommercial webcasters. this evidence. of nonpayment; protection against large numbers of uncompensated Rubinfeld CWDT ¶¶ 33, 246. The only The SoundExchange/CBI settlement contemporary agreements in evidence agreement imposes a flat $500 fee on performances; reduced costs of processing usage data—that the that cover noncommercial webcasters NEWs that transmit up to 159,140 ATH are the two settlement agreements per month. Any NEW that exceeds that NRBNMLC proposal does not. To pluck out a single element of the deal, the flat between SoundExchange, on the one threshold loses its eligibility to operate hand, and CBI and NPR, respectively, under the settlement, and thus becomes royalty rate, and cite it as support for the NRBNMLC rate proposal simply on the other hand. As discussed in the subject to the CRB rate for preceding section, there are a number of noncommercial webcasters for the lacks credibility. The musical works rate under the elements of the SoundExchange/NPR remainder of the year.212 The § 118 statutory license suffers from a agreement that render it a poor NRBNMLC concludes that ‘‘no usage similar lack of comparability to the rates benchmark for setting noncommercial fees apply under the agreement’’ for a the Judges must set in this proceeding. rates generally. The SoundExchange/ NEW that exceeds the threshold, and Rates under § 118 are in a different CBI agreement lends support for some cites the agreement as support for a flat- market, with different sellers and for elements of SoundExchange’s rate rate structure with no usage fees. different copyrighted works. The proposal (e.g., a flat $500 rate for NRBNMLC PFF ¶ 93. The NRBNMLC’s NRBNMLC has presented no evidence noncommercial webcasters that transmit interpretation of the agreement is not to demonstrate how a rate structure in up to 159,140 ATH), but not for the credible. The parties’ decision not to that market, and with those sellers, proposed rate for usage beyond the ATH specify usage fees in the agreement does reflects what a willing buyer and a threshold. not mean that they contemplated that a willing seller would agree to in the That does not mean, however, that NEW that exceeded the ATH threshold sound recordings market. SoundExchange’s rate proposal is would not pay any usage fees. The Finally, SoundExchange’s position on entirely without evidentiary support. As existing CRB rates provide for usage fees legislation has little or no bearing on discussed, supra section V.B.1, expert above 159,140 ATH, and CBI could what constitutes a market rate. The economic testimony supports treating reasonably assume that compromises and tradeoffs that parties transmissions by noncommercial are prepared to make in the legislative webcasters above a certain ATH 212 The NEW may operate under the settlement in the following year, provided it takes affirmative arena have only the remotest threshold the same as transmissions by steps (e.g., imposes listening caps) to ensure that it resemblance to the give and take of the commercial webcasters. This is what the will not exceed the threshold again. marketplace. The record industry does SoundExchange proposal seeks to

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00080 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26395

achieve. The rates that SoundExchange 2099–100 (Rubinfeld) (25–30 willingly license unlimited use of its proposes for transmissions above the noncommercial licensees pay lower sound recordings for a fixed fee of ATH threshold are the same that rates under settlement agreements). $1,500. The Judges reject the SoundExchange proposes for These facts do no support the NRBNMLC’s proposed royalty cap. commercial webcasters. NRBNMLC’s conclusions. In itself, the 4. IBS’s Additional Arguments fact that more than 97% of ii. Inappropriateness of Specific Usage noncommercial webcaster payments do IBS did not direct any criticism Rates Proposed by SoundExchange not include usage fees could just as directly at either the SoundExchange or The NRBNMLC pursues two lines of easily support the conclusion that the the NRBNMLC rate proposal. IBS’s rate- attack against the specific usage rates vast majority of noncommercial related arguments were directed (or that SoundExchange proposes. The first, webcasters—like the noncommercial misdirected 214) at the SoundExchange/ concerning Professor Rubinfeld’s webcasters that testified in this CBI settlement agreement. Nevertheless, interactive benchmark analysis, proceeding—operate well below the had IBS applied those arguments to the essentially repeats the licensee services’ 159,140 ATH threshold. Without rate proposals before the Judges, the criticisms of SoundExchange’s proposal evidence that a substantial number of Judges would have rejected them. for commercial webcasting rates. See noncommercial webcasters are a. Lobbying Prohibition NRBNMLC PFF ¶ 122. The Judges operating near the threshold, or are discuss those arguments supra. The imposing listening caps, the Judges Captain Kass testified that many IBS Judges, in fact, do not adopt the specific cannot conclude that the threshold members are a part of state-funded rates that SoundExchange proposes, operates as a significant constraint or educational institutions that are barred precisely because they find that the usage fees are excessive. by state law from providing funds to SoundExchange’s benchmark analysis The evidence that most organizations that lobby. IBS argues that lacking in certain respects. Rather, the noncommercial webcasters that paid these laws prevent certain IBS members Judges adopt the same rates for usage fees did so under an alternative from paying royalties to transmissions in excess of the 159,140 rate structure also does not support the SoundExchange. ATH threshold by noncommercial NRBNMLC’s conclusions. These This argument is unavailing for webcasters as they do for commercial webcasters made a rational choice to several reasons. First, IBS failed to webcasters. pay an available lower rate. That tells provide any legal authority or expert The second line of attack is that the Judges nothing about their testimony to support Captain Kass’s Professor Rubinfeld’s benchmark willingness to pay the higher statutory interpretation of these state laws. Even analysis is inapplicable to rate in the absence of settlement. if the Judges accept as true the assertion noncommercial webcasters because Conversely, though, it strongly suggests that these state laws prohibit certain IBS none of the licensees under any of the that nearly all of the webcasters that members from remitting funds to benchmark agreements were opted for the statutory rate structure or lobbying organizations, it is far from noncommercial webcasters. Id. ¶ 123. the NEW settlement expected that they clear whether those laws would prevent As discussed, supra section V.B.1, the would not exceed the threshold. the same IBS members from paying Judges apply commercial rates to statutory license royalties to an 3. Cap on Royalties noncommercial webcasters above the organization designated by regulation as ATH threshold because economic logic The NRBNMLC proposes that the total a collective under a Federal statute. dictates that outcome, not because it obligation of a noncommercial Second, there is no evidence in the was observed in benchmark agreements. webcaster to pay royalties should be record concerning SoundExchange’s iii. Small Number of Noncommercial capped at $1,500, regardless of the lobbying activities, vel non. The Judges Webcasters Paying Usage Fees Confirms number of sound recordings the have no basis for concluding that the That the Fees Are Excessive webcaster performs. As with the other state laws to which IBS refers even elements of its rate proposal, the apply to SoundExchange. The NRBNMLC notes that few NRBNMLC contends that the cap on Third, and most fundamentally, the noncommercial webcasters pay usage fees is supported by marketplace fees and, of those that do, most pay a entire question is not relevant to the evidence. Neither of the two Judges’ task of setting rates for lower settlement rate in lieu of the rates noncommercial agreements in evidence set by the Judges for commercial noncommercial webcasters. If IBS employs the cap that the NRBNMLC contends that its members may webcast webcasters. NRBNMLC PFF ¶ 131. proposes. The SoundExchange/CBI Based on this evidence, the NRBNMLC sound recordings but are forbidden settlement imposes a flat royalty rate, under state law to pay royalties to concludes that the commercial but caps eligibility for that rate at webcaster rates are excessive, and that SoundExchange, that is an argument 159,140 ATH. Beyond that threshold, that must be resolved by a Federal noncommercial webcasters are imposing the noncommercial webcaster must pay listener caps or taking other affirmative District Court in an infringement action. under the noncommercial rate structure It has no bearing on the particular rate steps to avoid paying them. determined in this proceeding. The Of the 3,917 documented payments structure that the Judges must determine SoundExchange/NPR settlement for noncommercial webcasters. by noncommercial webcasters between agreement employs a flat-fee structure 2010 and 2014, 112 included payments (which serves as a cap on royalties), but b. Lack of ‘‘Proportionality’’ for usage above the ATH threshold. also imposes a cap on music usage. See IBS argues that royalty payments for NAB Ex. 4141; NAB Ex. 4149. Of these, 80 FR at 15961. 13 were at the commercial rate noncommercial webcasters must be There is no other evidence of any proportional to their use of sound determined by the Judges and 99 were kind that a copyright owner would at a lower rate established under a WSA recordings. While IBS’s argument has a settlement.213 Id.; see also 5/6/15 Tr. at superficial appeal, it suffers from Judges are not permitted to take into account the several shortcomings. rate structure, fees, terms and conditions of that 213 The noncommercial webcasters’ WSA agreement in setting rates in this proceeding. 17 settlement agreement is ‘‘nonprecedential.’’ The U.S.C. 114(f)(5)(C). 214 See supra note 204.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00081 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26396 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

IBS does not and cannot cite any A. Commercial Webcasters for a webcaster with 100 or more statutory authority for its argument. The stations or channels. Id. 1. Parties’ Proposals statute directs the Judges to set willing 2. Analysis and Conclusion buyer/willing seller rates.17 U.S.C. a. SoundExchange All participants that proposed a 114(f)(2)(B). Willing buyers and willing SoundExchange proposes a $500 per sellers may, and often do, agree to rates minimum fee for commercial webcasters station or channel annual minimum fee. asked the Judges to retain the current that are not strictly proportional to The minimum fee would be annual minimum fee that the Judges usage. The SoundExchange/NPR and nonrefundable, but would be credited adopted in Web III pursuant to a SoundExchange/CBI agreements are against royalties incurred during the settlement. See Web III Remand examples of agreements that incorporate applicable year. The minimum fee Decision, 79 FR at 23104. The minimum a flat-rate structure where royalties are would be capped at $50,000 annually fee settlement in Web III kept in place not strictly proportional to use. for a webcaster with 100 or more a settlement of the minimum fee for The statutory requirement of a stations or channels. SoundExchange commercial webcasters that the parties minimum fee also runs counter to IBS’s Rate Proposal at 2 (October 7, 2014). reached in Web II. See Digital argument. By definition, a minimum fee b. Pandora Performance Right in Sound Recordings and Ephemeral Recordings, final rule, (whatever its level) is not proportional Pandora does not make an explicit to usage. 75 FR 6097 (February 8, 2010) (Web II proposal for a minimum fee. Pandora Minimum Fee Settlement). That IBS also fails utterly to provide any does, however, propose that, apart from settlement, in turn, retained a $500 evidentiary basis for concluding that the those terms for which it proposes minimum fee that was determined by a rates proposed by SoundExchange or changes, ‘‘the terms currently set forth CARP, and upheld by the Librarian, in the NRBNMLC are so disproportional to in 37 CFR 380 be continued.’’ Proposed Web I, see Determination of Reasonable noncommercial webcasters’ usage as to Rates and Terms of Pandora at 2 (Oct. Rates and Terms for the Digital be unreasonable. To be sure, some 7, 2015). Those terms include the Performance of Sound Recordings and noncommercial webcasters transmit a current minimum fee of $500 per station Ephemeral Recordings, Final Rule and very small number of performances of or channel (capped at $50,000) for Order, 67 FR 45240, 45262–63 (July 8, recorded music. See Kass WDT at 3 commercial webcasters. 2002), but added a $50,000 cap for a (‘‘instantaneous listenership to such c. iHeartMedia webcaster with 100 or more stations or music on member stations is typically channels. See Web II Minimum Fee on the order of five listeners, with the iHeartMedia does not propose a Settlement, 75 FR at 6098. minimum fee. exception of course-related music While there is no settlement of the . . .’’). Noncommercial webcasters— d. Sirius XM minimum fee issue in the current proceeding, the convergence of the even those that are IBS members—are a Sirius XM does not make an explicit heterogeneous group, with some parties’ proposals on the existing $500 proposal for a minimum fee. Sirius XM minimum fee (capped at $50,000) operating above SoundExchange’s does, however, propose that ‘‘other than counsels strongly in favor of its proposed 159,140 ATH threshold. See the royalty rate, the terms currently retention. In addition, the Judges follow supra, section V.B.1. IBS has not even applicable to commercial webcasters be their earlier determination that proposed, much less provided an retained in their current form.’’ commercial and noncommercial evidentiary basis to adopt, subcategories Introductory Memorandum to the webcasters alike should have to pay a of noncommercial webcasters. Written Direct Statement of Sirius XM at minimum fee that at least defrays a 1–2 (Oct. 7, 2014). Those terms C. Conclusion portion of SoundExchange’s costs to presumably include the current administer the statutory licenses. See For the rate period 2016–2020 the minimum fee of $500 per station or Digital Performance Right in Sound Judges adopt an annual rate of $500 per channel (capped at $50,000) for Recordings and Ephemeral Recordings, station or channel for all transmissions commercial webcasters. Final Determination after Second by noncommercial webcasters up to a e. NAB Remand, 79 FR 64669, 64672 (Oct. 31, threshold of 159,140 ATH. For 2014). Mr. Jonathan Bender, transmissions in excess of 159,140 ATH, NAB proposes a $500 annual SoundExchange’s Chief Operating minimum fee for each terrestrial AM or noncommercial webcasters shall pay Officer, testified that ‘‘SoundExchange FM radio station that a broadcaster royalties for 2016 at the commercial rate does not track its administrative costs webcasts. For purposes of calculating (i.e., $0.0017 per-performance), and for on a licensee-by-licensee, station-by- the minimum fee, each individual station, or channel-by-channel basis such transmissions in excess of 159,140 stream (e.g., primary radio station, HD and, as a result, there is no precise way ATH in the remainder of the statutory multicast radio side channels, different to determine exactly’’ how much term, at the commercial rate as adjusted stations owned by a single licensee) is SoundExchange spends on that basis. annually for changes in the Consumer to be counted as a separate radio station, Bender WDT at 16–17. The costs to Price Index, as set forth in the except that identical streams for SoundExchange vary depending on regulations. simulcast stations will be treated as a such factors as the quality of the data a VI. Minimum Fee single stream if the streams are available service submits. Id. at 16. In 2013, the at a single Uniform Resource Locator average administrative costs per Sections 112 and 114 of the Act (URL). NAB Proposed Rates and Terms licensee (i.e., the total administrative require the Judges to establish minimum at 4. costs divided by the number of fees as part of any rate structure under The minimum fee would be licensees) were $11,778. Id. at 17. the respective statutory licenses. 17 nonrefundable, but would be credited SoundExchange’s average U.S.C. 112(e)(3)–(4) and 114(f)(2)(A)– against royalties incurred during the administrative cost per licensee is (B). applicable year. The minimum fee substantially higher than the minimum would be capped at $50,000 annually fee it proposes to charge each licensee.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00082 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26397

While a higher minimum fee could be FR 58201, 58206 (Sept. 28, 2015) (37 [REDACTED] and [REDACTED] for the justified on this record, no party has CFR 380.22(a)). [REDACTED] service). requested anything higher than the Although WHRB and IBS do not SoundExchange contends that no current level of $500. attack the SoundExchange and participant offered evidence of a The current $500 minimum fee for NRBNMLC minimum fee proposals benchmark agreement that does not commercial webcasters has been in directly, they argued against adoption of bundle performance rights and the right force for more than a dozen years,215 the SoundExchange/CBI settlement to make ephemeral copies. SX PFFCL and has been voluntarily re-adopted by which incorporates the same $500 ¶ 1375. SoundExchange further licensors and licensees on two minimum fee, and they repeat those contends that ‘‘[a]s of the Web III occasions. It has been proposed by arguments in this proceeding. The proceeding, recording artists and record licensors and licensees in this Judges addressed their objections to the companies had reached an agreement proceeding. SoundExchange’s SoundExchange/CBI settlement in the that five percent of the ‘payments for administrative costs (which the Federal Register notice adopting the activities under Section 112(e) and 114 minimum fee is intended to defray, in settlement terms. See id. at 58203–04. should be allocated to Section 112(e) part) exceed the proposed minimum fee The Judges have also addressed WHRB’s activities.’ ’’ SX PFFCL ¶ 1377, quoting by a wide margin. The Judges find the and IBS’s objections in the context of Dr. Ford. According to SoundExchange, proposed minimum fee (including the the SoundExchange and NRBNMLC rate no participant has presented evidence $50,000 cap) to be reasonable and proposals. For the same reasons in support of a different allocation supported by record evidence, and will articulated in the Federal Register between artists and record companies. therefore adopt it. notice and supra, section V.B.4, the SX PFFCL ¶ 1377. SoundExchange Judges reject WHRB’s and IBS’s concludes that ‘‘[b]ecause B. Noncommercial Webcasters objections as they may apply to the SoundExchange’s Board represents both 1. Parties’ Proposals proposed minimum fee for artists and copyright owners, its noncommercial webcasters. proposed rate of 5% for ephemeral a. SoundExchange copies is appropriate evidence and The current $500 annual minimum ‘credibly represents the result that SoundExchange proposes a $500 per fee for noncommercial webcasters has would in fact obtain in a hypothetical station or channel annual minimum fee been in force since Web I. See 37 CFR for noncommercial webcasters. The marketplace negotiation between a 261.3(e)(1) (2003). It was adopted by willing buyer and the interested willing minimum fee would be nonrefundable, SoundExchange and CBI in a settlement but would be credited against royalties sellers under the relevant constraints.’ ’’ agreement covering the rate period of SX PFFCL ¶ 1378, quoting Dr. Ford. incurred during the applicable year. this proceeding. It has been proposed by Other participants that address the SoundExchange Rate Proposal at 4. SoundExchange and the NRBNMLC, the rate for the Section 112 license do not b. NRBNMLC only noncommercial webcaster to file a contradict SoundExchange’s assertions. rate proposal in this proceeding. It See iHeart Reply PFFCL at 203 NRBNMLC proposes a $500 per constitutes a small (but nontrivial) (‘‘iHeartMedia supports the current station or channel annual minimum fee. fraction of the costs that bundling of the § 112 and § 114 The minimum fee would be SoundExchange incurs in administering royalties’’); Sirius XM PFF ¶ 2 (‘‘Sirius nonrefundable, but would be credited the statutory license. The Judges find XM maintains that the Section 112 against royalties incurred during the the proposed minimum fee to be ephemeral license has no value applicable year. reasonable and supported by record independent of the Section 114 c. IBS and WHRB evidence, and will therefore adopt it. performance license, and consequently VII. Ephemeral License Rate and Terms proposed that the royalty for the Section As discussed supra, IBS and WHRB 112 license be deemed included within did not submit rate proposals. Section 112(e) grants entities that the Section 114 royalty payment. Sirius transmit performances of sound 2. Analysis and Conclusion XM takes no position at this time as to recordings a statutory license to make what, if any, percentage of the Section Both the SoundExchange and ephemeral recordings. SoundExchange 114 royalty should be deemed attributed NRBNMLC rate proposals include a proposes that the Judges bundle the to the Section 112 ephemeral license.’’); $500 annual per station or channel royalties for Section 114 and 112 and NRBNMLC PFFCL ¶ 151 (‘‘[t]here is no minimum fee for noncommercial allocate five percent (5%) of the Section dispute between SoundExchange and webcasters—i.e., retention of the current 114 performance right royalty deposits the NRBNMLC regarding how the minimum fee. No other participant to the Section 112(e) ephemeral royalties for the ephemeral recording proposed a minimum fee for recording right, a rate structure that statutory license specified in 17 U.S.C. noncommercial webcasters,216 although would continue the extant arrangement. 112(e) should be set. Both participants CBI and SoundExchange agreed to SX PFFCL ¶ 1369. SoundExchange propose that those royalties for retain the existing $500 minimum fee as contends that its proposal regarding the ephemeral reproductions used solely to part of their settlement covering bundled rate for the Section 112 license facilitate transmissions made pursuant noncommercial educational is supported by the designated to the 17 U.S.C. 114(f) statutory license broadcasters. See Digital Performance testimony of Dr. Ford. SX PFFCL at be deemed to be ‘included within, and Right in Sound Recordings and 1370 & n.64. SoundExchange also cites constitute 5% of’ the § 114(f) statutory Ephemeral Recordings, Final Rule, 80 as support for its Section 112 proposal license payments made by a particular certain license agreements that were service’’ quoting the respective 215 The $50,000 cap has been in force since 2010 introduced into evidence. SX PFFCL proposals of SoundExchange and (applicable to the rate period beginning January 1, ¶ 1374 (citing agreements between NRBNMLC); NAB PFFCL ¶ 226 (‘‘no 2006). [REDACTED] and [REDACTED], dispute between SoundExchange and 216 As noted supra, neither of the other two noncommercial webcasters that participated in this [REDACTED] agreements with NAB regarding how the royalties for the proceeding (WHRB and IBS) submitted a rate [REDACTED] and [REDACTED], [Section 112(e) license] should be set.’’) proposal. [REDACTED]’s agreements with and Pandora PFFCL ¶ 416 (‘‘[c]onsistent

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00083 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26398 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

with past proceedings and the Merlin current amended regulation: ‘‘A thereto, the Judges issued a separate Agreement (which has no separate Licensee is authorized to make more order detailing SoundExchange’s ephemeral recording fee), Pandora than one Ephemeral Recording of a requests and the Judges’ conclusions.220 proposes that the royalty payable for sound recording as it deems necessary In the interest of making this final ephemeral recordings be included to make noninteractive digital audio Determination a complete and cohesive within the Section 114 royalty. There is transmissions pursuant to 17 U.S.C. record of the Judges’ findings and no dispute on this point: 114.’’ Pandora PFFCL ¶ 417. Pandora conclusions in this proceeding, the SoundExchange has proposed the contends that such ‘‘as necessary’’ Judges include additional material in same.’’). language is consistent with industry this section to reflect their rehearing The Judges accept SoundExchange’s practice. Id. ¶ 418. SoundExchange ruling. proposal to continue the current proposes that the current regulation be In addition to the proposed terms bundling of the Section 112 and 114 carried over into the new rate period but concerning licensing ephemeral rates. The Judges find persuasive the appears to acknowledge that authorizing recordings discussed in the preceding designated testimony of Dr. Ford and the making of more than one ephemeral section of this Determination, the Judges the license agreements that copy is not inconsistent with current have weighed the proposals and the SoundExchange cites in its PFFCL that industry practice.218 arguments of the parties in support of or willing buyers and willing sellers would The Judges adopt Pandora’s proposed opposed to various regulatory prefer that the rates for the two licenses language and do not carry forward the provisions and, after due consideration be bundled and that they would be language ‘‘for which it pays royalties’’ in of the rehearing papers, adopt the Terms agnostic with respect to the allocation of the current regulation because they as detailed below this Supplementary those rates to the Section 112 and 114 believe that the phrase could be Information section. The parties’ license holders.217 The Judges also find construed in a way that would limit the proposals—and the Judges’ rulings— that the minimum fee for the Section application of the Section 112 license to include the following.221 112 license should be subsumed under certain transmissions made consistent the minimum fee for the Section 114 with Section 114 that are not royalty A. Section 380.1—Scope and license, 5% of which shall be allocable generating, such as skips. The Judges Compliance to the Section 112 license holders, with also are sympathetic to the Services’ 1. Legal Compliance—§ 380.1(c) the remaining 95% allocated to the contention that, in certain Section 114 license holders. circumstances (e.g., where different file a. Sound Recording Performance SoundExchange and the services format requirements may necessitate the Complement disagree, however, on the terms with creation of multiple copies), it may be iHeart proposed changes to the respect to the Section 112(e) license. necessary to make more than one statutory definition of ‘‘sound recording CRB Rule 380.3(c), which addresses ephemeral copy to facilitate performance complement’’ to reflect the ephemeral recordings, states: ‘‘The transmissions made pursuant to Section practice of waiving the statutory royalty payable under 17 U.S.C. 112(e) 114. Nevertheless, the circumstances performance complement in private for the making of Ephemeral Recordings must be necessary and commercially agreements, IHM PFF ¶ 425. The used by the Licensee solely to facilitate reasonable. The language the Judges provision would ‘‘ensure[ ] that transmissions for which it pays royalties adopt includes this standard. Broadcasters do not need to alter the shall be included within, and constitute content of their radio broadcasts simply VIII. Terms 5% of, the total royalties payable under because they have elected to simulcast 17 U.S.C. 112(e) and 114.’’ 37 CFR One of the purposes of this those broadcasts over the Internet’’. IHM 380.3(c), emphasis added. proceeding is to establish terms for the Rate and Terms Proposal at 2–3. Pandora proposes that the Judges administration of the rates the Judges According to iHeart, because programs strike the italicized language and determine for the rate period 2016 to on terrestrial radio stations can play replace it with the phrase ‘‘made 2020. The parties proposed changes to entire albums, iHeart should be allowed pursuant to 17 U.S.C. 114.’’ Pandora Subchapter E of Chapter III, title 37 to simulcast the programs without believes the current language ‘‘creates CFR, relating to royalty rates and terms. altering them to satisfy the performance the possibility (likely unintended) that The Judges adopted some changes and complement requirement, and the ephemeral copies of sound recordings rejected others in the initial Judges have the authority to modify that are used by a service for non- Determination. In its Petition for such ‘‘background terms of the statutory compensable performances under Rehearing (Rehearing Motion), license’’ where willing buyers and Section 114 might not be authorized SoundExchange raised several issues sellers would negotiate such terms under the regulations.’’ Pandora PFFCL relating to the Judges’ determinations absent the statute. IHM COL ¶ 34–35. ¶ 416. Pandora also proposes that the regarding proper regulatory language to SoundExchange argued that statutory Judges add the following sentence to the effect their conclusions in the changes can only be made by Congress. Determination. After considering the 217 SX Ex. 1931 (designated testimony of Dr. 219 George S. Ford). Dr. Ford testifies that ‘‘in the Rehearing Motion and the responses increases, if any. SoundExchange also listed marketplace deals between record companies and (without sufficient analysis) several other regulatory webcasters for non-statutory forms of licenses, it is 218 Compare SX Reply PFFCL ¶ 1247 concerns. The Judges permitted SoundExchange to typical for ephemeral copy rights to be expressly (‘‘SoundExchange believes that Pandora’s proposed detail the other regulatory concerns in a included among the grant of rights provided to the changes [to CRB regulations] should be rejected Supplemental Motion (Supplement). The Judges webcasters . . . [incorporating the rate for the outright’’) with SX PFFCL ¶ 1374 (referencing solicited and received responses from the Licensees ephemeral copy] into the overall rate that the agreements between labels and services wherein to all issues in the original Rehearing Motion and webcaster pays for the ephemeral copy rights and services are authorized to create and store a the Supplement. performance rights.’’ Id. at 10–11. He also reasonable, limited number of ephemeral copies). 220 See Order Denying in Part SoundExchange’s concluded that ‘‘recording artists and record 219 In the Rehearing Motion, SoundExchange Motion for Rehearing and Granting in Part companies have reached an agreement that five analyzed its concerns regarding several substantive Requested Revisions to Certain Regulatory percent (5%) of the payment for activities under determinations, including the provision for annual Provisions (Feb. 10, 2016), issued in PUBLIC Section 112(e) and 114 should be allocated to royalty rate adjustments. With regard to the version on February 22, 2016. Section 112(e) activities [and] that appears to be a regulations, SoundExchange challenged the stated 221 Section references are to the section numbers reasonable proposal.’’ Id. at 15. method of calculation of annual royalty rate in the regulations adopted by this Determination.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00084 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26399

The Judges agreed. The Judges did not for a regulation because it already sends Statements of Account. iHeart proposed adopt this change. reminders. It also argued that an changes that would allow Licensees to acknowledgment email would be recoup overpayments. SoundExchange b. Waiver of Requirement to Destroy challenging because it does not have argued that the proposals are Ephemeral Recordings After Six Months current email addresses for each of its unreasonable because of, inter alia, the iHeart proposed to add a provision licensees, and the cost would outweigh window of time within which, and the that exempts Broadcasters from the the benefit. SoundExchange countered number of occasions upon which, a statutory six-month limitation on the that it will soon have an online payment Licensee could make adjustments. In retention of ephemeral recordings portal, a fact that NRBNMLC points out addition, SoundExchange complained subject to certain conditions. shows that SoundExchange realizes that that the administrative burden of such SoundExchange argued that the Judges the receipts would be useful. The Judges a proposal could be excessive. are not authorized to make changes to found that the online portal should SoundExchange also noted that the the statute by enacting regulations, and address the receipt concern and that the money may not be recoupable once it is the Judges agreed. The Judges cannot practice of sending reminders does not paid to artists. Pandora argued that and did not adopt this proposal. warrant a regulation. Therefore, the making good faith adjustments are part B. Section 380.2—Making Payment of Judges did not adopt this proposed of the normal course of business and Royalty Fees change. that SoundExchange’s technological 2. Late Fees—§ 380.2(d) advances will make the administration 1. Monthly Payments—§ 380.2(b) of adjustments manageable. Pandora a. Payment Period a. A Single Late Fee RFF at 192–93. iHeart pointed out that SoundExchange proposed shortening Pandora proposed a single late fee for SoundExchange has a method for the payment period from 45 days to 30 both a late payment and a late Statement reversing its own inadvertent days. Pandora and Sirius did not oppose of Account. It argued that a late fee for overpayments. IHM PFF ¶ 433; IHM RFF the change, but the NAB, NRBNMLC, each of these is duplicative and ¶ 202; see PAN PFF ¶ 1300. and IHM did. SoundExchange argued unnecessary. SoundExchange countered The Judges agreed with that the shorter term would allow them that it incurs duplicative costs when SoundExchange. The burden of to distribute payments more quickly and both items are late and that it is fair to submitting accurate payments is on the that the majority of agreements in the hold a late payor accountable for such Licensee, and the Licensee bears the risk industry have payments terms of 30 costs. In addition, SoundExchange’s of overpayment. In addition, the record days. The NAB and IHM argued that ability to enforce compliance and make contained no evidence to guide the because of the unique character of their efficient distribution relies on late fees Judges in determining a reasonable respective business models, shortening for each of these. The Judges agreed that period for, or a reasonable number of, the term would cause additional such fees encourage compliance for adjustments. Therefore, the Judges did burdens and create inaccuracies and each required item. As a result, the not adopt this proposed change. The parties also raised the issue of overpayments that potentially would Judges did not adopt this proposed royalty fee payment adjustments in the not be refunded. The Judges also are change. context of audits. See discussion considering this issue in a rulemaking b. Late Fee Rate regarding overpayments and proceeding that is currently pending iHeart, the NAB, and NRBNMLC underpayments discovered at audit before them. The Judges do not believe proposed that the late fee rate be under section 380.6 below. the record before them in this rate- reduced from 1.5% (the equivalent of setting proceeding supports the change 2. Signature Attestation—§ 380.3(a)(8) 18% per year) to a more ‘‘reasonable’’ that SoundExchange seeks, and fee; that is, one similar to statutory Pandora proposed adding a sentence therefore decline to adopt it. The Judges interest rates on judgments and tax to the required language in a Statement can perceive the costs to the Services underpayments. iHeart pointed out that of Account—just below the sentence that the shortened reporting period its agreements with the Indies contain where the signatory attests to the would impose, and it is less clear that no late fee provision and that Warner statement’s accuracy and the benefits identified by has never asked them to pay the late fee completeness—that would allow SoundExchange from such a change when they have submitted a late Licensees to amend their Statements of would justify those costs. Nevertheless, payment. SoundExchange argued that Accounts. This proposal was related to the Judges will consider revisiting this the high fee provides an incentive for iHeart’s proposal regarding issue in the broader context of the timely payments and covers costs due to overpayment and corrections to pending rulemaking proceeding. late payments. The evidence shows that payments. The proposed sentence b. Emails Acknowledging Receipt of late fees in market agreements range contained no time limit for making Payment from no fees up to the proposed fee of amendments to the Statements of NRBNMLC proposed that 1.5%. The 1.5% rate is an accepted rate Accounts and is therefore an SoundExchange send emails (similar to in the market, and the services unreasonable addition to the Statement those that the musical works collectives produced no evidence of actual of Account. The Judges did not adopt send) with reminders that annual hardship from the current rate of 1.5%. this proposed change. payments are due, which would serve a For this reason, the Judges did not adopt D. Section 380.4—Distributing Royalty function similar to an invoice. this proposed change. Fees NRBNMLC also proposed a provision C. Section 380.3—Delivering Statements 1. Best Efforts to Identify and Locate— requiring SoundExchange to email of Account acknowledgments of receipt of payment, § 380.4(a)(2) which would function like a receipt and 1. Adjustments to Statements of In this proceeding, the Licensees which is a common business practice, Account—§ 380.3(a) proposed, and the Judges adopted, including in the nonprofit arena. Pandora proposed a change to allow additional regulatory language regarding SoundExchange argued there is no need Licensees to make adjustments to their the Collective’s duty to locate parties

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00085 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26400 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

entitled to receive royalty upcoming rate period, the Judges recognized that the language of section distributions.222 SoundExchange adopted the more neutral drafting that 380.2(e) may be unclear, and that the objected to the added language. A Pandora proposed to ensure that the amendment that SoundExchange SoundExchange executive testified that Collective’s use of unclaimed funds requested would clarify the regulation the Collective maintains an extensive comports with applicable law. in a manner consistent with the Judges’ database and can locate distributees In the Rehearing Motion, intent. Therefore, the Judges accepted without the due diligence suggested by SoundExchange further objected to the the SoundExchange proposal and the new language. See SX Ex. 23 at 18– Judge’s insertion of language to define clarified the regulatory language 19, SX Ex. 2 at 5–11. As SoundExchange the three-year holding period for accordingly: The three-year escrow conceded, however, the regulations unclaimed funds. The extant regulations period for undistributable royalties shall contain similar language in section contain an internal ambiguity be three years from the date of first 370.5(d) regarding best efforts to find concerning the measurement of the distribution of relevant royalty deposits copyright owners in order to make period for holding unclaimed funds. from a Licensee. available reports of use. When the Judges suggested If SoundExchange is able to make— reorganization of the Part 380 3. Designation of the Collective—§ 380.4 and amenable to making—records regulations, they highlighted this issue (d)(1) searches to assure proper distribution of for the parties. See Judges’ letter to The Judges designated reports of use, the Judges should assure participants dated April 2, 2015. For SoundExchange as Collective.224 that SoundExchange makes no less of an example, in § 380.4 of the current SoundExchange participated as the effort to locate copyright owners when regulations, the Collective is required to existing and presumed Collective. the time comes to distribute royalty hold funds if it is ‘‘unable to locate a SoundExchange indicated its funds. It would seem even more Copyright Owner . . . within 3 years willingness to continue as the appropriate for SoundExchange to from the date of payment by a Licensee Collective. See Bender WDT at 14–15. engage in best efforts when distributing . . . .’’ 37 CFR 380.4(g)(2) (emphasis No party objected to SoundExchange royalties to avoid any appearance of added). If the Collective is unable to continuing in the role of Collective. The impropriety or conflict of interest, in locate the rightful payee, then the funds Judges acknowledged the administrative light of section 380.4(b), which may become subject to § 380.8, which and technological knowledge base permit retention of unclaimed funds by requires the Collective to retain developed by SoundExchange over its SoundExchange. This minimal ‘‘unclaimed’’ funds for ‘‘a period of 3 years of service as the Collective. additional due diligence can do little years from the date of distribution.’’ See, Finding no reason to change the other than assure the currency and e.g., 37 CFR 380.8 (emphasis added). designation, the Judges re-named integrity of SoundExchange’s The Collective may apply those funds to SoundExchange to serve as the distribution database. offset its costs at the end of the three- Collective for purposes of collecting, Further, SoundExchange outlined its 223 year holding period. Id. monitoring, managing, and distributing search capabilities, but did not object On its face, the ‘‘date of payment by sound recording royalties established by expressly to the due diligence language a Licensee’’ is not the same as the ‘‘date this Part 380. proposed by NAB and NRBNMLC. The of distribution,’’ the latter of which is Judges adopted the proposal of NAB and ambiguous, at best. Despite the Judges’ E. Section 380.5—Handling Confidential NRBNMLC. invitation, no party offered explanation Information 2. Unclaimed Funds—§ 380.4(b) for the current regulatory discrepancy or 1. Disclosure of Confidential suggested clarifying language to Information—§ 380.5(c) Pandora proposed that the provision eliminate the ambiguity. In section in the regulations dealing with the 380.2(e) of the regulations adopted by Upon review of the supplemental Collective’s use of unclaimed funds may the Judges as part of this proceeding, the papers, the Judges made an additional not be consistent with state escheatment Judges sought to resolve the ambiguity change to the language regarding laws. SoundExchange opposed changes by specifying that the three-year holding handling of confidential information, to this provision, which allows the period commences on ‘‘the date of final anticipating a claim of ambiguity. In its Collective, under certain circumstances, distribution of all royalties.’’ discussion of the new regulatory to use unclaimed funds for SoundExchange averred that the Judges’ requirements for, inter alia, written administrative purposes. introduced uncertainty into the confidentiality agreements, SoundExchange argued that the changes regulation because it is unclear when a SoundExchange referred to Pandora had proposed, which would ‘‘final distribution of all royalties’’ takes confidentiality obligations arising by have required the Collective to use ‘‘operation of law.’’ Supplement at 3. unclaimed funds in a manner consistent place when a copyright owner cannot be located and the funds that copyright The Judges acknowledged that a with applicable law, could impose an Qualified Accountant and any attorney unnecessary regulatory burden on the owner may be entitled to cannot be admitted to a state’s bar is under a Collective. distributed. professional ethical obligation 225 The Judges adopted the changes SoundExchange requested that the to substantially as proposed by Pandora. Judges amend the regulation to specify 224 In the provision relating to the potential Although the Judges do not believe the that the three-year holding period commences on the date of the first dissolution of SoundExchange as the Collective, unclaimed funds provision in the Pandora and SoundExchange agreed that the phrase current regulations runs afoul of any distribution of royalties from the ‘‘that have themselves authorized the Collective’’ in state law, in abundance of caution and relevant payment by the service. current CRB Rule 380.4(b)(2)(i) is unnecessary and should be deleted. See SX Reply PFFCL ¶ 1231 to avoid potential confusion in the Rehearing Motion at 10. No other party responded to SoundExchange’s n.74. Accordingly, the applicable provision the Judges adopted, § 380.4(d)(2)(i), does not retain that 222 In their post-Determination review, the Judges requested amendment. The Judges unnecessary language. noted that the due diligence language was 225 These obligations might or might not arise by misplaced in § 380.2(e), which is concerned with 223 Similar language is repeated in subparts B ‘‘operation of law’’ depending upon the payment of royalty fees by Licensees. The Judges (§§ 380.13(i)(2), 380.17) and C (§§ 380.23(h)(2), jurisdiction, but any party aggrieved by a breach of have deleted the language from § 380.2. 380.27) of the extant regulations. these professional obligations is likely nonetheless

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00086 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26401

maintain confidentiality of his or her confidentiality agreements and (2) 380.6(b) could be interpreted as limiting client’s confidential information. The restriction of use of confidential SoundExchange to a single audit of a Judges, therefore, eliminated ‘‘attorney’’ information to purposes ‘‘directly’’ single service each year. Id. from the list of potential viewers of related to collection and distribution of SoundExchange asked the Judges to confidential information required to royalties. Id. (citing, e.g., SX Exs 110 at clarify that it is not restricted to auditing sign a confidentiality agreement. The 11 (iHeart-Concord agreement) and 33 at only one licensee per year; rather that Judges added ‘‘outside counsel’’ to 30 (iHeart-Warner agreement)). iHeart’s the limit is one audit per year for each ‘‘Qualified Auditor’’ in subsection (c)(2) citation to the record illustrated the licensee. No party responded in of section 380.5, as eligible to receive Judges’ ability to look to ‘‘comparable opposition to this clarification request. confidential information without circumstances under voluntary license As SoundExchange’s proposed executing a separate confidentiality agreements’’ in setting rates under § 114. clarification was consistent with the agreement. The Judges specified SoundExchange’s objection was too intent of the language originally adopted ‘‘outside counsel’’ as some entities little, too late. The Judges declined to by the Judges, but was not subject to involved in these complex proceedings change the confidentiality language. misinterpretation, the Judges amended may employ in-house counsel, whose the regulatory language accordingly. duties would not necessitate their 3. Safeguarding Confidential seeing information relating to the Information—§ 380.5(d) 2. The Audit—§ 380.6(d) Judges’ rate proceedings. In-house SoundExchange objected to use of the a. Binding Nature counsel are deemed to be included in phrase ‘‘distributees of the collective’’ in The NAB proposed the Judges modify the term ‘‘employees’’ in the list of section 380.5(d) as creating an uncertain the audit regulation by removing the persons required to sign the standard, contending that the provision requirement that the Qualified Auditor’s confidentiality agreement.226 could be interpreted to require results be binding on the parties. recipients of confidential information to 2. Written Agreements—§ 380.5(c)(1) SoundExchange objected to the Judges’ ‘‘adhere to the unknowable standards adoption of the NAB proposal. NAB and NRBNMLC proposed, and employed by SoundExchange’s tens of Supplement at 4. As the NAB noted, the Judges adopted, additional verbiage thousands of distributees.’’ Supplement SoundExchange 227 witness, Dr. Thomas for the regulation (section 380.5(c) (1) in at 4. SoundExchange proposed to clarify Lys, testified that requiring an audit the newly-revised regulations) regarding that recipients of confidential report be dispositive would be confidential information shared by information are bound by the standard ‘‘unreasonable.’’ NAB/Pandora Supp. participants in webcasting proceedings of care that they employ with their own Opp. at 3, citing 5/4/15 Tr. at 1507–08 that: (1) Required confidentiality confidential information by substituting (Lys). agreements to be in writing; and (2) the phrase ‘‘Person authorized to receive limited disclosure of confidential confidential information’’ for The Judges credited Dr. Lys’s information to those performing ‘‘distributees of the collective.’’ Id. No testimony and agreed that the subject of activities ‘‘related directly’’ to collection other party raised an issue with the any audit should be permitted to contest and distribution of royalty payments. language of the newly-revised audit results. SoundExchange offered no SoundExchange did not indicate that it regulation; nor did any party object to record support for its proposal that the ever addressed these proposed changes SoundExchange’s requested change. regulations return to the current to the regulations. It was not until SoundExchange correctly discerned language, albeit made reciprocal in SoundExchange sought rehearing that it the intended meaning of the language nature. The ‘‘binding’’ language has raised a specific challenge to this added that the Judges adopted. The Judges did been excised from the newly-revised 228 confidentiality language. Supplemental not view the potential misinterpretation regulations. Petition for Rehearing . . . at 4 that SoundExchange feared to be a b. Acceptable Verification Process (Supplement). reasonable reading of the section In their joint opposition to the 380.5(d). The Judges also did not view SoundExchange proposed removing Supplement, NAB and Pandora objected SoundExchange’s proposed amendment this provision because it allows audits to allowing SoundExchange to raise a as likely to clarify the Judges’ intent. to be routine financial audits instead of new issue on rehearing. See NAB and Nevertheless, to remove all doubt the specialized ‘‘royalty examinations.’’ SX Pandora’s Opposition to . . . Judges amended section 380.5(d) by PFF ¶ 1285–86. Although the services Supplement [ ] . . . at 5 (NAB/Pandora deleting everything after the second-to- did not oppose this change, Supp. Opp.). iHeart further pointed to last comma and substituting the SoundExchange offered no evidence of record evidence to support the following: ‘‘but no less than the same the ineffectiveness of the audits to date additional language relating to handling degree of security that the recipient uses due to the existence of the provision, confidential information during the to protect its own Confidential and therefore the Judges did not adopt process of royalty collection and Information or similarly sensitive the proposed change. A Service’s recent distribution. See iHeart Opposition to information.’’ financial audit need not preclude a . . . Supplement[ ] at 2–3 (iHeart Supp. business audit that focuses on the Opp.). iHeart cited direct license F. Section 380.6—Auditing Payments agreements that were in evidence in this and Distributions 227 In drafting, the Judges inadvertently included language the NAB proposed to make the choice of proceeding as support for the reasonable 1. Frequency of Auditing—§ 380.6(b) addition of requirements for (1) written a Qualified Auditor binding, in addition to adopting SoundExchange argued that the the NAB proposal to drop the requirement that the audit results be binding. The Judges found that entitled to a legal or equitable remedy from a court Judges’ newly-revised regulatory language making the choice of a Qualified Auditor of competent jurisdiction. language regarding audit frequency binding is unnecessary, and have removed it. 226 The Judges understand that in-house counsel included an unintended ambiguity 228 Accordingly, any attempt to seek a remedy admitted to the bar carry the same professional regarding the frequency with which the based upon an auditor’s findings, and any attempt ethical obligations as outside counsel. Admission to to challenge those findings, must be made in a court the bar alone, however, is not sufficient to grant in- Collective may audit Licensees. Motion of competent jurisdiction, or through any private house counsel unnecessary access to confidential at 10. In its Supplement, alternative dispute resolution procedure to which information of a business competitor. SoundExchange contended that section the affected parties may have agreed.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00087 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26402 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

Service’s royalty policies and connection with adjustments based on non-profit organization that makes procedures. revised Statements of Account. distributions directly to a multiplicity of Rehearing Motion at 10. In the then- 3. Audit Results; Underpayment or artists and record companies from each extant regulations, the provisions Overpayment of Royalties—§ 380.6(g) royalty deposit. SoundExchange is not regarding audits and audit findings did in the same position that an individual a. Terms for Restitution of not address the question of financial Licensor might be with regard to Underpayment adjustment,231 either restitution for management of its funds. Pandora suggested that Licensees and underpayment or recoupment of overpayment. In this proceeding, the The Judges thus adopted for audit SoundExchange be permitted to agree findings the same rationale as that 229 Services introduced evidence of the on acceptable terms regarding the applicable to Statements of Account: time for restitution of underpayments by practice of ‘‘truing’’ accounts. See e.g., The burden of accurate reporting and Licensees.230 SoundExchange did not SX Ex. 33 at 18 (¶ 4(c) of document) payment is on the Licensee. oppose Pandora’s proposal in its Reply (Licensee to make immediate restitution PFF/PCL. In its opposition to the of any underpayment discovered by Accordingly, the Judges’ regulations SoundExchange Supplement, iHeart audit), IHM Ex. 3351 at 11 (¶ 7(b), p. 10 continue to require immediate suggested that agreed terms for of document) (Licensee may withhold restitution in the case of underpayment, reconciliation are consistent with royalties prospectively in certain but no right of recoupment for market terms allowing for agreement on circumstances), IHM Ex. 3340 at 3 overpayment. As with any untimely the identity of an auditor and the scope (¶ 1(b), p. 2 of document) (same). payment, a Licensee that is obligated to of an audit. iHeart Supp. Opp. at 2, Reconciliation of accounts should be no remedy an underpayment is liable to citing, e.g., SX Ex. 38 at 40 (re timing less a practice in the context of statutory pay reasonable interest thereon. licensing. See 17 U.S.C. 114(f)(2)(B)(II) and scope of audit). 4. Other Audit Related Proposals The legislative emphasis in the Act on (in establishing terms, Judges may voluntary, negotiated settlements, consider ‘‘comparable circumstances a. Notice and Cure should, without clear, contrary evidence under voluntary license agreements’’). or authority, extend to permitting The Licensees participating in this The NAB proposed adding a notice agreement regarding the timing for proceeding proposed an open-ended and cure provision to apply in case of account reconciliation. SoundExchange term that would permit them to amend breach because it is customary in failed to show that permission to resolve SOAs and make concomitant financial contracts and is included in some of the a conflict by agreement is without adjustments (with interest). The Judges agreements in evidence. evidentiary support or contrary to any rejected this proposal because of the SoundExchange wanted the option to legal requirements in the Act. The open-ended nature of the proposal, use informal methods of dealing with Judges did not err in adding this which could result in an excessive breach, but the NAB argued this provision to the revised regulations. administrative burden on provision would not preclude such However, the regulatory language the SoundExchange. The Judges concluded, efforts; it would only be required in case Judges adopted might be construed as rather, to allocate the burden of of a material breach that requiring, rather than permitting accuracy in reporting to the Licensees. SoundExchange planned to assert. Such SoundExchange and Licensees to agree In allocating that administrative a provision is not necessary merely on acceptable terms of payment. burden, however, the Judges were not because it is customary, and informal or Accordingly, the Judges clarified section opining on the propriety of or need for formal methods of notice are always 380.6(g). a balancing of accounts after an audit. available to the parties. Therefore, the SoundExchange may audit Licensees b. Recoupment of Overpayment Judges did not adopt this proposed annually, but the period audited may be change. The parties raised the issue of up to three years. No party offered underpayment collection and evidence of past audit practices or b. Completion of Audit Within Six overpayment recoupment (with interest) results. The Judges were unaware Months in the context of monthly royalty whether any audit findings had ever deposits. A periodic audit may also resulted in cost-shifting, for example, let The NAB and NRBNMLC proposed reveal underpayments and alone what remedies, if any, the parties augmenting the audit notice provision overpayments. SoundExchange objected had employed to reconcile under- or with what they termed a reasonable to new language in section 380.6(g) that over-payments. Further, a sampling of deadline for completion of audits, gives licensees a credit, with interest, direct license agreements did not reveal arguing the potential for abuse and the for overpayments that are revealed in an a standard regarding recoupment of burden that lengthy audits place on audit, arguing that the provision is overpayments detected by audit. Broadcasters. They point to comments inconsistent with the Judges’ rejection Nonetheless, even if directly- in a rulemaking proceeding regarding of a similar proposal by the services in contracting parties negotiated reciprocal the burden. SoundExchange argues that reconciliation of payments in any the length of an audit is in the control 229 The Judges addressed elsewhere whether circumstance, the Collective is in a of the services more than of the auditor those terms shall include interest. different business posture than its and that the NAB and NRBNMLC point 230 SoundExchange complained that Pandora members making direct license deals. ‘‘sneaked’’ in these changes. The record did not to no such provisions in private support SoundExchange’s allegation. Pandora As SoundExchange pointed out, it is a agreements. The comments in the included its request for this regulatory change rulemaking procedure are not evidence twice—once with its written rebuttal statement and 231 The only reference to a financial issue in the in this proceeding. What is reasonable is again with its proposed findings of fact and current audit regulations relates to restitution of an conclusions of law. Pandora First Amended Rates underpayment and allocation of the cost of the the ultimate finding of fact. The parties and Terms (Feb. 22, 2015) (submitted concurrently audit in the event the auditor finds an submitted no evidence on what would with Pandora Written Rebuttal Statement); Pandora underpayment discrepancy of 10% or more. See, be a reasonable time within which to Second Amended Rates and Terms at 3, 13 (Jun. 24, e.g. 37 CFR 380.6(g), 380.7(g). No regulation complete an audit. The Judges do not 2015) (submitted concurrently with Pandora PFF/ addresses underpayment of less than 10% or PCL). overpayment at any amount. adopt this proposal.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00088 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26403

G. Section 380.7—Definitions 232 phrase ‘‘Public Broadcasting Entities’’ to require a license. More substantively, ‘‘Covered Entity under Subpart D’’ to the NAB and NRBNMLC also added two 1. Definition of Aggregate Tuning Hours conform the terminology with that exclusions to the definition, one (ATH) adopted in Subpart D of Part 380, regarding performances of very short The NAB and NRBNMLC proposed to pursuant to the settlement duration and one very technical one redefine ATH to allow for a reduction in SoundExchange reached with The regarding second connections from the reported ATH for broadcast time Corporation for Public Broadcasting same IP address. SoundExchange argued devoted to talk radio. SoundExchange (CPB) and National Public Radio (NPR). that rights owners should be countered that NRBNMLC provided no By its terms, the CPB/NPR settlement is compensated for all uses of their works, evidence to justify a reduction different by and between SoundExchange on the and thus that services should pay for from the one established (and used) by one hand and, on the other hand, NPR performances even if they are of brief NPR stations. SoundExchange pointed and CPB, on behalf of themselves and duration or the service deems them to out that all the rates would have to be on behalf of American Public Media, be ‘‘skips.’’ SoundExchange also recalculated if the basic assumption Public Radio International, and certain pointed out that the proposed rates were regarding ATH is changed at this point. public radio stations, together calculated based on the current The Judges agreed. If the definition designated the Covered Entities. statutory definition of ‘‘performance’’ changed, the threshold would need to No participant in the hearing self- and that any narrowing of the definition change as well, and there was no basis identified as a public broadcasting would require adjustments to the in the record for making those changes. entity. Presumably, if there were an proposals. The second exclusion is not The Judges did not adopt this change. entity satisfying the statutory definition necessary because SoundExchange’s of a public broadcaster that was witness, Mr. Bender, agreed that 2. Definition of Broadcast excluded by agreement from the reconnections are not performances Retransmission settlement memorialized in Subpart D of under the current regulations, which The NAB and iHeart proposed a the revamped regulations, the excluded specify that a ‘‘performance’’ requires a change in the definition of broadcast entity would be treated as a listener. retransmission (simulcast) to cover noncommercial webcaster or a The definition of performance in the anything that is at least 51% identical noncommercial educational webcaster, regulations has long been established. to its antecedent terrestrial broadcast. as the case may be.234 As the Judges did The NAB and NRBNMLC argued that This proposal was a companion not define ‘‘public broadcaster’’ in this performances of very short duration are proposal to the NAB’s proposal of iteration of their regulations, however, of no value to the listener or the service, separate royalty rates for simulcasters. the request from SoundExchange to and they pointed out that listeners The Judges declined to establish clarify the reference was well taken. cannot skip songs on their services. The separate rates for simulcasters and The Judges have added a definition of Judges agreed that performance as it has therefore did not include a definition of ‘‘public broadcaster’’ to section 380.7, been defined should continue to apply. ‘‘broadcast retransmission’’ in the new cross-referencing Subpart D. The Judges did not adopt these changes. regulations. In its Supplement, SoundExchange 5. Definition of Performance objected to the Judges’ ‘‘linguistic 3. Definition of Broadcaster To Include In the current regulations, a changes’’ to the definition of ‘‘Affiliate of’’ ‘‘performance’’ is defined as ‘‘each ‘‘performance’’ in section 380.7. The NAB and NRBNMLC proposed to instance in which any portion of a Supplement at 5. The Judges accepted change the definition of Broadcaster, but sound recording is publicly performed SoundExchange’s concern that the new did not provide a reason for the change. to a listener . . . .’’ See, e.g., 37 CFR language may harbor an ambiguity. No The Judges determined not to establish 380.2. The Services proposed various party objected to SoundExchange’s separate royalty rates for simulcasts by changes to the definition of request for modification of the over-the-air broadcasters, obviating the performance. Parties can and do alter definition. The Judges made the need for a definition of ‘‘broadcaster’’ in the definition of ‘‘performance’’ and requested modification. the regulations. The Judges did not, change other DMCA provisions in 6. Definition of Qualified Auditor therefore, adopt this proposed change. directly negotiated licenses. The Judges SoundExchange proposed that the cannot, however, make regulations that 4. Definition of Commercial Webcaster regulations allow non-CPAs to perform are contrary to the requirements of the audits if they have the requisite In the Rehearing context, Act. industry-specific expertise, arguing that SoundExchange asked the Judges to Pandora sought to add ‘‘in the United it is difficult to find CPAs with the change the definition of ‘‘Commercial States’’ to the definition. The NAB and needed expertise and that other actors Webcaster.’’ Motion at 10. As written in NRBNMLC asked for an alternate in the market allow content owners to the original ‘‘Exhibit A’’ to the parenthetical description and a audit royalty payments. The NAB and Determination, the definition of reference to the section in the Copyright NRBNMLC countered with the Commercial Webcaster excluded ‘‘an Act regarding performances that do not Educational Webcaster,233 a argument that CPAs inspire confidence in the audit results because of the Noncommercial Webcaster, or Public 234 Under section 118 of the Act, a ‘‘public Broadcasting Entities . . . .’’ broadcasting entity’’ means a noncommercial standards of their profession and that SoundExchange sought to change the educational webcaster as defined in 47 U.S.C. 397, they can rely on experts in the industry viz., ‘‘[CPB], any licensee or permittee of a public to assist them if necessary. broadcast station, or any nonprofit institution 232 The Judges included two sections numbered engaged primarily in the production, acquisition, SoundExchange had argued in past 380.6 in the initial iteration of the regulatory distribution, or dissemination of educational and proceedings for a change to allow in- language, one of which was the definitions section. cultural television or radio programs.’’ Not all house auditors to perform audits. The The Judges corrected that error and relabeled the noncommercial webcasters are public broadcasters. Judges had rejected that change. Final definitions section § 380.7. Not all educational webcasters are public 233 The Judges noted that the reference to broadcasters. The appellation ‘‘public broadcaster’’ Rule and Order, Docket No. 2005–1 CRB Educational Webcaster in this definition was appears to be reserved to those stations that receive DTRA (‘‘Web II’’), 72 FR 24084, 24109 misplaced and therefore removed it. funding by or through the CPB. (May 1, 2007). For the same reasons,

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00089 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26404 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

they did not adopt in this proceeding a simulcasters and therefore did not adopt IX. Royalty Rates Determined by the change to the requirement that the the proposed revision. Judges auditor be a CPA. The Judges further 2. Annual Royalty Fee Adjustment— A. Annual Rates and Price Level inserted the qualifier ‘‘independent’’ § 380.10(c) Adjustments into the definition of ‘‘Qualified Auditor’’ for the sake of regulatory While the Judges rejected The Judges will set statutory rates for efficiency. The Judges did not adopt SoundExchange’s objections to the the year 2016. For the years 2017 SoundExchange’s proposed change. royalty fee adjustment adopted in the through 2020, the rates shall be adjusted The Judges did, however, adopt Determination, the Judges to reflect any inflation or deflation, as language proposed by the NAB and acknowledged that the regulation measured by changes in a particular NRBNMLC concerting the licensing of should be clarified so that, in rounding Consumer Price Index (the CPI–U) an auditor. In its Rehearing Motion, to the nearest fourth decimal place, it is announced by the Bureau of Labor SoundExchange objected to the addition not understood to create a meaningful Statistics (BLS), in November of the of a requirement that a Qualified deviation from the unrounded real rate. immediately preceding year, as Auditor be licensed in the jurisdiction Accordingly the Judges adopted a described in the new regulations set in which it conducts the audit. Motion change to the regulation providing for forth in this determination. In this at 8–9. The NAB had requested this annual royalty fee adjustment in order regard, the Judges concur with Dr. additional requirement to qualify an to clarify the Judges’ intent with regard Shapiro, who testified that a regulatory auditor as part of its proposed terms. to, and provide examples of, calculating provision requiring an annual price the indexed increase, if any. NAB Proposed Rates and Terms at 3 level adjustment is preferable to an (Tab B to NAB CWDS Vol. 1). 3. Third Party Programming implicit or explicit prediction of future SoundExchange asserted that the The NAB proposed a waiver of census inflation (or deflation). 5/19/15 Tr. additional jurisdictional licensure reporting on any material that is 4608–10 (Shapiro). requirement was not supported by the transmitted by a simulcaster that is The Judges shall also adjust any record. This requirement provides programmed by a third party, i.e., not effective benchmark rate on which they assurance that the auditor will be the station owner/operator whose rely in this proceeding to reflect accountable and amenable to local broadcasts are retransmitted. The NAB inflation (or deflation) as measured by governance in the jurisdiction in which proposed estimating ATH for third party the CPI–U in the calendar years between it operates. Differences in ethical programming because the stations are the last calendar year in which the data standards and sanctions for CPAs unable to get the necessary data from was collected for the benchmark and among jurisdictions might be small, but the program originators. 2016, as reflected in the applicable the requirement that the auditor submit SoundExchange argued that some November announcement by the BLS. itself to the jurisdiction of the local CPA broadcasters use a lot of third party governing bodies and local courts is B. Commercial Rates material and that they should be significant. The NAB’s suggestion is required to get that data in order to 1. Commercial Subscription Rates supported by the testimony of Professor make accurate reporting to Roman Weil and, therefore, was not Based on the analysis in this SoundExchange. If broadcasters use without support in the record. See Weil determination, the Judges shall set two third party programming, WRT at 11–13. The Judges rejected separate rates for commercial SoundExchange should not have to bear SoundExchange’s objection. noninteractive webcasting. One rate the risk of inaccurate reporting. In shall apply to performances on H. Section 380.10 (Subpart B)—Royalty addition, the broadcaster is in the best subscription-based commercial Fees for the Public Performance of position to incorporate costs of census noninteractive services. A separate rate Sound Recordings and the Making of reporting into their negotiated payments shall apply to performances on Ephemeral Recordings with the third-party programmers. The nonsubscription (advertising-supported Judges did not adopt this change. 1. Minimum Fee—§ 380.10(b) free-to-the-listener) services. The NAB proposed a revision to the I. Miscellaneous—Proposed Relief From The Judges have identified two usable minimum fee provision that removed Reporting Requirement benchmark rates for commercial fees for individual channels, leaving The NAB and NRBNMLC proposed noninteractive subscription services for only fees for individual stations. that the regulation regarding 2016. SoundExchange argued that this is not distribution of royalties provide relief The first is the steering-adjusted rate necessary because of the annual cap on from reporting requirements for small derived from the benchmark developed total amount of minimum fees that any broadcasters and those noncommercial by Dr. Rubinfeld on behalf of licensee must pay; that fees would no webcasters that are ‘‘exempt from the SoundExchange. Dr. Rubinfeld longer be in proportion to report of use requirements contained in established a subscription-based SoundExchange’s costs; and that § 370.4’’. NAB Proposed Terms at 6; benchmark rate of $0.002376. SX Ex. 59 stations would game the system by NRBNMLC Amended Proposed Rates (Rubinfeld CWDT Ex. 16(a); see also SX streaming on multiple channels in order and Terms at 6. This is an argument the PFF ¶¶ 344; 393. to reduce fees. The NAB explained that NAB and NRBNMLC make in the As noted in this determination, the its rate proposal and terms applied only pending rulemaking proceeding and did Judges apply a steering adjustment to to stations that simulcast and that side not make in this proceeding other than this benchmark rate to reflect the rate- channels would have different rates and to add the language to their proposed reducing effect of steering as indicated terms. According to the NAB, this terms. SoundExchange’s response is in the Pandora/Merlin Agreement.235 In proposed change was a ‘‘conforming lodged in the rulemaking proceeding. the present case, the steering adjustment change’’ that presumably would bring See Docket No. 14–CRB–0005 (RM). The forum for that request is the rulemaking, this term in line with the NAB’s 235 Dr. Shapiro’s rate data covered a period proposed rate for simulcasters. The not this proceeding. The Judges did not through the third quarter of 2014. Shapiro WDT at Judges did not set a separate rate for adopt these proposals. 32.

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00090 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26405

derived from the evidence is 12%, Based on the foregoing, the Judges preceding year, as set forth in the calculated as follows: determine, in their discretion, that the regulations. (1) The unsteered subscription service appropriate per-play rate for royalties 3. Ephemeral Recording Rate rate for 2015 in the Pandora/Merlin paid by licensees to licensors in the Agreement is $0.[REDACTED]. See Pan noninteractive subscription market In accordance with the Judges’ Ex. 5014, ¶ 3(a)(ii). under § 114 for the year 2016 is $0.0022. analysis supra, section VII, the royalty (2) Pandora’s effective rate at the As discussed supra, the rate for the rate for ephemeral recordings under 17 [REDACTED]% (low end) of steering for remainder of the statutory term—2017– U.S.C. 112(e) applicable to commercial 2016, as derived by Dr. Shapiro, is 2020—shall reflect the foregoing rate of webcasters shall be included within, $0.002238. See Shapiro WDT at 35. $0.0022 per performance, as adjusted and constitute 5% of the royalties such (3) Dr. Shapiro’s $0.002238 steered annually upward or downward to reflect webcasters pay for performances of rate for 2016 includes a 2.2% changes in the CPI–U over the preceding sound recordings under § 114 of the Act. anticipated inflation factor that the year, pursuant to the applicable C. The Noncommercial Rates Judges do not apply. See id. regulations. 1. NPR–CPB/SoundExchange Settlement (4) Backing out that 2.2% inflation 2. Commercial Nonsubscription Rates factor indicates a 2015 steered rate of The Judges have previously adopted $0.002189 (i.e., $0.002238/1.022). The Judges have identified two usable the settlement agreement between (5) Adjusting for the actual inflation benchmark rates for commercial SoundExchange, on one hand, and in 2015 of 0.5% (announced by the BLS noninteractive nonsubscription services National Public Radio and the on December 15, 2015 236) increases the for 2016. First, the Judges have Corporation for Public Broadcasting, on above steered rate marginally to identified the adjusted, effective average the other, for simulcast transmissions by $0.002194, which the Judges round to per-play rate derived from the iHeart/ public radio stations. See Digital $0.0022. Warner Agreement. That rate, as Performance Right in Sound Recordings (6) The unsteered 2015 subscription developed, supra, is $0.[REDACTED] and Ephemeral Recordings, Final Rule, service rate of $0.[REDACTED] (step 1) per play. 80 FR 59588 (Oct. 2, 2015). The rates Second, the Judges have identified the minus the steered rate of $0.0022 equals and terms governing transmissions and effective per-play rate in the Pandora/ $0.0003. ephemeral recordings by the entities Merlin Agreement (with steering at (7) The percentage change in the that are covered by that settlement [REDACTED]%) as a usable benchmark. subscription service rate for 2015 is agreement for the period 2016–2020 The effective benchmark rate from that 12% (i.e., $0.0003/$0.[REDACTED]). shall be as set forth in the agreement agreement is $0.[REDACTED]. and codified at 37 CFR 380.30–380.37 Accordingly, Dr. Rubinfeld’s Thus, the Judges identify a zone of (subpart D). proposed benchmark rate of $0.002376 reasonableness in this market segment must be reduced by 12% to reflect an as well. That is, the zone embraces a 2. CBI/SoundExchange Settlement effectively competitive rate. A reduction low effective rate of $0.[REDACTED] The Judges have previously adopted of 12% brings that subscription service and high effective rate of rate to $0.0021 (rounded). the settlement agreement between $0.[REDACTED]. As noted earlier in this SoundExchange, and College However, Dr. Rubinfeld’s data determination, it would be improper covered the period 2011 through 2014. Broadcasters, Inc., for transmissions by based on the present record, to set Noncommercial Educational Webcasters As noted supra, the Judges reject Dr. separate rates for Indies and Majors. Rubinfeld’s linear $0.0008 year-over- (NEWs). See Digital Performance Right However, as the Judges have also in Sound Recordings and Ephemeral year increase. Instead, the Judges apply explained, supra, a fundamental the CPI–U inflation adjustment of 0.5% Recordings, Final Rule, 80 FR 558201 difference between these two (Sep. 28, 2015). The rates and terms to reflect the inflation announced by the benchmarks is that the iHeart/Warner BLS on December 15, 2015. That governing transmissions and ephemeral benchmark reflects an effective rate recordings by NEWs for the period adjustment raises the rate derived from between a Major and a noninteractive 2016–2020 shall be as set forth in the Dr. Rubinfeld’s proposed steering- service, whereas the Pandora/Merlin agreement and codified at 37 CFR adjusted benchmark marginally, to Agreement reflects an effective rate 380.20–380.27 (subpart C). $0.0021105, which the Judges round to between Indies and a noninteractive $0.0021. service. The evidence at the hearing 3. All Other Noncommercial Webcasters The second steering-based indicated that the Majors’ sound In accordance with the Judges’ subscription rate that the Judges credit recordings comprise 65% of analysis supra, section V, the royalty is the rate in the Pandora/Merlin noninteractive streams, and the Indies’ rate for webcast transmissions by all Agreement, which already incorporates sound recordings comprise 35% of other noncommercial webcasters during a steering adjustment. That proposed noninteractive streams. See, e.g., SX Ex. the 2016–2020 rate period shall be $500 benchmark rate (at 12.5% steering) is 269 at 73. annually for each station or channel for $0.002238, rounded to $0.0022. See Based on the foregoing factors, the all webcast transmissions totaling not Shapiro WDT at 35. Judges find that the appropriate more than 159,140 Aggregate Tuning Thus (and perhaps not surprisingly), statutory rate within this zone of rates, Hours (ATH) in a month, for each year the steering and inflation-adjusted for nonsubscription, ad-supported (free- in the rate term. In addition, if, in any subscription rates under both proposed to-the-listener) services is $0.0017 per month, a noncommercial webcaster benchmarks establish an extremely tight performance, as adjusted annually makes total transmissions in excess of zone of reasonableness, separated by upward or downward to reflect changes 159,140 ATH on any individual channel only $0.0001.237 in the Consumer Price Index over the or station, the noncommercial webcaster shall pay per-performance royalty fees 236 See Bureau of Labor Statistics, Economic and a hypothetical noninteractive service would for the transmissions it makes on that News Release (Dec. 15, 2015) (available at bls.gov). have a WTP of $0.0022. In such a hypothetical 237 From an economic perspective, these rates market, the parties could consummate a contract at channel or station in excess of 159,140 suggest that a hypothetical willing seller would any price point between $0.0021 and $0.0022 per ATH at the rate of $0.0017 per have a WTA of $0.0021 in this subscription market, play. performance, as adjusted annually

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00091 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26406 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

upward or downward to reflect changes 380.4 Distributing royalty fees. (d) Late fees. A Licensee must pay a in the Consumer Price Index over the 380.5 Handling Confidential Information. late fee for each payment and each preceding year. 380.6 Auditing payments and distributions. Statement of Account that the Collective 380.7 Definitions. receives after the due date. The late fee 4. Ephemeral Recording Rate § 380.1 Scope and compliance. is 1.5% (or the highest lawful rate, The royalty rate for ephemeral (a) Scope. Subparts A and B of this whichever is lower) of the late payment recordings under 17 U.S.C. 112(e) part codify rates and terms of royalty amount per month. The late fee for a applicable to noncommercial webcasters payments for the public performance of late Statement of Account is 1.5% of the shall be the same as the rate applicable sound recordings in certain digital payment amount associated with the to commercial webcasters; that is, transmissions by certain Licensees in Statement of Account. Late fees accrue royalties for ephemeral recordings shall accordance with the applicable from the due date until the date that the be included within, and constitute 5% provisions of 17 U.S.C. 114 and for the Collective receives the late payment or of the royalties such webcasters pay for making of Ephemeral Recordings by late Statement of Account. performances of sound recordings under those Licensees in accordance with the (1) Waiver of late fees. The Collective § 114 of the Act. provisions of 17 U.S.C. 112(e), during may waive or lower late fees for X. Conclusion the period January 1, 2016, through immaterial or inadvertent failures of a Licensee to make a timely payment or On the basis of the foregoing analysis December 31, 2020. (b) Limited application of terms and submit a timely Statement of Account. and full consideration of the record, the (2) Notice regarding noncompliant definitions. The terms and definitions in Judges propound the rates and terms Statements of Account. If it is Subpart A apply only to Subpart B, described in this Determination. The reasonably evident to the Collective that except as expressly adopted and applied Register of Copyrights may review the a timely-provided Statement of Account in subpart C or subpart D of this part. Judges’ Determination for legal error in is materially noncompliant, the (c) Legal compliance. Licensees resolving a material issue of substantive Collective must notify the Licensee relying upon the statutory licenses set copyright law. The Librarian shall cause within 90 days of discovery of the forth in 17 U.S.C. 112(e) and 114 must the Judges’ Determination, and any noncompliance. correction thereto by the Register, to be comply with the requirements of this published in the Federal Register no part 380 and any other applicable § 380.3 Delivering statements of account. later than the conclusion of the 60-day regulations. (a) Statements of Account. Any review period. (d) Voluntary agreements. payment due under this Part 380 must So ordered. Notwithstanding the royalty rates and be accompanied by a corresponding terms established in any subparts of this Issue Date: March 4, 2016. Statement of Account that must contain part 380, the rates and terms of any the following information: Suzanne M. Barnett, license agreements entered into by Chief Copyright Royalty Judge (1) Such information as is necessary Copyright Owners and Licensees may to calculate the accompanying royalty Jesse M. Feder, apply in lieu of these rates and terms. Copyright Royalty Judge payment; (2) The name, address, business title, David R. Strickler, § 380.2 Making payment of royalty fees. Copyright Royalty Judge telephone number, facsimile number (if (a) Payment to the Collective. A any), electronic mail address (if any) List of Subjects in 37 CFR Part 380 Licensee must make the royalty and other contact information of the payments due under subpart B to Copyright; sound recordings. person to be contacted for information SoundExchange, Inc., which is the or questions concerning the content of For the reasons set forth in the Collective designated by the Copyright preamble, amend part 380 of title 37 of the Statement of Account; Royalty Board to collect and distribute (3) The signature of: the Code of Federal Regulations as royalties under this part 380. follows: (i) The Licensee or a duly authorized (b) Monthly payments. A Licensee agent of Licensee; must make royalty payments on a PART 380—RATES AND TERMS FOR (ii) A partner or delegate if the monthly basis. Payments are due on or Licensee is a partnership; or TRANSMISSIONS BY ELIGIBLE before the 45th day after the end of the NONSUBSCRIPTION SERVICES AND (iii) An officer of the corporation if month in which the Licensee made the Licensee is a corporation. NEW SUBSCRIPTION SERVICES AND Eligible Transmissions. FOR THE MAKING OF EPHEMERAL (4) The printed or typewritten name (c) Minimum payments. A Licensee of the person signing the Statement of REPRODUCTIONS TO FACILITATE must make any minimum annual THOSE TRANSMISSIONS Account; payments due under Subpart B by (5) If the Licensee is a partnership or ■ 1. The authority citation for part 380 January 31 of the applicable license corporation, the title or official position continues to read as follows: year. A Licensee that as of January 31 held in the partnership or corporation of any year has not made any eligible by the person signing the Statement of Authority: 17 U.S.C. 112(e), 114(f), nonsubscription transmissions, 804(b)(3). Account; noninteractive digital audio (6) A certification of the capacity of ■ 2. Revise the title of Part 380 to read transmissions as part of a new the person signing; as set forth above. subscription service, or Ephemeral (7) The date of signature; and ■ 3. Revise Subpart A to read as follows: Recordings pursuant to the licenses in (8) An attestation to the following 17 U.S.C. 114 and/or 17 U.S.C. 112(e), effect: Subpart A—Regulations Of General but that begins making such Application I, the undersigned owner/officer/partner/ transmissions after that date must make agent of the Licensee have examined this Sec. any payment due by the 45th day after Statement of Account and hereby state that 380.1 Scope and compliance. the end of the month in which the it is true, accurate, and complete to my 380.2 Making payment of royalty fees. Licensee commences making such knowledge after reasonable due diligence and 380.3 Delivering statements of account. transmissions. that it fairly presents, in all material respects,

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00092 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26407

the liabilities of the Licensee pursuant to 17 Inc., as the Collective to receive an appropriate written confidentiality U.S.C. 112(e) and 114 and applicable Statements of Account and royalty agreement, who are engaged in the regulations adopted under those sections. payments from Licensees and to collection and distribution of royalty (b) Certification. Licensee’s Chief distribute royalty payments to each payments hereunder and activities Financial Officer or, if Licensee does not Copyright Owner and Performer (or related directly thereto who require have a Chief Financial Officer, a person their respective designated agents) access to the Confidential Information authorized to sign Statements of entitled to receive royalties under 17 for the purpose of performing their Account for the Licensee must submit a U.S.C. 112(e) or 114(g). duties during the ordinary course of signed certification on an annual basis (2) If SoundExchange, Inc. should their work; attesting that Licensee’s royalty dissolve or cease to be governed by a (2) A Qualified Auditor or outside statements for the prior year represent a board consisting of equal numbers of counsel who is authorized to act on true and accurate determination of the representatives of Copyright Owners behalf of: royalties due and that any method of and Performers, then it shall be replaced (i) The Collective with respect to allocation employed by Licensee was for the applicable royalty term by a verification of a Licensee’s statement of applied in good faith and in accordance successor Collective according to the account pursuant to this part 380; or with U.S. GAAP. following procedure: (ii) A Copyright Owner or Performer (i) The nine Copyright Owner with respect to the verification of § 380.4 Distributing royalty fees. representatives and the nine Performer royalty distributions pursuant to this (a) Distribution of royalties. (1) The representatives on the SoundExchange part 380; Collective must promptly distribute board as of the last day preceding (3) Copyright Owners and Performers, royalties received from Licensees to SoundExchange’s cessation or including their designated agents, Copyright Owners and Performers that dissolution shall vote by a majority to whose works a Licensee used under the are entitled thereto, or to their recommend that the Copyright Royalty statutory licenses set forth in 17 U.S.C. designated agents. The Collective shall Judges designate a successor and must 112(e) and 114 by the Licensee whose only be responsible for making file a petition with the Copyright Confidential Information is being distributions to those who provide the Royalty Judges requesting that the supplied, subject to an appropriate Collective with information as is Judges designate the named successor written confidentiality agreement, and necessary to identify and pay the correct and setting forth the reasons therefor. including those employees, agents, recipient. The Collective must distribute (ii) Within 30 days of receiving the consultants, and independent royalties on a basis that values all petition, the Copyright Royalty Judges contractors of such Copyright Owners performances by a Licensee equally must issue an order designating the and Performers and their designated based upon the information provided recommended Collective, unless the agents, subject to an appropriate written under the Reports of Use requirements Judges find good cause not to make and confidentiality agreement, who require for Licensees pursuant to § 370.4 of this publish the designation in the Federal access to the Confidential Information to chapter and this subpart. Register. perform their duties during the ordinary (2) The Collective must use its best course of their work; efforts to identify and locate copyright § 380.5 Handling Confidential Information. (4) Attorneys and other authorized owners and featured artists in order to (a) Definition. For purposes of this agents of parties to proceedings under distribute royalties payable to them part 380, ‘‘Confidential Information’’ 17 U.S.C. 8, 112, 114, acting under an under § 112(e) or 114(d)(2) of title 17, means the Statements of Account and appropriate protective order. United States Code, or both. Such efforts any information contained therein, (d) Safeguarding Confidential must include, but not be limited to, including the amount of royalty Information. The Collective and any searches in Copyright Office public payments and the number of person authorized to receive records and published directories of Performances, and any information Confidential Information from the sound recording copyright owners. pertaining to the Statements of Account Collective must implement procedures (b) Unclaimed funds. If the Collective reasonably designated as confidential by to safeguard against unauthorized access is unable to identify or locate a the party submitting the statement. to or dissemination of Confidential Copyright Owner or Performer who is Confidential Information does not Information using a reasonable standard entitled to receive a royalty distribution include documents or information that of care, but no less than the same degree under this part 380, the Collective must at the time of delivery to the Collective of security that the recipient uses to retain the required payment in a is public knowledge. The party seeking protect its own Confidential Information segregated trust account for a period of information from the Collective based or similarly sensitive information. three years from the date of the first on a claim that the information sought distribution of royalties from the is a matter of public knowledge shall § 380.6 Auditing payments and relevant payment by a Licensee. No have the burden of proving to the distributions. claim to distribution shall be valid after Collective that the requested (a) General. This section prescribes the expiration of the three-year period. information is in the public domain. procedures by which any entity entitled After expiration of this period, the (b) Use of Confidential Information. to receive payment or distribution of Collective must handle unclaimed funds The Collective may not use any royalties may verify payments or in accordance with applicable federal, Confidential Information for any distributions by auditing the payor or state, or common law. purpose other than royalty collection distributor. The Collective may audit a (c) Retention of records. Licensees and distribution and activities related Licensee’s payments of royalties to the and the Collective shall keep books and directly thereto. Collective, and a Copyright Owner or records relating to payments and (c) Disclosure of Confidential Performer may audit the Collective’s distributions of royalties for a period of Information. The Collective shall limit distributions of royalties to the owner or not less than the prior three calendar access to Confidential Information to: performer. Nothing in this section shall years. (1) Those employees, agents, preclude a verifying entity and the (d) Designation of the Collective. (1) consultants, and independent payor or distributor from agreeing to The Judges designate SoundExchange, contractors of the Collective, subject to verification methods in addition to or

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00093 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26408 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

different from those set forth in this underpaid royalties, the payor or the service’s ATH would equal 10 section. distributor shall remit the amount of hours. (b) Frequency of auditing. The any underpayment determined by the Collective means the collection and verifying entity may conduct an audit of auditor to the verifying entity, together distribution organization that is each licensee only once a year for any with interest at the rate specified in designated by the Copyright Royalty or all of the prior three calendar years. § 380.2(d). In the absence of mutually- Judges, and which, for the current rate A verifying entity may not audit records agreed payment terms, which may, but period, is SoundExchange, Inc. for any calendar year more than once. need not, include installment payments, Commercial Webcaster means a (c) Notice of intent to audit. The the payor or distributor shall remit Licensee, other than a Noncommercial verifying entity must file with the promptly to the verifying entity the Webcaster or Public Broadcaster, that Copyright Royalty Judges a notice of entire amount of the underpayment makes Ephemeral Recordings and intent to audit the payor or distributor, determined by the auditor. If the auditor eligible digital audio transmissions of which notice the Judges must publish in determines the payor or distributor sound recordings pursuant to the the Federal Register within 30 days of overpaid royalties, however, the statutory licenses under 17 U.S.C. 112(e) the filing of the notice. Simultaneously verifying entity shall not be required to and 114(d)(2). with the filing of the notice, the remit the amount of any overpayment to Copyright owners means sound verifying entity must deliver a copy to the payor or distributor, and the payor recording copyright owners who are the payor or distributor. or distributor shall not seek by any entitled to royalty payments made (d) The audit. The audit must be means to recoup, offset, or take a credit under Part 380 pursuant to the statutory conducted during regular business for the overpayment, unless the payor or licenses under 17 U.S.C. 112(e) and 114. hours by a Qualified Auditor who is not distributor and the verifying entity have Digital audio transmission has the retained on a contingency fee basis and agreed otherwise. same meaning as in 17 U.S.C. 114(j). is identified in the notice. The auditor (h) Paying the costs of the audit. The Eligible nonsubscription transmission shall determine the accuracy of royalty has the same meaning as in 17 U.S.C. payments or distributions, including verifying entity must pay the cost of the verification procedure, unless the 114(j). whether an underpayment or Eligible Transmission means a overpayment of royalties was made. An auditor determines that there was an underpayment of 10% or more, in subscription or nonsubscription audit of books and records, including transmission made by a Licensee that is underlying paperwork, performed in the which case the payor or distributor must bear the reasonable costs of the subject to licensing under 17 U.S.C. ordinary course of business according to 114(d)(2) and the payment of royalties generally accepted auditing standards verification procedure, in addition to paying or distributing the amount of any under this part. by a Qualified Auditor, shall serve as an Ephemeral recording has the same underpayment. acceptable verification procedure for all meaning as in 17 U.S.C. 112. parties with respect to the information (i) Retention of audit report. The Licensee means a Commercial that is within the scope of the audit. verifying party must retain the report of Webcaster, a Noncommercial Webcaster, (e) Access to third-party records for the audit for a period of not less than or any entity operating a noninteractive audit purposes. The payor or distributor three years from the date of issuance. Internet streaming service that has must use commercially reasonable § 380.7 Definitions. obtained a license under Section 112 or efforts to obtain or to provide access to 114 to transmit eligible sound any relevant books and records Aggregate Tuning Hours (ATH) means recordings. maintained by third parties for the the total hours of programming that the New subscription service has the same purpose of the audit. Licensee has transmitted during the meaning as in 17 U.S.C. 114(j). (f) Duty of auditor to consult. The relevant period to all listeners within Noncommercial webcaster has the auditor must produce a written report to the United States from all channels and same meaning as in 17 U.S.C. the verifying entity. Before rendering stations that provide audio 114(f)(5)(E). the report, unless the auditor has a programming consisting, in whole or in Nonsubscription has the same reasonable basis to suspect fraud on the part, of eligible nonsubscription meaning as in 17 U.S.C. 114(j). part of the payor or distributor, the transmissions or noninteractive digital Performance means each instance in disclosure of which would, in the audio transmissions as part of a new which any portion of a sound recording reasonable opinion of the auditor, subscription service, less the actual is publicly performed to a listener by prejudice any investigation of the running time of any sound recordings means of a digital audio transmission suspected fraud, the auditor must for which the Licensee has obtained (e.g., the delivery of any portion of a review tentative written findings of the direct licenses apart from 17 U.S.C. single track from a compact disc to one audit with the appropriate agent or 114(d)(2) or which do not require a listener), but excludes the following: employee of the payor or distributor in license under United States copyright (1) A performance of a sound order to remedy any factual errors and law. By way of example, if a service recording that does not require a license clarify any issues relating to the audit; transmitted one hour of programming (e.g., a sound recording that is not Provided that an appropriate agent or containing Performances to 10 listeners, copyrighted); employee of the payor or distributor the service’s ATH would equal 10 (2) A performance of a sound reasonably cooperates with the auditor hours. If three minutes of that hour recording for which the service has to remedy promptly any factual error[s] consisted of transmission of a directly- previously obtained a license from the or clarify any issues raised by the audit. licensed recording, the service’s ATH Copyright Owner of such sound The auditor must include in the written would equal nine hours and 30 minutes recording; and report information concerning the (three minutes times 10 listeners creates (3) An incidental performance that cooperation or the lack thereof of the a deduction of 30 minutes). As an both: employee or agent. additional example, if one listener (i) Makes no more than incidental use (g) Audit results; underpayment or listened to a service for 10 hours (and of sound recordings including, but not overpayment of royalties. If the auditor none of the recordings transmitted limited to, brief musical transitions in determines the payor or distributor during that time was directly licensed), and out of commercials or program

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00094 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations 26409

segments, brief performances during Webcaster must pay is $50,000. The such month occurs, pay royalties in news, talk and sports programming, minimum fee is nonrefundable. accordance, and otherwise comply, with brief background performances during (c) Annual royalty fee adjustment. the provisions of Part 380 Subparts A disk jockey announcements, brief The Copyright Royalty Judges shall and B applicable to Noncommercial performances during commercials of adjust the royalty fees each year to Webcasters; sixty seconds or less in duration, or reflect any changes occurring in the cost (2) The Minimum Fee paid by the brief performances during sporting or of living as determined by the most Noncommercial Educational Webcaster other public events; and recent Consumer Price Index (for all for such calendar year will be credited (ii) Does not contain an entire sound consumers and for all items) (CPI–U) to the amounts payable under the recording, other than ambient music published by the Secretary of Labor provisions of Part 380 Subparts A and that is background at a public event, and before December 1 of the preceding B applicable to Noncommercial does not feature a particular sound year. The adjusted rate shall be rounded Webcasters; and to the nearest fourth decimal place. To recording of more than thirty seconds (3) The Noncommercial Educational account more accurately for cumulative (as in the case of a sound recording used Webcaster shall, within 45 days after the changes in the CPI–U over the rate as a theme song). end of each month, notify the Collective period, the calculation of the rate for Performers means the independent if it has made total transmissions in each year shall be cumulative based on administrators identified in 17 U.S.C. excess of 159,140 Aggregate Tuning a calculation of the percentage increase 114(g)(2)(B) and (C) and the parties Hours on a channel or station during in the CPI–U from the CPI–U published identified in 17 U.S.C. 114(g)(2)(D). that month; pay the Collective any in November, 2015 (237.336), according Public broadcaster means a Covered amounts due under the provisions of to the formula (1 + (Cy ¥ 237.336)/ Entity under subpart D of this part. Part 380 Subparts A and B applicable to 237.336) × R , where C is the CPI– Qualified auditor means an 2016 y Noncommercial Webcasters; and independent Certified Public U published by the Secretary of Labor before December 1 of the preceding provide the Collective a statement of Accountant licensed in the jurisdiction account pursuant to part 380, subpart A. where it seeks to conduct a verification. year, and R2016 is the royalty rate for 2016 (i.e., $0.0022 per subscription (c) Royalties for other Noncommercial Transmission has the same meaning Educational Webcasters. A as in 17 U.S.C. 114(j). performance or $0.0017 per nonsubscription performance). By way Noncommercial Educational Webcaster ■ 4. Revise subpart B, consisting of of example, if the CPI–U published in that is not eligible to pay royalties under § 380.10, to read as follows: November 2016 is 242.083, the adjusted paragraph (a) of this section shall pay rate for nonsubscription services in royalties in accordance, and that Subpart B—Commercial Webcasters otherwise comply, with the provisions and Noncommercial Webcasters 2017 will be computed as (1 + (242.083 ¥ 237.336)/237.336) × $0.0017 and will of subparts A and B of this part § 380.10 Royalty fees for the public equal $0.00173 ($0.0017 when rounded applicable to Noncommercial performance of sound recordings and the to the nearest fourth decimal place). If Webcasters. making of ephemeral recordings. the CPI–U published in November 2017 * * * * * (a) Royalty fees. For the year 2016, is 249.345, the rate for nonsubscription ■ 6. In § 380.23, revise paragraph (b)(1) Licensees must pay royalty fees for all services for 2018 will be computed as (1 to read as follows: Eligible Transmissions of sound + (249.345 ¥ 237.336)/237.336) × recordings at the following rates: $0.0017 and will equal $0.00179 § 380.23 Terms for making payment of (1) Commercial Webcasters: $0.0022 ($0.0018 when rounded to the nearest royalty fees and statements of account. per performance for subscription fourth decimal place). The Judges shall * * * * * services and $0.0017 per performance publish notice of the adjusted fees in the (b) Designation of the Collective. (1) for nonsubscription services. Federal Register at least 25 days before The Copyright Royalty Judges designate (2) Noncommercial webcasters. $500 January 1. The adjusted fees shall be SoundExchange, Inc., as the Collective per year for each channel or station and effective on January 1. to receive statements of account and $0.0017 per performance for all digital (d) Ephemeral recordings royalty fees. royalty payments from Noncommercial audio transmissions in excess of The fee for all Ephemeral Recordings is Educational Webcasters due under 159,140 ATH in a month on a channel part of the total fee payable under this § 380.22 and to distribute royalty or station. section and constitutes 5% of it. All payments to each Copyright Owner and (b) Minimum fee. Licensees must pay ephemeral recordings that a Licensee Performer, or their designated agents, the Collective a minimum fee of $500 makes which are necessary and entitled to receive royalties under 17 each year for each channel or station. commercially reasonable for making U.S.C. 112(e) or 114(g). noninteractive digital transmissions are The Collective must apply the fee to the * * * * * Licensee’s account as credit towards any included in the 5%. ■ additional royalty fees that Licensees 5. In § 380.22, revise paragraphs (b)(1) Subpart D—Public Broadcasters may incur in the same year. The fee is through (3) and (c) to read as follows: payable for each individual channel and § 380.22 Royalty fees for the public ■ 7. Revise the heading of Subpart D to each individual station maintained or performance of sound recordings and for read as set forth above. operated by the Licensee and making ephemeral recordings. ■ 8. In § 380.33, revise paragraph (b)(1) Eligible Transmissions during each * * * * * to read as follows: calendar year or part of a calendar year (b) * * * during which it is a Licensee. The (1) The Noncommercial Educational § 380.33 Terms for making payment of maximum aggregate minimum fee in Webcaster shall, for such month and the royalty fees and statements of account. any calendar year that a Commercial remainder of the calendar year in which * * * * *

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00095 Fmt 4701 Sfmt 4700 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2 26410 Federal Register / Vol. 81, No. 84 / Monday, May 2, 2016 / Rules and Regulations

(b) Designation of the Collective. (1) designated agents, entitled to receive Dated: April 19, 2016. The Copyright Royalty Judges designate royalties under 17 U.S.C. 112(e) or Suzanne M. Barnett, SoundExchange, Inc., as the Collective 114(g). Chief Copyright Royalty Judge. to receive statements of account and * * * * * Jesse M. Feder, royalty payments for Covered Entities Copyright Royalty Judge. under this subpart and to distribute David R. Strickler, royalty payments to each Copyright Copyright Royalty Judge. Owner and Performer, or their Approved By: David S. Mao, Librarian of Congress. [FR Doc. 2016–09707 Filed 4–29–16; 8:45 am] BILLING CODE 1410–72–P

VerDate Sep<11>2014 20:22 Apr 29, 2016 Jkt 238001 PO 00000 Frm 00096 Fmt 4701 Sfmt 9990 E:\FR\FM\02MYR2.SGM 02MYR2 mstockstill on DSK5VPTVN1PROD with RULES2