IN RE PANDORA MEDIA, INC. 317 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

225 F.Supp.2d 190, 213 (N.D.N.Y.2002) sonable fees and terms for through-to-the- (‘‘In order to hold an individual liable un- audience (TTTA) blanket license to per- der [the aiding and abetting provision], TTT form musical compositions in repertoire of plaintiff must also show that the individual performing-rights organization (PRO), aided or abetted a primary violation of the pursuant to consent decree. The court pro- [NYS]HRL committed by another employ- hibited PRO from withdrawing rights from ee or the business itself.’’ (alterations in petitioner to perform any compositions original) (internal quotation marks omit- over which PRO retained any licensing ted)). Because plaintiff’s underlying retal- rights, 2013 WL 5211927. iation claim under NYSHRL has been dis- missed or otherwise abandoned, any claim Holdings: After conducting a bench trial, she seeks to assert against Heifferon as an the District Court, Denise Cote, J., held aider and abettor of such alleged retaliato- that: ry conduct fails as a matter of law. (1) TTTA blanket license rate of 1.85% of Plaintiff’s claims under NYSHRL revenue was reasonable for every year against Heifferon in her individual capacity of license term and are dismissed. (2) service was not entitled to rate of IV. CONCLUSION 1.70% of revenue. For all of the foregoing reasons, Defen- Ordered accordingly. dants’ motion for summary judgment dis- missing the Amended Complaint in its en- tirety is granted. 1. Copyrights and Intellectual Property IT IS SO ORDERED. O48.1 ‘‘Non-interactive’’ digital music ser- , vices are eligible for a compulsory or stat- utory licensing fee set by the Copyright Royalty Board (CRB) made up of Copy- In re Petition of PANDORA right Royalty Judges appointed by the Li- MEDIA, INC. brary of Congress, whereas interactive services must independently negotiate Related to United States of rates for sound recording licenses. 17 America, Plaintiff, U.S.C.A. § 114. v. 2. Copyrights and Intellectual Property American Society of Composers, O Authors, and Publishers, 48.1 Defendant. Through-to-the-audience (TTTA) blan- Nos. 12 Civ. 8035(DLC), ket license rate of 1.85% of revenue of 41 Civ. 1395(DLC). customized service was rea- sonable for every year of compulsory li- United States District Court, cense term for performance of musical S.D. New York. compositions in repertoire of performing- Signed March 14, 2014. rights organization (PRO), and service also Filed March 18, 2014. was entitled to take deduction for any Background: Customized Internet radio direct payments made to publishers follow- service petitioned for determination of rea- ing their partial withdrawals from PRO, 318 6 FEDERAL SUPPLEMENT, 3d SERIES since, among other things, there was tion (PRO) to a customized Internet radio strong basis to recognize presumption that service. 17 U.S.C.A. § 114(i). 1.85% rate would be reasonable rate, adop- 7. Copyrights and Intellectual Property tion of escalating rate would have been out O48.1 of step with historical practice, and other Customized Internet radio service was licenses that had been used for comparison not entitled to rate of 1.70% of revenue were not competitive, fair market rates. that commercial radio stations had to pay 3. Copyrights and Intellectual Property (RMLC rate) for performance of musical O 48.1 compositions in repertoire of performing- When determining the reasonableness rights organization (PRO), since RMLC of a compulsory licensing fee, a rate-set- rate applied to large-scale compulsory li- ting court must attempt to approximate cense agreement that bound variety of li- the fair market value of a license, i.e., what censees in both terrestrial and internet a license applicant would pay in an arm’s radio sphere, internet radio sphere was length transaction; in so doing, the court very small, governing consent decree for- must take into account the fact that a bad discrimination only among licensees performing rights organization (PRO), as a and service was not similarly situated to monopolist, exercises market-distorting any RMLC licensee, and service was simi- power in negotiations for the use of its larly situated to internet music services music. covered by license at rate of 1.85%. 4. Copyrights and Intellectual Property O48.1 When a rate-setting court determines the reasonableness of a compulsory licens- Kenneth L. Steinthal, Joseph R. Wetzel, ing fee, fair market value is a hypothetical Jason Blake Cunningham, Katherine matter; the appropriate analysis ordinarily Merk, King & Spalding, LLP, San Fran- seeks to define a rate or range of rates cisco, CA, Jeffrey Scott Seddon, King & that approximates the rates that would be Spalding, LLP, New York, NY, Mary set in a competitive market. Katherine Bates, King & Spalding, LLP, 5. Copyrights and Intellectual Property Atlanta, GA, Marc Brian Collier, Fulbright O48.1 & Jaworski LLP, Austin, TX, for applicant In rate court compulsory licensing Pandora Media, Inc. proceedings, a determination of the fair Jay Cohen, Eric Alan Stone, Darren W. market value often is facilitated by the use Johnson, Amy E. Gold, Lynn Beth Bayard, of a benchmark, i.e., reasoning by analogy Paul, Weiss, Rifkind, Wharton & Garrison, to an agreement reached after arm’s LLP, New York, NY, Richard H. Reimer, length negotiation between similarly situ- Christine A. Pepe, American Society of ated parties. Composers, Authors and Publishers, New 6. Copyrights and Intellectual Property York, NY, for the American Society of O48.1 Composers, Authors and Publishers. A rate-setting court may not take the rates set by the Copyright Royalty Board OPINION & ORDER (CRB) into account when determining the DENISE COTE, District Judge. fair market rate for a public performance license from a performing rights organiza- TABLE OF CONTENTS IN RE PANDORA MEDIA, INC. 319 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

INTRODUCTIONTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT320

FINDINGS OF FACT TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT322

I. The American Society of Composers, Authors and Publishers TTTTTTTTTTTT322 A. ASCAP Background TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT322 B. The ASCAP Consent DecreeTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT323

II. The Evolution of the Radio Industry TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT323

III. The RMLC–ASCAP License Agreement for the Period 2010–2016 TTTTTTTT325

IV. PandoraTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT327 A. Pandora’s Music Genome ProjectTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT327 B. Pandora Premieres TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT328 C. Pandora’s Comedy ProgrammingTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT328 D. Pandora’s Revenue TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT328 E. Pandora’s Competitive Environment TTTTTTTTTTTTTTTTTTTTTTTTTTTTTT328

V. Pandora’s Licensing History with ASCAP TTTTTTTTTTTTTTTTTTTTTTTTTTTTT330

VI. The April 2011 ASCAP Compendium Modification TTTTTTTTTTTTTTTTTTTTTT331 A. Overview and Context TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT331 B. Public Performance Rights for Compositions versus Sound Recordings TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT332 C. ASCAP–Publisher Negotiations Prior to the Compendium Modification TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT333 D. The Compendium Modification Allowing New Media Withdrawals is Enacted TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT336 E. ASCAP Provides Administrative Services for Withdrawing PublishersTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT337

VII. A Second Compendium Modification in December 2012 the ‘‘Standard Services’’ AgreementTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT338

VIII. Pandora Negotiates Direct Licenses with EMI, Sony, and UMPG and Fails to Negotiate an Agreement with ASCAP TTTTTTTTTTTTTTTTTTTTTTT339 A. The Pandora–EMI License Negotiations TTTTTTTTTTTTTTTTTTTTTTTTTT339 B. The Pandora–ASCAP License NegotiationsTTTTTTTTTTTTTTTTTTTTTTTT340 C. The Pandora–Sony License Negotiations TTTTTTTTTTTTTTTTTTTTTTTTTT342 D. The Pandora–UMPG License Negotiations TTTTTTTTTTTTTTTTTTTTTTTT347

IX. September 17 Partial Summary Judgment OpinionTTTTTTTTTTTTTTTTTTTTTT350

X. Other Licensing Agreements Put Forth as Benchmarks TTTTTTTTTTTTTTTTT351 A. The Pandora–SESAC License TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT351 B. Apple’s iTunes Radio Licenses with Publishers and PROs TTTTTTTTTTT351

CONCLUSIONS OF LAW TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT353

I. ASCAP’s Rate Proposal of 1.85% for 2011 and 2012 TTTTTTTTTTTTTTTTTTTTT355

II. ASCAP’s Rate Proposal of 2.50% for 2013 and 3.00% for 2014 and 2015 TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT355 A. Presumption of a Single Rate TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT356 B. Pandora’s Direct Licenses with Sony and UMPG TTTTTTTTTTTTTTTTTTT357 1. ASCAP and Publisher Coordination TTTTTTTTTTTTTTTTTTTTTTTTTTT357 2. The Pandora–Sony License TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT358 3. The Pandora–UMPG License TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT360 320 6 FEDERAL SUPPLEMENT, 3d SERIES

C. ASCAP’s Secondary Benchmarks: the SESAC and Apple Licenses TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT361 1. The Pandora–SESAC licenseTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT361 2. The Apple Licenses TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT362 D. ASCAP’s Theoretical Arguments and Motivations TTTTTTTTTTTTTTTTTT363 1. An Increase in Competition TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT363 2. Demand for VarietyTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT364 3. Disparity Between Sound Recording and Composition FeesTTTTT366 4. Cannibalization of Music SalesTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT367 5. Music IntensityTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT368 6. Pandora’s Success TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT368

III. Whether Pandora is Entitled to the RMLC 1.70% RateTTTTTTTTTTTTTTTTTT369

IV. Publisher Concerns Regarding the Consent Decree and the Rate Court TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT372

CONCLUSION TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTT372

INTRODUCTION was entered into after EMI purported to Pandora Media Inc. (‘‘Pandora’’) has ap- withdraw its new media 1 licensing rights plied for a through-to-the-audience blanket from ASCAP in 2011. license to perform the musical composi- ASCAP proposes a variety of bench- tions in the repertoire of the American marks, including the direct licensing Society of Composers, Authors and Pub- agreement into which Pandora entered lishers (‘‘ASCAP’’) for the period of Janu- with EMI, as well as Pandora’s direct li- ary 1, 2011 through December 31, 2015. The parties having been unable to reach censes with Sony/ATV Music Publishing agreement on an appropriate licensing fee, LLC (‘‘Sony’’) and Universal Music Pub- pursuant to Article IX of the consent de- lishing Group (‘‘UMPG’’) in the wake of cree under which ASCAP operates— those publishers’ putative withdrawals of known as the Second Amended Final new media licensing rights from ASCAP. Judgment (‘‘AFJ2’’), see United States v. ASCAP also puts forward other agree- ASCAP, Civ. No. 41–Civ–1395, 2001 WL ments between music rights holders and 1589999 (S.D.N.Y. June 11, 2001)—Pando- music users as secondary benchmarks. ra requested on November 5, 2012 that The parties have proposed the following this Court set a rate for that licensing fee. rates, expressed as a percentage of reve- The parties disagree as to which are the nue: ASCAP proposes a rate of 1.85% for most appropriate benchmarks for the li- the years 2011 and 2012, 2.50% for 2013, cense rate here. Pandora asserts princi- and 3.00% for the years 2014 and 2015. pally that it is similarly situated to radio Pandora proposes a rate of 1.70% for all stations licensed through a 2012 agree- five years. This Opinion sets the rate for ment between the Radio Music License all five years at 1.85%. Committee (‘‘RMLC’’), which represents commercial radio stations, and ASCAP, The task at hand is to determine the fair and is therefore entitled to the rate in that market value of a blanket license for the license. Pandora also points to a direct public performance of music. As this license agreement between Pandora and Court explained in a prior rate court pro- EMI Music Publishing Ltd. (‘‘EMI’’) that ceeding:

1. New media is defined below, but the term refers generally to internet transmissions. IN RE PANDORA MEDIA, INC. 321 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

The challenges of [determining a fair mento; Vice President for New Media and market rate for a blanket music license] Technology Matthew DeFilippis; Director include discerning a rate that will give of Licensing Vincent Candilora; and Sen- composers an economic incentive to keep ior Vice President Seth Saltzman. AS- enriching our lives with music, that CAP’s music publisher witnesses were Pe- avoids compensating composers for con- ter Brodsky, the Executive Vice President tributions made by others either to the of Business and Legal Affairs at Sony; creative work or to the delivery of that and Zach Horowitz, the Chairman and work to the public, and that does not CEO of UMPG. ASCAP’s composer-wit- create distorting incentives in the mar- ness was Brett James. ASCAP’s experts ketplace that will improperly affect the were Dr. Kevin Murphy, Dr. Orley Ashen- choices made by composers, inventors, felter, and Robin Flynn. ASCAP also pro- investors, consumers and other economic vided designated deposition excerpts from players. six witnesses.3 In re Application of MobiTV, Inc., 712 Pandora provided affidavits constituting F.Supp.2d 206, 209 (S.D.N.Y.2010), aff’d the direct testimony of eight witnesses, sub nom. ASCAP v. MobiTV, Inc., 681 four of whom are current or former Pan- F.3d 76 (2d Cir.2012). dora employees and four of whom are ex- A bench trial was held from January 21 perts. The current or former Pandora through February 10, 2014. Without ob- employees were founder and Chief Strate- jection from the parties, the trial was con- gy Officer Timothy Westergren; former ducted in accordance with the Court’s cus- CEO Joseph Kennedy; Chief Technology tomary practices for non-jury proceedings, Officer Thomas Conrad; and Chief Mar- which includes taking direct testimony keting Officer Simon Fleming–Wood. from witnesses under a party’s control Pandora’s experts were Dr. Leslie Marx, through affidavits submitted with the Joint Dr. Roger Noll, William Rosenblatt, and Pretrial Order. The parties also served Fred McIntyre. Pandora also provided with the Joint Pretrial Order copies of all the designated deposition testimony of a exhibits and deposition testimony that they number of witnesses.4 intended to offer as evidence in chief at The parties also offered deposition des- trial. ignations of certain witnesses as joint ex- Prior to trial, ASCAP presented affida- hibits.5 At the trial, the parties waived vits constituting the direct testimony from their right to cross-examine several of the ten witnesses, including four ASCAP em- witnesses. The witnesses who testified at ployees, two music publishing executives, trial were Brodsky, LoFrumento, DeFilip- one composer, and three experts.2 The pis, Murphy, Flynn, Marx, Rosenblatt, ASCAP employees are CEO John LoFru- Horowitz, Saltzman, Noll, Conrad, Kenne-

2. ASCAP initially submitted testimony from a 4. The witnesses who had excerpts of their fourth expert, Timothy Hanlon, on the issue of depositions offered by Pandora were Orley a proper advertising costs deduction, but the Ashenfelter, Martin Bandier, Peter Boyle, Pe- parties settled this issue before trial. ter Brodsky, Vincent Candilora, Matthew De- Filippis, Roger Faxon, Wayland Holyfield, 3. The witnesses who had excerpts of their John LoFrumento, Seth Saltzman, Raymond depositions offered by ASCAP were Tom Con- Schwind, and Paul Williams. rad, Simon Fleming–Wood, Michael Herring, 5. Those were the deposition designations for John Kennedy, John Trimble, and Timothy Richard Conlon, J.D. Connell, Zach Horowitz, Westergren. Robert Rosenbloum, and William Velez. 322 6 FEDERAL SUPPLEMENT, 3d SERIES

dy, Fleming–Wood, and McIntyre. In ad- ably among the multiple owners of the dition, Pandora called its outside counsel, public performance copyrights in each Robert Rosenbloum, as a witness. work. The ability of ASCAP and other This Opinion constitutes the Court’s performing rights organizations (‘‘PROs’’) findings of fact and conclusions of law to grant licenses covering a large number following that trial. The factual findings of compositions creates significant econo- are principally set forth in the first section mies of scale in the market for music of this Opinion, but appear as well in the licensing. second section. ASCAP offers the option of blanket li- censes to users. A blanket license is a FINDINGS OF FACT license that gives the music user the right I. The American Society of Composers, to perform all of the works in ASCAP’s Authors and Publishers repertoire, the fee for which does not vary depending on how much of the music the A. ASCAP Background user actually uses. These blanket licenses ASCAP is an unincorporated member- reduce the costs of licensing copyrighted ship organization of music copyright hold- musical compositions. They eliminate ers created and controlled by music writ- costly, multiple negotiations of the vari- ers and publishers.6 Its function is to ous rights and provide an efficient coordinate the licensing of copyrighted means of monitoring the use of musical musical works, and the distribution of roy- compositions. They also allow users of alties, on behalf of its nearly 500,000 mem- copyrighted music to avoid exposure to bers. ASCAP members grant ASCAP the liability for copyright infringement. non-exclusive right to license non-dramatic Buffalo Broadcasting Co. v. ASCAP, 744 public performances of their music. AS- F.2d 917, 934 (2d Cir.1984) (Winter, J., CAP licenses these works on behalf of the copyright holders to a broad array of mu- concurring). sic users, including television networks, ra- ASCAP’s board of directors is com- dio stations, digital music services, col- prised of an equal number of composers leges, restaurants, and many other venues and music publishers. The head of the in which music is performed. ASCAP board is typically a songwriter. The present head of ASCAP is composer Employing ASCAP to perform these Paul Williams. functions is efficient for both music users and copyright holders. A music user can ASCAP competes with two other United license an enormous portfolio of copyright- States PROs: Broadcast Music, Inc. ed music through the execution of a single (‘‘BMI’’) and SESAC, LLC. (‘‘SESAC’’), license without having to contact each each of whom also offers blanket licenses. copyright holder. Copyright holders bene- BMI, which is slightly smaller than AS- fit from ASCAP’s expertise and resources CAP, operates under a consent decree that in policing the market, negotiating licens- is similar to the one that governs ASCAP’s es, and distributing the revenue from a licenses. SESAC is a PRO that is not vast array of licenses promptly and reli- currently bound by any consent decree.7

6. A music publisher is an entity which coordi- cordings of performances of copyrighted com- nates licensing and other logistics pertaining positions. to copyrighted compositions. Music publish- ers are distinguished from record labels, 7. There was reference at trial to ongoing anti- which coordinate licensing of the sound re- trust litigation concerning SESAC. IN RE PANDORA MEDIA, INC. 323 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

B. The ASCAP Consent Decree II. The Evolution of the Radio Industry Since 1941, ASCAP has operated under Much of the focus at trial was on the a consent decree stemming from a Depart- question of whether Pandora can be prop- ment of Justice antitrust lawsuit. This erly classified as ‘‘radio.’’ A description of consent decree has been modified from the evolution of the radio industry will time to time. The most recent version of provide context in understanding Pando- the consent decree was issued in 2001 and ra’s features and its place within the music 8 is known as ‘‘AFJ2.’’ AFJ2 governs here. business. In an attempt to ameliorate the anti- Radio is a form of media in which a competitive concerns raised by ASCAP’s provider transmits audio programming to consolidation of music licenses, AFJ2 re- a listener, where the programming is not stricts how ASCAP may issue licenses in a directly selected by the listener but is variety of ways. First, AFJ2 provides a mechanism whereby a court, known as the programmed by the provider. As a result, rate court, will determine a reasonable fee in the context of a music station, the listen- for ASCAP licenses when ASCAP and an er does not choose the songs and does not applicant for a license cannot reach an know what composition will be played next. agreement. AFJ2 § IX. This Court is This radio experience has remained con- presently the ASCAP rate court. Second, stant through the years, regardless of AFJ2 requires ASCAP to grant a license whether radio programming is transmitted to perform all of the musical compositions by broadcasting, through a cable, from a in ASCAP’s repertoire to any entity that satellite, or over the internet. requests such a license. AFJ2 §§ VI, Radio made its debut approximately a IX(E). And third, AFJ2 prevents ASCAP century ago and has been a dominant force from discriminating in pricing or with re- in the music industry ever since. The first spect to other terms or conditions between commercial radio station in the United ‘‘similarly situated’’ licensees. AFJ2 States was located in Pittsburgh, Pennsyl- § IV(C). ASCAP members agree to be vania and was licensed in 1920. The origi- bound in the exercise of their copyright nal technological means for delivering ra- rights by the terms of AFJ2. For example, dio programming was by broadcasting an the 1996 Agreement Between Sony and ‘‘amplitude modulation,’’ or ‘‘AM’’ signal. ASCAP provides that ‘‘[t]he grant [of By the 1930s, ‘‘frequency modulation’’ or rights to ASCAP] TTT is modified by and ‘‘FM’’ signal technology was developed. subject to the provisions of [AFJ2].’’ FM broadcasting offered better audio In addition to operating under a consent quality but over a smaller range. decree, ASCAP is governed by a series of internal rules and contracts. The most Another important moment in the histo- important internal rule set for purposes of ry of radio occurred with the passage of this litigation is the ASCAP Compendium. the Telecommunications Act of 1996. The ASCAP Compendium can be modified Pub.L. No. 104–104, 110 Stat. 56 (1996). by the ASCAP Board and reflects many of Empowered by that legislation, the FCC the important rules that govern ASCAP’s eliminated most caps on the number of obligations to its copyright holder mem- stations that a single company could own. bers and vice versa. Following that change in the law, there

8. For background discussion of AFJ2, see 2014 WL 812795, at *11–12 (S.D.N.Y. Mar. 3, generally Meredith Corp. v. SESAC LLC, 09 2014). Civ. 9177(PAE), 1 F.Supp.3d 180, 196–98, 324 6 FEDERAL SUPPLEMENT, 3d SERIES

was a large-scale expansion of group own- Clear Channel and CBS Radio, two ma- ership of stations. Many terrestrial radio jor commercial radio companies, launched stations are now owned by large conglom- their own internet radio services in 2008 erates, such as Clear Channel Communica- and 2010, respectively. Clear Channel’s tions, Inc. (‘‘Clear Channel’’), which owns internet radio service is called iHeartRa- over 800 stations. dio, and began as a vehicle to simulcast In the 1990s, the first successful national Clear Channel’s own stations. cable radio network was launched, using The arrival of the internet as a radio cable TV transmission lines. Over time, delivery platform has also permitted radio what came to be known as digital TV radio providers to introduce a level of instanta- was transmitted through means of cable, neous user interactivity for the first time. 9 satellite, and telephone-company lines. With the internet, each listener’s device Some of the major competitors in this gets its own data stream, in contrast to the market are Music Choice, SiriusXM Satel- broadcasting of a common signal across a 10 lite Radio, Muzak, and DMX. geographic area. As will be explained in Also in the 1990s, the nascent internet greater detail below, this permits internet provided a new means of radio transmis- radio services to offer customized music sion. The introduction in the early 1990s programming based on user feedback. of MP3 digital audio encoding and com- Thus, while a listener to a customized ra- pression format permitted music to be dio service cannot select and does not compressed in way that facilitated distri- know what song will be played next, that 11 bution over the internet. In 1994, the listener can often give feedback to the first simulcast of an AM/FM broadcast customized station to shape the nature of occurred over the internet. As of today, the music that will be played. over 10,000 AM and FM stations stream online. Internet radio includes not just As of today, Pandora is the most suc- the simulcasting of signals broadcast by cessful customized radio service. But it AM and FM stations, but also the creation was not the first. Prior to Pandora’s of internet-only radio stations. Over time, launch in 2005, LAUNCHcast and Last.fm, some independent companies built directo- two customized radio services, began in ries of internet radio stations. These di- 1999 and 2002, respectively.12 Recently, rectories can contain tens of thousands of three major competitors have emerged as radio stations. Thus, the internet has en- challengers to Pandora’s dominance. In abled providers to present listeners with a 2011, Clear Channel launched a customized vast library of radio programming, the radio offering within its iHeartRadio ser- likes of which has never been available vice, called ‘‘Create Station.’’ Spotify before. launched a customized radio feature called

9. Satellite radio permits coast-to-coast nation- purchased in physical form, whether a vinyl wide programming. It is primarily directed LP or CD, to a digital world in which digital at the automotive market. downloads and digital music streaming are major forces. Apple’s iTunes Store was 10. High Definition radio—a digital radio launched in 2003 and established a main- technology which piggybacks on existing stream market for the purchase of digital mu- AM/FM signals but cannot be received by sic files. traditional radios—was approved by the FCC in 2002. 12. These services have been acquired, respec- 11. With the digital age, the music world has tively, by MTV, Yahoo! and by CBS. transitioned from one in which music must be IN RE PANDORA MEDIA, INC. 325 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

Spotify Radio in late 2011. And in Sep- cludes music delivered over radio stations tember 2013, Apple launched its custom- playing through TV cable systems and ized radio service called ‘‘iTunes Radio.’’ 13 over the internet. Almost half of radio In addition to programmed and custom- listening occurs while the listener is in an ized radio, the presence of digital technolo- automobile. The other 20% or so of music gy and the internet have allowed for the listening in the United States is experi- emergence of a third means of delivering enced by listeners who seek more control music: ‘‘on-demand’’ streaming services. over the music that they hear, whether These services provide users with access through the purchase and playing of a to large libraries of songs, from which they record album or a CD, or the subscription can select exactly which song to play at to an on-demand digital music service such any time. A leading on-demand service is as Spotify. Spotify, which had 24 million active users globally as of March 2013. Launched in III. The RMLC–ASCAP License Agree- 2008 in Europe and in the United States in ment for the Period 2010–2016 2011, Spotify has a library of over 20 mil- Much of the radio industry obtains its lion songs.14 Other popular services with license for the public performance of AS- on-demand offerings include Rhapsody and CAP music through the RMLC, which is a Grooveshark. The most popular on-de- trade association that represents the com- mand services offer both advertising sup- mercial radio industry. Between 2003 and ported and subscription options, but seek 2009, the RMLC paid ASCAP for public to persuade consumers to elect the sub- performance licensing rights in the form of scription model. a ‘‘fixed fee’’ agreement.15 As a result of Through its century of existence, radio’s the deep recession that hit the country in popularity has remained robust. The ra- 2008, the RMLC’s members’ revenues con- dio industry is a $15 billion industry. It is tracted and the fixed fee license began to understood by those in the music business constitute an increasingly high percentage to account for roughly 80% of the music of RMLC member revenue. Consequent- listening experience in the United States. ly, in 2009, the RMLC began to negotiate This percentage has remained roughly new licensing terms to apply to commer- constant despite the rapid evolution in cial radio stations effective January 1, 2010 technology. Thus, while the 80% figure through December 31, 2016. The RMLC was once confined to listening to music wanted a return to the 1.615% rate at over AM/FM radios, that figure now in- which it had paid ASCAP before the fixed-

13. Spotify Radio and iHeartRadio’s ‘‘Create tain conditions not otherwise relevant here. Station’’ use personalization technology from This licensing fee has marginal relevance to a company called The Echo Nest. CBS’s this trial, however, since that fee is a compo- Last.fm uses technology called ‘‘scrobbling.’’ nent of a federal regulatory license rate of Although not part of the trial record, it has 10.5%, which Spotify must pay to obtain both been recently reported that Spotify has ac- public performance rights and mechanical quired The Echo Nest. rights under 17 U.S.C. § 115. See 37 C.F.R. § 385.12(b)(2). Any fluctuation in the public 14. Despite the introduction of a customized performance licensing fee has no impact on radio feature in late 2011, Spotify remains an the overall 10.5% rate which a service like overwhelmingly on-demand service. Under Spotify is required to pay. Spotify’s licensing agreement with ASCAP, which covers the period of [REDACTED], 15. The RMLC allocated the fees among indi- Spotify pays between [REDACTED] and [RE- vidual radio stations according to a formula DACTED] of its revenue, depending on cer- that it developed. 326 6 FEDERAL SUPPLEMENT, 3d SERIES fee arrangement was adopted. It also broadcasting and from internet transmis- wanted a license not only for new media sions by RMLC members. In addition, transmissions by RMLC member stations, the 1.70% rate applies not only to simul- but also new media transmissions by Clear cast radio stations that are streamed over Channel which aggregates many stations the internet by terrestrial broadcasting for delivery over the internet and mobile RMLC members but also to programmed devices. and customized internet radio stations At one point in those negotiations, the owned by RMLC members. One RMLC RMLC offered ASCAP a dual rate struc- member, Clear Channel, is licensed at this ture for new media and terrestrial broad- rate under a ‘‘Group License’’ form, which casts, under which the RMLC would pay a covers new media royalty payments for proposed rate of [REDACTED] of revenue revenues not associated with an individual for terrestrial broadcasting and simulcast- station. Thus, iHeartRadio’s customized ing of terrestrial radio over the internet, radio Create Station feature, which com- and a separate rate of [REDACTED] of petes head on with Pandora, is licensed at revenue for other new media uses of the the 1.70% rate. Until July of 2012, howev- ASCAP repertoire. The revenue for the er, there was no revenue generated by the latter category of uses was miniscule in Create Station service. comparison with the revenue to be covered In 2011, the year of Create Station’s at the [REDACTED] rate. RMLC licen- launch, as well as the year that ASCAP sees still derive almost all of their revenue and the RMLC agreed to license terms, from traditional broadcasting services. Create Station constituted just [REDACT- ASCAP and the RMLC ultimately ED] of the listener hours on iHeartRadio. reached an agreement on terms for a li- For the month of March of 2013, which is cense that established separate rates for the last month for which there was data at radio stations depending on their intensity trial, it had grown to comprise approxi- of music use. The rate for Music Format mately [REDACTED] of its total listener Stations 16 was set at 1.70% of all revenue, hours on iHeartRadio for that month. including revenue derived from new media Thus, a rapidly increasing share of the 17 uses. The RMLC and ASCAP memorial- iHeartRadio listeners are choosing the ized their agreement in a binding letter of customized Create Station feature when December 21, 2011. And on January 27, listening to iHeartRadio. But from its 2012, this Court approved the agreement inauguration in September of 2011 until and its terms became public. June of 2012, Clear Channel ran Create It is of significance to the issues litigated Station as an advertising-free service. It in this trial, that the 1.70% rate applies began to contribute revenue for the first both to revenue derived from terrestrial time in July of 2012. Despite its growth in

16. The category ‘‘Music Format Station’’ is station that does not fall within the Music broad. A station is defined as a Music For- Format Station definition because it uses AS- mat Station if it has a featured performance CAP music less frequently pays at a rate start- of ASCAP music in more than 90 of its ing at .0296% of revenue, plus a supplemental ‘‘Weighted Program Periods’’ in a given week. fee. A Weighted Program Period is of 15 minutes duration, and there are 318 Weighted Pro- 17. The agreement provided for a deduction of gram Periods in a week. So, a station is a 12% for gross revenues from ‘‘Radio Broad- Music Format Station whether it plays ASCAP casting’’ and a 25% deduction for gross reve- music anywhere between approximately 28% nues from ‘‘New Media Transmissions,’’ sub- to 100% of the Weighted Program Periods. A ject to a cap. IN RE PANDORA MEDIA, INC. 327 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

audience, the contribution to digital reve- thumbs-down when a composition is nue from the Create Station feature [RE- played, or by signaling that a song should DACTED]. For the first three months of be skipped. 2013, it contributed substantially less than The MGP contains a wealth of data for [REDACTED] to Clear Channel’s digital every composition in its database. revenue, which itself is only a small compo- Trained music analysts, many of whom nent of Clear Channel’s entire revenue have music related degrees or are musi- stream. cians, listen to the compositions selected for inclusion in the database and register IV. Pandora the composition in reference to as many as Pandora is the most successful internet 450 characteristics.19 For pop and rock radio service operating in the United songs, for example, Pandora analyzes be- States today. It is estimated to have ap- tween 150 to 200 musical traits. Rap has proximately 200 million registered users about 350 ‘‘genes,’’ and classical works worldwide 18 and an approximately 70% have between 300 to 500 MGP-defined at- share of the internet radio market in the tributes. As Pandora’s Conrad testified, United States. Pandora launched its in- ‘‘[b]ecause Pandora utilizes trained musi- ternet radio service in 2005. Roughly cologists to analyze songs, the MGP is able eight years later, it had achieved great to differentiate not only between an alto popularity, streaming an average of 17.7 and tenor saxophone, but also between billion songs per month in the fiscal year various styles of playing a tenor saxo- 2013. phone.’’ When a Pandora listener seeds a A. Pandora’s Music Genome Project station or registers a thumbs-up reaction, Pandora records that feedback and draws Pandora’s exponential growth and popu- upon the MGP to locate other compositions larity can be directly attributed to its sub- that the listener is likely to enjoy. Con- stantial investment in its proprietary Mu- versely, when the feedback is a ‘‘thumbs sic Genome Project (‘‘MGP’’) database and down,’’ the song will not reoccur in the associated algorithms. Pandora uses the user’s playlist, and songs sharing its at- MGP database to create customized inter- tributes will appear less frequently. net radio stations for each of its customers. A Pandora customer creates a station by Besides listening to as many as 100 of ‘‘seeding’’ it with a song, artist, genre, or their own customized stations, Pandora composer. That seed serves as a starting users can opt to listen to programmed point to which Pandora then applies the ‘‘genre’’ stations. The most popular Pan- information in its MGP database to match dora genre stations include ‘‘Today’s Hits,’’ that seed with other songs that Pandora’s ‘‘Today’s Country,’’ and ‘‘Today’s Hip Hop algorithms predict that the listener is like- and Pop Hits.’’ These genre stations are ly to enjoy. The listener continues to give populated by songs which are hand select- feedback by giving a thumbs-up or ed by Pandora curators.

18. Pandora provides streaming internet mu- cold start problem exists when the recom- sic services in the United States, New Zea- mendation system cannot draw on adequate land, and Australia. inferences for a composition because the item is new or obscure. Because of this limitation, 19. The use of human beings to classify each the service may play the composition for lis- composition is unique to Pandora and has the teners who do not like it or fail to play it for advantage of ameliorating what the industry those who might. recognizes as the ‘‘cold start’’ problem. A 328 6 FEDERAL SUPPLEMENT, 3d SERIES

Pandora has a catalog of between ap- Pandora’s revenue has grown exponen- proximately 1,000,000 to 2,000,000 songs, tially since its inception. For fiscal year somewhat less than half of which are li- 2009, Pandora reported revenue of approx- censed through ASCAP. This number is imately $19 million. By fiscal year 2013, considerably lower than the catalog size of Pandora’s revenue had risen to over $400 an on-demand service like Spotify, which million. As of today, however, Pandora must have the ability to play virtually any has yet to demonstrate sustained profita- composition any customer might select. bility. Successful on-demand services have cata- Pandora’s payment of licensing fees for logs in the range of 20 million songs. the use of music consumes a very signifi- B. Pandora Premieres cant portion of its revenue. In 2013, Pan- dora’s content acquisition costs were close Pandora has a small on-demand music to $260 million, or over 60% of its revenue service, but it is not part of this license for that fiscal year. A very substantial application. The Pandora on-demand ser- portion of these costs are for the fees paid vice is called Pandora Premieres. It fea- to record companies for licenses for sound tures at any one point in time a few doz- recordings, as described in more detail en songs, each available for listening on- below. demand for a limited period of time. This is music that artists and music pub- E. Pandora’s Competitive Environ- lishers provide to Pandora for promotion- ment al purposes, typically before the commer- cial release of an album. This on-demand Pandora’s competitive environment is component of Pandora’s service is not a dictated by the nature of its service. Pan- significant part of Pandora; it constitutes dora is a radio service, albeit a customized a ‘‘barely measurable portion of Pandora radio service. Unlike traditional broadcast listening.’’ Pandora secures the rights to AM/FM radio, in which one program is play the songs on Pandora Premiers by played for many listeners, Pandora’s digi- negotiating direct licenses with the copy- tal radio service provides the opportunity right holders. to have a unique program created for the enjoyment of each listener. This distinc- C. Pandora’s Comedy Programming tion between programmed and customized While Pandora is overwhelmingly a ser- radio has been referred to as the one to vice that plays music, in 2011 it introduced many, versus the one to one distinction. comedy offerings. The comedy content But, despite that differentiation, made pos- constitutes a very small percentage of Pan- sible by digital technology, Pandora is ra- dora’s played content. dio. The listener does not control what song will next be played and doesn’t know D. Pandora’s Revenue what that next song will be. As with other Pandora derives revenue from two prin- forms of radio, the listener may be intro- cipal sources: advertising and subscription duced to new music she has not heard fees. As of today, Pandora derives ap- before. There is an industry term for proximately 80% of its revenue from the distinguishing among types of listening ex- sale of display, audio and visual advertis- periences: lean-back versus lean-forward. ing, and the remaining 20% or so from a Like radio, Pandora is a lean-back service, paid subscription service without advertis- in contrast to the on-demand lean-forward ing called ‘‘Pandora One.’’ services like Spotify. IN RE PANDORA MEDIA, INC. 329 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

Not surprisingly, therefore, Pandora hired a large in-house local advertising competes aggressively with other radio sales force. And to improve its ability to stations for listeners. It competes directly compete for advertising dollars with ter- with internet radio stations, whether they restrial radio stations, Pandora contracts are programmed music streaming services with third party Triton Digital, a firm that or customized radio stations. But, because collects radio audience data on both a local the internet radio market is comparatively and national level, in order for radio adver- small and because Pandora already holds a tising buyers to better understand the significant share of that market, Pandora reach of advertisements run on Pandora. expects its increased audience, listening hours, and advertising revenue to come Despite this intensive effort to build ad- largely at the expense of terrestrial radio. vertising revenue, Pandora is still unable While Pandora has a 71% share of the to play as many minutes of advertising per internet radio market, it has less than an hour as its broadcasting competitors. 8% share of the overall radio market. Therefore, as of today, Pandora plays on average approximately 15 songs per hour Pandora is attempting to make itself as compared to terrestrial radio’s roughly ubiquitous, so that its listeners have Pan- 11 songs per hour.20 While Pandora’s free dora available to them throughout their radio service now runs less audio advertis- day, whether they are at home, at work, in ing per hour than do terrestrial radio sta- the car, or somewhere else. As Pandora tions, this gap may lessen as Pandora’s explained in a 2013 SEC 10–Q filing, ‘‘[o]ne business matures. key element of our strategy is to make the Pandora service available everywhere that In accordance with the above, in its pub- there is internet connectivity.’’ Pandora lic filings with the SEC Pandora identifies has consequently expanded its service to its principal competitors as broadcast radio smartphones, tablets, and television providers, including terrestrial radio pro- streaming devices. And because almost viders such as Clear Channel and CBS, half of radio listening takes place in cars, satellite radio providers such as Sirius XM, Pandora has negotiated agreements to in- and online radio providers such as CBS’s tegrate its service into new cars built by a Last.fm and Clear Channel’s iHeartRadio. number of auto companies. But, while programmed radio and custom- Besides competing with traditional radio ized radio are Pandora’s primary competi- for listeners, Pandora also competes with tion, Pandora also competes with interac- traditional radio for advertising dollars. tive,21 on-demand internet music services.22 Most terrestrial radio advertising revenue Its identified competitors in this market comes from local advertising. To compete include Apple’s iTunes Store, RDIO, Rhap- for these advertising dollars, Pandora has sody, Spotify, and Amazon.

20. While there is general agreement about 22. Pandora points out a cost advantage that these numbers, they may not be an altogether broadcast and satellite radio have over Pan- accurate description of the difference in a dora. Broadcast radio pays no royalties for listener’s experience of music. Listeners to terrestrial broadcasts of sound recordings; broadcast radio stations frequently switch sta- satellite radio pays 9% of revenue for satellite tions in search of more music when an ad transmission of sound recordings; and Pan- begins. Moreover, even ads contain music. dora paid 55.9% of its revenue for internet 21. The music industry’s use of the term ‘‘in- transmission of sound recordings in 2012. teractive’’ is explained below. 330 6 FEDERAL SUPPLEMENT, 3d SERIES

V. Pandora’s Licensing History with The 5.0 License allowed non-interactive ASCAP users to choose between three rate sched- 25 On July 11, 2005, Pandora first entered ules. Schedule A of the 5.0 License, into an agreement with ASCAP for a blan- which Pandora chose, required it to pay ket license to publicly perform the compo- the higher of 1.85% of revenue or a per- sitions in the ASCAP repertoire. This session rate. The 1.85% rate represented license was in effect from 2005 to 2010, an increase in ASCAP’s form license rate when Pandora exercised its option to can- from the previous rate. The predecessor cel it. The license which ASCAP issued to to the 5.0 License had an equivalent rate 26 Pandora during this span of years was a for this schedule of 1.615%. ASCAP’s form license. form license for interactive services pro- vided for a substantially higher license Pandora was licensed by ASCAP from rate of 3.0%.27 2005 to 2010 under the ASCAP Experi- mental License Agreement for Internet The interactive/non-interactive distinc- Sites & Services—Release 5.0 (‘‘5.0 Li- tion in the ASCAP form license agree- cense’’). ASCAP first adopted the 5.0 Li- ments is borrowed from 17 U.S.C. § 114’s cense in 2004.23 As of 2004, internet radio (‘‘Section 114’’) use of the term interactive had been in existence for roughly ten in the context of the licensing of sound years, customized internet radio had been recording rights (Section 114 and sound in existence for approximately five years, recording rights are discussed below). and on-demand services had been in exis- Because ASCAP considers its music to be tence at least three years.24 With its more valuable to the services it classifies adoption of this form license, ASCAP as interactive, it has licensed them at a made two important distinctions. First, it higher rate than non-interactive services. raised the rate for new media licenses, As noted, under the 5.0 License Pandora reflecting a judgment that those services was required to pay the greater of either made more intensive use of music than the percentage of revenue corresponding broadcast radio. Second, it made a dis- to its applicable rate, or a fee based on a tinction between interactive and non-inter- concept known as ‘‘sessions.’’ A ‘‘session’’ active new media services. It did not, is defined in the license as ‘‘an individual however, make any distinction between visit and/or access to [the] Internet Site or programmed and customized internet mu- Service by a User.’’ Any visit that exceed- sic services. ed one hour in length began a new ses-

23. ASCAP created a ‘‘New Media’’ depart- ular musical compositions, unless such com- ment in 1995 to address the licensing of mu- positions are sixty (60) seconds or less in sic over the internet. Although the meaning duration.’’ of the term new media may depend on the context, ASCAP generally considers new me- 26. The 1.615% rate that ASCAP had applied dia to include any music user that operates to internet radio before its adoption of the 5.0 ‘‘primarily over the Internet, through wireless License was ASCAP’s rate for a terrestrial devices, or through other emerging digital radio license in the era before the RMLC technologies.’’ license adopted a flat rate. 24. The Rhapsody on-demand service was in- 27. augurated in 2001. Interactive services are defined as those which ‘‘transmit[ ] and/or provide[ ] access to 25. The 5.0 License defined non-interactive transmissions of content comprising or con- services as ‘‘site[s] TTT from which ‘Users’ taining music to ‘Users’ at their request or may not download or otherwise select partic- direction.’’ IN RE PANDORA MEDIA, INC. 331 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

sion.28 At some point in 2010, Pandora gotiation, and spurred by Sony’s impend- recognized that it had been calculating ses- ing withdrawal from ASCAP (as discussed sions incorrectly, and that it had substan- below), on November 5, 2012, Pandora tially underpaid ASCAP. It paid ASCAP filed with this Court a petition for determi- over $1 million to account for that error. nation of reasonable licensing fees pursu- If the payments Pandora was required to ant to AFJ2. See AFJ2 § IX. make to ASCAP, when measured by the per session rate, are converted into a flat VI. The April 2011 ASCAP Compendium percentage of Pandora’s annual revenue, Modification the effective rate for the years 2005 A. Overview and Context through 2010 ranged from a high of 3.63% in 2006 to a low of 1.91% in 2007. But In 2011, ASCAP modified its Compendi- there is no evidence that any party be- um to permit its members to selectively lieved that Pandora was obligated to pay withdraw from ASCAP the right to license above the 1.85% rate until 2010. works to new media entities. This was an unprecedented event. Never before had Pandora’s systems do not track its cus- ASCAP granted partial withdrawal rights tomers’ use of its services with any meas- ure that corresponds to the 5.0 License to its members. As this Court would hold definition of a session, and it was a com- in 2013, the modification violated the terms plex undertaking for Pandora to calculate of AFJ2. AFJ2 requires that ASCAP li- the amount it owed to ASCAP using that cense to any applicant all of the works in measure.29 As a result of its dissatisfac- its repertoire, and consequently if a pub- tion with this sessions component of its lisher leaves a composition in ASCAP’s license, on October 28, 2010, Pandora sent repertoire for some licensing purposes AS- a letter to ASCAP terminating its license CAP is required to license that work to and applying for a new license, pursuant to any applicant. See In re Pandora Media, the terms of AFJ2, to run from January 1, Inc., 12 Civ. 8035(DLC), 2013 WL 5211927, 2011 through December 31, 2015. at *1 (S.D.N.Y. Sept. 17, 2013). In the year and a half that followed the adoption Upon making a written request to AS- of the modification of the Compendium, CAP, Pandora obtained, pursuant to AFJ2 three of the four largest music publishers § V’s requirement that ‘‘ASCAP is hereby withdrew their new media rights from AS- ordered and directed to issue, upon re- quest, a through-to-the-audience license to CAP. EMI’s withdrawal was followed by TTT [inter alia ] an on-line user,’’ the right the withdrawal of Sony and then UMPG. to perform all of the compositions in AS- Each of these publishers thereafter negoti- CAP’s repertoire for that period, with only ated direct licenses with Pandora. Those the proper payment rates to be deter- negotiations and licenses have been a cen- mined, either through negotiation or by tral feature of this litigation, and are dis- the rate court. Having been unable to cussed in detail below. agree with ASCAP on the proper price for To place the Compendium modification the license after roughly two years of ne- in broader context, it was simply one of

28. For example, if a customer used a licensed 29. As a consequence, Pandora and ASCAP service once for forty minutes and once for agreed that Pandora could use a sample of fifteen minutes, that was measured as two usage to arrive at a reasonable estimate of the sessions. Similarly, if a customer used the amount it owed. service for 61 minutes, that was counted as two sessions. 332 6 FEDERAL SUPPLEMENT, 3d SERIES the many ripple effects that have followed time a public performance copyright in the onset of the digital age in the music sound recordings. See 17 U.S.C. §§ 106, business, and the industry’s attempt to 114 (‘‘Section 114’’). Section 114 did not recover from the concomitant decline in require all music users to obtain a license some types of music sales. The modifica- to perform a sound recording, but only tion of the Compendium came in response services that ‘‘perform the [sound record- to pressure from ASCAP’s largest music ing] publicly by means of a digital audio publishers. These publishers were fo- transmission.’’ 17 U.S.C. § 106(6) (em- cused principally on the disparity between phasis added). Section 114 differentiates the enormous fees paid by Pandora to among services that are, in the meaning of record companies for sound recording the statute, ‘‘interactive’’ and ‘‘non-interac- rights and the significantly lower amount tive.’’ An interactive service is defined as it paid to the PROs for public performance a service ‘‘that enables a member of the rights to compositions. The modification public to receive a transmission of a pro- was enacted despite significant concern gram specially created for the recipient, or about the impact of this change on AS- on request, a transmission of a particular CAP, its writers and its independent pub- sound recording TTT which is selected by lishers. or on behalf of the recipient.’’ 17 U.S.C. B. Public Performance Rights for § 114(j)(7). If a digital service does not Compositions versus Sound Re- provide users with this level of control it is cordings non-interactive. The distinction between interactive and non-interactive services is A brief overview of the distinction be- meaningful because ‘‘non-interactive’’ digi- tween public performance rights in sound tal music services are eligible for ‘‘a com- recordings and public performance rights pulsory or statutory licensing fee set by in compositions is necessary to provide the Copyright Royalty Board [‘‘CRB’’] context for the discussion of the motiva- made up of Copyright Royalty Judges ap- tions of the largest publishers in effectuat- pointed by the Library of Congress,’’ see ing the Compendium modification. A Arista Records, LLC v. Launch Media, right to the public performance of a sound Inc., 578 F.3d 148, 151 (2d Cir.2009), recording is the right to control the per- whereas interactive services must indepen- formance of one recording of a perform- dently negotiate rates for sound recording ance of a song. By contrast, a right of public performance in a composition is the licenses. Pandora is a non-interactive ser- right to control the use of the underlying vice within the meaning of Section 114. musical composition itself. The latter Importantly for purposes of this pro- right has been long recognized; but the ceeding, Congress also provided that this right of public performance of a sound rate court (and the BMI rate court) may recording is a relatively new phenomenon not take into account sound recording li- and is restricted to digital services. The censing fees in setting a rate for the licens- licensing fees for sound recordings are ing of the compositions themselves. The paid to an entity called SoundExchange, DPSRA provides that ‘‘[l]icense fees pay- which collects and distributes these fees to able for the public performance of sound the holders of sound recording copyrights. recordings TTT shall not be taken into ac- [1] In 1995, Congress passed the Digi- count in any TTT proceeding to set or tal Performance in Sound Recordings Act adjust the royalties payable to copyright (‘‘DPSRA’’), which provided for the first owners of musical works for the public IN RE PANDORA MEDIA, INC. 333 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

performance of their works.’’ 30 17 U.S.C. ing entirely from ASCAP. EMI Chief Ex- § 114(i). ecutive Roger Faxon has explained that Ultimately, the CRB decided that the EMI wanted to withdraw because it be- market for sound recording rights was ma- lieved that it was inefficient to license each terially different from the market for the right in the musical works and recordings public performance rights to musical com- it administered through different institu- positions, and set rates for compulsory li- tions. Faxon wanted EMI to be able to cense fees for sound recordings at rates ‘‘unify the rights in the compositions that many times higher than the prevailing we represented so that a single negotiation rates for the licensing of the public per- with TTT a customer who wanted the rights formance of the compositions. Conse- could encompass all rights TTT necessary quently, Pandora pays over half of its rev- to empower their business.’’ Faxon also enue to record companies for their sound said that EMI was dissatisfied with the recording rights, and only approximately ‘‘delays’’ in ASCAP’s procedures and AS- four percent to the PROs for the public CAP’s high operational costs. performance rights to their songs. The disparity between rates for the pub- Spurred by the potential loss of one of lic performance of compositions versus the four largest music publishers, ASCAP sound recordings does not exist for most of began in 2010 to explore a proposal to ASCAP’s revenue streams since, as just amend its Compendium to allow members explained, the need to acquire sound re- to withdraw from ASCAP only the right to cording licenses only applies to services license works to new media users. It was, who conduct digital audio transmissions. after all, only new media users who needed Thus, there is no disparity at all when it to acquire both a public performance and comes to most of ASCAP’s business, in- sound recording license. Not all of the cluding its general licensing program and ASCAP board was in agreement on this its licensing of cable TV, broadcast TV, proposal. Large publishers were in gener- and terrestrial radio. Because only new al enthusiastic about such a change, but media music services must acquire sound the songwriters and independent publish- recording licenses, the PROs end up re- ers were less so. ceiving far more money from public per- formance rights license fees for composi- As noted, the largest publishers were tions than do the record companies from fixated on the higher rates that record public performance license fees for sound companies—often their corporate affili- recordings. ates—were receiving from internet music providers, of which Pandora was the most C. ASCAP–Publisher Negotiations prominent, for sound recording rights un- Prior to the Compendium Modifi- der Section 114’s compulsory license re- cation gime when compared to the rates that Against this backdrop, music publishers Pandora and others were paying for public assessed their options. In September performance rights under AFJ2. The ma- 2010, music publisher EMI advised AS- jor publishers viewed AFJ2 as preventing CAP that it was contemplating withdraw- them from closing the gap between the

30. Publishers lobbied for this provision in and drag down the latter. ASCAP also sup- Congress because they were concerned that ported the enactment of the provision, for the the sound recording rates would be set below same reason. the public performance rates for compositions 334 6 FEDERAL SUPPLEMENT, 3d SERIES

composition rates and the sound recording had conversations with Brodsky about the rates. In the words of Sony’s Brodsky: same issue ‘‘about 2 months ago’’ and that We were struck by the vast disparity ‘‘[h]e seemed to understand the basic rea- between what record companies received sons, i.e., CRB set rates, ASCAP/BMI from digital music services for the sound market share.’’ 31 recording rights that they conveyed and The publishers believed that AFJ2 stood what was paid for the performance in the way of their closing this gap. They right. believed that because the two PROs were This concern over the discrepancy be- required under their consent decrees to tween the revenues generated for record issue a license to any music user who labels and those generated for music pub- requested one, they could not adequately lishers is repeated in many of the commu- leverage their market power to negotiate a nications related to the adoption of the significantly higher rate for a license to Compendium modification and the subse- publically perform a composition. quent withdrawals by the publishers of On occasion the publishers offered a sec- new media rights from ASCAP. In many ond rationale for executing the new media of those exchanges they focus their atten- withdrawals: the high cost of litigating a tion directly on Pandora. For example, an case in rate court. Martin Bandier, the email from ASCAP General Counsel Joan CEO of Sony, cited this rationale in an McGivern to LoFrumento of July 30, 2010, email to Sony employees, in which he con- brought up the possibility of permitting partial withdrawals from ASCAP as a tended that ‘‘litigating a rate is an expen- means to attempt to close the gap in Pan- sive and inefficient way to license our rep- dora’s payments: ertoire to new digital services.’’ I spoke to Peter Brodsky at Sony late Against the backdrop of the urgency felt yesterday. He would like to meet with by the largest music publishers to close us TTT to discuss—why publishers are the gap between payments for composition not receiving as much as their record rights and sound recording rights, other labels from Pandora and what options ASCAP members had their own separate Sony might have, such as trying to li- concerns. Songwriters, and at least some cense Pandora directly, withdrawing its independent music publishers, were con- rights, etc. cerned about the damage that might be Similarly, an email of July 31, 2010 from wrought from the Compendium modifica- McGivern to other top ASCAP officials tion and the partial withdrawal of rights confirms Sony’s focus on the gap in the from ASCAP. Songwriters trusted AS- payments made by Pandora: CAP to account reliably and fairly for the Peter Brodsky at Sony, asked that we revenues ASCAP collected and to distrib- meet with him and two [outside lawyers] ute the portion of revenues owed to writ- to discuss our Pandora license, and in ers promptly and fully. Songwriters were particular, why publishers and writers concerned about the loss of transparency are not receiving as much from Pandora in these functions if publishers took over as it is paying to SoundExchange. the tasks of collection and distribution of ASCAP’s DeFilippis responded to this licensing fees. They were concerned as email by explaining that he had already well that the publishers would not manage

31. The references in these two communica- from and set the rate for sound recording tions to SoundExchange and the CRB are rights licenses. references to the entities that distribute fees IN RE PANDORA MEDIA, INC. 335 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) with as much care the difficult task of stood by writers to be motivated primarily properly accounting for the distribution of by profits, and that writers would not look fees to multiple rights holders, and might positively on ASCAP becoming a clearing- even retain for themselves certain monies, house for processing direct licensing royal- such as advances, in which writers believed ties.’’ [REDACTED] concluded by ex- they were entitled to share. Overall, they pressing his opinion about ‘‘the vital role were concerned about the increasing con- ASCAP plays in protecting writers from centration of the publishing industry and the shark-infested waters of the music the willingness by some, particularly Sony, business.’’ 32 to engage in direct licensing outside the Tension between the major publishers framework of the PROs. These concerns and the writers of ASCAP is not surpris- ripened as the writers learned that Sony ing given that the two groups’ interests intended to follow EMI’s lead and take are not perfectly aligned. To balance their advantage of the Compendium modifica- competing interests, ASCAP’s internal tion to partially withdraw from ASCAP. rules are premised on equality in decision- Some of this tension is captured in an making between writers and publishers. email sent by ASCAP-member and com- As LoFrumento testified: poser [REDACTED] to LoFrumento on [ASCAP’s] rules are geared towards September 6, 2012. In that email, [RE- equality between writers and publishers. DACTED] explained the conflicts that he [For example] [o]ur rules say that if you perceived between the major publishers get a stream of revenue that was an and writers of ASCAP: adjustment for the past, we normally go [W]riters and (the major) publishers dif- back to the past and make that adjust- fer. Writers, I believe are concerned ment. There is not necessarily the same with the health and well being of AS- credibility that a publisher who got ex- CAP. As small business owners we are tra money would say, oh, well, this is for dependent upon ASCAP for our suc- cessTTTT Today’s publishers (the majors) two years ago, well, let’s go back and are executives not owners. Their focus make that distribution. It’s not a cer- is on the well being of their company, tainty what they would actually do with their investors and their own perceived it. performance all of which is reflected in As significantly, ASCAP provides writers the quarterly bottom line. In their vi- with transparency. Again, in the words of sion of the future, ASCAP plays an in- LoFrumento: consequential role. Major, major driving issue is [that] with [REDACTED] was not alone among ASCAP [the writers] get transparency writers in his concern about the publish- TTTT [They] know our rules and we take ers’ plan for new media withdrawals. the money that we collect, take off our Writer [REDACTED] wrote in an email of overhead and split it fifty-fifty. Our August 28, 2012 to LoFrumento, that writers get that part of the fifty percent, there was a ‘‘disintegration of trust be- the publishers get the other parts. It’s tween writers and publishers,’’ and that an equal division TTTT The writer’s ‘‘the new breed of publishers was under- greatest fear is that in the world of

32. The [REDACTED] and [REDACTED] Capstar Acquisition Corp., 756 F.Supp.2d 516, emails addressed as well their concern with 552 (S.D.N.Y.2010), aff’d sub nom. BMI v. the direct licensing activities of Sony, exem- DMX Inc., 683 F.3d 32 (2d Cir.2012). plified by its license with DMX. See In re THP 336 6 FEDERAL SUPPLEMENT, 3d SERIES

publishers collecting the money the who may see this as an ASCAP death splits will not be reflective of how AS- knoll TTTT [W]e are in fact giving [the CAP splits the money. major publishers] the right to negotiate. Finally, the writers were concerned that The end result being that they will set a to the extent that the major publishers higher market price which will give us pulled their significant resources out of bargaining power in rate court. ASCAP, the writers would have to shoul- As an internal debate swirled, the AS- der a larger burden in paying for activities CAP Board authorized management on like licensing, advocacy, and litigation. In September 16, 2010 to ‘‘examine alterna- that vein, [REDACTED] urged LoFru- tive means of licensing digital media and to mento to ‘‘not let ASCAP become the ‘ac- engage antitrust counsel.’’ In March 2011, counting firm’ for the publishers who want ASCAP notified the Department of Justice to withdraw rights.’’ of its consideration of a proposal to allow The large publishers were well aware of the withdrawals of new media licensing 33 the discomfort that at least some writers rights from ASCAP. felt with the new media withdrawals and D. The Compendium Modification Al- made the following argument to convince lowing New Media Withdrawals is them to come on board: if the major pub- Enacted. lishers could get higher license rates by direct negotiations with new media compa- On April 27, 2011, the ASCAP Board nies outside of ASCAP then those rates adopted a resolution to amend its Compen- could be used in rate court litigation to dium to allow a member to withdraw from raise the ASCAP license fees. The pub- ASCAP its rights to license music to new lishers found an ally on this issue in writer media outlets, while allowing ASCAP to and ASCAP chairman Williams, who retain the right to license those works to agreed with the new media rights with- other outlets. Six songwriter members of drawal strategy. His email illustrates the the Board abstained from the vote, but strategy he pursued to get writers to sup- there was no vote in opposition. port the publishers’ partial withdrawal of The Compendium modification was exe- rights from ASCAP: cuted by creation of Compendium Rule My job is to make this transition as 1.12. It allowed any ASCAP member, on smoothly as possible in the board room six months notice,34 to ‘‘modify the grant of TTT to assuage the fears of the writers rights made to ASCAP TTT by withdraw-

33. The ASCAP submission to the DOJ focused were problems created by the rise of ‘‘carve on three issues which ASCAP believed might out’’ licenses, and other challenges associated require amendments to AFJ2. The third of the with online licensing. issues it mentioned was the proposal to amend the Compendium to allow a publisher 34. With respect to the timing of a publisher’s (the submission focused on EMI) to ‘‘reserve withdrawal of rights, the modified Compendi- exclusively to itself the right to license partic- um provided that: ular on-line users.’’ In an oblique reference A [new media withdrawal] TTT will be effec- to the higher rates for a license for the public tive on the first day following the last day of performance of a sound recording, the sub- the calendar quarter in which the anniver- mission disclosed that EMI had concluded sary date of the Member’s election falls (the that ‘‘the consent decree is not giving it ade- ‘‘Effective Date’’), upon submission of an quate value for its repertory, especially as executed copy of such [withdrawal] to AS- compared to revenues they derived from oth- CAP no more than nine months nor less er, similar rights.’’ The first two issues, to than six months from the calendar quarter which it devoted a far lengthier discussion, in which the anniversary date falls. IN RE PANDORA MEDIA, INC. 337 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) ing from ASCAP the right to license the risk to the publisher if a withdrawal right of public performance of certain New proved to be a bad idea. Section 1.12.6 of [M]edia Transmissions.’’ The modified the Compendium provided that: ‘‘[a]ny Compendium defines ‘‘New Media Ser- Member may terminate its Membership vices’’—i.e., entities which make ‘‘New Me- Modification at any time upon written no- dia Transmissions’’ and which would be tice to ASCAP, and thereby grant back to purportedly subject to a decrease in their ASCAP the rights previously withdrawn.’’ ASCAP rights as the result of publisher (Emphasis added.) withdrawals—as To manage the withdrawal process, the any standalone offering by a ‘Music Compendium modification mandated, in TTT User’ by which a New Media Trans- Section 1.12.4, ASCAP’s creation of a list mission of musical compositions is made of works subject to any publisher’s with- available or accessible (i) exclusively by drawal by ‘‘[n]o later than ninety days means of the Internet, a wireless mobile before the Effective Date’’ of the with- telecommunications network, and/or a drawal. The publisher was required to computer network and (ii) to the public, notify ASCAP of any errors or omissions whether or not, in exchange for a sub- ‘‘within ten days of receipt of the List of 35 scription fee, other fee or charge. Works.’’ Thus, ASCAP and the publisher In the Compendium modification, ASCAP would both have a list of works that would provided that ASCAP would be affected by the withdrawal well in ad- continue to have the right to license vance of the effective date of the with- such works only to those New Media drawal. Services which are licensed under Li- censes–in–Effect on the Effective Date E. ASCAP Provides Administrative of the Membership Modification, and Services for Withdrawing Publish- only for the duration of such Licenses– ers. in–Effect. EMI publicly announced in early May of One effect of the Compendium modifica- 2011, within days of the adoption of the tion was that major publishers could pull a Compendium modification, that it would be writer’s works out of the PRO that the withdrawing new media rights from AS- writer had decided to join. Although pub- CAP. The turmoil caused by EMI’s deci- lishers had in the past considered a work sion was widespread. Confronted with the to belong to the repertoire of the PRO to reality of losing this major publisher, on which the writer of the work belonged, in May 5, ASCAP’s LoFrumento made a pro- fact, it was a publisher that generally had posal to EMI. He offered ASCAP’s ser- contractual control over the licensing deci- vices in distributing the EMI revenues to sions for the work. With the withdrawal ASCAP members and to other songwriters of rights from the PRO, the withdrawing and publishers who would be entitled to publisher unilaterally removed the work share in the revenues. LoFrumento ar- from the PRO insofar as new media licens- gued at the time that ing rights were concerned. ASCAP is uniquely positioned to handle The Compendium modification also al- the distribution of these rights because lowed the withdrawing publishers to re- it already distributes royalties from the join ASCAP at any point, eliminating any online licensees; its operating ratio re-

35. At trial, the publishers agreed that the CAP do not appear to have a position yet on Compendium modification does not apply to whether it applies to satellite radio. digital downloads. The publishers and AS- 338 6 FEDERAL SUPPLEMENT, 3d SERIES

mains one of the lowest in the world and remaining ASCAP members to pay for all certainly the lowest in the US; its tech- of the other functions that ASCAP per- nology is leading edge and its databases forms for its members, including in Lo- are authoritative; and finally, its staff is Frumento’s words at trial, ‘‘membership, truly professional. legislative, legal, senior management, [and] LoFrumento also advised EMI’s Faxon international costs,’’ increased. On the that ASCAP had been flooded with inqui- other hand, because ASCAP continued to ries since EMI’s announcement from both administer the distribution of licensing foreign and domestic rights holders and revenues, the writers could continue to organizations. As he explained, many have confidence that they would actually writers were concerned that EMI would receive the monies owed them by the with- not distribute royalties as carefully, accu- drawing publishers. Finally, the Adminis- rately, and promptly as they had relied tration Agreements meant that the with- upon ASCAP to do. drawing publishers faced little downside in Ultimately, EMI and other withdrawing withdrawing new media rights. They publishers agreed to let ASCAP handle could continue to enjoy the benefits of the distribution of royalties collected for having ASCAP perform burdensome back- the new media direct licenses that they office tasks while licensing internet music negotiated. They executed ‘‘Administra- entities directly. tion Agreements’’ for this purpose. AS- CAP charged a fee of [REDACTED] for VII. A Second Compendium Modifica- this service, which represented a very sub- tion in December 2012: the ‘‘Stan- stantial discount from its ordinary charge dard Services’’ Agreement to members. Essentially, ASCAP set a At the urging of Sony, another change rate based on the direct costs associated to the Compendium, executed in December with these functions. ASCAP was con- 2012, further reduced the burdens on with- cerned that without a low rate, the with- drawing publishers. The modification al- drawing publishers would be tempted to lowed the publishers to target large new use competing PROs to perform the ad- media entities for direct licensing negotia- ministration services. tions and to effect withdrawals of rights As a result of the publishers’ partial from ASCAP solely with respect to those withdrawals from ASCAP, the burden on large licensees.36

36. By the Fall of 2012, there was increasing sions that set rates for ASCAP and BMI li- dissent within ASCAP about the wisdom of censes with DMX that were comparable to the 2011 Compendium modification. At a DMX’s direct license rates, with certain ad- September Board meeting, the writer mem- justments. DMX Inc., 683 F.3d at 47. The bers of the Board urged that ASCAP reverse decision also required ASCAP to provide a the modification and reject Sony’s pending blanket license that was subject to carve-outs request that the option to withdraw new me- to account for an applicant’s direct licensing dia rights be further modified to allow a pub- program. Id. at 44. DMX is a commercial lisher to target its withdrawal and limit it to music service provider that supplies back- the rights needed by large music users only. ground and foreground music to public ven- The writers did not want ASCAP to assist ues such as restaurants, frequently through Sony as it weakened ASCAP. Much of this the transmission of music via satellite trans- anger was engendered by Sony’s direct licens- mission. To succeed in its direct licensing ing program generally, including the licensing campaign with music composers and publish- of DMX. ers, DMX decided that it was necessary to In June of 2012, the Court of Appeals for the sign at least one major music publisher. In Second Circuit had affirmed rate court deci- 2007, it entered into such a license with Sony IN RE PANDORA MEDIA, INC. 339 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

The December 2012 amendment permit- VIII. Pandora Negotiates Direct Licens- ted a member that was withdrawing under es with EMI, Sony, and UMPG Section 1.12 of the Compendium to indi- and Fails to Negotiate an Agree- cate that it wished to leave to ASCAP the ment with ASCAP. right to license certain new media services A. The Pandora–EMI License Negoti- that paid to ASCAP license fees of less ations than $5,000 per year. Where the with- Upon learning in May 2011 of EMI’s drawing member indicated that it was only withdrawal of its new media licensing withdrawing new media rights ‘‘in part,’’ rights from ASCAP, Pandora immediately ASCAP continued to license new media began to negotiate with EMI for a license services for the member that were defined to its catalog. The negotiations were not in the Compendium as ‘‘Standard Ser- contentious and the contours of the license vices.’’ As a consequence, smaller new were quickly settled. Indeed, in their very first substantive discussion, which oc- media entities could avail themselves of an curred on June 6, EMI confirmed that it ASCAP license so long as they accepted would be using 1.85% as the headline rate, ASCAP’s 5.0 License (or its successor li- and hoped to have the agreement effective censes) without negotiation. as of January 1, 2012. The collegial tone ASCAP’s DeFilippis offered the follow- is reflected in handwritten notes by Pando- ra’s Rosenbloum. Rosenbloum colorfully ing explanation for the adoption of the recorded that EMI was ‘‘not looking to Standard Services exception for the with- screw anyone.’’ drawal of new media licensing authority: EMI’s Faxon testified that the rate in Given the rapidly changing marketplace the EMI–Pandora license was ‘‘freely and the low barriers to entry, new digi- agreed to.’’ EMI was less concerned with tal music services launch quite frequent- the precise rate than the flow of revenue ly. Many will never gain traction with into EMI. Because the ASCAP deductions listeners or generate substantial reve- from gross receipts would be smaller, EMI nue. From the perspective of the with- viewed the license terms with Pandora as a drawing music publishers, they lacked ‘‘substantial improvement.’’ the necessary staff and infrastructure to The 1.85% rate in the Pandora–EMI track the thousands of small music users agreement was the same rate that was that wished to license their music. available to Pandora under ASCAP’s 5.0 License for a non-interactive service. A Sony’s Brodsky stated that Sony wanted July term sheet with EMI reflected this this revision to the Compendium so that rate and an expectation that the agree- Sony’s withdrawal could be limited ‘‘to just ment would have a two year term. It also the music services that we wanted to enter reflected calculations premised on EMI’s into direct deals with.’’ 37 estimate that it had approximately a 20% market share at the time.

by offering a substantial advance and an ad- 37. During the negotiations over this Compen- ministrative payment. Id. at 38. dium amendment, ASCAP’s counsel commu- nicated to Sony that there were antitrust con- cerns with the carve-out proposal. 340 6 FEDERAL SUPPLEMENT, 3d SERIES

During the ensuing months, the parties The reference to the 1.70% rate was discussed the size of an advance that Pan- prompted by the recently announced set- dora would pay to EMI, among other tlement of rate court litigation between things. Meanwhile, Pandora continued to ASCAP and the RMLC in January 2012. pay its licensing fees to ASCAP. As described above, that license provided for a blanket license rate of 1.70% of reve- The licensing agreement, although not nue and a 25% deduction for advertising executed until March 16, 2012, covered the expenses in connection with new media. two-year period January 1, 2012 to Decem- ber 31, 2013. Pandora agreed to a license B. The Pandora–ASCAP License Ne- that provided EMI with a pro-rata share gotiations of 1.85% of Pandora’s revenues.38 The EMI agreement did not contain any ‘‘per- As noted above, Pandora had terminated session’’ component like that included in its license with ASCAP on October 28, the ASCAP 5.0 License. It required Pan- 2010 because of its concern over the calcu- dora to pay a non-refundable advance of lation of the per-session rate in the 5.0 [REDACTED], which Pandora was confi- License, and had applied at that time for a dent would be exceeded by its payments to new license for the calendar years 2011 EMI over the course of the license. In through 2015. It remained an applicant addition, the EMI agreement permitted for such a license throughout 2011 and Pandora to take up to a [REDACTED] 2012, as ASCAP adopted its modification adjustment to revenue for advertising ex- to the Compendium and as EMI withdrew penses.39 This adjustment included com- new media rights from ASCAP. missions paid to internal advertising sales On September 16, 2011, Pandora execut- personnel, but only if Pandora were able to ed an interim license agreement with AS- obtain an adjustment for internal advertis- CAP effective as of January 1, 2011. It ing expenses in connection with an agree- adopted the 5.0 License rate of 1.85% with- ment from another major music publisher out any per session fee. The agreement or PRO. noted the parties’ competing positions on Finally, the agreement included a most- several issues, including Pandora’s position favored-nation clause, or ‘‘MFN,’’ for the that the adjustment for advertising ex- benefit of Pandora. The agreement con- penses should apply to its internal adver- templated a prospective decrease in the tising expenses. headline rate from 1.85% to as low as Roughly a year later, on September 28, 1.70% if Pandora succeeded in obtaining a 2012, Pandora learned that Sony was also lower rate for licensing a repertoire as withdrawing its new media rights from large or larger than EMI’s catalog. It ASCAP. With its discussions with ASCAP similarly allowed for an increase in the ‘‘languish[ing]’’, and with Sony’s withdraw- advertising expense adjustment up to [RE- al from ASCAP due to take effect at year DACTED]. end, which was just weeks away, Pandora

38. EMI and ASCAP estimated ASCAP’s mar- 47% market share. This calculation required ket share as 47%. The parties determined the EMI to provide Pandora with a list of its revenue base against which the 1.85% would works, which it did monthly until Sony ac- be applied by calculating Pandora’s revenue, quired EMI. multiplying it by the percentage of tracks played that embodied EMI’s catalog, and mul- 39. [REDACTED]. tiplying that number by ASCAP’s estimated IN RE PANDORA MEDIA, INC. 341 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

filed this rate court petition on November (‘‘NMPA’’), which is a music industry trade 5. group based in Washington, D.C. LoFru- Pandora’s filing in rate court angered mento assured Horowitz that he was ap- some in the ASCAP community, particu- proaching Pandora with the mindset Horo- larly the major publishers. They ex- witz advocated. pressed their outrage not only to Pandora, Horowitz continued to apply pressure on but also to its outside counsel, the law firm Pandora. He called Pandora’s outside Greenberg Traurig, LLP. The day after counsel a second time, about a week after the rate court filing, UMPG’s Horowitz his first call, and reported once more to called one of Pandora’s attorneys at LoFrumento. As Horowitz explained to Greenberg Traurig. As Horowitz prompt- LoFrumento on November 14, Pandora’s ly memorialized in an email to ASCAP’s outside counsel ‘‘has been spending hours LoFrumento, Horowitz on fallout from their repping Pandora. told [Pandora’s outside counsel], as a They are embarrassed. [Pandora’s coun- ‘‘friend’’ of the firm, that I thought both sel] said they will withdraw from repping the firm and Pandora are completely Pandora in the next few weeks if the [rate tone deaf. That whether his firm has court litigation with ASCAP] doesn’t set- the legal right to rep Pandora in litiga- tle.’’ tion, the firm has lost huge goodwill with Not surprisingly, given the fallout from writers and artists by doing so. And Pandora’s filing of the rate court petition, that filing now for a rate court proceed- and with the deadline for Sony’s withdraw- ing against ASCAP TTT had the effect of al from ASCAP approaching, the negotia- unifying artists, writers, and PROs tions between Pandora and ASCAP inten- against Pandora. sified. Had those negotiations succeeded, Horowitz also gave some advice to Lo- of course, this rate court action would have Frumento regarding ASCAP’s negotiating become moot. stance with Pandora. His advice boiled down to two words: be strong. Horowitz By the end of November, Pandora be- wrote: lieved that it had reached an agreement on terms with ASCAP, although it understood My take: [Pandora’s outside counsel] and Pandora are scared. They just that the agreement needed final approval want to settle with ASCAP and settle from ASCAP. Pandora emailed a term fast. Be strong. Time is on your side. sheet to Pandora on November 29. AS- Pandora is now under intense pressure CAP had assured Pandora that if they to settle with ASCAP. They have to put finalized their agreement before the end of this behind them. You can really push 2012, the license would cover the Sony Pandora and get a much better settle- repertoire since the Sony withdrawal from ment as a result. They are reeling. ASCAP was only effective as of January 1, They will pay more, a lot more than they 2013. originally intended, to do that. LoFrumento decided to reject the li- Horowitz forwarded this same email to cense that his team had negotiated with other ASCAP board members, including Pandora. He knew that either way he Sony’s Martin Bandier, and BMG Music faced litigation. He knew that if he exe- Publishing’s Laurent Hubert. Besides cuted the license, Sony would sue ASCAP. these ASCAP Board members, Horowitz Sony had threatened to sue ASCAP in the sent the email to David Israelite of the event any license agreement with Pandora National Music Publishers Association that encompassed the Sony repertoire was 342 6 FEDERAL SUPPLEMENT, 3d SERIES

executed before the end of 2012.40 Sony rejoin ASCAP for all purposes. LoFru- had also notified ASCAP that it might not mento also had to consider the writers who use ASCAP for administration services if had become restive and were doubtful ASCAP issued a license to Pandora. Lo- about the supposed benefits of the publish- Frumento was already facing rate court er withdrawals. In the midst of all of this, litigation with Pandora. Given the pres- LoFrumento cast the lot of ASCAP with sure being exerted on him by both Sony the withdrawing major publishers and and UMPG, LoFrumento was only willing chose to let the rate court decide the dis- to execute a license with Pandora that pute between Pandora and ASCAP. On included a headline rate of at least 2.5%, December 14, ASCAP surprised Pandora and he knew Pandora was not willing to and rejected the terms they had negotiat- pay that much. ed. LoFrumento advised the Law and Li- censing Committee of ASCAP’s Board of C. The Pandora–Sony License Negoti- Directors on December 12 that he intend- ations ed to reject the terms Pandora and AS- Since the Fall of 2010, Sony had been CAP had negotiated. Everyone under- discussing with ASCAP the possibility of a stood that that meant that the rate court withdrawal of rights so that it could direct- proceeding would go forward. None of ly negotiate with Pandora. In July 2012, the Committee members asked for a de- Sony notified ASCAP that it would exer- scription of terms Pandora and ASCAP cise its right under the modified Compen- had negotiated or to discuss LoFrumento’s dium to withdraw new media rights. In decision.41 late September, Pandora (along with the Thus, in mid-December 2012, ASCAP rest of the world) learned that Sony would set itself on a course to have its rate for be withdrawing new media rights from licensing Pandora set in this rate court ASCAP effective January 1, 2013. As al- proceeding, despite the cost associated ready described, Sony worked with AS- with that litigation. The decision was CAP during late 2012 to effect a second made in the midst of great turmoil, uncer- change to the Compendium that would tainty and pressure. The partial with- permit a partial withdrawal of new media drawals of new media rights by major rights from ASCAP. Under the Standard publishers, who collectively controlled Services exception, Sony allowed ASCAP about 50% of ASCAP’s music, threatened to retain licensing authority for smaller to make ASCAP a weaker organization. new media services while assuming re- Sony and UMPG had also made clear to sponsibility for the direct licensing of larg- LoFrumento that they wanted to negotiate direct licenses with Pandora and opposed er entities such as Pandora. ASCAP entering into a final license with As of the Fall of 2012, Sony was the Pandora. There was, of course, a chance world’s largest music publisher. It owned that by placating the major publishers, or controlled between 25% and 30% of the they might later exercise their option to market. It had taken this frontrunner

40. Sony’s attitude to a negotiated Pandora– 41. Prior to that meeting, LoFrumento had ASCAP license had been clear for months. In discussed with a few committee members the an email of October 4, Sony (which by that terms on which Pandora and ASCAP had time controlled EMI) refused Pandora’s re- agreed in principle and which he intended to quest to disclose the terms of the Pandora– reject. EMI license to ASCAP. IN RE PANDORA MEDIA, INC. 343 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) position in the summer of 2012, when it works and remove them from its reper- became responsible for licensing EMI’s ca- toire by January 1, Pandora had to obtain talog.42 Combined, the Sony and EMI a license from Sony or face crippling copy- catalogs contain roughly 3 million songs. right infringement claims. Sony was in While the effective date of the withdraw- the driver’s seat and the clock was ticking. al came as a surprise to Pandora, Pandora The remainder of the conversation was had been aware that the withdrawal was a largely devoted to Sony’s statement of possibility ever since ASCAP adopted the the reasons why it needed Pandora to Compendium modification. Indeed, in the pay for the public performance of music Spring of 2012, Pandora wrote to the Fed- at a substantially higher rate. The prin- eral Trade Commission in opposition to cipal reason was the ‘‘massive unfair dis- Sony’s acquisition of EMI and referred to parity’’ between what Pandora was paying this very possibility. Noting that a Sony the record labels for sound recording withdrawal from the PROs would require rights and what it was paying the music Pandora to negotiate directly with Sony publishers for composition rights. Brod- and that Pandora would be faced with a sky explained that if the labels were get- choice of either paying higher rates ‘‘or ting 50% of Pandora’s revenue, then it continuing to operate without Sony’s would be ‘‘fair’’ for music publishers to songs,’’ Pandora’s Kennedy expressed con- get 12% of the revenue, although Brodsky cern that the combination of the Sony and acknowledged that Pandora could not af- EMI catalogs would give Pandora ‘‘no ford to pay that much. As Brodsky em- choice’’ but to enter into a direct license phasized, it was the ‘‘differential’’ between for the content. While Pandora ‘‘could the rates paid to the labels and the pub- survive without access to Sony’s musical lishers that was the problem, and that content,’’ it ‘‘could not survive without ac- Pandora was really just caught in the cess to the combined Sony and EMI cata- logues.’’ middle of a tug of war between the labels and publishers. Brodsky admitted that if The first substantive discussion between the labels were getting only 25% of Pan- Pandora and Sony occurred in a telephone dora’s revenue, then Pandora’s current in- call on October 25 between Sony’s Brodsky dustry-wide rate of 4% for the licensing and Pandora’s Rosenbloum.43 Sony of rights to publicly perform compositions promptly set the tenor for the negotiations would probably be alright and there with a not-too-veiled threat. Brodsky stat- wouldn’t be any need to increase it. ed ‘‘[i]t’s not our intention to shut down Pandora.’’ In his many years of negotiat- Brodsky identified a second, subsidiary ing music licenses, Rosenbloum testified reason for needing Pandora to pay more. that had never before heard such a threat. Referring to the writers’ skepticism over In some ways, this threat put on the table the motives of the publishers in withdraw- no more than what was obvious. Sony’s ing from the PROs, Brodsky added that works were already being played on Pan- Sony had to show the writer-members of dora; they were incorporated in the MGP. the PROs that there was some ‘‘reasonable Unless Pandora could do without those justification’’ for Sony’s withdrawal. At

42. Sony Corporation and other investors pur- 43. EMI’s Michael Abitbol was also a partici- chased EMI Music Publishing companies in pant in the call. June 2012. With that purchase, Sony/ATV undertook the administration of EMI, which remains a separate entity. 344 6 FEDERAL SUPPLEMENT, 3d SERIES the end of the call, Rosenbloum threw out In their telephone conversations during the possibility that Pandora might pay a the month of November, Rosenbloum reit- ‘‘modest’’ increase to Sony for a year as erated the request for a list of works on they all waited to see what happened to several occasions but never got any re- the rates Pandora was paying the labels. sponse. Rosenbloum repeated the request Following this conversation, Pandora de- once more at a breakfast meeting that he cided on a two-prong strategy. It would and Pandora’s Kennedy had with Sony’s intensify its efforts to get an ASCAP li- Brodsky and Bandier on November 30. cense before the end of the year. To Again, Sony did not respond. bring ASCAP to the negotiating table it The list of Sony works was potentially filed its petition in this rate court for an important for several purposes, and Pan- ASCAP license on November 5. Secondly, dora referred to those several purposes in Pandora attempted to obtain leverage in its discussions with Sony. In addition to its negotiations with Sony. It requested a wanting to be able to remove the Sony list of the Sony catalog so that it could works from its service if Pandora and Sony take the Sony works off, or at least threat- could not come to terms, Pandora needed en to take them off, of the Pandora service the list so that it could understand how to if no deal could be reached. In his years apportion any payments between the EMI of negotiating licenses, this was the first time that Rosenbloum had ever requested and Sony catalogues since the payments a list of works from a publisher. would apparently be made at two different rates. Pandora also wanted the list so it Pandora’s first request for the list came could evaluate whether the substantial, on November 1, 2012, in an email from non-refundable advance that Sony was de- Rosenbloum to Brodsky. Rosenbloum ad- manding would likely be recouped. vised Brodsky that: I wanted to follow up with you about our Sony had a list readily at hand, since the conversation last week regarding Pando- Compendium required that a publisher ra. As I mentioned, given the uncer- and ASCAP work together during the 90 tainties around Sony/ATV’s and EMI’s day period before the effective withdrawal position with respect to webcasting date to confirm precisely which works rates, Pandora has decided that it needs were being withdrawn. Sony understood to be prepared to take down all Sony/ that it would lose an advantage in its nego- ATV and EMI content in the event we tiations with Pandora if it provided the list are unable to agree on rates by the end of works and deliberately chose not to do of this year. In that regard, please let so. Brodsky’s explanation at trial that he me know if you can provide us with an did not provide the list because he believed electronic listing of Sony/ATV and EMI that negotiations were proceeding smooth- repertoire. ly and did not want to impose an unneces- On a related note, as the end of the year sary ‘‘burden’’ on Sony’s staff is not credi- is rapidly approaching, we look forward ble. The negotiations were not going to receiving a rate proposal as soon as smoothly; the list had already been pre- possible (to the extent that EMI and pared and its production imposed no bur- Sony/ATV are still interested in moving den. As Brodsky recognized in his testi- forward with a direct license agree- mony, the list was ‘‘necessary’’ to Pandora ment). in the event the parties did not reach a Brodsky received this request for a list deal. Sony decided quite deliberately to of the Sony works, but never responded. withhold from Pandora the information IN RE PANDORA MEDIA, INC. 345 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

Pandora needed to strengthen its hand in quest with an accurate list of the Sony its negotiations with Sony. works. But, ASCAP, like Sony, stone- Ultimately, Sony made an offer to Pan- walled Pandora and refused to provide the dora in early December. Still hoping to list. reach an agreement with ASCAP which In making the request to ASCAP, Pan- would obviate the need for license from dora’s counsel wrote that ‘‘Pandora must Sony, Pandora did not respond to the offer prepare for the possibility of being unli- or to a follow-up email of December 6. censed by Sony/ATV or ASCAP for [Sony’s] works effective January 1st, so it On Friday, December 14, with two is important that we get this information weeks left in the year, and one week re- from ASCAP as soon as possible.’’ This maining before the music industry took its request set off a flurry of emails within annual holiday break, ASCAP notified ASCAP. ASCAP ultimately decided to Pandora that it would not execute the contact Sony to see if it would give its agreement they had negotiated. The fol- permission to share the list of works. On lowing Monday, Pandora urgently made Wednesday, December 19, ASCAP notified two renewed written requests for the list Sony of Pandora’s request and that it of Sony’s works, one to Sony and another would be providing Pandora with the list of to ASCAP. Sony works that ASCAP had previously Since the repeated requests from Pan- given to Sony in connection with its with- dora’s outside counsel Rosenbloum had drawal of rights. Not surprisingly, given gone unanswered, Pandora’s general coun- its own refusal to share the list with Pan- sel Delida Costin sent her own email to dora, Sony did not give ASCAP permission Brodsky on December 17 requesting the to provide the list.46 As a result, neither list of works.44 Not wishing to empower Sony nor ASCAP provided the list of Pandora, Sony never responded.45 works to Pandora. That same day, Pandora also asked AS- If either Sony or ASCAP had provided CAP for the list of Sony works in ASCAP’s Pandora with a list of the Sony works, repertoire. It would have taken ASCAP Pandora would have been able to remove about a day to respond to Pandora’s re- Sony’s compositions from its service within

44. Costin wrote: list of works because we were very close to While we remain hopeful that we will reach finalizing a deal.’’ Rosenbloum denies ever mutually acceptable terms, we also find telling Brodsky any such thing. Brodsky also ourselves in a position where we must pre- testified that Rosenbloum did not make oral pare for the possibility that we are unable requests for the list of works in between the to obtain a license prior to January 1, 2013, November 1 written request and the request which is the date that has been signaled as during the breakfast meeting on November the effective date of the Sony/ATV with- 30. While Rosenbloum was entirely credible drawal of certain of its compositions for in his testimony on these issues, Brodsky was certain uses. I am writing, therefore, to not. request that Sony/ATV identify the specific musical compositions that it intends to 46. ASCAP personnel shared their amusement withdraw from each of [the PRO’s] license with each other over Sony’s decision to with- authority effective as of January 1, 2013. hold the list from Pandora. In one email, 45. Brodsky testified that Sony did not provide DeFilippis asked ASCAP’s counsel Richard Pandora with a list of works because, when Reimer ‘‘Why didn’t Sony provide the list to he contacted Rosenbloum regarding Ms. Cos- Pandora,’’ to which Reimer replied ‘‘Ask me tin’s request, Rosenbloum replied that ‘‘there tomorrow,’’ to which DeFilippis responded was no need for Sony/ATV to provide such a ‘‘Right. With drink in hand.’’ 346 6 FEDERAL SUPPLEMENT, 3d SERIES about a week.47 Although ASCAP at- right license fees under Licensee’s agree- tempted at trial to show that Pandora ments with other major music publishers’’ could have used public sources of informa- or PROs. tion to identify the Sony catalog, it failed On a December 21 draft of the agree- to show that such an effort would have ment, Rosenbloum wrote to Sony that, to produced a reliable, comprehensive list, the extent Pandora was willing to conclude even if Pandora had made the extraordi- the license without receiving ‘‘actual data’’ nary commitment necessary to try to com- from Sony, it ‘‘at least’’ needed confirma- pile such a list from public data. tion of the approximate percentage of the The terms of the Pandora license with ASCAP repertoire that consisted of Sony Sony were negotiated in four business and EMI compositions. Sony’s Brodsky days during the single week that ran be- responded to this request not by giving a tween ASCAP’s rejection of the Pandora list but with a rough estimate that the term sheet and the start of the holiday Sony/EMI share of the ASCAP repertoire break. On December 18, Brodsky sent was 30%. Rosenbloum a term sheet. As proposed in Although the agreement was predomi- that document, the license term would be nately on Sony’s terms, the December 21 one year, starting January 1, 2013. It draft agreement did include a change in required Pandora to pay a non-refundable Pandora’s favor regarding the adjustment but recoupable advance of [REDACTED] for advertising sales from a previous draft and a non-refundable [REDACTED] ad- of the agreement. Unlike a draft deliv- vance as an administrative fee. The royal- ered from Sony to Pandora on December ty rate was set at Sony’s pro-rata share of 18 which only allowed for a deduction from an industry-wide rate of 5%. Sony under- internal advertising costs if Pandora got a stood this to be a 25% increase over the similar deduction from another PRO or then prevailing industry rate of approxi- publisher, the December 21 agreement al- mately 4%. In his March 2013 report to his lowed for a reduction of up to [REDACT- Board of Directors, Sony’s Bandier ED] that included both outside commis- bragged that Sony had leveraged its size sions and direct internal costs of such sales to get this 25% increase in rate. without reference to another agreement The term sheet also allowed Pandora to with a PRO or publisher. The parties take an adjustment for advertising ex- executed a Binding Heads of Agreement penses of up to [REDACTED]. This on December 21, 2012. would include a deduction for Pandora’s By mid-January 2013, and despite the internal advertising sales personnel ‘‘to the existence of a confidentiality agreement, extent deducted from revenue in connec- Sony leaked the key terms of the Pandora tion with the calculation of performance license to the press.48 The headlines in

47. Pandora needed the publishers’ list of 48. Although Brodsky denied knowing that works before it could take any steps to re- anyone at Sony had leaked the terms of the move them from the Pandora service. Once license to the press, the evidence is that Sony Pandora had the list, it could quite quickly did just that. Despite reporting dutifully that remove any song with an identical title and Sony had ‘‘declined’’ to comment on the eliminate the copyright infringement risk. It terms of the deal, the articles referred to would take Pandora more than a week, how- anonymous industry insiders as their source ever, to identify which songs with identical and quoted Bandier’s analysis of the deal. titles but from other publishers could be rein- While Pandora had absolutely no interest in seeing the 25% hike in its rates known to troduced into the Pandora playlist. other licensors, Sony hoped that its rate IN RE PANDORA MEDIA, INC. 347 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

three articles said it all: ‘‘Sony/ATV ‘Now learned that UMPG was scheduled to with- Has the Power to Shut Pandora Down draw its new media licensing rights from TTT’ ’’; ‘‘Sony/ATV gets 25 percent in- ASCAP effective July 1, 2013. crease in Pandora royalties’’; and ‘‘Sony/ UMPG’s Horowitz had notified ASCAP’s ATV’s Martin Bandier on new ‘quite rea- LoFrumento at the end of November 2012 sonable’ Pandora deal.’’ A New York Post that UMPG intended to withdraw new me- article featured a photograph of Sony’s dia rights from ASCAP. Horowitz told Bandier in shirt sleeves with a large cigar LoFrumento that the ASCAP rate for in his mouth, as it reported that Sony had Pandora was ‘‘too low.’’ In making this ‘‘wrangled a 25 percent increase in royal- ties’’ for a one year license. Bandier was assertion he referred to Spotify’s 10.5% 49 quoted as saying that ‘‘[a]t the end of the rate. Horowitz added that he believed day, we got a terrific deal for our songwri- Sony would be getting a much better rate ters. Our thinking has been vindicated.’’ from Pandora than ASCAP would achieve. In his interview with Billboardbiz, report- Horowitz asked ASCAP for a waiver of the ed on January 18, Bandier explained that Compendium’s notice period; he wanted to the rates ‘‘are quite reasonable. When withdraw effective January 1, 2013. AS- you compare it to the rate record compa- CAP denied the waiver request. nies are getting, it was really miniscule.’’ The negotiations between UMPG and One article reported: ‘‘[m]any other pub- Pandora were even more contentious than lishers were rooting for Sony to deliver a the negotiations between Sony and Pando- TTT higher rate so that if [the PRO’s] deal ra. After difficult conversations in March with Pandora heads to rate court, the in which UMPG asked for an industry- judge will consider the Sony rate the mar- wide headline rate of 8%, Pandora essen- ket rate and raise performance royalties tially placed the negotiations on hold. accordingly.’’ The press coverage focused While a license agreement was executed in on Sony’s leverage in negotiations due to June, it was for a six month term only and its outsize market power: ‘‘Look a little was contingent on several events. closer, and this is ultimately a very lopsid- ed negotiation TTTT Pandora absolutely The negotiations between these parties needs Sony’s catalog to run an effective were conducted principally by Horowitz radio service. And if they don’t pay what for UMPG and Rosenbloum and Kennedy Sony/ATV wants, they can’t use it, by for Pandora. Kennedy and Horowitz law.’’ knew each other fairly well. They had been dealing with each other for years in D. The Pandora–UMPG License Ne- connection with sound recording rights. gotiations Horowitz had run the ‘‘label’’ side of Uni- Pandora did not have to wait long for versal’s business until April 2012, when he the next publisher to leave ASCAP and was transferred to the music publishing demand a yet higher rate for a direct side of the organization. Horowitz license. In February 2013, Pandora brought into these 2013 negotiations with

would be a jumping off point for the next 49. Unlike Pandora, Spotify is an on-demand publisher’s negotiations with Pandora, and it service. As an on-demand music service, its was. Pandora had its attorneys call Sony to 10.5% rate is set by a licensing board and complain of the breach of their confidentiality covers mechanical rights and a public per- agreement. formance right. The public performance right rate constitutes an offset from the over- all 10.5% rate. 348 6 FEDERAL SUPPLEMENT, 3d SERIES

Pandora, therefore, a thorough grasp of Horowitz said he was ‘‘not sure’’ he was the history behind the requirement that able to provide Pandora with a list, and Pandora pay a substantial portion of its indicated that Pandora should just make a revenue to obtain sound recording rights, deal based on UMPG’s representation of and perhaps as significantly, a desire to its overall market share. show that he could be similarly effective in achieving an enhanced payment from Pan- Rosenbloum had his own detailed con- dora for composition rights. versation with Horowitz about a week la- ter, on March 29. Horowitz and David In their first substantive meeting, which Kokakis of UMPG asked what Rosenbl- occurred on March 22, Horowitz quizzed oum thought the next steps were in com- Kennedy at some length about the state of Pandora’s business. Horowitz then moved mencing formal license discussions. Ro- to a discussion of Pandora’s need for a senbloum responded that UMPG needed license from UMPG, uttering what Kenne- to provide a proposal and a list of UMPG dy took to be an implicit threat. Horowitz works so that Pandora could ‘‘better un- said ‘‘we want Pandora to survive.’’ derstand the scope of rights at issue’’. Horowitz responded that UMPG was pre- Like Sony, Horowitz justified a substan- pared to provide a list so long as it was tial increase in the rate Pandora needed to pay by stressing the disparity between the covered by non-disclosure agreement rates at which Pandora paid for sound (‘‘NDA’’). recording rights and public performance In their conversation, Horowitz ex- rights for compositions. Showing confi- pressed amazement at the Sony rate for dence that he knew the material terms of Pandora, and indicated that he felt an the Sony–Pandora license, Horowitz re- industry-wide rate of 8% of revenue would peatedly asked Kennedy, (as Kennedy be reasonable, particularly in light of what paraphrased) ‘‘how did you get Marty Pandora was paying to the record labels [Bandier] at Sony to agree to such a low for sound recording rights. Rosenbloum payment?’’ was aghast. He told Horowitz that in his Avoiding invitations to discuss the spe- 20 years in the music industry he had cific terms of the Sony license, Kennedy never encountered a situation in which a explained to Horowitz that Pandora felt it licensor suggested that rates should effec- should be treated just like entities covered tively double overnight, going from 4% to by the ASCAP–RMLC license. Kennedy 8%. Rosenbloum observed that Sony ‘‘was argued that Pandora was competing for apparently more willing to adopt a busi- listeners with, and taking listeners from, nesslike approach’’ and that Sony’s Bandi- radio companies covered by the RMLC er ‘‘understood Pandora’s realities.’’ Skip- deal. Those radio services, including ping over the fact that UMPG wanted its iHeartRadio, would be paying the PROs rate to serve as a benchmark for all future for many years into the future at a rate PRO licenses, Horowitz responded that the lower than the roughly 4% range that Pan- 8% rate would not have such a significant dora had been paying the PROs. impact on Pandora because UMPG’s mar- Kennedy indicated a preference for ne- ket share was only about 17%. gotiating with the PROs, but added that, if UMPG wanted to negotiate directly with Horowitz bluntly reminded Rosenbloum Pandora, then UMPG should provide Pan- that Pandora did not have much negotiat- dora with a list of the withdrawn composi- ing leverage. Rosenbloum described tions and UMPG’s proposal for a rate. Horowitz as asking IN RE PANDORA MEDIA, INC. 349 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

what Pandora would do if we could not [Pandora] agrees not to use any Confi- reach an agreement as to rates (suggest- dential Information for any purpose ex- ing that they have all of the leverage). I cept to evaluate and engage in discus- told him if UMPG is unwilling to move sions concerning a potential business from its 8% figure it might be creating a relationship between the Parties. situation where Pandora would have no Pandora correctly interpreted this provi- choice other to take down all UMPG sion as forbidding it from using the list to repertoire. [Horowitz] indicated that he remove the UMPG works from its ser- definitely was not seeking such a result vice.51 TTTT It was at that point in the conver- On May 21, Pandora’s Rosenbloum and sation that the tone began to change a Horowitz met. While Horowitz expressed bit and [Horowitz] became somewhat his admiration for Pandora and assured its less ‘‘positional’’ in his approach. representatives that UMPG wanted it to In late April 2013, UMPG provided to thrive, he did not move much from his Pandora a complete list of the UMPG initial proposal for a 8% industry rate, only revising it downward to 7.5%. Rosenbloum works in the ASCAP repertoire, but in a responded by reminding Horowitz that way that prevented Pandora from using ASCAP had agreed to a 1.70% ASCAP the information to remove UMPG composi- rate with a generous advertising deduction tions from its service. The list was subject for Pandora’s competitor, iHeartRadio. to an NDA. The list itself was the very information that the NDA deemed confi- Pandora believed that UMPG’s rate re- dential.50 The NDA provided that: quest was unreasonable, and that UMPG would be inflexible in any negotiations. [Pandora] has requested that Universal Therefore, instead of engaging further provide to [Pandora] titles of songs in with UMPG, Pandora went on the offen- Universal’s music catalog controlled by sive. First, Pandora purchased KXMZ– ASCAP, corresponding writer names FM, a terrestrial radio station in Rapid and corresponding shares owned or con- City, . With this purchase, trolled by Universal and such writers, all Pandora hoped to shoehorn itself into the of which Universal deems to be confi- ASCAP–RMLC license. Then, Pandora dential (‘‘Confidential Information’’). believed, it would be in a position to argue The NDA then restricted Pandora’s use of that it was entitled to the RMLC 1.70% the list. It provided that rate.52 The agreement of purchase is dat-

50. Upset by Sony’s public disclosure of the Pandora the ability to use the list of works to Pandora license terms, Pandora requested an remove the works from its service, then there NDA that would bar UMPG from revealing would have been no need for any NDA. The the terms of any license. UMPG refused to only confidential information described in the include any such restriction in the NDA. NDA was the list of works. Nor was the creation of the NDA a trivial matter. The 51. While Horowitz took the position at trial negotiations over the NDA were carefully that he had believed in 2013 that Pandora was free to use the UMPG list of works to managed. At trial, Horowitz opined that remove those works from the Pandora ser- UMPG had no legal obligation to provide a vice, despite the requirement that Pandora list of works to Pandora. execute the NDA, that testimony was not credible. Horowitz was intimately involved 52. Pandora’s purchase of KXMZ–FM remains in these negotiations and is a strong execu- pending. ASCAP has petitioned the FCC to tive. He had no desire to strengthen Pando- deny the transfer of the station’s FCC license ra’s hand. If Horowitz had intended to give to Pandora. 350 6 FEDERAL SUPPLEMENT, 3d SERIES ed June 5, 2013. Second, on June 11, ASCAP at whatever rate the rate court Pandora moved in this Court for partial decides. summary judgment. Its motion argued The parties memorialized a six month that any purported new media withdrawals license agreement on July 1, 2013. The by publishers following the 2011 ASCAP agreement provided for an industry rate of Compendium modification did not affect 7.5%, with no deduction for advertising the scope of the ASCAP repertoire subject expenses, which would be contingent on to Pandora’s application for an ASCAP the two contingencies outlined in the June license. 13 email from Pandora. The agreement During this interim period, as it worked provided that ‘‘in the event that a final on these two projects, Pandora did not decision not subject to any further appeal respond to emails from Horowitz. In his is rendered in the pending ASCAP Rate emails to Pandora, Horowitz expressed in- Court TTT [that] UMPG’s July 1, 2013 creasing levels of anxiety and exasperation withdrawal from ASCAP of [New Media 53 about Pandora’s ‘‘radio silence.’’ Then, licensing rights] TTT is not effective’’ or if on the heels of announcements of its pur- ‘‘Pandora’s acquisition of the KXMZ–FM chase of a radio station and the filing of qualifies Pandora for the RMLC–ASCAP the summary judgment motion, Pandora license’’ then the agreement would ‘‘be of reached out to UMPG. In an understate- no further force or effect.’’ ment, Pandora observed in a June 13 email UMPG refused Pandora’s request that that ‘‘[a]s you may have read or heard, this the agreement reflect that it was non- week Pandora made a couple of announce- precedential and could not serve as a ments that are related to our discussions regarding a direct license with UMPG.’’ benchmark in rate court proceedings. In- Pandora expressed optimism that it would stead, both parties reserved their rights on win the summary judgment motion and the question of whether the agreement recognition of its entitlement to the RMLC could serve as a benchmark in this rate license, all before July 1. But, it added, court proceeding. In the unlikely event we don’t have a IX. September 17 Partial Summary decision on either of these points by July Judgment Opinion 1, it is our preference to continue to perform works in the UMPG catalog. On September 17, 2013, Pandora’s mo- To help facilitate that, we propose ac- tion for partial summary judgment was cepting UMPG’s 7.5% of revenue offer granted. See In re Pandora Media, Inc., on a provisional basis starting July 1, 2013 WL 5211927. The Opinion held, in- 2013, pending the Court’s rulings, with ter alia, that AFJ2 prohibited ASCAP the understanding that if the ASCAP from withdrawing from Pandora the rights rate court subsequently rules in Pando- to perform any compositions over which ra’s favor that Pandora will immediately ASCAP retained any licensing rights. thereafter—and on a retroactive basis Consequently, the publishers’ purported back to July 1, 2013—license the right to withdrawals of only new media rights un- works in the UMPG repertory through der the Compendium modification were

53. On June 6, Horowitz wrote to Rosenbloum not to respond to our proposal’’ and that that he was ‘‘calling [about] Pandora. We ‘‘[w]e are confused by Pandora’s unwilling- haven’t heard anything TTT There’s almost no ness to respond in any way to our repeated time left. Very odd process.’’ And on June inquiries for direction, even if to simply ad- 11, Horowitz wrote to Rosenbloum that he vise us that it no longer desires to license our was ‘‘disappointed that Pandora has chosen music.’’ IN RE PANDORA MEDIA, INC. 351 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

held inoperative. The Court found that rate escalation has been justified as a AFJ2 prohibited a regime in which pub- mechanism to account for SESAC’s grow- lishers allowed ASCAP to license a compo- ing repertoire. The rate in the SESAC sition to some music users but not others. license started at [REDACTED] of Pando- AFJ2 required each work that was in the ra’s revenue in 2007, and the rates from ASCAP repertoire to be available to any the period of 2011 to 2015 (assuming the user who requested a blanket license. The [REDACTED] increase) therefore start at publishers, of course, remained free to [REDACTED] in 2011 and escalate to withhold works from ASCAP entirely.54 [REDACTED] in 2015. [REDACTED]. Calculating the implied rate applicable X. Other Licensing Agreements Put to ASCAP depends on what SESAC’s mar- Forth as Benchmarks ket share in compositions is—a figure that There are two other sets of licenses is impossible to know with certainty. SE- which ASCAP argues are ‘‘confirmatory’’ SAC does not publicly report its revenue benchmarks for the reasonableness of AS- or its catalogue of compositions. There is CAP’s proposal for its license with Pando- no public consensus as to what share of the ra. One is Pandora’s own license with total number of musical compositions are SESAC. The other are licenses held by in the SESAC repertoire. Pandora’s Ken- Pandora’s competitor Apple iTunes Radio. nedy testified that based on SESAC’s rep- A description of these licenses will con- resentations during negotiations, he under- clude this section of the Opinion. stood the SESAC PRO share to be 10% in 2007.55 Pandora had no ability to confirm A. The Pandora–SESAC License that number, but it is a number that has Since 2007, Pandora has had a blanket appeared in court decisions. See, e.g., Mob- license from the PRO SESAC for the right iTV, Inc., 712 F.Supp.2d at 221; United to publicly perform musical compositions States v. ASCAP (Application of Youtube, in the SESAC repertoire. Pandora and et al.), 616 F.Supp.2d 447, 453 (S.D.N.Y. SESAC each have the option to terminate 2009). Using the 10% figure for SESAC and a 45.6% figure for ASCAP, and mak- the license each year, but neither has exer- ing no adjustments for an increase in SE- cised that option. SAC’s share and a concomitant decline in SESAC is the smallest of the three ASCAP’s share, the parties calculate an PROs. It is an invitation-only organization. implied ASCAP rate of [REDACTED] in SESAC proclaims that it is a selective 2011, [REDACTED] in 2012, [REDACT- organization, taking pride in having a rep- ED] in 2013, [REDACTED] in 2014, and ertory based on ‘‘quality, rather than [REDACTED] in 2015. quantity.’’ Pandora’s SESAC license rate increases B. Apple’s iTunes Radio Licenses with annually by the greater of [REDACTED] Publishers and PROs percent or an amount tied to the percent In September of 2013, Apple launched increase in [REDACTED]. The annual its iTunes Radio service. From a user

54. After the Opinion was rendered, the pub- 55. SESAC’s vice president for new media li- lishers sought intervention nunc pro tunc for censing confirmed in his deposition that in the sole purpose of appeal. The motion was the context of unrelated negotiations SESAC granted but the publishers were limited in the had used a 10% figure to describe its market arguments they could raise. See In re Pando- share. ra Media, Inc., 12 Civ. 8035(DLC), 2013 WL 6569872, at *10 (S.D.N.Y. Dec. 13, 2013). 352 6 FEDERAL SUPPLEMENT, 3d SERIES

perspective, iTunes Radio operates like velopers Conference in June 2013. Its Pandora. Both are customized radio. license agreement with Sony is dated June Users seed an iTunes Radio station by 6, 2013, and is identical in its material identifying a song, artist, or genre. They terms to the other licenses that Apple provide feedback by signaling ‘‘Play More entered with publishers. The whereas Like This’’ and ‘‘Never Play This Song’’. clauses of the agreement with Sony em- And like Pandora, iTunes Radio uses an phasize the complementary relationship algorithm to select songs that users are between iTunes Radio and the sale of mu- likely to enjoy in light of their initial selec- sic through the iTunes Store. The clauses tion and feedback. explain inter alia that the parties to the Apple’s iTunes Radio is available in a agreement wish to deter piracy and com- free, advertising-supported format, and pensate songwriters appropriately for the through a program called ‘‘iTunes Match,’’ digital distribution of their compositions, in which users pay an annual fee for a and that Apple wants to create an adver- bundle of services, one of which is access tising-supported internet radio service to to iTunes Radio without advertising. enhance ‘‘recommendations features of the Through a subscription to iTunes Match, iTunes Store for the purpose of promoting subscribers have access to their entire per- sales of eMasters.’’ 58 sonal music library or ‘‘locker’’,56 allowing Apple simultaneously negotiated a li- them to stream music wherever they are. cense agreement for the public perform- iTunes Radio is only available within the ance rights to the ASCAP repertoire. The Apple ecosystem. ASCAP license covers a [REDACTED]. Pandora considers iTunes Radio a major Based on an industry-wide rate of 10%, the competitor. Upon its launch, Pandora parties agreed that Apple would pay AS- tracked the impact of iTunes Radio on CAP a share of [REDACTED] of the Pandora closely. While it appears that the iTunes Radio advertising revenue, as that launch of iTunes Radio in the Fall of 2013 term was defined in their agreement. had a measurable (albeit relatively small) The revenue base for the Apple license impact on Pandora, after a short period of fee includes none of the subscription reve- time that impact appeared to decline. nue from the iTunes Match service. In Pandora has continued to grow despite the addition, the revenue base does not include presence of iTunes Radio. This may be any contribution from the sale of Apple due to several reasons, including unique products promoted through iTunes Radio, characteristics of Pandora’s service, the including the sale of music tracks sold availability of a Pandora app on Apple through the iTunes Store. The revenue 57 devices, and the fact that iTunes Radio is base does not include any attributed value only available on Apple devices. for that advertising on iTunes Radio that Apple negotiated licenses for its iTunes promotes Apple’s music products.59 In the Radio service in order to announce the license agreement, however, Apple com- launch of the service at its Worldwide De- mitted that it would ‘‘use good faith, com-

56. A locker service stores a customer’s digital 58. eMasters are defined as sound recordings music in the cloud, and permits the customer available for download from the iTunes Store. to access the music through multiple devices. 59. [REDACTED]. 57. Approximately forty percent of Pandora’s listeners access Pandora through their Apple devices. IN RE PANDORA MEDIA, INC. 353 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

mercially reasonable efforts to sell the Ad- only included the right to publicly perform vertising to third parties,’’ and also agreed works in their repertories, the Apple li- to attribute revenue at a reasonable rate censes with the publishers provided Apple for the advertising of Apple non-music with both the right to publicly perform the products that appeared on iTunes Radio. compositions in the publishers’ reperto- For several reasons, it would be a diffi- ries, as well as the right to ‘‘[e]ncode, re- cult task to make the necessary adjust- produce, and otherwise use the Publisher ments to the terms of the Apple license to Materials solely to the extent reasonably calculate an equivalent rate for an ASCAP necessary to effectuate, implement and fa- license issued to Pandora, and none of the cilitate the foregoing Performances of trial experts attempted to do so.60 Be- Publisher Materials.’’ cause iTunes Radio launched only a short time before trial, data about the service is CONCLUSIONS OF LAW scarce. Moreover, the differences in the [2–4] Pandora requests that this rate revenue bases; the use of iTunes Radio to court set a fee for its license with ASCAP. promote sales of Apple products, which Section IX of AFJ2 requires the rate has no equivalent for Pandora; and the court to set a ‘‘reasonable’’ fee for a re- absence any means of capturing imputed quested license, but that term is not de- revenue for the advertising of Apple’s mu- fined in AFJ2. Governing precedent dic- sic products, all add to the difficulty of the tates, however, that in determining the task.61 reasonableness of a licensing fee, a court The ASCAP license had other features ‘‘must attempt to approximate the ‘fair as well. The parties agreed to a minimum market value’ of a license—what a license fee amount of [REDACTED] in the event applicant would pay in an arm’s length that iTunes Radio failed, or ran primarily transaction.’’ MobiTV, Inc., 681 F.3d at Apple advertisements. The agreement 82. ‘‘In so doing, the rate-setting court contained an advertising deduction of [RE- must take into account the fact that AS- DACTED] for external advertising ex- CAP, as a monopolist, exercises market- penses only. Finally, the license provided distorting power in negotiations for the for a ‘‘Most Favored Nations’’ clause for use of its music.’’ Id. The Second Circuit the benefit of ASCAP.62 has recognized that, because music per- Using this same industry-wide rate of formance rights are largely aggregated in 10%, Apple also negotiated direct licenses the PROs which operate under consent with publishers Sony, Warner/Chappell, decrees, ‘‘there is no competitive market EMI, and BMG. Unlike Pandora’s agree- in music rights.’’ ASCAP v. Show- ments with music copyright holders, which time/The Movie Channel, 912 F.2d 563,

60. ASCAP’s expert did attempt to make one 61. There is a suggestion in the record that the ‘‘rough’’ adjustment to address the absence Apple negotiations over a public performance from the Apple revenue base of any subscrip- licensing fee may have been influenced by its tion income from iTunes Match. Using the overall obligation to pay for music content, percentage of revenue derived from Pandora’s including its ability to negotiate more favor- subscription service to estimate the amount of able rates with record labels. There was in- iTunes Match subscription revenue, Dr. Mur- sufficient evidence, however, to permit any phy concluded that even if Apple’s share of reliable finding in this regard. subscriber hours was twice as large as Pando- ra’s, the Apple license rate implies an ASCAP 62. [REDACTED]. rate for the Pandora license which exceeds what ASCAP is seeking here. 354 6 FEDERAL SUPPLEMENT, 3d SERIES

577 (2d Cir.1990). Consequently, fair In choosing a benchmark and determin- market value is a ‘‘hypothetical’’ matter. ing how it should be adjusted, a rate Id. at 569. In such circumstances, ‘‘the court must determine the degree of appropriate analysis ordinarily seeks to comparability of the negotiating parties define a rate or range of rates that ap- to the parties contending in the rate proximates the rates that would be set in proceeding, the comparability of the a competitive market.’’ Id. at 576. rights in question, and the similarity of Helpfully, both ASCAP and Pandora the economic circumstances affecting the have endorsed the same definition of ‘‘fair earlier negotiators and the current liti- market value,’’ drawn from a recent text- gants, as well as the degree to which the book: assertedly analogous market under ex- amination reflects an adequate degree of A widely used description of fair market competition to justify reliance on agree- value is the cash equivalent value at ments that it has spawned. which a willing and unrelated buyer would agree to buy and a willing and United States v. BMI (In re Application of unrelated seller would agree to sell TTT Music Choice), 426 F.3d 91, 95 (2d Cir. when neither party is compelled to act, 2005) (‘‘Music Choice IV’’ ) (citation omit- and when both parties have reasonable ted); accord DMX Inc., 683 F.3d at 45. knowledge of the relevant available in- ‘‘[T]he burden of proof [is] on ASCAP to formationTTTT Neither party being com- establish the reasonableness of the fee it pelled to act suggests a time-frame con- seeks.’’ AFJ2 § IX(B). ‘‘Should ASCAP text—that is, the time frame for the not establish that the fee it requested is parties to identify and negotiate with reasonable, then the Court shall determine each other is such that, whatever it hap- a reasonable fee based upon all the evi- pens to be, it does not affect the price at dence.’’ AFJ2 § IX(D). which a transaction would take ASCAP and Pandora have each pro- placeTTTT The definition also indicates posed a set of benchmarks for assessing the importance of the availability of in- the appropriate rate for an ASCAP license formation—that is, the value is based on to Pandora. Interestingly, they both an information set that is assumed to agree that the Pandora license with EMI contain all relevant and available infor- is a valid benchmark. Their sets of pro- mation. posed benchmarks share no other common Robert W. Holthausen & Mark E. Zmijew- element. ski, Corporate Valuation 4–5 (2014). As already noted, ASCAP relies princi- [5] In rate court proceedings, a deter- pally on the three direct licenses negotiat- mination of the fair market value ‘‘is often ed between Pandora and EMI, Sony, and facilitated by the use of a benchmark— UMPG in the wake of the April 2011 Com- that is, reasoning by analogy to an agree- pendium modification. ASCAP arrives at ment reached after arm’s length negotia- proposed rates of 1.85% for 2011–2012 (the tion between similarly situated parties.’’ Pandora–EMI license rate), 2.50% for United States v. BMI (In re Application of 2013, and 3.00% for 2014–2015. This is the Music Choice), 316 F.3d 189, 194 (2d Cir. first time that ASCAP has sought a license 2003) (‘‘Music Choice II’’ ). Rate courts rate of over 1.85% from any non-interac- have been provided with guidance in their tive internet music service. analysis of the parties’ proposed bench- Pandora recognizes the Pandora–EMI marks: license agreement as a suitable bench- IN RE PANDORA MEDIA, INC. 355 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) mark, as well as the historical ASCAP– I. ASCAP’s Rate Proposal of 1.85% for Pandora license rate of 1.85% under the 2011 and 2012 5.0 License. But, in addition to its analy- For the years 2011 and 2012, ASCAP sis of appropriate benchmarks, Pandora proposes a rate of 1.85%. ASCAP’s bench- argues that it is ‘‘similarly situated’’ to the mark for this proposal is the Pandora– RMLC licensees and is accordingly enti- EMI license (which is for the years 2012 tled by the terms of AFJ2 to the RMLC and 2013), which provided for a headline 1.70% rate. rate of 1.85%. For confirmation that 1.85% In summary, ASCAP has carried its is a reasonable rate, ASCAP relies on the burden of demonstrating that its rate pro- fact that it is the same rate under which posal of 1.85% is reasonable for the years Pandora was licensed under the 5.0 Li- 2011 and 2012. It has failed to carry its cense from 2005 to 2010. burden of demonstrating that its rate pro- Pandora agrees. It admits that a head- posals of 2.50% and 3.00% for the years line rate of 1.85% is within a range of 2013 and 2014–2015, respectively, are rea- reasonable rates in the event that Pandora sonable. Pandora has failed to show that is not entitled to the 1.70% rate in the it is entitled to the 1.70% RMLC rate as ASCAP–RMLC license. According to the result of being similarly situated, with- Pandora, the 1.85% rate is the ‘‘upper in the meaning of AFJ2, to the RMLC bound of a range of reasonable rates for member radio stations. Pandora.’’ Since AFJ2 only requires AS- In conducting an independent inquiry CAP to demonstrate that its rate proposal into a reasonable rate, this Court is guided is ‘‘reasonable,’’ Pandora’s concession by the following parameters. First, hav- makes further discussion unnecessary.63 ing determined a reasonable rate for the first years of the five-year license period, II. ASCAP’s Rate Proposal of 2.50% for there is a presumption that that rate will 2013 and 3.00% for 2014 and 2015 continue to be a reasonable rate for the ASCAP proposes a rate of 2.50% for entire license period. Second, the histori- 2013, and 3.00% for 2014 and 2015. AS- cal division between interactive and non- CAP predicates these rates principally on interactive internet music services requires the Pandora–Sony license, which covers that Pandora be licensed well below the the year 2013 and yields an industry wide 3.0% rate at which ASCAP licenses inter- rate of 5.0% and an ASCAP implied rate of active music services. Third, the circum- 2.28%; and on the Pandora–UMPG li- stances under which Sony imposed upon cense, which covered the six month period Pandora an implied ASCAP headline rate from July 1 to December 31, 2013, and of 2.28% confirm that any reasonable rate yielded an industry wide rate of 7.50% and for an ASCAP–Pandora license is below an implied ASCAP rate of 3.42%. ASCAP 2.28% by a measurable margin. For these also puts forth, as confirmatory bench- and the other reasons described below, the marks, the SESAC–Pandora license and 1.85% license rate is the reasonable rate Apple’s licenses with the PROs and pub- for the entirety of the five year term of the lishers in connection with its iTunes Radio ASCAP–Pandora license. service.64 In addition to these benchmarks

63. Pandora’s argument that it is entitled to benchmark, but it did not press this bench- the rate in the RMLC license is addressed mark at trial. In all events, the Spotify li- below. cense is a manifestly poor benchmark because it is a license for an overwhelmingly on-de- 64. At points in this litigation ASCAP also cit- mand service and the public performance rate ed the Spotify–ASCAP license as a potential 356 6 FEDERAL SUPPLEMENT, 3d SERIES

there have been other justifications offered 1.85% would also be a reasonable rate for at trial for a license rate that exceeds the the last three years of the Pandora license 1.85% rate dictated by the 5.0 License and (2013 through 2015). Indeed, the two the EMI license. ASCAP has offered a benchmarks that support adoption of a two-part theoretical argument in the form headline rate of 1.85% continued beyond of Dr. Murphy’s opinions regarding in- the year 2012. The successor form license creasing competition among internet radio to ASCAP’s 5.0 License still has the head- providers and the need for variety in mu- line rate of 1.85%, and there was no evi- sic. There were also two other theoretical dence offered at trial to suggest that AS- arguments raised in the record and at trial CAP is planning to alter that rate for non- (although not initially put forward by AS- interactive new media music services. As CAP) for an elevated rate: the potential for the EMI license, it was for the period for cannibalization of music sales by Pan- 2012 and 2013. Since ASCAP agreed that dora, and the gap between what Pandora the EMI rate was reasonable for the year pays record labels for sound recording 2012, it is presumptively reasonable for rights and what it pays the PROs and 2013 as well. Indeed, ASCAP agrees that publishers for composition rights. Finally, it remains a reasonable benchmark. Pandora’s success is a factor that has been Also, adoption of an escalating rate over present, implicitly, throughout the trial. the term of a five year license would be ASCAP has not carried its burden of out of step with historical practice. AS- showing that its proposed rates for 2013, CAP has never negotiated nor issued a five 2014, and 2015 are reasonable. To begin year license with an escalating rate, and with, rate court precedent and ASCAP’s rate court jurisprudence is devoid of any own licensing history establish a presump- example of an escalating ASCAP rate for a tion that a five-year license should have a single license term. The sole example in single rate. ASCAP has not rebutted that this record of an escalating rate is the presumption. ASCAP has also failed to SESAC license with Pandora. In that demonstrate that Pandora’s direct licenses case, however, SESAC’s escalating rate with Sony and UMPG constitute fair mar- was justified by a mutual assumption that ket benchmarks. The infirmities in these SESAC’s market share would increase proposed benchmarks are not overcome by over the term of the license. Even accept- reliance on either the Pandora–SESAC li- ing that the SESAC license, with its cense or the Apple licenses for its iTunes unique features, could be informative Radio service. Finally, none of ASCAP’s about a reasonable rate for an ASCAP theoretical arguments support an upward license, the justification for an escalating departure from the 1.85% rate to the rate for SESAC suggests that the ASCAP 2.50% and 3.00% rates that ASCAP also rate should be a declining rate since SE- seeks. SAC’s growth would come at the expense of ASCAP and BMI. A. Presumption of a Single Rate There appear to be good reasons why Having accepted ASCAP’s proposal of a ASCAP and the industry generally adopt a rate of 1.85% as a reasonable rate for the single rate for the term of a license. Ab- first two years of the Pandora license sent some unusual circumstances, the val- (2011 and 2012), there is a strong basis to ue of music to a user is assumed to remain recognize a presumption that the rate of constant through the term of a license.

need not be closely negotiated since it is sim- mechanical rights. See 17 U.S.C. § 115; 37 ply a component of an overall 10.5% rate for C.F.R. § 385.12(b)(2). IN RE PANDORA MEDIA, INC. 357 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

And an escalating rate is not necessary for 1. ASCAP and Publisher Coordination the licensor to share in the success of the Pandora has shown that the Sony and licensee: with a single rate as a percent- UMPG licenses were the product of, at the age of revenue a joint interest is created very least, coordination between and between the parties in the growth of the among these major music publishers and licensee’s business. Adoption of a single ASCAP. Sony and UMPG justified their rate facilitates business planning, encour- withdrawal of new media rights from AS- ages reliance on historical data, and dis- CAP by promising to create higher bench- courages resort to contested projections. marks for a Pandora–ASCAP license and Likely for these reasons, and others, there purposefully set out to do just that. They is a well developed practice that supports also interfered with the ASCAP–Pandora the adoption here of a headline rate of license negotiations at the end of 2012. 1.85% for not just the first two years, but UMPG pressured ASCAP to reject the also for the last three years of the license. Pandora license ASCAP’s executives had ASCAP has failed to overcome any pre- negotiated, and Sony threatened to sue sumption that exists in favor of a unitary ASCAP if it entered into a license with rate. But, even without such a presump- Pandora. With only a few business days tion it has not carried its burden to estab- remaining in the year 2012, ASCAP re- lish that the rates of 2.50% and 3.00% are fused to provide Pandora with the list of reasonable. Sony works without Sony’s consent, which Sony refused to give. Without that list, B. Pandora’s Direct Licenses with Pandora’s options were stark. It could Sony and UMPG shut down its service, infringe Sony’s ASCAP has not shown that either the rights, or execute an agreement with Sony Pandora–Sony or the Pandora–UMPG li- on Sony’s terms. Then, despite executing censes are good benchmarks for its license a confidentiality agreement with Pandora, with Pandora. Sony and UMPG each ex- Sony made sure that UMPG learned of all ercised their considerable market power to of the critical terms of the Sony–Pandora extract supra-competitive prices. The license. And LoFrumento admitted at tri- UMPG agreement is a particularly flawed al that ASCAP expected to learn the terms benchmark, for the several reasons dis- of any direct license that any music pub- cussed below. In addition, the evidence at lisher negotiated with Pandora in much trial revealed troubling coordination be- the same way. tween Sony, UMPG, and ASCAP, which There is no need to explore which if any implicates a core antitrust concern under- of these actions was wrongful or legiti- lying AFJ2 and casts doubt on the proposi- mate. Nor is there any reason to explore tion that the ‘‘market under examination here the several justifications that ASCAP, reflects an adequate degree of competition Sony, and UMPG have given for at least to justify reliance on agreements that it some of this conduct.65 What is important has spawned.’’ Music Choice IV, 426 F.3d is that ASCAP, Sony, and UMPG did not at 95 (citation omitted). act as if they were competitors with each

65. Among other things, ASCAP asserts that it and Sony, he rejected the term sheet negotiat- was pure of heart. It points out that it denied ed with Pandora because of his own indepen- UMPG the waiver from the notice period that dent judgment. And Sony’s Brodsky denied UMPG sought when it gave notice of its intent at trial being involved in or knowing who had to withdraw new media rights from ASCAP. leaked the confidential license terms that ap- LoFrumento also explained that, while he peared in the early 2013 press reports. considered the pressure exerted by UMPG 358 6 FEDERAL SUPPLEMENT, 3d SERIES other in their negotiations with Pandora. new media rights from ASCAP effective Because their interests were aligned January 1, 2013, however, the identity of against Pandora, and they coordinated the Sony works suddenly became signifi- their activities with respect to Pandora, cant to Pandora.66 Because of the nature the very considerable market power that of its music service, Pandora had more of each of them holds individually was magni- an ability to substitute one work for anoth- fied. But, since the UMPG and Sony li- er than many other music services. It cense agreements constitute poor bench- certainly had more flexibility than an on- marks even in the absence of coordination, demand service, which needed to play vir- it is not necessary to engage more deeply tually any composition its listeners de- with the implications of this evidence. manded. It even had, at least theoretical- 2. The Pandora–Sony License ly, more flexibility than many programmed The Pandora–Sony license was for the radio services. For instance, it would be year 2013. It was premised on a 5% in- difficult for a terrestrial Top 40 radio sta- dustry-wide rate, which implies a 2.28% tion to thrive without access to each week’s headline rate for an ASCAP license. AS- top 40 hits. Thus, with a list of the Sony CAP has not shown that this rate reflects works, Pandora would have information the fair market value of an ASCAP license necessary to remove Sony works from its with Pandora. service or steer listeners away from Sony When Pandora inaugurated its service in works, or at least to threaten to do so.67 2005, it obtained a blanket license from Both Pandora and Sony treated knowl- ASCAP and fully expected to continue to edge of Sony’s catalogue as a significant be able to do so throughout the life of its bargaining chip in their license negotia- business. It was entitled under that AS- tions. Pandora repeatedly asked for it, CAP license to full access to the ASCAP repertoire and to use any composition in orally and in writing, and Sony pointedly the ASCAP repertoire as frequently as it ignored those requests and stopped AS- wished. This included compositions to CAP from providing the list to Pandora. which Sony held public performance Sony knew that it held the upper hand, as rights. Pandora had no incentive there- it acknowledged when it conveyed to Pan- fore to identify Sony works or to steer its dora that it was not its intention to ‘‘shut listeners toward or away from those down’’ Pandora. works. By withholding the list, Sony deprived Once the Compendium modification had Pandora of significant leverage in their been adopted, and Sony had withdrawn negotiations.68 Pandora was faced with

66. ASCAP suggests that Pandora could have would have been difficult for Pandora to oper- begun to pester Sony for a list of its works as ate for any length of time without access to soon as Pandora learned of the Compendium any Sony composition. Nonetheless, the modification. There is no reason to find that threat of being removed, substantially re- any business executive would have considered moved, or even incrementally removed from a that wise. service as popular as Pandora would be a risk that Sony would need to weigh with care. 67. After it executed the license with Sony, Pandora had no incentive to steer listeners Sony’s determined refusal to provide the list away from Sony works; Sony had demanded despite repeated requests over the license ne- a sizable but recoupable advance. gotiation period is testament to the impor- tance Sony itself placed on this bargaining 68. Since Sony controlled about 30% of the chip. market (counting the EMI repertoire), it IN RE PANDORA MEDIA, INC. 359 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) three options: shut down its business, face made no or more limited use of Sony mu- crippling copyright infringement liability,69 sic. or agree to Sony’s terms. Accordingly, For any economic model to be useful the agreement fails the parties’ agreed- here, it must account for the circumstances upon definition of fair market value: that that created Pandora’s need for Sony’s neither party to the negotiation be ‘‘com- repertoire information. Even if Sony had pelled to act.’’ provided the list of its works to Pandora, Dr. Murphy argues that the Sony li- Sony would have retained enormous bar- 70 cense nonetheless can constitute a compet- gaining power; by withholding the list, itive price because in a competitive mar- Sony deprived ASCAP of a chance to ar- ketplace, a copyright owner would not be gue in any persuasive way that the Sony– likely to offer Pandora information that Pandora license reflects a fair market price. would enable it to operate without the publisher’s works. Dr. Murphy reasons ASCAP next argues that, even if Pando- that, given this expectation of common ra’s possession of the Sony list of works business practice, Sony’s refusal to provide was necessary to create a valid bench- the list of works to Pandora should not be mark, Pandora could have obtained this read as constituting undue compulsion information from sources other than AS- such that the Sony 2.28% rate is not a fair CAP and Sony. But, Sony did not act in market rate. This analysis is too divorced 2012 as if Pandora had a reliable alterna- from the world in which Pandora and Sony tive source of information available to it were actually functioning to be helpful. (other than ASCAP), and ASCAP failed at trial to prove that such an alternative ex- First, their marketplace is not the isted. ‘‘atomistic’’ marketplace from which Dr. ASCAP also asserts that Pandora did Murphy’s theoretical framework is de- not actually require a list of Sony works to rived. In a competitive, atomistic market, negotiate a fair market rate license with if one of many rights holders refuses to Sony since it had no list of EMI works share critical information, then the music when it negotiated the EMI license, and user can see if a competitor will be more the EMI license is endorsed by both par- cooperative. Instead, Pandora and Sony ties to this litigation as a suitable bench- operated in a highly concentrated market. mark. The negotiations that Pandora con- Second, Pandora had built its business ducted with EMI are not comparable. with the understanding that it could obtain Pandora did not need a list of EMI works a blanket license from ASCAP. It had since it learned in its first substantive already, therefore, incorporated the Sony meeting with EMI that EMI was not seek- repertoire into its MGP. Unlike a new ing an increase in Pandora’s license rate. entrant into a market, it was not free (at EMI immediately offered to let Pandora least, without a list of the Sony works) to pay for an EMI license at the 1.85% head- attempt to create a business model that line rate in the ASCAP 5.0 License.

69. Statutory copyright damages are up to works to Pandora in 2012. In judging the $150,000 per work. See 17 U.S.C. extent to which any future benchmark pro- § 504(c)(2). vides guidance about fair market value, the totality of the circumstances that surround 70. This Opinion does not take a position on the creation of that future license will have to whether a fair market price would have re- be considered. sulted if Sony had provided a timely list of its 360 6 FEDERAL SUPPLEMENT, 3d SERIES

ASCAP makes two additional arguments compulsion renders the 2.28% rate a poor in support of its contention that the Pando- benchmark. Since Sony achieved the ra–Sony license reflects a fair market val- 2.28% rate in such circumstances, it is ue for a Pandora license. First, as evi- reasonable to infer that the fair market dence that the license was the product of value for Pandora’s license is materially meaningful give and take between the par- lower than 2.28%. ties, ASCAP contends that Sony made a 3. The Pandora–UMPG License meaningful concession regarding the ad- vertising deduction when it allowed Pando- UMPG and Pandora executed a six- ra to deduct internal advertising costs. month license for the last half of 2013 that There was little evidence offered at trial had an industry-wide rate of 7.5%, and an on negotiations over this term of the li- implied ASCAP rate of 3.42%. ASCAP has cense. ASCAP certainly did not show that failed to show that this license is a useful Sony resisted it or was reluctant to agree benchmark for an ASCAP license with to it. After all, EMI had already agreed Pandora. to such a deduction in the event another As already described, there were virtu- major publisher or a PRO accepted it. ally no meaningful negotiations between Given this record, this single modification Pandora and UMPG because UMPG, con- of a draft agreement in the course of a trolling roughly 20% of the music market, four-day negotiation period does not alter began with and insisted upon a demand the conclusion that Pandora was compelled that bore no relation to the then-existing to enter into a license on Sony’s terms. market price. One of Pandora’s principal ASCAP also argues that Pandora’s wit- competitors was covered by the RMLC at nesses have admitted that the effective a rate of 1.70%; Pandora had been cov- rate for Pandora of the Sony license, after ered under the 5.0 License and had recent- the deduction of internal advertising costs ly executed a license with EMI that en- is taken into account, was only a ‘‘modest’’ compassed the year 2013 at a 1.85% rate; increase over the effective rate of the AS- and Sony had obtained a hike to an implied CAP license and was not ‘‘out of con- ASCAP rate of 2.28%. But, UMPG’s 7.5% trol.’’ 71 But even if this were true, it industry-wide rate implied an ASCAP rate would not cure the primary concern with of 3.42%. This was even higher than the the Sony license as a benchmark, which is ASCAP rate for interactive music services, the coercive process by which it was nego- tiated. In any event, some of Pandora’s which was set at 3.00%. If there was one characterizations of the Sony rate were principle regarding rate structure on made to contrast it with the exorbitant which the parties agreed at trial it was UMPG license rate, and not as an indepen- that the rate for customized radio should dent assessment of its reasonableness. be set below the rate for on-demand inter- active services. In sum, the combination of the looming January 1, 2013 deadline and the lack of UMPG’s leap in rate, demanded within a information about the Sony catalogue matter of weeks following the Sony negoti- meant that Pandora was compelled to con- ations, was so astounding that it drove clude a licensing agreement with Sony at Pandora to buy a radio station and to file a the end of 2012. The presence of such summary judgment motion challenging the

71. While the implied ASCAP headline rate of calculated in light of the [REDACTED] deduc- the Sony–Pandora license is 2.28%, the net tion for internal advertising costs, is [RE- effective rate of the Sony–Pandora license, DACTED]. IN RE PANDORA MEDIA, INC. 361 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) legality of the Compendium modification. C. ASCAP’s Secondary Benchmarks: Given UMPG’s bargaining stance, includ- the SESAC and Apple Licenses ing its unwillingness in Pandora’s eyes to ASCAP has offered the Pandora license proceed in a businesslike manner, Pandora with SESAC and the Apple iTunes Radio agreed to a contingent, short-term license, licenses as confirmatory of the Sony and and placed its fate in the hands of the UMPG license rates. Without the Sony ongoing rate court proceedings.72 In such and UMPG license agreements as bench- circumstances, this license rate cannot be marks, however, these confirmatory bench- said to represent a bargain arrived at by a marks lose their utility. Nonetheless, willing buyer and seller. since the parties devoted attention to these Moreover, UMPG not only demanded an licenses, they will be discussed. In short, extraordinarily steep increase above the there is insufficient data about the SESAC prevailing market rate, but also deprived repertoire and the Apple iTunes Radio Pandora, as Sony had, of critical leverage business model to make the adjustments in their negotiations. Although UMPG required to support an increase of rate provided Pandora with a list of its works, above 1.85%. UMPG insisted on doing so under the 1. The Pandora–SESAC license umbrella of an NDA. The NDA accompa- nying the list would appear to any reason- The Pandora license with SESAC had able reader to prohibit Pandora from using an escalating rate that runs from an im- the list to take down UMPG works from plied rate for ASCAP of [REDACTED] to its service.73 At the very least, the NDA [REDACTED], assuming that SESAC has raised the specter that UMPG would sue a 10% market share and ASCAP has a Pandora if it used the list to do so. For 45.6% market share.75 Of course, even at this additional reason, the UMPG license its upper range, the SESAC license rate is rate is not a useful benchmark.74 lower than the implied Sony rate of 2.28%

72. As described above, the UMPG agreement clusively to the list of works was reasonably was contingent on the outcome of (1) Pando- taken by Pandora as further evidence that any ra’s summary judgment motion that the pub- efforts to negotiate with UMPG would be fu- lisher withdrawals had no effect on its ASCAP tile. license application, and (2) a determination that Pandora is entitled to the RMLC rate as 75. ASCAP contends that the implied rate for the result of its purchase of a terrestrial radio ASCAP should be even higher. ASCAP uses a station. Pandora succeeded with the first 7% figure rather than a 10% figure, relying on motion in this Court, which means that the materials obtained in discovery, to show that UMPG license will have no effect unless that SESAC’s share of PRO revenue (as opposed to decision is reversed on appeal. its share of compositions) in 2011 was 7%. Holding this measure of SESAC’s share 73. Pandora is correct that the NDA unambig- steady, and again making no adjustments for uously prohibited Pandora from using the list any growth of SESAC’s share and concomi- to take down UMPG works from its service. tant decline in ASCAP’s, ASCAP calculates ‘‘Whether a contract is ambiguous is a ques- that the 7% figure yields an implied ASCAP tion of law.’’ VAM Check Cashing Corp. v. rate of [REDACTED] in 2011, [REDACTED] Fed. Ins. Co., 699 F.3d 727, 729 (2d Cir.2012) in 2012, [REDACTED] in 2013, [REDACTED] (citation omitted). in 2014, and [REDACTED] in 2015. But it is not the industry practice to use share of PRO 74. ASCAP contends that Pandora’s failure to revenue as the relevant number in calculating object to the NDA on this ground is evidence implied rates. Moreover, the relevant figure that the complaint about the NDA’s terms is is what Pandora thought the SESAC market an after-the-fact concoction. To the contrary, share constituted at the time it entered into UMPG’s insistence on an NDA addressed ex- the agreement. 362 6 FEDERAL SUPPLEMENT, 3d SERIES

and implied UMPG rate of 3.42% for AS- escalating rate was intended to recapture CAP licenses to Pandora. There are sev- the initial underpayment over time. eral reasons, however, that the SESAC Therefore, for each of these reasons, the license terms provide minimal guidance SESAC license with Pandora is of limited here. utility in assessing the appropriate rate for The SESAC license has historically been a Pandora–ASCAP license. If it were to a benchmark of limited value because the be used at all, it does not suggest a rate public knows little about the size of the above 1.85% for an ASCAP license. SESAC repertoire. See MobiTV, 712 2. The Apple Licenses F.Supp.2d at 254. As such, it is difficult to As it was announcing the inauguration of adjust a SESAC license rate to arrive with iTunes Radio, Apple entered into a set of confidence at an implied ASCAP rate. licenses with ASCAP and music publishers This problem is exacerbated by the fact premised on an industry-wide rate of 10%. that SESAC’s small size, when compared to ASCAP and BMI, not only amplifies This implies an ASCAP rate of [REDACT- any error in a projection, but also reduces ED]. This rate is substantially in excess the incentive to resist SESAC’s rate re- of the 1.85% rate of the Pandora–EMI quests. While the cost associated with license, as well as the implied rates of resistance may not be justified when a 2.28% and 3.42% for the Sony and UMPG license fee is relatively small, the willing- licenses with Pandora, respectively. ness to incur those costs will necessarily There are at least two reasons why the grow with the size of the anticipated pay- Apple licenses for its iTunes Radio service ments. provide little guidance for an ASCAP–Pan- dora license. Second, Pandora’s contemporaneous un- derstanding of the SESAC repertoire and First, iTunes Radio is a service offered the rate undercut any suggestion that it by Apple to complement its iTunes Store supports an ASCAP rate above 1.85%. SE- and iTunes Match. Through these latter SAC argued that an escalating rate in the services, Apple sells digital downloads and SESAC license was appropriate to account operates a locker system so that subscrib- for SESAC’s anticipated growth in market ers may access their music from any of share.76 There was no evidence presented their Apple devices. The integration of at trial to suggest that the ASCAP market these services, seamlessly, within the Ap- share is growing. Indeed, to the extent ple ecosystem generates synergies. As a the SESAC share is growing, ASCAP’s consequence, Apple conducted negotiations share would be presumed to be declining. for its licenses for the public performance Thus, if this principle were applied even- of compositions within the context of a handedly, for the latter years of the AS- business model that has no analogue for CAP–Pandora license, the ASCAP rate Pandora. might move below 1.85%. Second, the Apple revenue base for its In addition, Pandora believed that the licenses has several exclusions that may be initial SESAC rate reflected an approxi- important. Although Apple advertises its mate SESAC market share of [REDACT- music offerings over iTunes Radio, none of ED], despite SESAC’s representation that that imputed advertising revenue is cap- its market share figure was 10%. The tured in the revenue base. The revenue

76. Through this discussion of the SESAC li- endorse a rate structure in which an increas- cense, this Opinion should not be read to ing market share justifies an increase in rate. IN RE PANDORA MEDIA, INC. 363 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

base also excludes any contribution from the sale of music, and this justification the iTunes Match subscription fees. None arose at trial. Fourth, ASCAP has em- of the revenue from the sales of down- phasized a purported difference in the in- loads, purchases that can be made by click- tensity of music use between internet mu- ing on a buy button while listening to sic services like Pandora and terrestrial iTunes Radio, is captured. And the list radio services. Finally, although unstated, goes on. And because iTunes Radio the publishers and ASCAP appear to be- launched only a short time before trial, lieve that Pandora’s license rate should be data about the service is scarce and no one increased because Pandora is currently a was in a position to undertake the exceed- successful internet radio service. Each of ingly difficult task of making adjustments these reasons for a hike in rates will be to the terms of the Apple license to calcu- discussed in turn. late an equivalent rate for an ASCAP li- 1. An Increase in Competition cense issued to Pandora. Consequently, Dr. Murphy posits, on behalf of ASCAP, the Apple licenses do not provide a basis that an increase in the demand for public to find with any confidence that a rate performance rights in musical works on above 1.85% is a fair market rate for an the internet would lead to an increase in ASCAP license issued to Pandora. market prices in a competitive market. Dr. Murphy and ASCAP list a number of D. ASCAP’s Theoretical Arguments recent customized radio services which and Motivations have emerged in support of the relevance In addition to offering benchmarks, ei- of this theory here. They tender this ob- ther ASCAP or its witnesses presented servation to support an increase in the five arguments in support of either a high- Pandora licensing rate from 1.85% in 2012 er rate for a Pandora license than it had to 3.00% by 2015. This theory of economic historically paid, or an escalating rate behavior in a competitive market is so within a single Pandora license. These untethered to actual music industry mar- theoretical arguments seek to justify AS- ket conditions and historical evidence that CAP’s request for an otherwise hard-to- it provides minimal assistance when the explain sharp rate increase from 1.85% in task at hand is to set the rate for this five 2011 and 2012 to 2.50% and 3.00% in the year Pandora–ASCAP license. years between 2013 and 2015. First, Dr. First, Dr. Murphy does not grapple with Murphy identified increased competition the history of music on the internet and among internet music users, and listeners’ the licensing rates for that music, and the preference for variety in music, to support implications of that history for his theory. ASCAP’s license request. Second, UMPG There has been a sizable and growing in- and Sony both justified their demands for ternet radio industry since the mid–1990s. a higher rate from Pandora because of the At first, these were simulcast stations. In- extreme gap between the size of payments ternet-only radio arrived shortly thereaf- made by Pandora for rights to the public ter. Customized radio entered the arena performance of compositions and sound re- in the late 1990s.77 There has been in- cordings. Third, there is theoretical and creasing competition in the radio and in- historical support for imposing a higher ternet music spheres for a long time with- rate on a music service that cannibalizes out corresponding rises in licensing rates

77. On-demand services have also existed tors with radio, they are in the same general since at least 2001, when Rhapsody was market. launched. While not the most direct competi- 364 6 FEDERAL SUPPLEMENT, 3d SERIES

attributable to increased competition. AS- that market, an increase in demand does CAP adopted its 5.0 License for internet not necessarily result in an increase in music in 2004. Since that year, right rate. The rights holder participates in the through until today, ASCAP has utilized a growth of revenue by application of a sta- single rate—1.85%—for the most music- ble rate to an expanding revenue base. In intensive internet services. ASCAP li- light of all of the above, Dr. Murphy’s censed Pandora itself under the 1.85% rate competition theory does not persuasively for the entire period of 2005–2010, despite suggest that the rate for Pandora’s license the purported increase in competition should rise above 1.85% in the final three within the field of internet radio, including years of its license with ASCAP, nor that customized radio, over that period. EMI the Sony or UMPG license rates are nec- chose to adopt the 1.85% rate for Pandora essarily fair market rates. for the years 2012 and 2013. As recently 2. Demand for Variety as 2012, ASCAP accepted a blended rate Dr. Murphy offers a second theoretical of 1.70% for terrestrial and internet ra- argument in support of ASCAP’s request- 78 dio. This stability in rates over a decade ed fee structure. He contends that, all in the internet music market is unex- else being equal, ‘‘listeners prefer more plained by Dr. Murphy’s observation, and variety to less.’’ From this observation, he indeed runs contrary to it. In particular, concludes that the demand for variety in- ASCAP has made no showing that compe- creases the competitive market price of tition within the internet music market rights to publicly perform musical works increased so dramatically within the single and justifies an increased rate for a Pando- year in which the Sony and UMPG licens- ra license. es were negotiated, that an increase in As was true with Dr. Murphy’s first rate from 1.85% (which ASCAP agrees is theoretical assumption, this theory comes the correct rate for the year 2012) to undone when applied to the real world. 3.42% (the ASCAP rate implied by the Dr. Murphy’s claim that listeners prefer UMPG license in 2013) was reasonable. variety above all is unsupported and can- In addition, there are theoretical gaps in not form the basis for an upward depar- Dr. Murphy’s theory. For one, the laws of ture from a rate of 1.85%. Dr. Murphy did supply and demand teach us that the price not conduct any research or analysis into of a commodity will increase as demand consumer listening behavior to arrive at increases, but only to the extent that sup- his conclusion that listeners prefer variety ply is held constant. Dr. Murphy did not above all. And it is likely that once a attempt to address whether a growth in certain minimal variety threshold is music supply may also have contributed to reached listeners don’t actually prioritize a stability in rates over many years, and extra variety. The record evidence sug- whether any tendency to raise rates within gests that, as a general matter, listeners a competitive market is tempered by the are not so eclectic in their tastes that the expectation that supply would increase in addition of a song to a music service will such a circumstance. Moreover, Dr. Mur- necessarily provide added value. Listen- phy’s theory is also difficult to apply to a ers often like to hear music that they market in which the price of music is already know that they like, or music very expressed as a percentage of revenue. In similar to music they already like.

78. As one more example, for ten years, been set at the same level: 10.5%. See 37 through two five year terms, the rate for me- C.F.R. § 385.12(c). chanical rights for on-demand services has IN RE PANDORA MEDIA, INC. 365 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

Confronted with this fact, Dr. Murphy er demonstrated demand for an extensive emphasized instead the need a music ser- library of music than do on-demand ser- vice has for variety so that it can satisfy vices. Although Pandora has access to the the many distinct individual tastes of its full repertoire of each of the PROs, its listeners. But, even if the price of a li- MGP includes only about one million com- cense were driven by the extent to which a positions.79 In contrast, on-demand ser- music service requires a large library of vices need to be able to play roughly twen- compositions, ASCAP’s license fee request ty times as many compositions. Spotify would not be supported. First, there was has somewhere in the neighborhood of 20 no change in the nature of the Pandora million tracks. After all, on-demand ser- service between the years 2012 and 2013, vices essentially promise their members when measured by an exhibited demand that they can listen to any song they want for variety in its repertoire, to support the anytime they want. Consequently, even if shift in rate from 1.85% to 2.50% and one accepted Dr. Murphy’s theory of vari- higher, as proposed by ASCAP. This is ety, ASCAP’s rate would have to be set especially notable in light of the fact that well below the 3.0% rate applicable to on- Pandora, as a blanket licensee, faced no demand services under ASCAP’s form li- marginal cost from adding songs from the cense. ASCAP repertoire to its MGP and pre- Even when it comes to programmed ra- sumably would add every song in the AS- dio stations, ASCAP did not show that CAP repertoire to its service if demand for those stations could succeed with fewer variety were a primary driver. Instead, licenses or smaller libraries than Pandora. its business model has allowed it to use Take just one example: a Top 40 station. just a fraction of the repertoire. To be successful, a Top 40 station would need licenses from every holder of rights As significantly, even if Pandora were to Top 40 songs. This would in all likeli- shown to demand variety above all, the hood necessitate deals with all three PROs ‘‘variety’’ metric certainly does not support as well as any major publishers with works a fee for Pandora’s service that is as high outside of the PROs. Pandora, in contrast, as the 3.0% fee ASCAP charges on-de- has the ability to substitute songs.80 Its mand music users in its form license. And MGP enables it to play songs with charac- it is even unlikely to justify a higher rate teristics its listeners enjoy rather than than the fee ASCAP uses for programmed each popular song in a category. As a internet music. theoretical matter, this flexibility in pro- By a large order of magnitude, custom- gramming gives Pandora more flexibility ized music services like Pandora have low- in licensing negotiations.81 If the demand

79. There was evidence at trial that Pandora 81. ASCAP did not attempt to show at trial may recently have added as many as a million that Pandora’s need for an extensive library of more compositions to its music service for compositions was greater than, to take an some purposes. example, Clear Channel’s iHeartRadio, which ASCAP licenses at a rate of 1.70%. ASCAP did 80. For example, Pandora conducted experi- offer evidence from a surveyor of major com- ments in the Summer of 2013 called the ‘‘La- mercial radio stations, which it used to argue that Pandora performs a wider variety of bel Experiments,’’ in which it determined that songs from a wider selection of artists than it could substitute away from certain record terrestrial radio. Given Pandora’s ability to labels with little negative response from Pan- substitute songs within a genre, however, the dora users. survey may show little more than Pandora’s flexibility. 366 6 FEDERAL SUPPLEMENT, 3d SERIES

for variety were a driving factor in licens- been driven by either of the rationales ing negotiations, therefore, Pandora’s li- proffered by Dr. Murphy. There was no cense rate should probably not be set high- evidence that ASCAP, EMI, Sony, UMPG, er than broadcast radio’s 1.70%, and or any other licensor negotiated with any should perhaps be lower. music user on the ground that their ser- Dr. Murphy’s theory also has very little vice required a larger catalogue or that utility in the world in which ASCAP and there had been a recent surge in competi- Pandora operate. It does not help to dis- tion in the market.82 But, there was ample tinguish among types of music services in evidence of the actual driving force behind a concentrated industry, and where blan- the Sony and UMPG withdrawal of new ket licenses exist. To be successful, every media rights from ASCAP and their nego- music service needs a license with every tiations with Pandora. That driving force PRO. And where their works are not cov- was the music publishers’ envy at the rate ered by PRO licenses, every music service their sound recording brethren had ex- will ultimately need a license with both tracted from Pandora through proceedings Sony and UMPG, which together control before another rate setting body, the CRB. about half of the U.S. market. Thus, it is [6] ASCAP has not offered any theo- difficult to see how demand for variety retical support for raising the rate for could assist in distinguishing between rates for different types of music services public performance of a composition by a when all types of music services must comparison to the rate set for sound re- make the same licensing deals to survive. cording rights. There may be several rea- Moreover, with a blanket license a music sons for this, but first and foremost is the user receives the right to every work in statutory prohibition on considering sound the repertoire, whether it avails itself of recording rates in setting a rate for a the opportunity to play all the works or license for public performance of a musical not. As a result a music service with a work. See 17 U.S.C. § 114(i) (‘‘License high demand for variety can satisfy that fees payable for the public performance of demand by playing more of the songs cov- sound recordings TTT shall not be taken ered by its blanket license. into account in any TTT proceeding to set In sum, Dr. Murphy’s contention that a or adjust the royalties payable to copy- demand for variety in music can explain right owners of musical works for the pub- the upward adjustments in the Pandora– lic performance of their works.’’). Thus, ASCAP license rates must be rejected. It this Court may not take the rates set by does not fit the world in which Pandora the CRB into account in determining the operates or the Pandora business model. fair market rate for a public performance If anything, resort to this theory under- license from ASCAP to Pandora. cuts ASCAP’s request for an increase in a Despite this statutory prohibition, one license rate for Pandora. observation may be safely made. Unhap- 3. Disparity Between Sound Recording piness about the gap between what Pando- and Composition Fees ra pays record companies and what it pays It is worth observing that there is no the PROs drove the modification to the evidence in the record that any of the ASCAP Compendium, the publishers’ licensing negotiations in this industry have withdrawals from ASCAP, and the Sony

82. It is also worth noting that there is no programmed radio in any context outside of evidence that ASCAP has taken action to li- Pandora. cense customized radio at a higher rate than IN RE PANDORA MEDIA, INC. 367 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) and UMPG negotiations with Pandora. The parties have argued about the ex- The corporate rivalries over digital age tent to which Pandora and services like it revenues explain a great deal of this histo- are promotional or cannibalistic. There is ry. In any event, the record is devoid of apparently no industry consensus on this any principled explanation given by either question. It is worth noting, however, that Sony or UMPG to Pandora why the rate what evidence was presented at trial sug- for sound recording rights should dictate gests that Pandora is promotional. any change in the rate for composition To begin with, radio has traditionally rights. been considered promotional. The record 4. Cannibalization of Music Sales industry has long sought to have its music 83 There is agreement between the parties played on radio stations. Pandora is no exception. Record labels have taken ad- that it is appropriate to require a higher vantage of Pandora Premieres to feature licensing fee from a music service that acts new work in advance of release, with the as a substitute for the sale of a musical hope that that exposure will engender work, when compared to one that does not. sales. Pandora itself has buy buttons that To the extent that a music service is a permit listeners to buy digital downloads replacement for sales, it is said to canni- from Amazon and Apple, and they use balize the sales; to the extent it encour- them with some frequency.84 There is no ages sales, it is said to be promotional. evidence that artists have taken steps to The reasons for this distinction arise, at prevent Pandora from playing the artist’s least in part, from a separate stream of work. As significantly, one of Pandora’s rights belonging to composers. Compos- principal competitors—iTunes Radio—was ers have a copyright interest in the repro- created to complement Apple’s iTunes duction and distribution of musical works, Store and promote sales in that digital an interest that is referred to as ‘‘mechani- store.85 cal rights.’’ 17 U.S.C. §§ 106(1) & (3); 17 In contrast, on-demand streaming ser- U.S.C. § 115. The licensing regime for vices like Spotify are widely considered mechanical rights is complex, and its de- cannibalistic and are licensed at a higher tails need not be described here. It is rate accordingly. After all, a listener has sufficient to observe that when hard copies no need to purchase a digital download (e.g., vinyl records, CDs) or digital down- when the listener has any song that she loads of compositions are sold, the compos- wants to hear instantaneously available ers receive mechanical rights payments. through Spotify. For this very reason, On-demand services, as well, are required some prominent performers have acted to to pay mechanical rights. As described prevent Spotify from playing their record- earlier, Spotify pays a 10.5% fee for both ings. In sum, while this metric—whether mechanical rights and the right to publicly a service is promotional or cannibalistic— perform a musical work. could justify a differentiation of rates be-

83. There is a well-documented history of rec- 85. The preamble to Apple’s licenses with pub- ord promoters going so far as to use bribes, lishers for its iTunes Radio service provides or ‘‘payola,’’ to increase the number of times that ‘‘[w]hereas, Apple wishes to enter into songs are played on a radio station. this Agreement with Publisher to enable Ap- ple to create an advertising supported Inter- 84. Pandora’s ‘‘buy button’’ resulted in over $3 net radio service, which will further enhance million per month in music sales on Amazon the music discovery and recommendations and the iTunes Store during 2013. features of the iTunes Store for the purpose of promoting sales of eMasters.’’ 368 6 FEDERAL SUPPLEMENT, 3d SERIES

tween services, ASCAP failed to show that basis of music use. As described above, Pandora is anything other than promotion- the habits of terrestrial radio listeners and al of sales.86 the presence of music in advertising may 5. Music Intensity make any differential largely illusory. ASCAP argued at trial that Pandora’s Moreover, ASCAP does not have a history licensing fee should exceed the RMLC of fine-tuning its rates in the way suggest- rate of 1.70% because its channels use ed here. The RMLC license rate for mu- music more intensively than terrestrial ra- sic intensive services covers a far broader dio stations. Music intensive broadcast range of music usage than the differential stations play, on average, 11 songs per between 11 and 15 songs per hour. hour; Pandora’s stations play 15 or so. Finally, ASCAP agrees that Pandora’s This difference is attributable at least in rate for 2011 and 2012 should be 1.85%. part to the difficulty of placing advertising And it has presented no evidence that the on internet radio, which is a challenge that music intensity of Pandora’s services will Pandora is addressing through its substan- change in any material way for the last tial investment in an in-house advertising three years of the license term. For this department. In any event, ASCAP has reason, as well, the music intensity metric not shown that this current differential cannot provide a basis to justify the hike in justifies any increase in the last three rates to 2.50% and 3.00% that ASCAP years of the Pandora license above the seeks. 1.85% rate it has requested for the first 6. Pandora’s Success two years. There is one final motivation for AS- First, assuming that the purported mu- sic intensity differential justifies an up- CAP’s requested rates that must be ac- ward departure for Pandora’s rate, AS- knowledged. The backdrop for this rate CAP has already created that differential. court dispute is the arrival of the digital Its 5.0 License rate for internet music age in the music industry, the resulting services like Pandora is 1.85% whereas its disruption to the business models of the RMLC license rate for Music Format sta- music industry, and Pandora’s current suc- tions is 1.70%. And the form license rate cess in the digital radio market. for new media services escalated from the A rights holder is, of course, entitled to prior ASCAP form license rate of 1.615% a fee that reflects the fair value of its to reflect ‘‘the maturation of the new me- contribution to a commercial enterprise. dia marketplace and ASCAP’s observation It is not entitled, however, to an increased that many of these services were using a fee simply because an enterprise has found large amount of ASCAP [m]usic.’’ success through its adoption of an innova- Second, it is not clear, in all events, that tive business model, its investment in tech- any rate differential is justified on the nology, or its creative use of other re-

86. ASCAP relies on an annual 2012 study by vices versus customized or programmed ra- a firm called NPD which showed that Pando- dio services. Users of on-demand services ra users tend to purchase less music than do tend to be music ‘‘super fans’’ who know users of on-demand services like Spotify. what they want to listen to and use on-de- But this does not show that Pandora is more mand services as supplements to purchased cannibalistic of music sales than on-demand music collections. Users of customized radio services (or that it is cannibalistic at all). services like Pandora tend to be more casual, Correlation does not equal causation, and the or ‘‘lean back’’ music listeners, who are less disparity may be fully explained by the self likely to purchase music for their own collec- selection of music users into on-demand ser- tions. IN RE PANDORA MEDIA, INC. 369 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014) sources. It appears that Sony, UMPG, issue of whether such a separate rate and ASCAP (largely because of the pres- structure would be justified, much less sure exerted on ASCAP by Sony and what the spread in rates should be be- UMPG) have targeted Pandora at least in tween programmed radio and customized part because its commercial success has internet radio. Suffice it to say that AS- made it an appealing target. CAP has not shown that Pandora’s partic- Pandora has shown that its considerable ular success in expanding its audience and success in bringing radio to the internet is revenue justifies in any way an increase in attributable not just to the music it plays rate from the 1.85% rate which ASCAP (which is available as well to all of its seeks for the years 2011 and 2012 to the competitors), but also to its creation of the higher rates ASCAP seeks in the three MGP and its considerable investment in succeeding years. the development and maintenance of that innovation. These investments by Pando- III. Whether Pandora is Entitled to the ra, which make it less dependent on the RMLC 1.70% Rate purchase of any individual work of music [7] Before concluding that the rate for than at least some of its competitors, do an ASCAP license to Pandora for the five not entitle ASCAP to any increase in the years from 2011 through 2015 should be rate it charges for the public performance 1.85%, it is necessary to address Pandora’s of music. To the extent Pandora prospers contention that it is similarly situated to because of its innovations and because of the RMLC licensees and entitled to the its separate investment in an initiative to RMLC license rate of 1.70% under the develop advertising revenue, ASCAP and anti-discrimination provisions of AFJ2. See its members will prosper through the in- AFJ2 §§ IV(C), IX(G). Pandora has not creased revenue stream that is generated shown that a ruling in this Opinion that by the application of an appropriate rate to requires ASCAP to license Pandora at a Pandora’s revenue base. rate of 1.85% from 2011 to 2015 will violate Moreover, market share or revenue met- the anti-discrimination provisions of AFJ2. rics are poor foundations on which to con- There are several provisions in AFJ2 struct a reasonable fee. Internet radio that concern similarly situated licensees remains in its infancy. There is little like- and impose upon ASCAP the duty to treat lihood that the landscape of today will them in a non-discriminatory manner. remain unaltered. Indeed, remarkable AFJ2 § IV(C) enjoins and restrains AS- changes occur with lightning speed in the CAP from ‘‘[e]ntering into, recognizing, digital age. enforcing or claiming any rights under any As of today, ASCAP has two sets of license for rights of public performance rates for internet radio: those in the which discriminates in license fees or other RMLC license (which are available to own- terms and conditions between licensees ers of broadcast radio stations) and those similarly situated.’’ Section IX(G) of in its form licenses. Neither set of rates AFJ2 further provides that ‘‘[w]hen a rea- has a separate schedule for customized sonable fee has been determined by the radio. ASCAP may or may not wish to Court,’’ ASCAP must ‘‘offer a license at a explore the creation of a separate rate comparable fee to all other similarly situat- structure for customized radio services like ed music users who shall thereafter re- Pandora. It is unnecessary to reach the quest a license of ASCAP.’’ 87

87. See also AFJ2 §§ VIII(A) (prohibiting dis- crimination in the types of licenses offered); 370 6 FEDERAL SUPPLEMENT, 3d SERIES

AFJ2 defines ‘‘similarly situated’’ licen- is a direct competitor with Pandora and is sees as ‘‘music users or licensees in the a growing component of the iHeartRadio same industry that perform ASCAP music offering. and that operate similar businesses and Although Pandora contends that it is use music in similar ways and with similar similarly situated to all RMLC licensees, it frequency.’’ AFJ2 § II(R). It lists the emphasizes its similarity to Clear Chan- factors that are relevant to a determina- nel’s iHeartRadio generally, and more spe- tion of whether licensees are similarly situ- cifically to customized radio offerings by ated as including, but not limited to, ‘‘the RMLC members Clear Channel and CBS, nature and frequency of musical perform- the Create Station and Last.fm services, ances, ASCAP’s cost of administering li- respectively.88 Pandora has shown that its censes, whether the music users or licen- service is indistinguishable for licensing sees compete with one another, and the purposes from these components of Clear amount and source of the music users’ Channel and CBS. revenue.’’ Id. More generally, Pandora has shown that The licensees to which Pandora seeks it is radio and competes with programmed comparison are those covered by the radio, including terrestrial radio, for listen- RMLC license. The current RMLC li- ers and advertising dollars. Its most di- cense was approved in early 2012 and runs rect competitors within the radio industry through the year 2016. As described in are other internet radio services, especially more detail above, it allows RMLC mem- customized radio services. Although ter- bers to pay at a rate of 1.70% of the restrial radio stations generally play about revenue derived from radio stations that four fewer songs per hour than Pandora, principally play music. The RMLC mem- Pandora has shown that this difference is bers own commercial radio stations. With not material given the broad categories of the advent of the internet, many of the music use in the ASCAP–RMLC license. RMLC members have simulcast their pro- In any event, ASCAP has not offered any gramming for their terrestrial stations on evidence to suggest that the internet radio the internet. Some have also created in- ternet-only radio stations. Of the top 20 stations of RMLC members use music less internet radio services, as measured by the intensively than Pandora. Indeed, for that Triton Digital firm, 16 of the 19 services period of time in which Clear Channel ran that are not Pandora are owned by RMLC its Create Station feature as an ad-free entities. Clear Channel’s internet radio service, it was a more music intensive ser- offering, known as iHeartRadio, includes a vice than Pandora. And ASCAP has not customized internet radio service called suggested that there is any distinction Create Station. Create Station began op- among any of these services in terms of eration in 2011 and was initially run as a ASCAP’s cost of administering their li- commercial free service. By late 2012, it censes. was running with advertising and generat- Of the factors which AFJ2 lists as perti- ing revenue. The Create Station feature nent to an analysis of similarity, the only

IX(F) (presumption of setting interim rates at 1997), which questioned whether the compar- same level as those between ASCAP and licen- ison in a similarly-situated analysis in the sees similarly situated to an applicant). context of a multi-member license should be to the median licensee or to any given licen- 88. The parties debate the significance of dicta see. Because of the conclusion drawn herein, in In re Applications of Salem Media of Cali- it is unnecessary to resolve that question. fornia, et al., 981 F.Supp. 199, 201 (S.D.N.Y. IN RE PANDORA MEDIA, INC. 371 Cite as 6 F.Supp.3d 317 (S.D.N.Y. 2014)

factor which may provide a basis for dis- the internet radio sphere. Moreover, the tinguishing between Pandora and other revenues from terrestrial radio swamp customized internet radio services run by those from the internet services. Second, RMLC members is the amount of revenue. while Pandora’s service is, for the purpose But, while it appears that Pandora is earn- of this analysis, identical to services of- ing significantly more revenue than the fered by some RMLC members, AFJ2 for- RMLC customized radio services, all of the bids discrimination among licensees, and internet music services at issue here are Pandora has not shown that it is similarly run as commercial stations. situated to any RMLC licensee. Pandora relies heavily on comparison with Clear ASCAP emphasizes the degree to which Channel’s iHeartRadio’s customized Cre- Pandora markets itself as an improvement ate Station feature. But Clear Channel is on traditional radio. It is true that the the licensee, and the Create Station fea- digital delivery of music has permitted the ture constitutes a very small part of Clear creation of customized radio stations that Channel’s business at present. Third, Pan- are unique to individual listeners. But, dora is as similarly situated to internet despite that development, customized radio music services covered by the 5.0 License retains the essential characteristics of ra- at the rate of 1.85%. Since this Opinion dio. The radio service programs and de- sets the rate for the Pandora license at livers the music and the listener does not 1.85%, it is difficult, if not impossible, to know which song will be played next. For find that there is any violation of AFJ2 purposes of analyzing the non-discrimina- due to a discrimination in rates. tion prohibitions of AFJ2, of course, what matters is the degree to which the RMLC What this discussion may underscore is licensees are similarly situated to Pandora, a lack of coherence in the present rate and it is uncontested that RMLC licensees structure of ASCAP licenses. This is un- provide both programmed and customized derstandable given the evolving nature of internet radio services. the radio market. Any change in rate structure (for instance, to create a rate In light of these similarities, the ques- structure for customized music services) tion that is fairly presented by Pandora’s would have to be made with care based on application is whether it is entitled by a thorough understanding of the market AFJ2 to the RMLC rate.89 The answer to and the uses of music in the market, in- that question, while close, is no. Pandora formed by a desire not to discriminate is not entitled to the 1.70% RMLC rate for among similarly situated licensees or be- at least three reasons. tween similar services simply because of a First, the RMLC rate applies to a large- difference in the mode of distribution.90 scale license agreement that binds a vari- After all, if there is no commercially legiti- ety of licensees in both the terrestrial and mate reason for a distinction in rates, then

89. As AFJ2 requires, nothing in this Opinion any way affected or altered by such determi- may be construed as affecting the present nation for the term of such license agree- ASCAP–RMLC license. See AFJ2 § IX(G) ment.’’). (‘‘[A]ny license agreement that has been exe- cuted between ASCAP and another similarly 90. If ASCAP revises its rate structure it will situated music user prior to such determina- no doubt be attuned to the need to treat major tion by the Court shall not be deemed to be in competitors in a market fairly. 372 6 FEDERAL SUPPLEMENT, 3d SERIES the distinction would not survive in a com- payments to publishers made following petitive market. their partial withdrawals from ASCAP. SO ORDERED. IV. Publisher Concerns Regarding the Consent Decree and the Rate Court There is one remaining issue to address. , ASCAP, Sony, and UMPG witnesses ex- pressed frustration with the Consent De- cree and the rate court process, both in their communications with each other and in their trial testimony. LoFrumento ex- plained that this frustration arrived with PEEKSKILL CITY SCHOOL the digital age and reflects a fear that the DISTRICT, Plaintiff, record industry will grab all of the avail- v. able revenue from the digital transmission of music. According to ASCAP, AFJ2 and COLONIAL SURETY COMPANY, its processes, in particular the requirement Defendant. that ASCAP issue a license to any appli- No. 11 Civ. 341(SHS). cant, hamper ASCAP’s ability to negotiate a fair market rate. Sony and UMPG wit- United States District Court, nesses asserted that they had to withdraw S.D. New York. their licensing rights from ASCAP in or- Signed March 18, 2014. der to negotiate effectively with Pandora and achieve appropriate parity with sound Background: School district brought recording licensing rates. They expressed breach of contract claim against surety for skepticism that the rate court proceedings an electrical contractor, after the contrac- could determine a fair market value for a tor filed for bankruptcy and ceased work Pandora license. on a school’s construction. The surety moved for summary judgment. The Court is sensitive to ASCAP’s con- cerns and understands that the unique Holdings: The District Court, Sidney H. characteristics of the market for music Stein, J., held that: licensing and the Consent Decree regime (1) the limitations period was not tolled by produce challenges for all parties. But, the district’s motion to dismiss a prior for the reasons already discussed, ASCAP action that was voluntarily dismissed; did not show that the upshot of the negoti- (2) an order granting the district addition- ations conducted by either Sony or UMPG al time to file an answer in a prior with Pandora was a competitive, fair mar- action did not toll the limitations peri- ket rate. od; (3) the limitations period was not equita- CONCLUSION bly tolled; and The headline rate for the ASCAP–Pan- dora license for the years 2011 through (4) the district could not challenge the dis- 2015 is set at 1.85% of revenue for every missal of the prior action in the subse- year of the license term. Pandora is enti- quent action. tled to take a deduction for any direct Motion granted.