United States v. American Society of Composers, Authors & Publishers

09-0539-cv(L ) United States v. American Society of Composers, Authors and Publishers et al. 1 UNITED STATES COURT OF APPEALS 2 FOR THE SECOND CIRCUIT 3 4 August Term 2009 5 (Argued: March 3, 2010 Decided: September 28, 2010) 6 Docket Nos. 09-0539-cv (L), 09-0542-cv (con), 09-0666-cv (xap), 7 09-0692-cv (xap), 09-1572-cv (xap) 8 -----------------------------------------------------x 9 10 UNITED STATES of AMERICA, 11 12 Plaintiff-Appellee, 13 14 -- v. -- 15 16 AMERICAN SOCIETY OF COMPOSERS, AUTHORS AND 17 PUBLISHERS, 18 19 Defendant-Appellant-Cross-Appellee, 20 21 In the matter of Applications of 22 REALNETWORKS, INC., YAHOO! INC., 23 24 Applicants-Appellees-Cross-Appellants. 25 26 27 -----------------------------------------------------x 28 29 B e f o r e : JACOBS, Chief Judge, WALKER and LIVINGSTON, 30 Circuit Judges. 31 American Society of Composers, Authors and Publishers 32 (“ASCAP”) appeals the district court’s ruling that a download of 33 a digital file containing a musical work does not constitute a 34 public performance of that work. Yahoo! Inc. and RealNetworks, 35 Inc. (collectively, the “Internet Companies”) cross-appeal the 36 district court’s assessment of the fees for the blanket licenses 1 1 they seek to perform musical works in the ASCAP repertory. 2 We affirm the district court’s ruling that a download of a 3 musical work does not constitute a public performance of that 4 work, but we vacate the district court’s assessment of fees for 5 the blanket ASCAP licenses sought by the Internet Companies and 6 remand for further proceedings in light of this opinion. 7 AFFIRMED in part, VACATED in part, and REMANDED. 8 CATHERINE E. STETSON (Joshua D. 9 Hawley, Hogan & Hartson LLP, 10 Washington, DC, Ira M. Feinberg, 11 Chava Brandriss, Hogan & Hartson 12 LLP, New York, NY, Christopher J. 13 Glancy, I. Fred Koenigsberg, Stefan 14 M. Mentzer, White & Case LLP, New 15 York, NY, Joan M. McGivern, Richard 16 H. Reimer, Christine A. Pepe, 17 ASCAP, New York, NY on the brief), 18 Hogan & Hartson LLP, Washington, DC 19 for Defendant-Appellant-Cross- 20 Appellee. 21 22 THOMAS P. LANE (Michael S. Elkin, 23 Robert C. Turner on the brief), 24 Winston & Strawn LLP, New York, NY 25 for Applicant-Appellee-Cross- 26 Appellant Yahoo! Inc. 27 28 KENNETH L. STEINTHAL (Jonathan 29 Bloom, Gregory Silbert, Harris 30 Cohen on the brief), Weil, Gotshal 31 & Manges LLP, New York, NY for 32 Applicant-Appellee-Cross-Appellant 33 RealNetworks, Inc. 34 35 NICHOLAS BAGLEY (Tony West, 36 Assistant Attorney General, Philip 37 J. Weiser, Deputy Assistant 38 Attorney General, Scott R. 39 McIntosh, Attorney, Appellate 40 Staff, Civil Division, Department 41 of Justice, Catherine G. 2 1 O’Sullivan, David Seidman, 2 Attorneys, Appellate Section, 3 Antitrust Division, Department of 4 Justice, Washington, DC on the 5 brief), Attorney, Appellate Staff, 6 Civil Division, Department of 7 Justice, Washington, DC for United 8 States. 9 10 MICHAEL E. SALZMAN(Marvin L. 11 Berenson, Joseph J. DiMona, John 12 Coletta, Broadcast Music, Inc., New 13 York, New York on the brief), 14 Hughes, Hubbard & Reed LLP, New 15 York, New York for Amicus Curiae 16 Broadcast Music, Inc. 17 18 KEENAN POPWELL (John C. Beiter, 19 Zumwalt, Almon & Hayes PLLC, 20 Nashville, Tennessee on the brief), 21 SESAC, Inc., New York, New York for 22 Amicus Curiae SESAC, Inc. 23 24 DAVID LEICHTMAN (Hillel I. Parness, 25 Robins, Kaplan, Miller & Ciresi 26 LLP, New York, New York, W. Edward 27 Bailey, Eleanor M. Lackman, Lovells 28 LLP, New York, New York, David 29 Uwemedimo, Confederation 30 Internationale des Societes 31 d’Auteurs et Compositeurs, Neuilly 32 sur Seine, France on the brief), 33 Robins, Kaplan, Miller & Ciresi 34 LLP, New York, New York for Amicus 35 Curiae Confederation Internationale 36 des Societes d’Auteurs et 37 Compositeurs. 38 39 JAY COHEN (Lynn B. Bayard, Paul, 40 Weiss, Rifkind, Wharton & Garrison 41 LLP, New York, New York, Jay 42 Rosenthal, Kathryn E. Wagner, 43 National Music Publishers’ 44 Association, Inc., New York, New 45 York on the brief), Paul, Weiss, 46 Rifkind, Wharton & Garrison LLP, 47 New York, New York for Amici Curiae 48 Association of Independent Music 3 1 Publishers, Church Music Publishers 2 Association, Music Publishers’ 3 Association of the United States, 4 National Music Publishers’ 5 Association, Inc., and Production 6 Music Association. 7 8 CHARLES CUMMINGS (Carl W. Hampe, 9 Baker & McKenzie LLP, Washington, 10 DC, Charles J. Sanders, Attorney at 11 Law PC, Briarcliff Manor, New York 12 on the brief), Baker & McKenzie 13 LLP, New York, New York for Amici 14 Curiae The Society of Composers and 15 Lyricists, National Academy of 16 Recording Arts and Sciences, Inc., 17 Ad Hoc Coalition of Production 18 Music Company Owners, The Game 19 Audio Network Guild, and 20 Songwriters Guild of America. 21 22 AL J. DANIEL (Toby M.J. 23 Butterfield, Christopher J. Marino, 24 Cowan, DeBaets, Abrahams & Sheppard 25 LLP, New York, New York, C. Paul 26 Spurgeon, Society of Composers, 27 Authors, and Music Publishers of 28 Canada, Toronto, Ontario, Canada on 29 the brief), Cowan, DeBaets, 30 Abrahams & Sheppard LLP, New York, 31 New York for Amicus Curiae The 32 Society of Composers, Authors, and 33 Music Publishers of Canada 34 35 PAUL M. SMITH (Steven R. Englund, 36 Carrie F. Apfel on the brief) 37 Jenner & Block LLP, Washington, DC 38 for Amici Curiae Digital Media 39 Association, Entertainment Software 40 Association, and Motion Picture 41 Association of America, Inc. 42 43 JOHN T. MITCHELL, Interaction Law, 44 Washington, DC for Amici Curiae 45 National Association of Recording 46 Merchandisers, Inc., and 47 Entertainment Merchants 48 Association, Inc. 4 1 SUSAN CLEARY, Independent Film and 2 Television Alliance, Los Angeles, 3 California for Amicus Curiae 4 Independent Film and Television 5 Alliance 6 7 BRUCE G. JOSEPH, Wiley Rein LLP, 8 Washington, DC for Amicus Curiae 9 CTIA – The Wireless Association. 10 11 JOHN M. WALKER, JR., Circuit Judge: 12 13 This case presents two distinct questions that arise from 14 the transmittal of musical works over the Internet: First, 15 whether a download of a digital file containing a musical work 16 constitutes a public performance of that musical work; and, 17 second, whether the district court, acting in its capacity as the 18 rate court, was reasonable in its assessment of the blanket 19 license fees of Yahoo! Inc. and RealNetworks, Inc. (collectively, 20 “the Internet Companies”) to publicly perform any of the millions 21 of musical compositions in the American Society of Composers, 22 Authors and Publishers (“ASCAP”) repertory. 23 For the reasons set forth below, we affirm the district 24 court’s ruling that a download of a musical work does not 25 constitute a public performance of that work, but we vacate the 26 district court’s assessment of fees for the blanket ASCAP 27 licenses sought by the Internet Companies and remand for further 28 proceedings. 29 BACKGROUND 30 I. FACTS 5 1 The Internet Companies seek separate blanket licenses to 2 publicly perform the entirety of the ASCAP repertory for certain 3 of their websites and services. A blanket license is a license 4 that gives the licensee the right to perform all of the works in 5 the repertory for a single stated fee that does not vary 6 depending on how much music from the repertory the licensee 7 actually uses. United States v. Am. Soc’y of Composers, Authors 8 & Publishers, No. 41-1395 (WCC), 2001 WL 1589999, at *1 (S.D.N.Y. 9 June 11, 2001). ASCAP licenses the non-dramatic, public 10 performance rights in copyrighted musical works. More than 11 295,000 composers, songwriters, lyricists, and music publishers 12 in the United States participate exclusively in licensing their 13 music through ASCAP. ASCAP licenses approximately 45% of all of 14 the musical works that are played on-line. 15 The Internet Companies perform music in myriad audio and 16 audio-visual contexts. Yahoo! provides music content in various 17 ways across its website. For example, a user can enjoy the 18 specific song or music video he desires from an “on-demand” 19 stream in Yahoo! Search, listen to a radio-style webcast in 20 Yahoo! Music, view audio-visual clips from movies and television 21 shows in Yahoo! Movies and Yahoo! TV, or upload and share his own 22 videos using Yahoo! Video.1 However, only a small portion of the 1 Yahoo! provides audio and audio-visual content in the following areas of its website and through the following services: the Yahoo! homepage, Yahoo! Music, My Yahoo!, Yahoo! Movies, Yahoo! Video, Bix, Yahoo! Kids, Yahoo! TV, Yahoo! Games, 6 1 activity on Yahoo!’s website involves performances of musical 2 works, and not all of the areas on Yahoo!’s website offer audio 3 or audio-visual content. 4 RealNetworks performs music in audio and audio-visual 5 contexts through a number of websites and subscription services.2 6 Like Yahoo!, these sites and services publicly perform musical 7 works in numerous formats, including, inter alia, radio, 8 television, movie, game, and music-video formats. Also like 9 Yahoo!, only a portion of the content on RealNetworks’ sites and 10 services consist of performances of musical works. 11 In addition to performing music on websites and through 12 services, the Internet Companies offer to users copies of 13 recordings of musical works through download transmittals. A Yahoo! Tech, Yahoo! Autos, Yahoo! Finance, Broadway on Yahoo!, Yahoo! Food, Yahoo! Search, Yahoo! Toolbar, Yahoo! Messenger, and the Yahoo! music widget. 2 At the time of oral argument in this case, the RealNetworks sites and services that publicly performed music included, inter alia: the following: RealNetworks.com, Real.com (including its sub-domains such as Real Guide, and affiliated sites such as Rollingstone.com and Film.com), Rhapsody, Rhapsody.com, Rhapsodydirect.com, SuperPass, and Listen.com. Public filings, of which we take judicial notice, see Kavowras v. New York Times Co., 328 F.3d 50, 57 (2d Cir. 2003), report that RealNetworks has since spun off Rhapsody as an independent venture. See RealNetworks, Inc., Current Report (Form 8-K) (March 31, 2010), available at http://www.sec.gov/Archives/edgar/data /1046327/000095012310032214/v55452e8vk.htm. Because this case involves the setting of license rates for the period of January 1, 2004 to December 31, 2009, however, Rhapsody’s spin-off is not relevant to our analysis. 7 1 download is a transmission of an electronic file containing a 2 digital copy of a musical work that is sent from an on-line 3 server to a local hard drive. See United States v. Am. Soc’y of 4 Composers, Authors & Publishers (Application of Am. Online, Inc., 5 RealNetworks, Inc., and Yahoo! Inc.) (“RealNetworks and Yahoo! 6 I”), 485 F. Supp. 2d 438, 441 (S.D.N.Y. 2007). With a download, 7 the song is not audible to the user during the transfer. Id. at 8 442, 446. Only after the file has been saved on the user’s hard 9 drive can he listen to the song by playing it using a software 10 program on his local computer. Id. 11 The Internet Companies primarily generate revenue from 12 performances of musical works in two ways. On their websites, 13 they make available, at no cost to users, performances of music, 14 music videos, television programming, and the like that generate 15 revenue from advertisements on the web page or in the audio or 16 audio-visual player.3 The district court found that, in all of 17 the forms of website advertising it considered, one principle is 18 common: the larger the audience and the more times a site is 3 Advertising on websites can take numerous forms, including, inter alia, display advertising, rich-media advertising, and sponsorships. Display advertising may be displayed, inter alia, as an item on a web page, in a pop-up or pop-under window, on an interstitial page (an ad page that appears between two content pages), or in a floating window that moves across the user’s screen. Rich-media advertising, consisting of streaming audio or video, is generally played in the audio and audio-visual players in which the musical works are performed, either before, during, or after the performances of the musical works. 8 1 visited, the greater the revenue generated. See United States v. 2 Am. Soc’y of Composers, Authors & Publishers (Application of Am. 3 Online, Inc., RealNetworks, Inc., and Yahoo!) (“RealNetworks and 4 Yahoo! II”), 559 F. Supp. 2d 332, 338 (S.D.N.Y. 2008). For 5 example, advertisers typically pay for display advertising based 6 on the number of “impressions,” or views, of the advertisement by 7 users of the page on which an advertisement appears. The second 8 primary way that the Internet Companies generate revenue from 9 performing musical works is through subscription-based services. 10 II. PROCEDURAL BACKGROUND 11 Acting in its capacity as the rate court,4 the district 12 court issued rulings in April 2007, April 2008, and January 2009 13 resolving the issues presented on appeal. In its 2007 decision, 14 the district court held that a download of a digital file 15 containing a musical work does not constitute a public 4 In 1941, the United States brought a civil action against ASCAP for alleged violations of the Sherman Antitrust Act based on the fact that ASCAP’s members license their songs collectively, thereby enhancing their market power. The action was settled by the entry of a consent decree, see United States v. Am. Soc’y of Composers, Authors & Publishers, No. 13-95, 1941 U.S. Dist. LEXIS 3944 (S.D.N.Y. Mar. 4, 1941), which has subsequently been amended from time to time. The most recent version, entered on June 11, 2001, the Second Amended Final Judgment (“AFJ2”), currently regulates how ASCAP may participate in the music industry and gives the United States District Court for the Southern District of New York jurisdiction as the rate court to oversee the implementation of the AFJ2. United States v. Am. Soc’y of Composers, Authors & Publishers, No. 41-1395 (WCC), 2001 WL 1589999 (S.D.N.Y. June 11, 2001). In its capacity as the rate court, the district court determines a reasonable fee for an applicant’s ASCAP license. Id. at *6-*8 9 1 performance of that work. In its 2008 decision, the district 2 court determined a method for calculating the fees for the 3 blanket licenses payable to ASCAP for the Internet Companies’ 4 performances of musical works in the ASCAP repertory. In two 5 separate opinions issued in January 2009, the district court, 6 applying the method it determined in 2008, issued Final Fee 7 Determinations for Yahoo! and RealNetworks, respectively. 8 In its second opinion, issued in 2008, the district court 9 arrived at a license fee formula that multiplied a royalty rate 10 by the percentage of revenue attributable to the performance of 11 music. The district court applied a uniform royalty rate to the 12 Internet Companies’ varying music uses that did not fluctuate 13 over the different types of performances on the Internet 14 Companies’ sites and services. In ultimately determining a 15 royalty rate of 2.5% for both of the Internet Companies, the 16 district court relied upon several benchmark agreements, 17 including ASCAP’s agreements with Music Choice, terrestrial radio 18 stations, the broadcast television networks, and the cable 19 television networks. 20 For Yahoo!, because only a portion of the revenue generated 21 from its website is attributable to performances of musical 22 works, the district court decided to measure Yahoo!’s music-use 23 revenue by multiplying the company’s total revenue from its 24 licensed services – defined as those business units that publicly 10 1 perform music – less certain customary costs (such as for 2 advertising sales commissions and traffic acquisition expenses) 3 by a music-use-adjustment factor (“MUAF”). The MUAF was a 4 fraction that reflected the amount of time users spent streaming 5 performances of musical works relative to their overall time on 6 the website; its numerator was the number of hours of music 7 streamed from the licensed sites and services, and its 8 denominator was the number of hours that the company’s licensed 9 sites and services were utilized. 10 For RealNetworks, the district court at first accepted 11 ASCAP’s argument that it was unnecessary to apply a MUAF because, 12 unlike Yahoo!, “the vast majority of RealNetworks’s revenue 13 subject to fee is generated from subscription music services and 14 advertising-supported sites where music is the cental theme.” 15 RealNetworks and Yahoo! II, 559 F. Supp. 2d at 399; see id. at 16 411. The district court, however, did reduce RealNetworks’ 17 revenue figure by subtracting revenue attributable to 18 RealNetworks’ Technology Products and Solutions business unit, 19 which develops and markets software products and services that 20 enable wireless carriers, cable companies, and other media 21 communication companies to distribute media content to PCs, 22 mobile phones, and other non-PC devices. Id. at 359-60, 411-12. 23 In its 2009 Final Fee Determination, the district court altered 24 course and applied certain MUAFs to RealNetworks’ various sites 11 1 and services, but without explaining how it arrived at these 2 MUAFs. (RealNetworks Judgment Order 2-4) 3 DISCUSSION 4 I. Public Performance Right as Applied to Downloads 5 The Copyright Act confers upon the owner of a copyright “a 6 bundle of discrete exclusive rights,” each of which may be 7 transferred or retained separately by the copyright owner. N.Y. 8 Times Co. v. Tasini, 533 U.S. 483, 495-96 (2001) (internal 9 quotation marks omitted). Section 106 of the Copyright Act sets 10 forth these various rights, including the right “to reproduce the 11 copyrighted work in copies” and the right “to perform the 12 copyrighted work publicly.” 17 U.S.C. § 106(1), (4). In this 13 case, the Internet Companies offer their customers the ability to 14 download musical works over the Internet. It is undisputed that 15 these downloads create copies of the musical works, for which the 16 parties agree the copyright owners must be compensated. However, 17 the parties dispute whether these downloads are also public 18 performances of the musical works, for which the copyright owners 19 must separately and additionally be compensated. The district 20 court held that these downloads are not public performances, and 21 we agree.5 5 We review a district court’s grant of summary judgment based on its interpretation of the Copyright Act de novo. See Larry Spier, Inc. v. Bourne Co., 953 F.2d 774, 775 (2d Cir. 1992). 12 1 In answering the question of whether a download is a public 2 performance, we turn to Section 101 of the Copyright Act, which 3 states that “[t]o ‘perform’ a work means to recite, render, play, 4 dance, or act it, either directly or by means of any device or 5 process.”6 17 U.S.C. § 101. A download plainly is neither a 6 “dance” nor an “act.” Thus, we must determine whether a download 7 of a musical work falls within the meaning of the terms “recite,” 8 “render,” or “play.” 9 “As in all statutory construction cases, we begin with the 10 language of the statute. The first step is to determine whether 11 the language at issue has a plain and unambiguous meaning with 12 regard to the particular dispute in the case.” Barnhart v. 13 Sigmon Coal Co., 534 U.S. 438, 450 (2002) (internal quotation 14 marks omitted). “[U]nless otherwise defined, statutory words 15 will be interpreted as taking their ordinary, contemporary, 16 common meaning.” United States v. Piervinanzi, 23 F.3d 670, 677 17 (2d Cir. 1994) (internal quotation marks and alterations 18 omitted). “When the language of a statute is unambiguous, 19 judicial inquiry is complete.”7 Marvel Characters, Inc. v. 6 Section 101 of the Copyright Act, its definitional section, fully defines “[t]o ‘perform’ a work” as “to recite, render, play, dance, or act it, either directly or by means of any device or process or, in the case of a motion picture or other audiovisual work, to show its images in any sequence or to make the sounds accompanying it audible.” 17 U.S.C. § 101. 7 Because we see no ambiguity in the language of the Copyright Act, we need not reach ASCAP’s arguments regarding (i) 13 1 Simon, 310 F.3d 280, 290 (2d Cir. 2002) (internal quotation marks 2 omitted); accord Barnhart, 534 U.S. at 450. 3 The ordinary sense of the words “recite,” “render,” and 4 “play” refer to actions that can be perceived contemporaneously. 5 To “recite” is “to repeat from memory or read aloud esp[ecially] 6 before an audience,” Webster’s Third New International Dictionary 7 1895 (1981); to “render” is to “say over: recite, repeat,”8 id. 8 at 1922; and to “play” is to “perform on a musical instrument,” 9 “sound in performance,” “reproduce sound of recorded material” or 10 to “act on a stage or in some other dramatic medium,” id. at 11 1737. All three actions entail contemporaneous perceptibility. parallel provisions, (ii) legislative history, or (iii) secondary authorities. 8 The one definition that, if applicable, would support ASCAP’s position is the definition of “to render” that is “to hand over to another (as the intended recipient): deliver, transmit.” Id. We do not, however, find this definition to be applicable in the context of the Copyright Act’s definition of “to perform.” “To render” does not stand alone in the § 101 definition of “to perform”; it is contained within a list of words that, by association, give content to the term within the context of the statute. See Jarecki v. G.D. Searle & Co., 367 U.S. 303, 307 (1961) (noting that often “a word is known by the company it keeps”); see also City of New York v. Beretta U.S.A. Corp., 524 F.3d 384, 401 (2d Cir. 2008) (stating that “the meaning of one term may be determined by reference to the terms it is associated with” (internal quotation marks and alterations omitted)); Black’s Law Dictionary 1160-61 (9th ed. 2009) (defining noscitur a sociis as “[a] canon construction holding that the meaning of an unclear word or phrase should be determined by the words immediately surrounding it”). In addition to the fact that “to recite” and “to play” require contemporaneous perceptibility, the remaining terms in the § 101 definition of “to perform” - “to dance” and “to act” – are also actions that are necessarily perceptible by sight and sound. 14 1 These definitions comport with our common-sense 2 understandings of these words. Itzakh Perlman gives a “recital” 3 of Beethoven’s Violin Concerto in D Major when he performs it 4 aloud before an audience. Jimmy Hendrix memorably (or not, 5 depending on one’s sensibility) offered a “rendition” of the 6 Star-Spangled Banner at Woodstock when he performed it aloud in 7 1969. Yo-Yo Ma “plays” the Cello Suite No. 1 when he draws the 8 bow across his cello strings to audibly reproduce the notes that 9 Bach inscribed. Music is neither recited, rendered, nor played 10 when a recording (electronic or otherwise) is simply delivered to 11 a potential listener. 12 The final clause of the § 101 definition of “to perform” 13 further confirms our interpretation. It states that a 14 performance “in the case of a motion picture or other audiovisual 15 work, [is] to show [the work’s] images in any sequence or to make 16 the sounds accompanying it audible.” 17 U.S.C. § 101. The fact 17 that the statute defines performance in the audio-visual context 18 as “show[ing]” the work or making it “audible” reinforces the 19 conclusion that “to perform” a musical work entails 20 contemporaneous perceptibility. ASCAP has provided no reason, 21 and we can surmise none, why the statute would require a 22 contemporaneously perceptible event in the context of an audio- 23 visual work, but not in the context of a musical work. 15 1 The downloads at issue in this appeal are not musical 2 performances that are contemporaneously perceived by the 3 listener. They are simply transfers of electronic files 4 containing digital copies from an on-line server to a local hard 5 drive. The downloaded songs are not performed in any perceptible 6 manner during the transfers; the user must take some further 7 action to play the songs after they are downloaded. Because the 8 electronic download itself involves no recitation, rendering, or 9 playing of the musical work encoded in the digital transmission, 10 we hold that such a download is not a performance of that work, 11 as defined by § 101. 12 ASCAP, pointing to the definition of “publicly” in § 101, 13 argues that a download constitutes a public performance. Section 14 101 defines “[t]o perform or display a work ‘publicly’” as 15 follows: 16 (1) to perform or display it at a place open to the 17 public or at any place where a substantial number of 18 persons outside of a normal circle of a family and its 19 social acquaintances is gathered; or (2) to transmit or 20 otherwise communicate a performance or display of the 21 work to a place specified by clause (1) or to the 22 public, by means of any device or process, whether the 23 members of the public capable of receiving the 24 performance or display receive it in the same place or 25 in separate places and at the same time or at different 26 times. 27 Id. § 101. ASCAP argues that downloads fall under clause 28 (2) of this definition because downloads “transmit or 16 1 otherwise communicate a performance,” id., namely the 2 initial or underlying performance of the copyrighted work, 3 to the public. We find this argument unavailing. The 4 definition of “publicly” simply defines the circumstances 5 under which a performance will be considered public; it does 6 not define the meaning of “performance.” Moreover, 7 ASCAP’s proposed interpretation misreads the definition of 8 “publicly.” As we concluded in Cartoon Network LP v. CSC 9 Holdings, Inc., “when Congress speaks of transmitting a 10 performance to the public, it refers to the performance 11 created by the act of transmission,” not simply to 12 transmitting a recording of a performance. 536 F.3d 121, 13 136 (2d Cir. 2008). ASCAP’s alternative interpretation is 14 flawed because, in disaggregating the “transmission” from 15 the simultaneous “performance” and treating the transmission 16 itself as a performance, ASCAP renders superfluous the 17 subsequent “a performance . . . of the work” as the object 18 of the transmittal. See Duncan v. Walker, 533 U.S. 167, 174 19 (2001) (“It is our duty to give effect, if possible, to 20 every clause and word of a statute.” (internal quotation 21 marks omitted)). In contrast, our interpretation in Cartoon 22 Network recognizes that a “transmittal of a work” is 23 distinct from a transmittal of “a performance” – the former 24 being a transmittal of the underlying work and the latter 17 1 being a transmittal that is itself a performance of the 2 underlying work. See 536 F.3d at 134 (“The fact that the 3 statute says ‘capable of receiving the performance,’ instead 4 of ‘capable of receiving the transmission,’ underscores the 5 fact that a transmission of a performance is itself a 6 performance.”). 7 The Internet Companies’ stream transmissions, which all 8 parties agree constitute public performances, illustrate why 9 a download is not a public performance. A stream is an 10 electronic transmission that renders the musical work 11 audible as it is received by the client-computer’s temporary 12 memory. This transmission, like a television or radio 13 broadcast, is a performance because there is a playing of 14 the song that is perceived simultaneously with the 15 transmission. See, e.g., Twentieth Century Music Corp. v. 16 Aiken, 422 U.S. 151, 158 (1975). In contrast, downloads do 17 not immediately produce sound; only after a file has been 18 downloaded on a user’s hard drive can he perceive a 19 performance by playing the downloaded song.9 Unlike musical 20 works played during radio broadcasts and stream 21 transmissions, downloaded musical works are transmitted at 9 Our opinion does not foreclose the possibility, under certain circumstances not presented in this case, that a transmission could constitute both a stream and a download, each of which implicates a different right of the copyright holder. 18 1 one point in time and performed at another. Transmittal 2 without a performance does not constitute a “public 3 performance.” Cf. Columbia Pictures Indus., Inc., v. Prof’l 4 Real Estate Investors, Inc., 866 F.2d 278, 282 (9th Cir. 5 1989) (holding that renting videodiscs to a hotel guest for 6 playback in the guest’s room does not constitute the 7 “transmission” of a public performance). 8 ASCAP misreads our opinion in NFL v. PrimeTime 24 Joint 9 Venture, 211 F.3d 10, 11-13 (2d Cir. 2000), to hold that the 10 Copyright Act does not, in fact, require a contemporaneously 11 perceptible performance to infringe on the public 12 performance right. In NFL, defendant PrimeTime, a satellite 13 television provider, captured protected content in the 14 United States from the NFL, transmitted it from the United 15 States to a satellite (“the uplink”), and then transmitted 16 it from the satellite to subscribers in both the United 17 States and Canada (“the downlink”). PrimeTime had a license 18 to transmit NFL games to its subscribers in the United 19 States but not to Canada. The NFL sought to enjoin the 20 transmissions sent to Canada by arguing that the uplink in 21 the United States constituted unauthorized public 22 performances of the games in the United States. The 23 relevant issue was whether the uplink transmission was a 24 public performance even though the uplink was only to a 19 1 satellite and could not, itself, be perceived by viewers. 2 Id. at 12. We determined that PrimeTime’s uplink 3 transmission of signals captured in the United States 4 amounted to a public performance because it was an integral 5 part of the larger process by which the NFL’s protected work 6 was delivered to a public audience. Id. at 13. 7 ASCAP seizes on the fact that the uplink to the 8 satellite was not contemporaneously perceptible to argue 9 against a contemporaneous perceptibility requirement in this 10 case. ASCAP’s argument, however, fails to accord 11 controlling significance to the fact that the immediately 12 sequential downlink from the satellite to Canadian PrimeTime 13 subscribers was a public performance of the games. Id. at 14 11-13; see also David v. Showtime/The Movie Channel, Inc., 15 697 F. Supp. 752, 758-60 (S.D.N.Y. 1988) (finding that 16 because “Showtime and The Movie Channel both broadcast 17 television programming . . . to cable system operators,” 18 which, in turn, “pass[ed] the signal along to their 19 individual customers,” the initial transmissions constituted 20 public performances because they were a “step in the process 21 by which a protected work wends its way to its audience”); 22 Melville B. Nimmer & David Nimmer, 2 Nimmer on Copyright § 23 8.14[C][2] at 190.6 & n.63 (2009) (explaining that when a 24 transmission is made “to cable systems that will in turn 20 1 transmit directly to the public,” the earlier transmission 2 is a public performance despite the absence of any 3 contemporaneous perceptibility). In holding the 4 transmission in Cartoon Network not to be a public 5 performance, we distinguished NFL on the basis that in that 6 case the final act in the sequence of transmissions was a 7 public performance. See 536 F.3d at 137. That same 8 distinction applies here. Just as in Cartoon Network, the 9 Internet Companies transmit a copy of the work to the user, 10 who then plays his unique copy of the song whenever he wants 11 to hear it; because the performance is made by a unique 12 reproduction of the song that was sold to the user, the 13 ultimate performance of the song is not “to the public.” 14 See id. at 137, 138; see also United States v. Am. Soc’y of 15 Composers, Authors & Publishers (Application of Cellco 16 P’ship), 663 F. Supp. 2d 363, 371-74 (S.D.N.Y. 2009). 17 Accordingly, we affirm the district court’s grant of 18 partial summary judgment on the basis that downloads do not 19 constitute public performances of the downloaded musical 20 works.10 10 Several amici suggest that our obligations under the 1997 World Intellectual Property Organization Copyright Treaty (“WIPO Copyright Treaty”) require us to find that downloads of musical works constitute public performances. The WIPO Copyright Treaty provides authors with the following right: 21 1 II. Fee Determination for Using the ASCAP Repertory 2 We now turn to the district court’s determination of 3 the appropriate fees payable by the Internet Companies for 4 blanket licenses to publicly perform any of the millions of 5 musical compositions in the ASCAP repertory. The district 6 court determined these blanket license fees by applying a 7 uniform royalty rate of 2.5% to the Internet Companies’ 8 music-use revenue, which was calculated by multiplying the 9 total revenue from licensed services by a MUAF (music-use- 10 adjustment factor), a fraction that reflected the amount of 11 time users spent streaming performances of musical works [T]he exclusive right of authorizing any communication to the public of their works, by wire or wireless means, including the making available to the public of their works in such a way that members of the public may access these works from a place and at a time individually chosen by them. WIPO Copyright Treaty art. 8. Congress has recognized that this treaty does not “require any change in the substance of copyright rights,” see H.R. Rep. No. 105-551(I), at 9 (1998), in part because the Copyright Act already permits copyright holders to control the reproduction and distribution of their musical works over the Internet. To the extent that a download implicates these rights, the conclusion that a download does not also trigger the public performance right does not infringe on Article 8 of the WIPO Copyright Treaty. The other policy arguments raised by ASCAP and amici – regarding global harmony of doctrine, and adequate compensation – are better addressed to Congress, which has the power to amend the Copyright Act. See Eldred v. Ashcroft, 537 U.S. 186, 205-07 (2003). 22 1 relative to their overall time on the website. We conclude 2 that the district court’s analysis was flawed in two major 3 respects. 4 First, the district court did not adequately support 5 the reasonableness of its method for measuring the value of 6 the Internet Companies’ music use. Second, the district 7 court did not adequately support the reasonableness of the 8 2.5% royalty rate applied to the value of the Internet 9 Companies’ music use. Accordingly, we remand to the 10 district court so that it may redetermine reasonable fees 11 for the licenses in light of the following discussion. 12 A. Standards of Review 13 “In order to find that the rate set by the District 14 Court is reasonable, we must find both that the rate is 15 substantively reasonable (that it is not based on any 16 clearly erroneous findings of fact) and that it is 17 procedurally reasonable (that the setting of the rate, 18 including the choice and adjustment of a benchmark, is not 19 based on legal errors).” United States v. Broad. Music Inc. 20 (Application of Music Choice) (“Music Choice IV”), 426 F.3d 21 91, 96 (2d Cir. 2005). Fundamental to the concept of 22 “reasonableness” is a determination of what an applicant 23 would pay in a competitive market, taking into account the 23 1 fact that ASCAP, as a monopolist, “exercise[s] 2 disproportionate power over the market for music rights.” 3 Id. 4 “A rate court’s determination of the fair market value 5 of the music is often facilitated by the use of benchmarks – 6 agreements reached after arms’ length negotiation between 7 other similar parties in the industry.” Id. at 94. 8 Determinations by the district court that particular 9 benchmarks are comparable and particular factors are 10 relevant are questions of law reviewed de novo. Am. Soc’y 11 of Composers, Authors & Publishers v. Showtime/The Movie 12 Channel, Inc. (“Showtime”), 912 F.2d 563, 569-71 (2d Cir. 13 1990). However, factual findings as to each factor under 14 consideration or those underlying a proposed benchmark 15 agreement, as well as findings with respect to fair market 16 value, are reviewed for clear error. Music Choice IV, 426 17 F.3d at 96; Showtime, 912 F.3d at 569; United States v. Am. 18 Soc’y of Composers, Authors & Publishers (Application of 19 Capital Cities/ABC, Inc., CBS Inc., & Nat’l Broad. Co.), 157 20 F.R.D. 173, 195-96 (S.D.N.Y. 1994). 21 B. The District Court’s Determination of the Music- 22 Use-Adjustment Factor 23 Fundamental to the reasonableness of a fee for music 24 use under a license is the reasonableness of the 24 1 determination of the revenue attributable to the actual uses 2 by the applicant of the music to which the rate percentage 3 is to be applied. See United States v. Am. Soc’y of 4 Composers, Authors and Publishers (Applications of Capital 5 Cities/ABC, Inc. & CBS, Inc.) (“Capital Cities”), 831 F. 6 Supp. 137, 156-57 (S.D.N.Y. 1993); United States v. Am. 7 Soc’y of Composers, Authors and Publishers (Application of 8 Nat’l Cable Television Ass’n.), No. 41-CV-1395 (WCC) (MHD), 9 1999 WL 335376, at *12 (S.D.N.Y. May 26, 1999). In this 10 case, the value of the applicants’ uses could not be 11 premised on total revenue without an adjustment for the fact 12 that some revenues were not at all attributable to any use 13 of ASCAP music. The district court decided to make this 14 adjustment by using a MUAF that discounted the total revenue 15 to reflect only those revenues attributable to music use. 16 We have no quarrel with the use of a MUAF here, but we find 17 error in the district court’s method of determining its 18 components. 19 The district court began by calculating the total 20 revenue of the licensed sites and services – defined as 21 those business units that publicly perform music – less 22 certain customary deductions, to arrive at the revenue base 23 to which the MUAF was applied. The Internet Companies argue 24 that total revenue, when used as the revenue base, bears no 25 1 relation to the value they derive performing musical works 2 on the Internet. This argument is unavailing because it 3 overlooks the function of the MUAF, which can be understood 4 as accounting for the value of the Internet Companies’ music 5 use in relation to their overall revenue. We find nothing 6 wrong in concept with using a formula that reduces the total 7 revenue of the licensed sites and services by a factor that, 8 with substantial accuracy, accounts for music-use revenue, 26 1 as the MUAF can be understood to do here.11 With respect to 11 If designed and calibrated properly, a single formula incorporating a MUAF should produce a fee reasonably equivalent to the fee produced by a set of formulae that applies a different reasonable rate to the revenue derived from each of Yahoo!’s different music uses, as is attempted in Yahoo!’s agreement with BMI, infra. (Thus, the single formula will be sensitive to the different intensities of Yahoo!’s various music uses.) The MUAF (if suitably constructed) calculates and accounts for the contribution that music makes to the overall service that Yahoo! sells to its customers, and makes the rate formula sensitive to relative changes in that contribution over time. At the same time, the MUAF, because it is a third factor in a rate-formula structure that usually includes only two factors (revenue base and rate), complicates any benchmark analysis. The MUAF must be grouped with another of the factors to render the three-factor formula comparable to the two-factor benchmark formulae: [Grouping 1] (revenue base X MUAF) X 2.5% [Grouping 2] revenue base X (MUAF X 2.5%) In line with our discussion above, the district court at times grouped the MUAF with the revenue base factor; however, at other times, the district court grouped the MUAF with the 2.5% factor. There is no dollar difference in result, and which grouping is appropriate depends on what is being compared. For instance, if one is benchmarking a formula that multiplies licensed-services revenue by a rate (as most two- factor formulae do), one might not group the MUAF with the revenue base factor (Grouping 1 above). The benchmark rate cannot be compared to the 2.5% factor alone if the factors not being compared are not equivalent: [Benchmark] licensed-services revenue X rate [Yahoo!] (licensed-services revenue X MUAF) X 2.5% In such an instance, the problem disappears if the MUAF is grouped with the 2.5% factor (Grouping 2 above): 27 1 the components of the MUAF, however, we find error. 2 1. Yahoo! 3 The district court’s MUAF accounts for the value of 4 Yahoo!’s music use by using the amount of time that music is 5 streamed. Streaming time, however, neither drives nor 6 correlates with Yahoo!’s advertising revenue. The record 7 evidence makes plain that Yahoo!’s advertising revenue model 8 more accurately correlates with the number of times a 9 particular page is accessed by users than to the duration of 10 streaming time. To the extent that the district court’s 11 MUAF relied on an imprecise metric for determining 12 advertising revenue attributable to music use when a 13 superior metric was apparently available and practicable,12 [Yahoo!] licensed-services revenue X (MUAF X 2.5%) With this grouping, MUAF X 2.5% is the rate to which the benchmark rate is compared. As mentioned, the alternative groupings are mathematically equal. The reasonableness of the Yahoo! formula can be established by comparing benchmarks to either grouping. 12 One reason a district court may use a less precise metric is because it is impracticable to use a more precise one, for example if relevant statistics are unobtainable or unreliable. See Capital Cities, 831 F. Supp. at 156-57 (using the amount of music use in television programming as an “adequate proxy” for the value of the music to the network, “in the absence of any other yardstick”). However, the district court did not make a sufficient showing that this was the case. 28 1 the district court’s method for calculating the MUAF was 2 unreasonable.13 E.g., Capital Cities, 831 F. Supp. at 156- 3 57 (accepting amount of music used in television program as 4 an adequate measure of the value of the music to the program 5 because it was the only measure available). 6 Display advertising on the Internet is sold on a 7 cost-per-thousand model that counts the total number of page 8 impressions, i.e., how many times a particular page is 9 accessed. Pages that are accessed a greater number of times 10 occasion higher advertising rates because the advertisements 11 on these pages are viewed with greater frequency. It is, 12 thus, unreasonable to use streaming time, which has no 13 necessary correlation with page views, as a proxy for the 14 number of times a page is viewed; time spent on-line is not 15 reflective of how a user interacts with a particular page. 16 A user may have a page open that he is not viewing at 17 all, either because he has multiple pages open and is 13 The district court, recognizing the imprecision of its metric, offered, inter alia, the following rationale for using it: “Although the streaming time is increased by visitors who stream music as a ‘background’ while they are engaged in other activities on the website, such as searches or e-mailing, that effect is largely if not wholly offset by the myriad incidental performances of music in movies, advertisements, user-uploads and elsewhere, which are not counted as music streaming time (against Yahoo!).” RealNetworks & Yahoo! II, 559 F. Supp. 2d at 413. We find this sort of rough estimation, with no basis in the record, unreasonable because there appears to be a much more precise metric available to the district court. 29 1 interacting with another page, website, or computer program, 2 or because he has walked away from the computer altogether 3 but has left that page open. For example, user A, who is 4 listening to a four-minute song, may view only the page on 5 which the song is playing in that four minutes because his 6 exclusive focus is on the song, while user B, who is 7 listening to the same song on as background music, may be 8 simultaneously clicking on links and reading articles 9 throughout Yahoo!’s website, and thus may be seeing multiple 10 advertisements on multiple pages during that same four- 11 minute period. The streaming time for users A and B is the 12 same, but the advertising exposure of each differs widely. 13 The advertising marketplace takes account of the 14 foregoing consideration by applying different advertising 15 rates to different areas of the website. For example, 16 advertising for radio-style webcasts in Yahoo! Music is 17 priced at a rate lower than similarly placed ads on Yahoo!’s 18 homepage, because users normally have the Yahoo! Music 19 window minimized (and thus not viewable) for much of the 20 time the radio-style music is playing, in contrast to other 21 areas on Yahoo!, like the homepage, that command greater 22 viewer attention.14 The district court erred by 14 We are not suggesting that the amount of time music is played has no correlation to the revenue attributable to music on Yahoo!’s licensed sites and services. Streaming time likely 30 1 constructing a MUAF that failed to take into account these 2 various realities of Internet advertising.15 3 In sum, the district court erred by adopting an 4 imprecise metric - music streaming time rather than page 5 views - as the basis of its MUAF, without providing a 6 sufficient rationale for that decision. We will not specify 7 a particular method of developing a formula for determining 8 music-use revenue on remand; we also leave it to the 9 district court to determine whether it should proceed with a 10 variant of its current formula (revenue multiplied by a bears some relation to the importance of music on Yahoo!. Furthermore, we recognize that music can enhance a user’s experience on Yahoo! even when he navigates away from the streaming page to another Yahoo! webpage. For example, music may be driving advertising revenue on the non-music page to the extent that the music is making Yahoo! experience, as a whole, more appealing. The district court may take this into account in the formula it adopts on remand. We find the district court’s reasoning unreasonable not because streaming time bears no relation to the value of music to the revenue of the licenses sites and services, but because there is a more accurate metric available - page impressions - that the district court chose not to use without providing a sufficient reason. 15 We recognize that revenue is generated from music use by other methods as well, such as through site sponsorship or sponsored search results. However, the district court gave no rationale supporting the conclusion that streaming time is a better measure of the value created by these forms of advertising than it is for display advertising, nor did it indicate that these alternative forms of advertising were a significant part of its analysis. On remand, the district court may wish to address how its metric for calculating music-use revenue will interact with these alternative forms of advertising and whether they are significant in terms of its ultimate fee determination. 31 1 MUAF) or in some other way altogether, in light of the 2 foregoing discussion. 3 Yahoo! has also faulted the district court for using 4 statistics with differing methodologies in the MUAF’s 5 numerator and denominator. In calculating the numerator, 6 the district court used Yahoo!’s statistics for the number 7 of hours of music streamed, which give no effect to the 8 specific window engaged by the user. In calculating the 9 denominator, the district court used statistics based on a 10 different methodology provided by comScore Media Metrix for 11 the total hours of use of the licensed sites and services.16 12 Unlike Yahoo!, comScore measures hours used only for the 13 specific window that is engaged. The difference is that, in 14 an instance in which a user has multiple windows open at one 15 time, comScore will count the time for only the single 16 window that is in active use, while Yahoo! will count the 17 time for all the windows open. Yahoo!’s statistics thus 18 reflect considerably greater use time than comScore’s 19 because, anytime a user has multiple windows open, Yahoo! is 20 counting the time use for each of the Yahoo! windows open 21 while comScore is counting the time use for only the single 16 ComScore is an Internet audience measurement firm that measures traffic to, and time spent on, Internet sites and services. 32 1 window engaged by the user. Without addressing whether one 2 method is preferable to the other, we conclude that 3 constructing a MUAF by using Yahoo!’s statistics for the 4 numerator and comScore’s statistics for the denominator is 5 unreasonable because these statistics are not comparable, 6 with the result that their comparison overstates music- 7 streaming time. 8 The district court’s opinion states that, as of the 9 date of its issuance, Yahoo! had failed to supply any site- 10 hours data comparable to that supplied by comScore that 11 could have been used for the denominator, but that Yahoo! 12 was free to do so before the district court ordered its 13 Final Fee Determination. The parties disagree over whether 14 Yahoo! ultimately furnished adequate total site-hours data. 15 Because we remand for reconsideration of the MUAF, we refer 16 this dispute to the district court. We do, however, note 17 that in calculating any MUAF, the district court must strive 18 to use measurements that are as consistent and as precise as 19 practicable. 20 2. RealNetworks 21 Turning to RealNetworks, as noted previously the 22 district court in 2009 determined and applied differing 23 MUAFs to RealNetworks’ various licensed sites and services 33 1 despite its decision not to apply MUAFs in 2008, without any 2 explanation for the basis of these MUAFs. Accordingly, we 3 remand for explanation (or reconsideration if the current 4 MUAFs cannot be justified) because the district court’s 5 rationale was insufficient. 6 The district court applied the following MUAFs to 7 RealNetworks. For its Rhapsody subscription service,17 the 8 district court defined the MUAF’s numerator as total number 9 of plays of audio-music and music-video streams, and the 10 denominator as the total number of Rhapsody streams plus the 11 number of deliveries of conditionally downloaded music files 12 to subscribers. (RealNetworks Judgment Order 2-3) For its 13 SuperPass subscription service,18 the numerator was fixed as 17 The district court found that Rhapsody is an unlimited on-demand music subscription service that offers subscribers access to over four million songs. In addition to streaming music and selling permanent downloads of music, Rhapsody subscribers can also conditionally download music. A conditional download is a download that may be accessed by the user for a limited duration or number of uses. For a monthly fee, a Rhapsody subscriber can play as much or as little music as he wants. Regardless of the actual amount of music played, however, the Rhapsody subscriber must still pay the full subscription fee. If the subscriber continues to pay the subscription fee, then neither the amount nor the type of music actually played by the subscriber affects the amount of revenue received by RealNetworks. 18 The district court found that SuperPass is a subscription service that offers, for a monthly fee, access to news, sports, movies, games, music, and other entertainment content; short films, video clips, and music; music downloads (at $0.99 per song) and streaming previews of music; access to the 34 1 the total number of hours of streams to users in each month 2 by means of the music-radio portion of SuperPass, and the 3 denominator as the total number of SuperPass music hours 4 plus the total number of hours of all other streams to users 5 by means of SuperPass. (RealNetworks Judgment Order 3) And 6 for RealNetworks’ Music, and Media Software and Services 7 groups, the district court used as the numerator the total 8 number of hours of streams that non-subscription, on-demand, 9 and radio-music users receive from RealNetworks’ licensed 10 sites, and as the denominator the total number of hours 11 users spend on RealNetworks’ licensed sites. (RealNetworks 12 Judgment Order 4) 13 Because the district court failed to explain the basis 14 for these MUAFs, and in light of the issues we raised with 15 respect to the MUAF applied to Yahoo!, we remand for the 16 district court to explain or reconsider the MUAFs applied to 17 RealNetworks. In addition to consideration of any issues 18 that it deems appropriate, the district court should address 19 the following: (1) whether its method for calculating the 20 MUAF for the Rhapsody subscription service is more precise majority of RadioPass services; and CD burning and other features for the RealPlayer. As with Rhapsody, SuperPass subscribers pay the full monthly fee regardless of the amount of content they access, and the amount of subscription revenue that RealNetworks receives does not depend on the subscriber’s behavior or actual usage of the subscription’s offerings. 35 1 or practicable than the method used in the benchmark 2 agreements in the record; (2) whether there is a more 3 precise way, that is also practicable, to account for the 4 value of the music use for the SuperPass subscription 5 service in light of the fact that some components of the 6 subscription do not involve the streaming of content to 7 users; and (3) whether there is a more precise and still 8 practicable way to measure RealNetworks’ advertising 9 revenue, in light of the issues we raised in our discussion 10 of Yahoo!’s MUAF. 11 C. The Royalty Rate Applied by the District Court 12 1. Yahoo! 13 The district court arrived at Yahoo!’s royalty rate by 14 relying on benchmark agreements for blanket licenses that 15 ASCAP entered into with Music Choice, terrestrial radio 16 stations, the broadcast television networks, and cable 17 television providers, as well as the rates that Yahoo! 18 itself pays to the major record companies for music-video 19 rights. The district court’s factual findings that support 20 selecting these benchmarks were not clearly erroneous. 21 After reviewing the district court’s analysis of these 22 benchmarks in relation to this case, however, we hold that 23 the district court unreasonably arrived at its decision to 36 1 apply a uniform 2.5% royalty rate and that, in setting the 2 royalty rate, the district court must follow an approach 3 more tailored to the varying nature and scope of Yahoo!’s 4 music use. 5 Beginning with the Music Choice benchmark, the district 6 court found that Music Choice provides channels of music to 7 listeners on a subscription basis via cable and satellite 8 television and the Internet. Music Choice’s channels are 9 organized around genres of music, and, on the Internet, 10 listeners have the option to create up to ten personalized 11 audio channels. ASCAP’s current blanket license with Music 12 Choice, for the period January 1, 2006 through December 31, 13 2010, calls for payment of a royalty rate that is 2.5% of 14 Music Choice’s gross revenues, where gross revenue is 15 defined as all revenues derived from the licensed services. 16 We conclude that the royalty rate agreed to by Music 17 Choice provides strong support for applying a 2.5% royalty 18 rate to those Yahoo! sites and services that provide access 19 to music channels organized around music genre, similar to 20 those on Music Choice. Additionally, it provides a basis 21 for a 2.5% royalty rate, or higher, for Yahoo! sites and 22 services that permit an interactive music experience, in 23 which the user may control the selection of music he is 37 1 hearing, for example if a user tunes into a more customized 2 station or uses Yahoo! Search to listen to songs on-demand. 3 See RealNetworks and Yahoo! II, 559 F. Supp. 2d at 413. 4 However, the Music Choice benchmark does not justify 5 applying a 2.5% royalty rate to all of Yahoo!’s music uses, 6 because Yahoo! offers numerous sites and services that are 7 less music intensive than Music Choice’s offerings. The 8 district court finessed the Music Choice benchmark’s limited 9 relevance by concluding that there are other benchmarks in 10 the record that support applying a 2.5% royalty rate to all 11 of Yahoo!’s performances of the ASCAP repertory. Because we 12 conclude that these other benchmarks do not adequately 13 support an across-the-board 2.5% royalty rate, we remand for 14 reconsideration of a reasonable royalty rate. 15 Turning to the major broadcast television networks’ 16 agreements with ASCAP, the district court found that these 17 agreements differed from Yahoo!’s because the major networks 18 pay a flat rate for their ASCAP usage without regard to 19 revenue (as specified in Yahoo!’s application) and that “[a 20 similar] flat-fee structure is unsuitable for the online 21 music industry,” RealNetworks and Yahoo! II, 559 F. Supp. 2d 22 at 401. The district court nevertheless concluded that, by 23 comparing the percentages of broadcast revenue that the 38 1 television companies pay ASCAP under their flat royalty rate 2 to the percentage of licensed services revenue that Yahoo! 3 would pay if a 2.5% royalty rate (as well as a MUAF) were 4 applied to its license, the networks’ agreements could be 5 “useful to gauge the reasonableness of the fee range ASCAP 6 seeks from . . . Yahoo!, and [Yahoo!’s] ability to pay the 7 blanket fees rather than resorting to less efficient 8 licensing options or foregoing the use of ASCAP music 9 altogether.” Id. 10 The district court found that “[t]he three television 11 networks, ABC, CBS, and NBC, have annual revenues in the 12 range of $3 to $4 billion[,] comparable to current . . . 13 Yahoo! revenues” and that the networks pay a percentage of 14 broadcast revenue for their ASCAP license that is comparable 15 to the amount Yahoo! would have paid if a 2.5% royalty rate 16 (as well as a MUAF) was applied to the total domestic 17 revenue from the licensed services for the same period. 18 From this finding, the district concluded that, to the 19 extent that music is not much less important to Yahoo! than 20 it is to the television networks, a 2.5% royalty rate is 21 reasonable. 22 The district court similarly looked to percentage of 23 total revenue when assessing ASCAP’s agreements with the 39 1 general-entertainment and music-intensive cable television 2 networks. The district court found that the general- 3 entertainment cable television networks pay 0.375% of total 4 revenue derived from licensed services and that the music- 5 intensive cable television networks pay 0.9% of licensed- 6 services revenue derived from licensed services. As it did 7 with the major broadcast television networks, the district 8 court compared the percentage of total revenue that Yahoo! 9 would pay under the court’s formula to the 0.375% and 0.9% 10 paid by the cable television networks to conclude that a 11 flat 2.5% royalty rate, when multiplied by a MUAF, is 12 reasonable for Yahoo!’s license. 13 We are unconvinced that percentage of revenue 14 comparisons between broadcast and cable television networks, 15 and Yahoo! are useful in determining a reasonable fee for 16 Yahoo!’s public performances of the ASCAP repertory. Nearly 17 every program on a television station somehow utilizes 18 musical works. In contrast, only a fraction of the traffic 19 on Yahoo!’s website uses music; much of Yahoo!’s website 20 does not implicate any music use whatsoever. Given that 21 Yahoo!’s revenue base relies far less on ASCAP content than 22 the television networks’ revenue base, we believe that 40 1 comparing percentages of overall revenue bases is of little 2 probative value in this benchmark analysis.19 3 These comparisons, moreover, indicate a deeper flaw in 4 the district court’s analysis: its inclination to lump all 5 of Yahoo!’s varying musical uses together, instead of 6 looking to the nature and scope of Yahoo!’s different types 7 of uses and applying a rate that reflects (or rates that 8 reflect) the varying nature of Yahoo!’s music use. See 9 United States v. Am. Soc’y of Composers, Authors & 10 Publishers (Application of Nat’l Cable Television Ass’n.), 11 No. 41-CV-1395 (WCC) (MHD), 1999 WL 335376, at *12 (S.D.N.Y. 12 May 26, 1999). The district court’s finding that cable 13 television networks pay 0.375% of their revenue for their 14 ASCAP licences is a good example. The most direct 15 conclusion to be drawn from this benchmark is that providers 16 of cable-style television will pay 0.375% of their revenue, 17 in a competitive market, to license ASCAP music.20 However, 18 the district court looked past this conclusion in setting 19 the royalty rate and instead used this benchmark to justify 19 The district court’s analysis of the terrestrial radio stations’s licencing agreements with ASCAP suffers from the same flaw. 20 This conclusion is supported by the fact that the major broadcast television networks, which perform a blend of cable- television style programming, as well as less music-intensive programming such as news and sports, pay between 0.24%-0.34% for their ASCAP licenses. 41 1 applying a 2.5% rate to all of Yahoo!’s music use based on 2 an imprecise comparison of the percentages of overall 3 revenue. We think a better approach would be to attempt, if 4 practicable, to set a royalty rate that requires Yahoo! to 5 pay a rate for its cable-style programming that is similar 6 to that in the cable market.21 7 Our view that a reasonable royalty rate should reflect 8 the varying values of Yahoo!’s differing music uses is 9 supported by Yahoo!’s license with BMI.22 As previously 10 noted, BMI is ASCAP’s principal competitor in licensing the 11 performance rights for musical works. The two rights 12 organizations control approximately 90% of all on-line music 13 performances, with roughly equal shares of the market. 14 Because these two companies operate in the market in such 21 The district court’s analysis of Yahoo!’s agreement with the major record companies is erroneous for the same reason. The court did not use this benchmark to assess how much Yahoo! should pay for its right to perform ASCAP musical works in music videos. It instead used this benchmark to make a far more general conclusion concerning how much Yahoo! should pay for all of its various music uses. 22 The district court concluded that the royalty rates used in the BMI-Yahoo! license are not probative of what the market would yield for Yahoo!’s license in this case because, it concluded, the BMI-Yahoo! license is a “per-segment” license that confers a different set of rights than the “blanket” license Yahoo! seeks from ASCAP. Yahoo! contests this finding. In light of our remand, it is not necessary to rule on this issue. We leave it open to the district court, however, to revisit this finding on remand. 42 1 similar manners, BMI’s agreements are instructive. See 2 e.g., United States v. Am. Soc’y of Composers, Authors & 3 Publishers (Application of MobiTV, Inc.), Nos. 09-Civ-7071 4 (DLC), 41-Civ-1395 (DLC), 2010 WL 1875706, at *38 (S.D.N.Y. 5 May 11, 2010). 6 In its negotiated agreement with BMI, Yahoo! agrees to 7 pay the following license fees: (a) 1.75% of “Direct 8 Revenue” from “Preprogrammed Radio”; (b) 2.5% of “Direct 9 Revenue” from “On-Demand Streams”; (c) 2.15% of “Direct 10 Revenue” from “User-Influenced Programming”; (d) 1.0% of 11 “Direct Revenue” from “Audio-Visual Programming”; and (e) 12 0.6875% of “Yahoo! Music Advertising Revenue and Yahoo! 13 Music Run of Site Allocation.” This agreement, providing 14 for a so-called “bucket” for each different use, supports 15 the conclusion that the other benchmarks also suggest: the 16 market assigns different values to Yahoo!’s different music 17 uses and is capable of yielding a royalty rate that reflects 18 the varying intensity of Yahoo!’s music uses.23 23 ASCAP’s licensing agreement with Turner Broadcasting further supports the conclusion that the market is capable of yielding a royalty rate that reflects the varying intensity of Yahoo!’s music uses. Turner owns numerous types of television stations that use music differently from one another – including, by way of example, general entertainment cable television stations such as TNT and TBS, and a cable news station, CNN – all of which are covered under a single license with ASCAP. Similar to Yahoo!’s agreement with BMI, Turner’s license with ASCAP includes different rates on a per-network basis, thereby reflecting the economic reality that Turner’s different channels 43 1 The district court found that effectuating the BMI 2 structure is quite complex, because Yahoo! is required to 3 subdivide its sites and services into 17 separate revenue 4 categories that are apportioned into “5 buckets” and that, 5 to report the revenue under these categories, Yahoo! is 6 required to make dozens of calculations and collect numerous 7 data points. “Because most of these data [points] are not 8 ordinarily collected for other business purposes,” the 9 district court concluded that “their accuracy [i]s suspect 10 and auditing [Yahoo!’s] reports [i]s difficult and 11 expensive.” The district court found that the BMI method is 12 additionally complicated by how the parties decide what 13 revenue is “directly attributable” to the streaming of 14 music. RealNetworks and Yahoo! II, 559 F. Supp. 2d at 411. 15 We neither endorse nor challenge these conclusions, and we 16 leave it to the district court on remand to decide whether a 17 bucket system is feasible with alterations, or whether 18 another system is preferable in light of our guidance today. 19 We further acknowledge the requirement that fee 20 structures and the proceedings used to arrive at them 21 comport with the provisions of the Second Amended Final 22 Judgment, see, e.g., United States v. BMI (Application of use ASCAP’s repertory in manners that the market values differently. 44 1 Muzak LLC), 275 F.3d 168, 176-77 (2d Cir. 2001) (assessing 2 whether a rate structure that included "carve outs" for 3 songs licensed by the Applicant directly from the copyright 4 holder was within the rate-setting court's blanket-license 5 authority), and the district court’s concern that the 6 setting of different rates for the Internet Companies’ 7 various services would be in some tension with prior case 8 law regarding blanket licenses, see RealNetworks & Yahoo! 9 II, 559 F. Supp. 2d at 406-08. We note, however, that the 10 district court’s own proposed MUAF would have involved 11 “reporting or . . . keeping track of . . . music use,” 12 notwithstanding the court’s observation that such tracking 13 is generally not a feature of blanket licenses, see id. at 14 407, and that a recent rate-setting opinion has made use of 15 multiple rates as part of a blanket license, see MobiTV, 16 2010 WL 1875706, at *22, *37-38, *43. Moreover, our own 17 precedent indicates that non-traditional fee structures can 18 be compatible with blanket licenses. Muzak, 275 F.3d at 19 177. It is for the district court to consider, in the first 20 instance, what options are available to it in setting a 21 reasonable rate or rates in this case. 22 For the foregoing reasons, we remand to the district 23 court to redetermine the royalty rate (or rates) that Yahoo! 24 must pay ASCAP for its license, in light of our holding that 45 1 the district court’s valuation of Yahoo!’s use of the ASCAP 2 repertory must reflect, as well as practicable, the varying 3 nature and scope of Yahoo!’s music use. We do not, however, 4 suggest that the specific royalty rates set forth in the 5 BMI-Yahoo! agreement must be accepted by the district court 6 on remand, nor do we suggest that a bucket system is the 7 only method by which a reasonable fee could be calculated. 8 Instead, we believe that there are other ways to proceed, 9 including possibly using a “blended” uniform rate24 (e.g., 10 taking a snapshot of Yahoo!’s current music use, valuing the 11 different uses independently, averaging them into a blended, 12 uniform royalty rate, and then revisiting that rate 13 periodically), or perhaps using a modification of the BMI 14 bucket system to avoid some of the reliability problems 15 noted by the district court, or employing some third method 16 not yet discussed. We leave it to the district court to 17 determine the best way to proceed consistent with the 18 concerns we have discussed. 19 2. RealNetworks 24 We also do not mandate that, if the district court were to use a blended uniform rate, the rate would have to be lower than 2.5%. We only hold that the district court’s application of the benchmark agreements in its opinion does not support an across the board 2.5% rate. 46 1 The district court applied the Music Choice and radio 2 station benchmarks to RealNetworks. The court found that 3 Music Choice’s 2.5% royalty rate supported applying a 2.5% 4 royalty rate to RealNetworks because, inter alia, many of 5 RealNetworks’ music uses are equally or more music intensive 6 than Music Choice’s uses, specifically referring to 7 RealNetworks’ on-demand, music-video, and music-stream uses. 8 The district court also found that the radio station 9 benchmarks supported a 2.5% royalty rate for RealNetworks. 10 We do not take specific issue with the district court’s 11 analysis as it may pertain to uses by RealNetworks that are 12 analogous to those of Music Choice and the radio stations. 13 However, RealNetworks objects to the district court’s 14 analysis on the basis that RealNetworks’ services include 15 music uses that are less music-oriented than those of Music 16 Choice or the radio stations, such as uses in video games, 17 television shows, and ring tones. We find this objection 18 persuasive to the extent that it tracks the flaws we have 19 identified in the district court’s analysis underlying 20 Yahoo!’s royalty rate, and we remand so that the district 21 court may conduct a more complete analysis of the various 22 uses of ASCAP musical works by RealNetworks and may 23 determine, in light thereof, the appropriate method for 24 determining RealNetworks’ royalty rate (or rates). 47 1 D. Download Revenues 2 With respect to the district court’s fee calculation, 3 ASCAP cross-appeals one issue: whether download revenue 4 should be included in the revenue base, even though 5 downloads are not public performances, because public 6 performances of music are used to generate the downloads of 7 music. Although the district court recognized some 8 relationship between the ability to stream music and 9 download revenue, it made no formal factual findings about 10 the extent and implications of that relationship. In an 11 analogous context, the district court found that the ability 12 to preview ringtones (via streaming) “undeniably 13 increase[d]” the sale of ringtones (via download). In re 14 United States v. Am. Soc’y of Composers, Authors & 15 Publishers (Application of AT&T Wireless f/k/a Cingular 16 Wireless), 607 F. Supp. 2d 562, 565 (S.D.N.Y. 2009). We 17 leave it to the district court on remand to determine the 18 extent of any such relationship in this case. 19 Finally, ASCAP cross-appeals the district court’s 20 admission of the Internet Companies’ alleged late evidence. 21 This issue is moot in light of our remand to the district 22 court for further proceedings, at which additional evidence 23 may be considered. 48 1 CONCLUSION 2 For the foregoing reasons, we AFFIRM the district 3 court’s ruling that downloads of musical works do not 4 constitute public performances of those works, and we VACATE 5 the district court’s assessment of reasonable fees for the 6 blanket ASCAP licenses sought by the Internet Companies and 7 REMAND for further proceedings in light of this opinion. 49