Toward an Efficient Licensing and Rate-Setting Regime: Reconstructing § 114(i) of the Copyright Act
Why is Sony/ATV Music Publishing, the world’s largest music publisher,1 unhappy about its massive hit single “Happy”?2 According to CEO and Chairman Martin Bandier, the answer comes down to the math behind digital streaming revenues. In the first three months of 2014, the Internet radio titan Pandora streamed “Happy” more than forty-three million times,3 but this translated into only $2,700 in publisher and songwriter royalties.4 In contrast, Pandora paid twenty times that rate—$56,000—in sound-recording royalties.5
Happily for the songwriter, Pharrell Williams, he is also the recording artist on the smash-hit track and is thus entitled to a share of both types of royalties. Generally, however, songwriters and their publishers receive royalties under a rate-setting regime distinct from that governing artists and their recording labels. The resulting headline-grabbing disparity between the two types of royalties has become a source of major contention within the music industry.6
At the heart of the battle over fee parity is § 114(i) of the Copyright Act.7 This statutory provision prohibits the rate-setting court charged with determining a reasonable royalty rate for musical works (paid to songwriters and their publishers) from taking into consideration the much higher sound-recording royalties (paid to labels and recording artists) set by the Copyright Royalty Board (CRB). Although it is well understood that publishers favored § 114(i) before they opposed it, the rationale behind that initial support has been misconstrued. Most notably, when setting the musical work royalty rate in a closely watched 2014 decision, the designated rate-setting court declared that publishers pushed for the provision two decades ago because they were concerned that sound recording performance rates would be set lower than musical work performance rates—and thus drag musical work performance rates down.8
This Comment examines historical context to correct the record and to renew the call for reducing some of the inefficiencies deliberately built into our disjointed music-licensing and rate-setting regime. Part I explains that publishers supported § 114(i) twenty years ago for much the same reason they opposed the creation of sound recording performance rights fifty years ago: they feared that the newly created sound-recording royalties would cut into publishers’ existing royalty pie. Now that sound recording royalties far surpass musical-work royalties, the appeal of § 114(i) has flipped: publishers and songwriters support modifying or eliminating the statute on the theory that a court permitted to use sound-recording rates as a benchmark might be persuaded to raise composition rates. In contrast, music users who pay licensing fees support retaining § 114(i) because they fear the rise of musical-work royalties.
So told, the story of § 114(i) is significant in two respects. As Part II details, the provision’s simple but overlooked history helpfully illustrates why digital-streaming services and songwriters and publishers should support eliminating the provision. Although the industry currently sees § 114(i) as an obstacle to fee parity and thus beneficial to the music services interested in keeping musical-work royalties down, the provision is inherently problematic for music services in the long run. Part III then draws on the larger significance of § 114(i)’s passage and retention. Enacted out of the fear that newly created sound-recording royalties would cut into existing publisher and songwriter royalties, § 114(i) was a reactionary attempt to avoid a holistic revenue system that would take into account both types of royalties. Recognizing and rejecting the shortsighted motivations behind the provision’s enactment is an important step toward a more consolidated and more efficient licensing and rate-setting regime.
A digital music streaming service—be it Pandora, Apple Music, Spotify, or Music Choice—must obtain two kinds of copyright licenses for the “public performance” of any song.9 The musical-work copyright, owned by songwriters and their publishers, covers the song’s composition and its accompanying lyrics.10 The sound-recording copyright, owned by artists and their record labels, covers the performing artist’s recorded rendition of the composition.11
The story of how the second of these public-performance rights was belatedly born is key to understanding the purpose—and flaws—of § 114(i) and why music services should join the publishing world in advocating for its elimination.
Musical compositions became copyrightable in 1831,12and in 1897, Congress granted musical composition copyright owners the right to control the public performance of their copyrighted works.13 Because of the tremendous transaction costs that individual copyright owners would incur to personally enforce their rights, the member-owned performing rights organization (PRO) has long collected and distributed license fees on behalf of its members.14 The two largest and oldest PROs, the American Society of Composers, Authors, and Publishers (ASCAP) and Broadcast Music, Inc. (BMI),15 must offer a blanket license to all music users under the terms of two separate decades-old consent decrees with the Department of Justice.16 These blanket licenses grant licensees the nonexclusive right to perform a PRO’s entire repertory of works in exchange for a flat fee or a portion of the licensees’ revenues17 and require a designated rate-setting court to set a reasonable license fee should the parties fail to timely negotiate a rate.18
But sound recordings did not receive federal-copyright recognition until 1971—more than a century after musical compositions.19 Even then, the copyright was limited to reproduction and distribution rights and did not include an enforceable public-performance right. That changed in 1995, when Congress passed the Digital Performance Right in Sound Recordings Act (DPRA)20 in response to the emergence of music streaming, which threatened to displace sales of physical records and leave artists and their labels uncompensated for the widespread enjoyment of their work.21
The DPRA created a limited digital performance right for sound recordings by way of a complex licensing scheme.22 Under the statute, noninteractive Internet radio services like Pandora are eligible for compulsory licenses,23 which are administered today by the CRB using a “willing buyer, willing seller” standard meant to protect the licensee from above-market royalties.24 The CRB was created only after successive failures to establish a royalty regime that would not crush small webcasters under the combined weight of sound recording and composition-licensing fees.25 However, the CRB too has touched off its share of royalty controversy—a fact that demonstrates the problems associated with webcasters’ multipart and disjointed licensing obligations. For example, on June 26, 2007, Internet radio went dead as dozens of radio stations observed a “day of silence” to protest a CRB ruling mandating a major webcasting royalty rate increase.26
One notable characteristic of the CRB’s current rate-setting regime is that the CRB is permitted to consider the rates paid to publishers and songwriters when it determines the rates that should be paid to recording labels and artists. Indeed, the CRB has deliberately set compulsory-license fees for sound recordings at rates multiple times higher than the prevailing rates for composition-performance licenses, on the grounds that the markets for these two types of licenses are materially distinguishable.27 In contrast, thanks to § 114(i), the rate-setting courts are prohibited from considering sound-recording royalties when setting musical-work royalties.28 Although publishers and songwriters revile that legal prohibition today, as discussed below, half a century ago they were its primary proponents.
In the 1960s, some thirty years before Congress created a limited-performance right for sound recordings, broadcasters and music publishers formed an “unlikely alliance”: together they successfully opposed the creation of a general public-performance right in sound recordings, because they feared the effect on their revenue streams.29 Broadcast stations balked at the idea of paying performance royalties to record companies and recording artists on top of the performance royalties they already paid to music publishers and songwriters.30 Meanwhile, the PROs feared the consequences of a limited-royalty pie—that is, the PROs believed that broadcasters forced to pay royalties to record companies for sound recordings would have less money to pay publishers and songwriters for the underlying compositions.31
This fear of a finite overall “royalty pie” not only formed the basis for the PROs’ opposition to the creation of public-performance rights in sound recordings32 but also laid the groundwork for the creation of § 114(i) decades later. Indeed, the idea resurfaced in 1995 when bills proposing a limited performance right for certain digital transmissions of sound recordings were pending in the House and the Senate.33 In a House hearing that year, Marybeth Peters, Register of Copyrights and Associate Librarian for Copyright Services, explained in no uncertain terms that the inclusion of language that would eventually become § 114(i) in early versions of sound-recording legislation was the Senate’s attempt to address the concerns of music-composition owners: “These concerns related to what is sometimes referred to as the ‘pie’ theory: users might seek to reduce music performance fees to composers, songwriters and publishers because a new category of authors would be entitled to claim royalties from sound recording performance.”34
By this time, although ASCAP expressed its “support [for] the concept of a performance right in sound recordings,” it remained concerned that record companies would benefit from the new rights “on the backs of the songwriters and music publishers who created and own the underlying songs without which the sound recordings would not exist at all.”35 Specifically, ASCAP feared that the new sound-recording royalties would be subtracted from existing composition fees. Along with BMI,36 ASCAP thus urged the inclusion of statutory language that would “make it absolutely clear that the new rights being granted did not in any way derogate from existing performing rights in underlying musical compositions, and could not be used as a basis for diminishing the royalties paid for the performance of underlying musical works.”37
ASCAP and BMI prevailed: the 1995 DPRA amended § 114 of the Copyright Act to include subsection (i).38 The provision provides that sound recording license fees “shall not be taken into account in any administrative, judicial, or other governmental proceeding to set or adjust the royalties payable to copyright owners of musical works for the public performance of their works.”39 The legislative history of the provision clearly reflects the concerns associated with the “royalty pie” theory: a congressional report notes that § 114(i) was designed to ensure that “license fees for music performance shall not be reduced by reason of obligations to pay royalties under this bill.”40 Section 114(i) thus contrives a two-way mirror between sound-recording and musical-work rates, wherein composition rates can be considered in setting sound-recording rates but not vice versa.41
II. the problem with eliminating sound-recording fees from percentage-of-revenue rate-setting analysis
The “royalty-pie” rationale that drove publishers’ initial support for § 114(i) reveals that the provision was designed specifically to ensure sound royalties would be stacked on top of composition royalties. This design reflected no regard for the revenue constraints of music users or the highly interconnected ecology of the music-licensing regime. But as the current debate over § 114(i) reveals, even today, the fact that the provision was enacted at the expense of webcasters like Pandora, and could still work to their detriment in the long run, is not well understood.
The comments submitted by industry entities in response to the Copyright Office’s recent inquiry into the royalty rate-setting process42 demonstrate that even industry actors tend to understand the function of § 114(i) primarily in terms of the provision’s current role in preventing rate-setting courts from pursuing fee parity. Thus, the publishing world advocates modifying or eliminating § 114(i) to help secure higher musical-work rates.43 Most notably, this advocacy has included legislative efforts such as the proposed Songwriter Equity Act of 2014.44 Meanwhile, broadcasters and streaming services have pushed back with arguments for retaining § 114(i) to keep the musical-work royalties they pay from rising to the level of sound-recording royalties.45
Industry actors are not alone in reading § 114(i) as a prohibition primarily designed to prevent sound-recording rates from being used as a benchmark for musical-work rates. In March 2014, the same month that the Copyright Office launched its music-licensing study,46 the District Court for the Southern District of New York conducted a closely watched rate-setting proceeding over musical work license fees.47 There, to determine a reasonable royalty rate for Pandora’s use of the works in the ASCAP repertory, Judge Cote properly excluded sound-recording fees from her analysis on the grounds that § 114(i) barred the court from considering ASCAP’s argument for parity between the rate for performance of a composition and the rate for sound-recording rights.48 But in an unsupported footnote, the court summed up the history of § 114(i) as follows: “Publishers lobbied for this provision in Congress because they were concerned that the sound recording rates would be set below the public performance rates for compositions and drag down the latter. ASCAP also supported the enactment of the provision, for the same reason.”49
As is apparent from Part I, the court’s explanation for why publishers and ASCAP once supported § 114(i) is only partially correct. Undeniably, the publishing industry supported § 114(i) because it feared that sound-recording royalties might otherwise decrease the royalties already being paid to composition copyright owners.50 But this support was largely driven by the simple concern that the new royalties would cut into existing composition royalties—the “royalty pie” rationale.51
This history has not gone entirely unrecognized. For example, in their comment submissions to the Copyright Office, two PROs pointed out that § 114(i) was intended to prevent musical work royalties from being “cannibalized” by payments to owners of the newly created public-performance right in sound recordings.52 But Judge Cote’s suggestion that the provision was enacted out of fear that one rate would be set below and thus drag down the other suggests that the current fixation on § 114(i) as a barrier to fee parity has unduly influenced the contemporary understanding of the provision’s intended purpose. Indeed, a number of industry heavyweights have echoed Judge Cote’s reasoning to explain the change in the publishing world’s position on § 114(i)—arguing, for example, that “the only reason that the publishers now seek to excise this provision is that sound recording performance royalties have turned out to be higher than the publishers expected.”53
Why does it matter that publishers supported § 114(i) twenty years ago for reasons other than the supposed gravitational pull between sound-recording rates and composition rates? First, the court’s misunderstanding wrongly implies that publishers find themselves in a disadvantaged position today with respect to § 114(i) simply because they failed to correctly predict the direction in which the fee disparity would cut. But more significantly, misunderstanding § 114(i) as a buffer that prevents sound-recording rates from dragging down or pulling up composition rates distracts from the inherent illogic of the provision and the problems that it—and by extension, a severely fractured licensing and rate-setting regime—have always posed for music users.
The latter point is well illustrated by the comments submitted by Music Choice in response to the Copyright Office’s Notice of Inquiry soliciting public input on § 114(i). Unsurprisingly, Music Choice—which bills itself as the world’s first digital-music service—supports retaining § 114(i) because it equates the provision with preventing the rate-setting court from raising composition rates.54 Music Choice points out that forcing webcasters to pay publishers’ fees comparable to sound-recording royalties, currently estimated at forty to fifty-five percent of Music Choice revenue, could make for a combined licensing burden greater than one hundred percent of webcasters’ revenue.55 What Music Choice does not acknowledge, however, is that retaining § 114(i) could also bring about this absurd result in the long run. For § 114(i) does not merely prevent the courts from using sound-recording rates to raise composition rates; the provision prevents the courts from using sound-recording rates for any purpose with respect to setting composition rates. That could include using sound-recording rates as a reason for capping composition rates to ensure webcasters are not crushed under the combined licensing burden.56
In theory, if the court determined that a higher composition rate were warranted for reasons unrelated to current sound recording royalty rates, it would have to raise the rate without regard for whether the combined burden of composition royalties and sound-recording royalties destroyed streaming services’ profitability. Meanwhile, despite streaming services’ current support for retaining § 114(i), nothing in the text of § 114(i) guarantees perpetual low musical-work royalties relative to sound-recording royalties, nor does the provision’s modification or repeal automatically threaten to cause an increase in musical work royalty rates. Even if Congress repealed § 114(i), a rate-setting court could well decide to keep composition rates low relative to sound-recording rates, for the same reasons that the CRB cited in deliberately setting sound-recording rates far higher than composition rates in 2014.57
The significance of § 114(i)’s history extends beyond demonstrating the undesirability of the provision itself. Deliberately enacted to ensure the decoupling of royalty rates for musical works and sound recordings, the provision was by its very design at odds with the creation of a fair and sustainable overall licensing and rate-setting scheme. Against this backdrop, publishers’ push to enact the provision yesterday, and music services’ efforts to retain it today, together function as an ironic testament to the fundamental flaws of the fractured system we have inherited and the shortsighted biases that help perpetuate that system.
Moreover, this history underscores the inefficiencies and suboptimal outcomes virtually guaranteed by a system under which music users must obtain multiple rights at rates determined by separate licensing authorities. And by extension, this history supports the arguments of scholars such as Jonathan Cardi and Mary LaFrance who have long advocated for a consolidated licensing and rate-setting system.
First and most obviously, under the present system, users pay multiple licensing fees for the use of any single song and are subject to a rate-setting regime that systematically disregards their cumulative-fee burden. If, as the ASCAP and BMI consent decrees dictate, the court is to set a “reasonable” musical work royalty rate, the court should be permitted to look at a music-streaming service’s entire fee structure. As LaFrance has noted, this would ensure that the combined burden of sound-recording and musical-work royalties do not crush streaming services beneath the weight of cumulative and ever-rising fees or discourage new players from entering the market.58
Second, a more subtle consequence of the currently fractured licensing and rate-setting regimes may be higher-than-optimal individual royalty rates. As Cardi observed almost a decade ago, “[T]he separate licensing of complementary copyrighted works results in an inefficient price for those works—a price higher than if the works were licensed together as parts of a derivative whole.”59 Copyright owners are positioned to lose, too. Although their incentive is to increase their royalties, excessive total royalties could also reduce total sales,60 leading to overall lower profits in the long run.
Third, the burdens associated with a split licensing and rate-setting scheme have ramifications for the commercial availability of music and for innovative uses of music. Despite a decade of public demand, the music industry did not take steps to license online content for electronic distribution until 2001.61 This supply-demand gap has been attributed to the highly inertial and fractured state of music licensing and pricing.62 The system also discourages innovative music users who seek to create derivative works,63 as they must secure separate permissions from both sound recording and musical work copyright owners for use of a single sound recording.64
Finally, the biggest failing of the current licensing and rate-setting regime is that it prevents a coherent determination of the appropriate relative entitlements of songwriters and publishers vis-à-vis artists and recording labels.65 Fees are instead negotiated and renegotiated in piecemeal fashion through expensive and time-consuming adversarial proceedings that result in unpredictable attempts at royalty matching and royalty ratcheting.66
These myriad inefficiencies could be mitigated in at least two ways. First, commentators have suggested establishing a single rate-setting authority empowered to calibrate the two types of royalties,67 an idea recently endorsed by the Copyright Office.68 This more efficient joint rate-setting regime would openly recognize sound-recording rates and composition rates as part of an interconnected licensing universe.69 A joint regime could more coherently determine the appropriate relative entitlements of songwriters and publishers vis-à-vis artists and recording labels than the current separate and reactionary sound-recording and composition-rate regimes. A joint regime could also set rates with the assurance that the cumulative licensing burden placed on music users is equitable, sustainable, and consistent over time.70
Second, both as an alternative to and in addition to a single rate-setting scheme, commentators have proposed the consolidation of licenses or licensing administrators, such that a music user would need to acquire only one license rather than separate licenses from sound recording and musical composition copyright holders.71 To address the antitrust concerns that arise out of the overconsolidation of licensing rights, Congress could provide for a single licensing scheme that permits free negotiations between composition and sound recording copyright owners as to the appropriate division of royalties, with conventional provisions in place for arbitration or litigation where negotiations fail.72 In its recently issued study, the Copyright Office favorably cited a proposal put forth by the Recording Industry Association of America as a promising example of license bundling, wherein copyright holders could agree upfront on a public performance royalty split and then together participate in rate-setting proceedings against the licensee.73
Both of these avenues—vesting rate-setting power in a single authority and consolidating public-performance rights into a single license—offer clear advantages over the fractured licensing and rate-setting model, and they would go a long way toward mitigating the inefficiencies specified above. Joint rate-setting and joint-licensing regimes would provide for a way to account for the licensee’s cumulative-royalty burden, allay concerns about inflated overall musical work or sound-recording royalties, reduce uncertainty and complexity for potential licensees seeking to use licensed works, and express the relative entitlements of songwriters and publishers versus recording artists and labels.
This Comment has argued that efforts to create a copyright regime with long-term viability should account for the fact that § 114(i) was designed to force rate-setting courts to set composition fees without regard for music users’ total licensing burden. This context and analysis are significant because they not only help explain why industry actors should unite in seeking the provision’s elimination but also underscore the inherent inefficiencies of the present licensing and rate-setting system and point the way to change.