ISSN 1936‐5349 (print)     ISSN 1936‐5357 (online)  HARVARD   JOHN M. OLIN CENTER FOR LAW, ECONOMICS, AND BUSINESS  FELLOWS’ DISCUSSION PAPER SERIES        THE PRICE OF FAME: THE ANTITRUST LAW AND ECONOMICS OF BROADCASTING, MUSIC, AND ADVERTISING Ivan Reidel Discussion Paper No. 35 7/2010 Harvard Law School Cambridge, MA 02138 Contributors to this series are John M. Olin Fellows or Terence M. Considine Fellows in Law and Economics at Harvard University. This paper can be downloaded without charge from: The Harvard John M. Olin Fellow’s Discussion Paper Series: http://www.law.harvard.edu/programs/olin_center/ THE PRICE OF FAME: THE ANTITRUST LAW AND ECONOMICS OF BROADCASTING, MUSIC AND ADVERTISING  IVAN REIDEL ABSTRACT Why do we watch and listen to so many ads on radio and television when we would rather not? Why are most songwriters in the U.S. and elsewhere unable to make a living out of their creative efforts, but the few that can are often rewarded more than all others combined? Are these the features of efficient or otherwise desirable markets? Three separate strands of economic literature spanning over more than two decades on the economics of superstars, the structure and performance of broadcasting markets and the economics of music licensing suggest that these features are staples of rather well functioning markets, and that in fact there isn’t much to regret about this state of affairs. Superstardom is a direct consequence of talent, or perhaps consumer choice or bandwagon effects. Advertising likely provides valuable information, but even uninformative advertising is essential to fund the commercial radio or television programs that people do want to consume: “reduce it, and broadcasters won’t be able to afford high quality content” the argument goes. Courts, Congress, government agencies and legal scholars have long relied on many of these economic theories to decide antitrust cases, guide government oversight efforts, and engineer laws and regulations in broadcasting and music licensing markets. In this article I challenge the dominant economic analysis of these markets and show that more than six decades of antitrust oversight and enforcement efforts in these markets by U.S. courts, Congress and government agencies have been utterly misguided. It is in large part because of such misguided efforts, this article argues, that we presently experience inefficiently high levels of advertising, skewed income distributions for songwriters and reduced audience welfare. After examining one of the most far reaching and enduring regulatory efforts in U.S. history, I expand the economic understanding of these markets and leverage a novel analytical framework to show how legal scholarship, building on faulty economics, has evolved its own series of misguided policies. On this understanding I suggest that courts are indeed compelled by antitrust law to begin transforming these markets by declaring current licensing practices in broadcasting and music markets illegal and that the Antitrust Division of the Department of Justice is equally compelled to either abandon or radically modify its oversight over Performing Rights Organizations. I conclude with a proposal for market design that eludes most of the pitfalls of current market structure and likely results in vast welfare improvements all while substantially reducing unnecessary oversight expenditures.  S.J.D., Harvard Law School. For helpful comments and support, I am grateful to Einer Elhauge, William Fisher III, Stavros Gadinis, Jane Hopwood, Louis Kaplow, Katerina Linos, Vlad Perju, Yuval Procaccia, Holger Spamann, Steve Shavell, Elina Treyger and workshop participants at Harvard Law School, Yale Law School, the University of Michigan School of Law and Boston University School of Law. I also wish to thank the John M. Olin Center for Law, Economics and Business at Harvard Law School for its generous research support. THE PRICE OF FAME: THE ANTITRUST LAW AND ECONOMICS OF BROADCASTING, MUSIC AND ADVERTISING © 2010 IVAN REIDEL TABLE OF CONTENTS 1 INTRODUCTION.................................................................................................3 2 A THREE SIDED MARKET IN TWO SIDED SHOES: BROADCAST RADIO AND THE TAYLOR SWIFT PARADOX ........................................................................7 2.1 The Taylor Swift Paradox .........................................................................7 2.2 The Present State of Economic Analysis ..................................................9 2.2.1 Copyright Collectives and the Price of Fame ..................................10 2.2.2 Two-sided Markets (+1 Cartelized Significant Other) ....................12 2.2.3 Revisiting Superstardom in Three-sided Markets: Two Additional Explanations on Why Artists’ Revenues are Skewed in Welfare Decreasing Way ................................................................................24 2.2.4 Economic Analysis of Copyright Collectives ..................................28 3 THE LEGAL DETERMINANTS OF THE PRICE OF FAME ANOMALY ................29 3.1 PROs and the Blanket License: Testing Legality under the Rule of Reason ...............................................................................................30 3.1.1 Claimed Procompetitive Effects ......................................................31 3.1.1.1 Alleged Procompetitive Justification I: The “New Product” that Radically Lowers Transaction Costs ....................................32 3.1.1.1.1 Two Critical Flaws in the Analysis of Transaction Costs (Past and Present) ..........................................................................33 3.1.1.1.1.1 Myopic Balancing: The Unexamined Costs of Blanket Licenses ..................................................................................33 1 3.1.1.1.1.2 Less Restrictive Alternatives to Blanket Licenses: The Modern Transactional Platform Turns 15 ....................................35 3.1.1.2 Alleged Procompetitive Justification II: Blanket Licenses Optimally Increase Output (In a Way which is Superior to à-la-Carte Pricing) .......................................................................................36 3.1.1.2.1 Three Critical Flaws of the Optimal Output/Consumption/Price Claims .........................................................38 3.1.1.2.1.1 Negative Prices (Negated by Blanket Licenses) are Necessary for Achieving Optimal Creative Incentives (Dynamic Efficiency) and Optimal Consumption/Output Levels………………...........................................................................38 3.1.1.2.1.2 Consumption of Songs by Radio and Television Stations is often Rival ..........................................................................42 3.1.1.2.1.3 Songwriters are not Able to Compete Effectively against the Blanket License .................................................................45 3.1.2 Anticompetitive Harms ....................................................................52 3.1.2.1 Price, Output and Quality .........................................................52 3.1.2.1.1 Price .......................................................................................52 3.1.2.1.2 Output ....................................................................................55 3.1.2.1.3 Quality ...................................................................................56 3.1.2.2 Further Harms ...........................................................................57 3.1.3 The Need for a Proper Rule of Reason Analysis .............................58 3.2 Proposed Market Design: Music Licensing Markets 3.0 ........................59 4 CONCLUSION ..................................................................................................62 2 1 INTRODUCTION Of the many peculiar things we human beings are known to enjoy, watching television and listening to radio rank high. The average American, for example, spends the lion’s share of his or her leisure time watching about 28 hours of TV and listening to 19 hours of radio each week.1 Paradoxically however, much of this time is spent consuming ads, which most of us enjoy less than songs or other broadcasted content.2 Besides reducing the quality of perhaps the most important pastime in the life of millions of Americans (and billions of viewers and listeners elsewhere) ads take up a large portion of two very limited resources, airtime and available leisure time, which as a consequence cannot be devoted to (and therefore depresses demand for) content that audiences value more. The livelihoods of countless artists supplying such content are in turn profoundly affected by the ratio of advertising to content in commercial broadcasting, from which they derive a substantial part of their income. For hundreds of thousands of songwriters in the U.S. for instance, royalties collected from commercial broadcasting licenses are often the same as those they receive from all other income sources including record sales and song downloads. In this article I argue that this peculiar state of affairs is the result of poorly performing broadcasting and music licensing markets, and in the case of the U.S., of many decades of unwise regulation and misguided oversight by courts and government agencies—namely the Antitrust Division of the Department of Justice (DOJ) and the Federal Communications Commission (FCC)—whose actions are presently reducing the welfare of audiences, jeopardizing the livelihoods of most U.S. songwriters and squandering limited resources on costly and counterproductive policies. The primary source of such substantial market shortcomings and unwise policies however, runs deeper than the actions of courts and government agencies, and can be traced back to decades of misguided recommendations by influential works both in legal and economic scholarship. Focusing first on this long line of legal and economic inquiries, the analysis below begins the task of exposing their analytical shortcomings by eliciting novel answers from three rather traditional questions these literatures have repeatedly pursued: Why do we watch and listen to so many ads on radio and television when we would rather not? Why are most songwriters unable to make a living out of their 1 See Arbitron, Radio Today: How America Listens to Radio Today (2007 Edition), http://www.arbitron.com/downloads/radiotoday07.pdf.  2 Modeling ads as imposing nuisance costs on audiences is a common assumption in the literature and it provides an intuitive framework to explain the phenomena analyzed here. The proposal that this article puts forward, however, circumvents the issue of whether ads increase or reduce utility by advancing an alternative market design that allows markets to simply price annoyance costs more efficiently in all inputs used by broadcasters, be they ads or songs. For a survey on the literature on advertising and annoyance costs, see e.g., Anthony Dukes & Esther Gal-Or, Negotiations and Exclusivity Contracts, 22(2) MARKETING SCI. 222–45 (2003); see also, Gary S. Becker & Kevin M. Murphy, A Simple Theory of Advertising as a Good or Bad, 108-4 Q. J. ECON. 941, 961 (1993); GARY S. BECKER, ACCOUNTING FOR TASTES 223 (1996).   3 creative efforts, but the few that can are often rewarded more than all others combined? Are these superstars the end result of an efficient market? While these questions have been answered many times by analytically isolated lines of inquiry, presenting them under a common analytical framework quickly reveals that the answers presently offered within the economic and legal literatures are incapable of providing a unified view of broadcasting, advertising and music licensing markets which is both encompassing and coherent. How so? The economic analysis of TV and radio as two-sided markets has offered uneasy, slightly disheartening, learn-to-live-with-your-share-of-ads comfort in predicting advertising levels as a result of interdependent demands of audiences and advertisers mediated by broadcasting platforms. But the analysis has so far failed to consider the interdependent supply and demand of (cartelized) songwriters, squaring a three-sided market in two-sided shoes and therefore missing readily available policy tools capable of reducing advertising levels. Superstardom economists have focused on the talent and popularity of superstars, but have neglected to examine the prices of songs, which nearly every outlet that plays songs—from supermarkets to broadcasters, in every corner of the world—observes through the distorted goggles of blanket licenses, which price fame no differently than obscurity. Economists and lawyers dissecting the law and economics of copyright collectives have long suggested that the use of blanket licenses—just like pricing songs at marginal cost—results in an optimal output of songs, but they have misunderstood how high blanket license prices, by making advertising more attractive to broadcasters (i.e. suboptimally cheap) reduce the available airtime to songwriters (i.e. the output of songs) in ways that neither competition between songwriters or between collectives, nor rate courts have been able to correct. Indeed, they also missed that, regardless of the advertising market, marginal cost pricing is not necessarily output enhancing when negative prices can increase the profits of songwriters who sell not only songs, but bundles of products, such as concerts, song downloads, and t-shirts. What do these analytical oversights have in common? All of them, in some way or another, have overlooked “the price of fame,” that is, how song prices must vary from one songwriter to another, and why, when they don’t (and currently they don’t), we get higher advertising levels than desired, revenues more skewed than what is desirable or efficient, and why-oh-why most of us get to know that “15 minutes can save us 15% or more on car insurance.” Without better guidance from these strands of legal and economic analysis or the aid of a general theory of these markets, agencies have been unable to assess the full and unfortunate consequences of their actions. The DOJ has on the one hand enabled Performing Rights Organizations (PROs) to suppress price competition between songwriters (through their offerings of all-you-can-eat blanket licenses) and on the other prosecuted associations of broadcasters trying to limit the amount of advertising which high song prices invite. The FCC has been concerned about the amount of advertising on commercial broadcasting, and also about the livelihoods of independent songwriters trying to break into new media markets, but it has undermined its own goals by enforcing anti-payola regulations which put most 4 songwriters—implicitly overpricing their songs as a consequence of blanket licenses—at a disadvantage vis-à-vis advertisers and vis-à-vis few competing and implicitly better-priced songwriters—superstars. Additionally, courts which, with aid of better theories, should have constrained the actions of both the agencies and the firms these agencies seek to regulate, are instead helping to perpetuate the status quo. Correcting the analytical shortcomings which inform these actions results in the following novel legal and economic insights: (a) blanket licenses should presently be considered illegal under U.S. antitrust law and declared as such by courts and enforcement agencies; (b) doing away with them will likely lower advertising levels and de-skew—at least partially—revenues for hundreds of thousands of songwriters; (c) the large judicial and government expenditures that for more than six decades have been devoted to monitoring PROs through antitrust consent decrees, funding rate courts, and funding anti-payola enforcement should be understood as an unnecessary and entirely avoidable waste of resources, not only because songwriters can already price songs on their own without “collective” participation, but because declaring blanket licenses illegal can solve a chicken-and-egg problem that has likely prevented better transactional platforms (e.g. eBay like platforms) from supplanting PROs; (d) online transactional platforms, can allow markets to vastly outperform blanket licenses—quantitatively and qualitatively—by allowing different songwriters to employ several pricing strategies simultaneously (e.g. auctions or any arbitrarily set price). Payola, when properly examined in this context, is merely the tip of a much larger market that remains almost entirely submersed because it lacks the necessary transactional platforms that would make it viable. Because modern online platforms can accommodate simultaneously both negative and positive auction values (think of negative reservation values), for the first time in history, songwriters could also have access to a competitive advertising market for songs (i.e. a truly competitive payola market). Under such a system, the price of particular songs could be determined through auctions as either negative or positive, solely on the basis of competition between radio stations offering airtime (advertising spots) and songwriters offering songs—which are simultaneously an input for broadcasters and a promotional tool for songwriters wishing to stimulate the sale of CDs, song downloads, concerts, and the like. In other words, this is a problem of market design for which we now possess suitable technological solutions. The analysis below proceeds as follows: Section 2 examines the economic structure of broadcasting, advertising and music licensing markets and makes four principal contributions to economic analysis. First, it extends a recent strand of research examining commercial broadcasting as a two-sided market and corrects a fundamental shortcoming in the economic modeling of the market—the misunderstood significance of a cartelized side (songwriters) which is not presently modeled. This correction shows for the first time how music licensing practices increase the quantity of advertising in equilibrium in commercial broadcasting in a welfare decreasing way. Second, this section challenges the premature and misguided application of mainstream economic theories of superstardom to the music industry by advancing two novel alternative theories that 5 explain skewed revenue distributions in songwriter markets as partly attributable to decreased competition between songwriters and advertisers and decreased competition among songwriters, both caused by a blanket license pricing system that coalesces with uniform pricing systems also prevalent in record sales and downloaded music. Third, the section extends the economic analysis of payola by examining the practice in the framework of a multi-sided market and by exposing previously ignored positive externalities (information spillovers) associated with the practice that tend to raise the utility of audiences. Fourth, it extends the economic analysis of copyright collectives by assessing the performance of these institutions in conjunction with commercial broadcasting—by far their largest customers—and introducing a novel theory of competitive harms, which includes increased advertising levels, artificially skewed income distributions and reduced creative incentives. This analysis of the licensing behavior of collectives further examines why price and output effects related to songs, remain misunderstood by the economic literature, and why quality effects on songs have been missed entirely. The analysis of collectives is discussed in both Sections 2 and 3. Section 3 examines the legal determinants of the pricing anomalies uncovered in Section 2, and challenges mainstream analysis (both legal and economic) of the harms and pro-competitive benefits of blanket licenses and anti-payola regulations. The analysis of blanket licenses is structured as a rule of reason inquiry under U.S. antitrust law in order to show not only that mainstream legal analysis has failed to properly assess the welfare consequences of the use of blanket licenses, but indeed to suggest that proper judicial scrutiny presently compels a declaration of blanket licenses as illegal restraints of trade, and consequently that enforcement practices by the DOJ are in urgent need of reform. Section 3 further describes available remedies and examines alternative market design choices. The section then advocates for ELEGANCE in the market, that is, a novel market design where transactions take place within a network of Electronic Licensing Engines Giving Authors a Non- Collusive Environment. ELEGANCE, by pricing songs through auctions—although most individual pricing mechanisms are feasible—is able to eliminate pricing (and licensing) intermediaries such as PROs and the harms that they inflict through the collective pricing of songs. The system of competition between licensing engines or platforms—unconcerned with pricing songs—is shown to outperform current licensing practices by, among other things, allowing negative price auctions—making the payola/pay-for-play market truly competitive—and by enabling, for the first time in nearly a century of song licensing, the efficient trading of exclusive rights in public performance licenses, curving overuse and underuse of songs by commercial broadcasters. Section 4 offers concluding remarks. 6 2 A THREE SIDED MARKET IN TWO SIDED SHOES: BROADCAST RADIO AND THE TAYLOR SWIFT PARADOX 2.1 The Taylor Swift Paradox Fame and fortune are no strangers to songwriter and singer Taylor Swift. In the first week of August of 2009, her song “You belong to me” claimed the top spot in radio airplay and commanded, according to Mediaguide, 19,361 spins3 or plays on the more than 2500 radio stations they monitor in 150 U.S. markets.4 In fact, by August 9, 2009, “You Belong to Me” had spent four weeks at the number one spot of radio airplay, thousands of spins above the closest song.5 By industry standards, twenty year-old Swift is a phenomenon. Celebrated as “[o]ne of pop’s finest songwriters” by The New York Times,6 Swift was the biggest record selling artist of 2008 in the U.S., her album “Fearless” selling more copies than any other album since Santana’s “Supernatural” in 2000.7 When Swift goes on tour, things are not much different. According to Forbes, tickets for her May 22, 2009 concert at the Los Angeles Staples Center, for example, were sold out in the first two minutes.8 As much as audiences like Swift however, when counting the number of plays of all performances on those same radios, Swift hardly even makes the top ten list of those with the most “spins” or plays. During the first week of August 2009, it was insurance company “Geico” that took the number one spot, with 42,544 spins or more than twice the number of plays Swift received.9 Home Depot came in second with 41,371 and Mac Donald’s third with 34,593.10 In that week, in fact, Swift only scratched the number ten spot slightly behind AT&T which obtained 19,574.11 3 See Mediaguide, National Mainstream, http://charts.mediaguide.com/format/National_Mainstream_single.html (last visited Aug. 10, 2009).  4 See Mediaguide, http://www.mediaguide.com/about (last visited Dec. 26, 2009).  5 See Mediaguide, supra note 3.  6 See Jon Caramanica, Sounds of Swagger and Sob Stories, December 21, 2008, N.Y. TIMES, AR33, available at http://www.nytimes.com/2008/12/21/arts/music/21cara.html?_r=1.  7 See Katie Hasty, Taylor Swift Remains Atop Billboard 200, Feb. 25, 2009, http://www.billboard.com/news/taylor-swift-remains-atop-billboard-200- 1003944728.story#/news/taylor-swift-remains-atop-billboard-200-1003944728.story (last visited Dec. 26, 2009).  8 See CMA and AMA Award Winner Taylor Swift Expands Doll Line With Playsets, New Fashions, Dec. 1, 2009, http://www.forbes.com/feeds/businesswire/2009/12/01/businesswire132227701.html (last visited Dec. 26, 2009).  9 See Mediaguide, Top Brands & Advertisers on Radio, http://charts.mediaguide.com/ads/National_Advertiser.html (last visited Aug. 10, 2009).  10 Id.  11 Id.  7 This is an odd state of affairs, not only because people enjoy songs and overwhelmingly dislike listening to ads on the radio,12 or even because in this particular week the top song by one of this decade’s superstars barely even made it in the top most often played things on the radio, but because the first week of August 2009, was in fact an ordinary week for radio broadcasting in the U.S. and elsewhere. Indeed, that global leisure time is impoverished by ads also means that valuable talent that would have otherwise replaced those annoying commercials is instead squandered by societies’ inability to reward it. If ads could be replaced by the content audiences enjoy the most—songs for instance—the incomes lost and impoverished livelihoods of countless songwriters—the vast majority of which currently can’t make a living out of the public performance of their songs alone13— would be able to receive a substantial boost from all the freed airtime. In the U.S. alone for instance the average person listens to 19 hours of radio each week. Radio reaches 93% of U.S. consumers each week,14 and 72% every day.15 By 1999 Anderson and Coate report that non-programme minutes exceeded “20 min per hour on some network television programs and 30 min per hour on certain radio programmes.”16 Multiply those ad spins by the number of listeners times the number of hours they spent on such unpleasant an activity, and this massive waste of time by audiences provides not only dramatic measure of diminished audience welfare but a proxy for the large toll imposed by ads on songwriters in the case of radio, and an even bigger pool of artists in the case of television. The prevalence of ads, however, is not the only paradox afflicting songwriters, for regardless how big the pie is, it is only a few that get most of it. In many sectors of the economy, differences in the marginal productivity of capital or labor generally command proportional differences in interest or wages. But Taylor Swift has substantially higher number of spins than anyone else in the industry. Is that a reflection also of substantially higher talent? This question is not restricted to Taylor Swift: looking at the number of spins for all hits with the most radio air-time, the number “one” artists, and the few that follow them, generally command dramatically more spins than anyone else on radio. This metric tracks quite accurately the differences in income that artists derive from radio or TV airplay (as revenue is based 12 That advertising of products other than music generally annoys listeners is a common assumption in the economic literature in the field. See Sheila M. Campbell, Two-Sided Markets with a Negative Network Effect: Radio, Advertisers, and Audiences 6 (2006) (unpublished Ph.D. dissertation, Boston College)(on file with Boston College University Libaries).  13 See generally, Pew Internet & American Life Project, Artists, Musicians and the Internet, 46-47, http://www.pewinternet.org/Reports/2004/Artists-Musicians-and-the-Internet/The-musicians- survey.aspx?r=1; Martin Kretschmer & Philip Hardwick, Authors’ Earnings from Copyright and Non- copyright Sources: A Survey of 25,000 British and German Writers (2007), http://www.cippm.org.uk/downloads/ACLS%20Full%20report.pdf.  14 Age 12 and older.  15 See Arbitron, supra note 1.  16 See Simon P. Anderson & Stephen Coate, Market Provision of Broadcasting: A Welfare Analysis, 72 REV. ECON. STUD. 947, 947-48 (2005).  8