Journal of Competition Law and Economics Advance Access originally published online on June 20, 2008
Journal of Competition Law and Economics 2008 4(3):697-751; doi:10.1093/joclec/nhn019
| ||||||||||||||||||||||||||||||||||||||||||||||||
EVALUATING MARKET POWER WITH TWO-SIDED DEMAND AND PREEMPTIVE OFFERS TO DISSIPATE MONOPOLY RENT: LESSONS FOR HIGH-TECHNOLOGY INDUSTRIES FROM THE ANTITRUST DIVISION'S APPROVAL OF THE XM–SIRIUS SATELLITE RADIO MERGER
Correspondence: E-mail: jgsidak{at}aol.com
 |
ABSTRACT |
|---|
 TOP ABSTRACT I. INTRODUCTIONII. THE ANTITRUST DIVISION'S...III. MARKET DEFINITIONIV. COMPETITIVE EFFECTSV. PREEMPTIVE OFFERS TO...VI. CONCLUSION |
|---|
Can the standard merger analysis of the Department of Justice's and Federal Trade Commission's Horizontal Merger Guidelines accommodate mergers in high-technology industries? In its April 2007 report to Congress, the Antitrust Modernization Commission (AMC) answered that question in the affirmative. Still, some antitrust lawyers and economists advocate exceptions to the rules for particular transactions. In the proposed XM–Sirius merger, for example, proponents argue that the Merger Guidelines be relaxed to accommodate their transaction because satellite radio is a nascent, high-technology industry characterized by "dynamic demand." We argue that the AMC correctly refrained from recommending high-tech exceptions for defining markets in merger proceedings. Merger proponents naturally seek to expand the relevant product market as much as possible. But if alternative products are included in the relevant market without a showing of significant cross-price elasticities—that is, without evidence of buyer substitution between the two products in response to a relative change in prices—then market definition is unbounded. The XM–Sirius merger also follows a recent trend of prosecutorial inaction in merger reviews. The Antitrust Division's use of a higher standard for intervention than the incipiency standard in Section 7 of the Clayton Act increases the risk of false negatives. Finally, the XM–Sirius merger exemplifies the use of preemptive offers of merger conditions by the merger parties to gain political favor and to allocate postmerger rents to influential third-party intervenors. The most significant preemptive concessions were XM's and Sirius's offer to freeze the monthly subscription price at the premerger monthly rate of $12.95 and to offer a variety of new tiered program packages that XM and Sirius characterized as "à-la-carte." These offers presumably were intended to neutralize the traditional antitrust concerns that a merger among direct competitors leads to higher prices and to win the support of certain vital constituencies. To the contrary, we argue that the offer to freeze prices could reduce welfare and that the Federal Communications Commission and the Department of Justice lack the authority to create a rate-regulated monopoly for satellite radio. Furthermore, because the "à-la-carte" offering would not hold constant other nonprice factors, consumer surplus could fall.
|
I. INTRODUCTION |
|---|
|
TOPABSTRACTI. INTRODUCTIONII. THE ANTITRUST DIVISION'S...III. MARKET DEFINITIONIV. COMPETITIVE EFFECTSV. PREEMPTIVE OFFERS TO...VI. CONCLUSION |
|---|
Can the standard merger analysis of the Department of Justice's and Federal Trade Commission's Horizontal Merger Guidelines accommodate mergers in high-technology industries? The suitability of applying the Merger Guidelines to high-technology products was affirmed in the April 2007 report to Congress by the Antitrust Modernization Commission (AMC), which found that "[n]o substantial changes to merger enforcement policy are necessary to account for industries in which innovation, intellectual property, and technological change are central features."1 Despite this finding, some antitrust lawyers and economists argue that the sheer dynamism of a particular industry defies standard market definition analysis. An example is the proposed merger of XM Satellite Radio Holdings Inc. (XM) and Sirius Satellite Radio Inc. (Sirius), the only U.S. licensed providers of satellite digital audio radio services (SDARS).2 The horizontal combination of the only two SDARS providers would constitute a merger to monopoly—if SDARS constitutes a relevant antitrust product market.
The conflict between standard antitrust analysis and a more "dynamic" approach advocated by XM and Sirius is most apparent when defining the relevant product market. The Merger Guidelines specify the kind of evidence that can inform market definition: "Market definition focuses solely on demand substitution factors—i.e., possible consumer responses."3 Applied here, to expand the product market beyond satellite radio (the narrowest possible set of products), one must demonstrate that satellite radio subscribers shift their demand between satellite radio and other forms of audio entertainment (for example, terrestrial radio) in response to a relative change in prices of those services.4 XM and Sirius failed to demonstrate any evidence of buyer substitution in response to changes in relative prices. Through their economists, XM and Sirius argued that such evidence was hard to find due to the fact that satellite radio prices had not changed between 2005 and 2007. More importantly, they argued that dynamic demand considerations in the satellite radio industry frustrated the accepted demand-side test for market definition. The vast majority of XM's and Sirius's inferences were based on supply-side information, which the Merger Guidelines exclude when defining product markets, except in rare cases in which decisions by sellers can serve as a proxy for how buyers would react to a relative change in prices.5 The fact that entrepreneurs may be designing new audio devices in their garages does not inform the ultimate question of whether, over the next two years, satellite radio customers would substitute away from satellite radio to another audio device in response to a relative change in prices.
Defining markets and measuring postmerger market power are two sides of the same coin. If outside products constrain the price of the merged entity, then the market should be expanded and the merged firm will lack market power. Section 7 of the Clayton Act and the judicial decisions embracing the Merger Guidelines are the alternative to concocting new theories to permit the latest merger to pass muster. They also have the dual virtues of being the law and being correct. As the Department of Justice (DOJ) and the Federal Trade Commission (FTC) reaffirmed in March 2006: "The core concern of the antitrust laws, including as they pertain to mergers between rivals, is the creation or enhancement of market power ... . The Guidelines set forth the analytical framework and standards, consistent with the law and with economic learning, that the Agencies use to assess whether an anticompetitive outcome is likely. The unifying theme of that assessment is that mergers should not be permitted to create or enhance market power or to facilitate its exercise."6 In that regard, the "Guidelines' analytic framework has proved both robust and sufficiently flexible to allow the Agencies properly to account for the particular facts presented in each merger investigation."7
Section 7 of the Clayton Act prohibits any merger, "the effect of [which] may be substantially to lessen competition, or to tend to create a monopoly."8 The Division explained that it decided not to challenge the XM–Sirius merger "because the evidence did not show that the merger would enable the parties to profitably increase prices to satellite radio customers for several reasons."9 However, two of the four factors that the Division then listed are unrelated to the ability of a merged firm to raise price, such as "a lack of competition between the parties in important segments even without the merger" and "efficiencies likely to flow from the transaction that could benefit consumers." Thus, the Division's competitive-effects conclusion had to rest on two other factors: "the competitive alternative services available to consumers" and "technological change that is expected to make those alternatives increasingly attractive over time."10
The way in which XM and Sirius addressed the issue of competitive alternatives through a novel approach to market definition was not the only occasion in which the merger review process was undermined. The merger is also a case study of the strategic use of preemptive concessions to influence political constituencies. The proposed merger of XM and Sirius attracted the attention of many diverse interest groups that sought to extract valuable concessions that would benefit their constituents—but not necessarily satellite radio consumers—in return for endorsing the transaction. The merger parties could consent to these preemptive concessions only if they expected the profits earned as a regulated monopolist to exceed the current duopoly profits plus the costs of the concessions.
The merger application11 submitted to the Federal Communications Commission (FCC) can best be viewed as an invitation by XM and Sirius for the federal government and various third parties to partake in "rent extraction." That invitation was predicated on XM's and Sirius's creation of an entirely new price-regulated monopoly. In contrast to rent creation, rent extraction connotes the dissipation through government policy of either publicly created monopoly rents or privately created quasi-rents.12 Regulation can function as a process by which economic rents are created, perpetuated, and threatened with dissipation (and thus extracted by third parties).
Given the certainty that the proposed merger would create monopoly rent, politically sophisticated interest groups came out of the woodwork to dissipate that rent. They did so by conditioning their endorsement or approval of the proposed merger on the receipt of a share of the expected monopoly rent. Being astute about how the game of rent creation and rent extraction is played, XM and Sirius offered to award claims on their future monopoly rent, beginning with several concessions that their merger application portrayed—erroneously in our view—as "merger-specific benefits." Because XM and Sirius were the residual claimant to the monopoly rent that the merger would create, they stood to profit from consenting to these commitments to dissipate rent up until the point at which the value of the concessions exceeds the value of the expected monopoly rent.
In this article, we analyze in greater detail these issues of market definition and preemptive conditions. We begin, in Section II, by comparing the incipiency standard under Section 7 of the Clayton Act with the standard that the Antitrust Division has used when announcing its decisions not to prosecute mergers. The Division's decision to refrain from challenging the XM–Sirius merger, while consistent with the Division's recent reluctance to prosecute mergers, nevertheless deviates from the statutory language of Section 7 and the applicable case law interpreting it. That deviation weakens merger review.
In Section III, we examine whether satellite digital audio radio service (SDARS) is a relevant product market for antitrust purposes. The Merger Guidelines13 impose specific rules with regard to the type of evidence that may be considered for market definition and how to draw the contours of a product market. An application of those rules implies that SDARS is a distinct product market. In particular, the demand for SDARS appears to be insensitive to price increases based on (1) the fact that XM did not lose subscribers when it raised prices in 2005 and (2) the low churn rate for SDARS.
This finding is bolstered by other analyses. First, indecency standards legislated by Congress and interpreted by the FCC have segmented the market between broadcast content and subscription-based content. As a result, indecent video content has gravitated to cable networks and direct broadcast satellite (DBS), and indecent audio content has gravitated to SDARS. In 2004, the chief executive of XM said: "We want to be the HBO of radio."14 Since that comment was made, the most compelling content on SDARS has indeed been content that would be deemed "indecent," and thus unlawful, if presented in a terrestrial broadcast environment. We next assess market-based evidence on substitution possibilities to determine whether consumers perceive alternative audio services such as podcasts, mobile Internet radio, terrestrial-based advertiser-supported radio, and hybrid digital (HD) radio to be reasonably interchangeable with SDARS.
We then analyze the novel concept of "dynamic demand," advanced by XM's and Sirius's economic consultants. Because SDARS providers face this so-called "dynamic demand," XM and Sirius argued that the small-but-significant-and-nontransitory-increase-in-price (SSNIP) test must be altered to account for long-run profit considerations. Despite their extensive experience in merger cases, the lawyers and economists representing XM and Sirius failed to cite a single instance in which a court or an agency altered the SSNIP test in this way. Indeed, in the last six high-profile mergers reviewed by the FCC, the SSNIP test was applied without any alteration. Through their economists, XM and Sirius also relied on the concept of "dynamic demand spillover" to motivate an unprecedented efficiency justification, including the claim, which we refute, that the merger of XM and Sirius would accelerate investment in interoperable radios. However, it is not consistent to argue on the one hand that the other types of audio entertainment compete with SDARS, but on the other that the merger solves the problem of "dynamic demand spillover."
In Section IV, we analyze the likely competitive effects of the proposed merger of XM and Sirius. We show that the proposed merger would likely increase prices relative to a world in which the merger is not consummated. Because a monopolist charges more for a service than do oligopolists, the postmerger price would necessarily be higher (assuming no decrease in the merged firm's marginal cost). Moreover, this anticompetitive unilateral effect is not limited to the incremental out-of-pocket costs that subscribers would need to pay to get programming. It also should take account of the costs associated with enduring additional commercials, a planned strategy of XM and Sirius conditional on their obtaining merger approval. The proposed merger would also adversely affect the market for audio programming. Because a combined SDARS provider would have monopsony power over content, the amount of content should decline. One possible form of a reduction in quantity here would be a reduction in the variety of SDARS programming. Because consumers value variety, such a reduction would decrease consumer welfare. Section IV concludes with an analysis of the antitrust significance of the National Association of Broadcaster's (NAB's) opposition to the merger.
In Section V, we review the proposed conditions offered preemptively by the merger proponents. The preemptive conditions took the form of conduct remedies. We begin by reviewing the DOJ's guidelines on merger remedies generally and the use of conduct remedies in particular. Despite the DOJ's general criticism of conduct remedies, the FCC was willing in at least one previous merger to serve as a platform for the exchange of rents via conduct remedies. We explain why these preemptive offers, although attractive to certain political constituencies, would not remedy the likely anticompetitive effects of the proposed merger. The biggest preemptive concession was XM's and Sirius's offer to freeze the monthly subscription price at the premerger monthly rate of $12.95 and to offer a variety of new tiered program packages that XM and Sirius generously describe as "à-la-carte." We explain why these conditions represent a de facto regime of price-cap regulation that is antithetical to the deregulatory movement at the FCC over the past decade. A price freeze at the current monthly price of $12.95 would be welfare-reducing to the extent that the future price that emerges from continued oligopolistic competition between Sirius and XM in the absence of the merger would naturally cause the equilibrium price to fall below $12.95 per month. Even assuming that it is possible to calculate the appropriate price level and duration of price controls for the merged firm, no FTC or DOJ precedent supports such a requirement as part of an antitrust consent decree.15 Section V concludes by reviewing other preemptive concessions made by XM and Sirius.
|
II. THE ANTITRUST DIVISION'S DEVIATION FROM THE INCIPIENCY STANDARD SPECIFIED IN SECTION 7 OF THE CLAYTON ACT |
|---|
|
TOPABSTRACTI. INTRODUCTIONII. THE ANTITRUST DIVISION'S...III. MARKET DEFINITIONIV. COMPETITIVE EFFECTSV. PREEMPTIVE OFFERS TO...VI. CONCLUSION |
|---|
Section 7 of the Clayton Act prohibits mergers in which "the effect of such acquisition may be substantially to lessen competition, or to tend to create a monopoly."16 The Supreme Court construes Section 7 to seek "to arrest the anticompetitive effects of market power in their incipiency," such that "the core question is whether a merger may substantially lessen competition"—a question that "necessarily requires a prediction of the merger's impact on competition, present and future."17 It is important to distinguish the statute's controlling standard, embodied in the phrase "may substantially lessen competition," from the five-step analysis required by the Merger Guidelines.18 Under the Merger Guidelines, the Antitrust Division or FTC makes a prosecutorial decision on the basis of its evaluation of whether a proposed combination is "likely to create or enhance market power or to facilitate its exercise."19 The methodology of the Merger Guidelines is intended to ensure that the government does not deter transactions that are competitively beneficial or neutral.20 However, the Guidelines are not meant to act as an insurmountable bar to government action. Some might argue, however, that in recent years the Antitrust Division's horizontal merger review process has had such an effect.
The last time the Antitrust Division sued to block a horizontal merger was the unsuccessful lawsuit in 2004 to block Oracle's acquisition of PeopleSoft.21 One possible explanation for the Antitrust Division's reluctance to litigate may stem from its defeat in Oracle, in which the Division alleged that the combination of two of the three providers of high-function enterprise software22 "would likely have" several anticompetitive effects.23 Upon entering judgment for the defendants, the court said that in such a case the Antitrust Division "must show that a pending acquisition is reasonably likely to cause anticompetitive effects."24 The Oracle court noted that the Supreme Court emphasized in Brown Shoe Co. v. United States that "Congress used the words ‘may be substantially to lessen competition’ to indicate that its concern was with probabilities, not certainties."25 The Oracle court also quoted Judge Posner's statement in Hospital Corp. of America v. FTC that "Section 7 does not require proof that a merger or other acquisition [will] cause higher prices in the affected market. All that is necessary is that the merger create an appreciable danger of such consequences in the future."26 The Oracle court thus emphasized that liability under Section 7 arises from a reasonable probability, rather than the certainty, of a substantial lessening of competition. Notwithstanding the lower evidentiary requirements of that incipiency standard, the Antitrust Division in Oracle failed to prove its theory of market definition, and that failure in turn prevented the Division from establishing the evidentiary presumption, under United States v. Phila. Nat'l Bank, that the merger would be unlawful.27 The Division's failure to shift the burden of proof to the merging parties effectively derailed its case.28
In each of eleven mergers between 2004 and March 2008, the Antitrust Division concluded that the transaction was not likely to reduce competition, and therefore the agency did not sue to block the merger.29 One can debate whether or not the Division's leadership during the 2004–2008 period was more risk-averse than before the defeat in Oracle. Regardless of the answer, the statutory threshold for action to block a merger obviously did not change. Neither did the controlling case law. The Division retains the legal authority to challenge a merger that creates a reasonable possibility of lessening competition. Put more forcefully, the Antitrust Division's faithful execution of the law requires it not to nullify the incipiency principle in Section 7 through a policy of prosecutorial discretion that inclines toward inaction.
When it announced its decision not to challenge the XM–Sirius merger, the Antitrust Division invoked a different legal standard than the incipiency standard in Section 7: "After a careful and thorough review of the proposed transaction, the Division concluded that the evidence does not demonstrate that the proposed merger of XM and Sirius is likely to substantially lessen competition, and that the transaction therefore is not likely to harm consumers."30 By elevating the level of certainty of competitive harm required to justify intervention, the Antitrust Division provided itself a way to avoid challenging the merger. The "likelihood" standard that the Division uses is convenient for government litigators who fear the embarrassment of losing big cases.31 A sure way to avoid such losses for the Division in court is not to file lawsuits, ostensibly on the ground that the facts of a given case do not warrant intervention. Although most discussions of error costs in antitrust law concern false positives, this risk-aversion on the part of government antitrust litigators, particularly in the context of horizontal merger enforcement, raises the serious prospect that false negatives are plausible. This standard of prosecutorial discretion significantly weakens merger review.
|
III. MARKET DEFINITION |
|---|
|
TOPABSTRACTI. INTRODUCTIONII. THE ANTITRUST DIVISION'S...III. MARKET DEFINITIONIV. COMPETITIVE EFFECTSV. PREEMPTIVE OFFERS TO...VI. CONCLUSION |
|---|
In a March 6, 2007 filing with the Securities and Exchange Commission (SEC), XM argued that the "audio entertainment market"—which purportedly includes "free ‘over-the-air’ AM, FM, and HD radio, Internet radio, music subscription services, iPods and other MP3 players, CD players, and cell phones, as well as satellite radio"—is the relevant product market for antitrust analysis of the merger effects.32 That market definition is overly broad. A straightforward application of the Merger Guidelines test for market definition indicates that SDARS represent a distinct product market. The question then becomes whether the Merger Guidelines can accommodate market definition for a merger in a high-technology industry.
A. The Role of Demand-Side and Supply-Side Evidence in Product Market Definition
Economists agree what constitutes horizontal competition between two products, A and B.33 Product B should be included in the same market as product A if product B significantly constrains the price of product A. Such discipline can occur if, in response to an increase in the price of A, there is (1) a significant decline in the demand for A as consumers switch from A to B ("demand substitution"), or (2) a significant increase in the supply of A as firms switch production from B to A ("supply substitution"). Because supply substitution is less likely to occur in a timely fashion, it is appropriate for the Merger Guidelines to place more emphasis on demand substitution, which has a better chance of disciplining prices.
The Merger Guidelines direct that demand-side evidence shall be used to define product markets.34 This approach was reaffirmed in the recent Commentary on the Merger Guidelines released by the Department of Justice and the FTC.35 Numerous courts have used the Merger Guidelines to define product markets in this manner, and in this sense one can confidently say that these statements of prosecutorial discretion have become part of the judicially created law interpreting Section 7 of the Clayton Act.36
Although demand-side evidence is preferred to supply-side evidence under the Merger Guidelines, not all demand-side information is relevant. What matters most is evidence that buyers have altered (or would consider altering) their purchase decisions among products in response to relative changes in price.37 In the absence of direct evidence of buyer substitution, supply-side evidence may be used as a proxy for the preferences of buyers, but only to the extent that "sellers base business decisions on the prospect of buyer substitution between products in response to relative changes in price or other competitive variables."38 The term "other competitive variables" presumably connotes nonprice factors that could induce buyer substitution outside the purported product market.39 Stated differently, the Merger Guidelines dictate that supply-side evidence shall be considered in market definition only if suppliers' conduct is reflective of the reactions of consumers within the purported product market to a relative change in price or some "other competitive variables" as defined above. Thus, supplier decisions that are based on input costs or production technology are to be ignored. In addition, supplier decisions based on the expected reaction of buyers outside the purported market are also to be ignored.40
Moreover, the term "other competitive variables" must be interpreted narrowly to mean variables that reflect the demand for a product or collection of products, and not supply-side factors such as repositioning or entry. Otherwise, any supply-side information could be considered, which would undermine the broader purpose of Section 1.0 to focus on demand-side responses for purposes of market definition.
In an effort to support the proposition that SDARS customers would substitute to alternative audio sources in response to a price increase—that is, in an effort to expand the product market beyond SDARS—XM and Sirius largely relied on anecdotes of what suppliers of MP3 players, mobile telephones, terrestrial radio, and mobile Internet radio providers have been doing, allegedly in response to entry by SDARS providers.41 But these supply-side arguments say nothing about how consumers would react to a small but significant nontransitory increase in the price of SDARS. Because XM and Sirius bore the burden of proof, their repeated failure to introduce relevant evidence on this point led one to conclude that no such evidence exists.42
To support the proposition that supply-side evidence should inform market definition, XM and Sirius (through their economists) cited Jonathan Baker's recent article on the subject.43 Closer inspection of that article, however, reveals that Baker's analysis would likely reject the use of supply-side evidence to define the relevant product market in this proceeding. Indeed, Baker argues that the Merger Guidelines approach to market definition, which largely ignores supply-side evidence, is "preferable" to the methods employed by "some U.S. courts" that consider supply substitution:
Since the mid-1970s, some U.S. courts have also employed market definition to account for a second economic force, supply substitution. These courts expand markets even though a group of products and locations would appear to form a valuable monopoly after accounting for buyer substitution to outside alternatives, when the monopoly would likely not be profitable after also accounting for the incentive of outside sellers to begin producing and selling within the candidate market. The Merger Guidelines instead account for supply substitution in steps of merger analysis that take place after market definition, either in the identification of market participants or the evaluation of entry conditions. Accordingly, the argument as to whether to incorporate supply substitution in market definition is not about whether to recognize this economic force in antitrust analysis; it is over what stage of the analytical process at which to do so. The approach taken by the Merger Guidelines is preferable because it can be both difficult and confusing to ask one analytical step, market definition, to account for two economic forces, demand and supply substitution.44Baker provides an example of how supply substitution could inform market definition, but he warns that "a number of conceptual and practical pitfalls must be avoided" when doing so.45 His example involves producers of insulated aluminum conductor quickly and inexpensively switching a portion of their production capacity to the production of copper conductor. In other words, Baker's example involves entry into the same product by producers in related industries. In their merger filings, XM and Sirius did not argue that terrestrial radio broadcasters, or any alternative audio providers for that matter, were contemplating acquiring spectrum and offering satellite radio services. Thus, the supply-side evidence offered by XM and Sirius would not be consistent with Baker's interpretation of the Merger Guidelines.
There is no principled reason to abandon the unambiguous prescription of the Merger Guidelines to focus solely on demand-side factors when defining the relevant product market. Doing so would invite obfuscation of the determinative economic issues in this proceeding and disserve consumers in a wider range of telecommunications markets by creating a precedent that would confound proper antitrust analysis in future mergers. Indeed, in the last six high-profile mergers reviewed by the FCC, supply-side evidence did not inform market definition.46 This is not to say that supply-side evidence serves no purpose in merger analysis. Once the relevant market has been properly defined in accordance with the Merger Guidelines (based exclusively on demand-side evidence), supply-side factors can be used to identify firms that participate in the relevant market (and their shares) and to evaluate the likelihood and extent of entry.
B. Demand-Side Evidence in the XM–Sirius Merger Proceeding
In this section, we analyze the demand-side evidence that was introduced in the public record during the XM–Sirius merger proceeding.
1. How Sensitive Are Satellite Digital Radio Subscribers to Price Increases?
If the actual own-price elasticity of demand is less than the critical level, then SDARS represent a distinct product market according to the Merger Guidelines. The own-price elasticity of demand measures the availability of close substitutes: if there are few viable alternatives, then the own-price elasticity of demand is small in absolute terms. According to Bernstein Research, SDARS enjoy what economists call an "early mover's advantage" over its potential rivals:
XM and Sirius also have a considerable head start on any new service, and their established brands, distribution relationships, promotional and marketing clout, high customer satisfaction and relatively inexpensive price points are likely to limit the number of consumers who would choose a competing offering.47
To the extent that SDARS consumers would be highly reluctant to switch to alternatives in response to a price increase, the own-price elasticity of demand for SDARS is likely to be less than the critical level.
Several pieces of evidence suggest that the elasticity of demand for SDARS is not highly sensitive. On April 2, 2005, XM increased its monthly price from $9.99 to $12.95 to bring its price in line with the price of Sirius—an increase of nearly 30 percent.48 In the two quarters following the price increase, XM realized subscriber growth of 13 percent (third quarter 2005) and 20 percent (fourth quarter 2005).49 The fact that subscriber growth continued at such a rapid pace in the presence of 30 percent price increase underscores the low elasticity of demand faced by SDARS providers.50
Second, the churn rate for SDARS is less than two percent, which Sirius says is the lowest among all subscription-based services.51 Bernstein Research noted in July 2005 that XM "saw no increase in churn, despite a 30 percent price increase taken at the start of the [second] quarter [of 2005]."52 Sirius's chief executive attributed the low churn to the fact that "[Sirius's] programming is so compelling, and so sticky, and so strong."53 Another reason for the low churn rate is high switching costs for the closest available substitute. If a SDARS customer wishes to substitute to HD radio, he or she must purchase new hardware, which currently costs $200—or roughly the equivalent of fifteen months of SDARS at the current monthly price of $12.95.54 According to Bernstein Research, the churn rate for Sirius was 1.4 percent in 2006, while the churn rate for XM's self-paying customers was the same.55 The extremely low churn rate for SDARS suggests that substitution possibilities for SDARS customers are lacking, which implies highly inelastic demand.56
In addition to low churn rates, another indicator of inelastic demand for SDARS is the high "conversion rate."57 The conversion rate is defined as the percentage of customers who sign a contract with an SDARS provider after sampling the service for three months free of charge. During 2003, XM was able to convert nearly three-quarters of all customers who were on a three-month free trial.58 During 2004 through 2005, the conversion rate decreased to 60 percent,59 yet was still impressive. The high conversion rate suggests that SDARS customers would not easily substitute toward another radio service in response to a small price increase for SDARS.
As we demonstrate below, the marquee content offered by SDARS is generally prohibited on terrestrial broadcast radio due to indecency standards. The demand for indecent content is widely considered to be price inelastic.60 For example, evidence indicates that the demand for adult-oriented entertainment is highly price inelastic. Pay-per-view adult entertainment on cable systems, for instance, garners some of the highest profit margins of any programming. Some analysts claim margins for cable or direct broadcast satellite operators of up to 80 percent on each purchase.61 Other studies show price-inelastic demand for indecent content on the Internet.62 This inelastic demand means that most current consumers of indecent content are "inframarginal" consumers who will tolerate a price increase. Although such content may compete weakly against Playboy magazine and other indecent content consumed in the privacy of one's home, indecent content delivered over the radio is distinguishable because it can be consumed in the car, while driving, and in remote geographic locations.
2. Previous Regulatory and Antitrust Proceedings Regarding Subscriber-based Programming Markets
In contrast to how it has regulated terrestrial broadcast radio and television, the FCC has consistently declined to extend indecency enforcement to subscriber-based services like SDARS or cable television. This regulatory asymmetry facilitates market division between satellite radio and terrestrial radio. In this section, we analyze the current state of indecency regulation and the demand for SDARS that would be vulnerable to indecency enforcement if aired over terrestrial broadcast radio. The FCC's decision not to extend its indecency standards to SDARS has allowed for a market segmentation to occur between SDARS and terrestrial broadcast radio.63
a. The FCC's Indecency Standards
The FCC has the authority, under Section 1464 of the U.S. criminal code, to regulate "obscene, indecent, or profane language" transmitted "by means of radio communication."64 Commission regulations bar the terrestrial broadcast of indecent content between the hours of 6 a.m. and 10 p.m.65 The Commission defines indecency as material that, in context, depicts or describes "sexual or excretory activities or organs" and is "patently offensive" under "contemporary community standards" for the broadcast medium.66 The FCC is empowered to assess forfeiture penalties, and may initiate license revocation proceedings or deny license renewal for violations.67
In recent years, one of the most significant Commission actions in response to indecent content concerned not a video image, but instead the audio portion of a national television broadcast. In January 2003, Bono, lead singer of the band U2, used profanity during a live broadcast of the Golden Globe Awards.68 The Commission's Enforcement Bureau initially found that no violation of law had occurred; it ruled that an isolated or fleeting expletive, used as an intensifying adjective rather than as a noun or verb, will not render a broadcast indecent.69 The full Commission reversed the Enforcement Bureau in early 2004.70 Several aspects of the Commission's ruling are particularly salient to the application of indecency regulation to the terrestrial broadcast radio market. In particular, the Commission found that the "F-Word," even when used as an intensifying adjective or insult, carries inherently sexual connotations and thus will always satisfy the first prong of the indecency analysis.71 Second, the Commission expressly overturned prior law, finding that even isolated uses of the "F-Word" may violate the second "patently offensive" prong of indecency analysis.72 The full Commission in Golden Globe also recognized a new and independent ground for liability: that the use of expletives, irrespective of their sexual or excretory connotation, may be a "profane" broadcast under Section 1464.73 This statutory interpretation by the FCC substantially expands potential liability for terrestrial radio broadcasters, especially for "shock jock" talk radio.
Since Golden Globe, the FCC has increased indecency enforcement against broadcasters. From 1995 to 2002, total annual notices of apparent liability (NALs) never exceeded $100,000.74 In 2003, NALs increased to $440,000.75 By 2004, they reached $8 million.76 In 2006, NALs reached almost $4 million.77 In addition, the Commission entered into three consent decrees in 2004, totaling almost $3.5 million.78 Significant actions during this period included a $550,000 fine for broadcast of Janet Jackson's performance during the Super Bowl XXXVIII Halftime Show, a $1.2 million fine for an episode of Married By America on the Fox Television Network, and the $3.6 million fine for an episode of Without A Trace on CBS—the largest in Commission history.79
Broadcast radio has also faced sizeable fines as recently as April 8, 2004, including a $495,000 NAL against Clear Channel Communications for an episode of the Howard Stern Show, a $755,000 NAL again against Clear Channel for a broadcast by radio host "Bubba the Love Sponge," and $357,000 in liability against Infinity Broadcasting for an episode of the Opie & Anthony Show.80 Notably, all of these controversial radio hosts are now offered on satellite radio.81 In late 2006, Congress passed the Broadcast Decency Enforcement Act, which raised potential fines to $325,000 per violation, or per day for a continuing violation.82 The legislation provides for a maximum fine of $3 million for a continuing violation. Broadcasters have responded with tough internal indecency guidelines and have invested in time-delay technology that allows them to censor potentially indecent broadcasts.83
SDARS providers, however, are not subject to these—or any other—indecency rules. The FCC has repeatedly refused to extend its indecency regime to subscription-based programming, such as direct broadcast satellite (DBS) service.84 In evaluating DBS, the FCC concluded that the rationales justifying regulation of indecent content distributed over traditional broadcast television do not apply. Indeed, DBS is more closely analogous to cable television than it is to broadcast television. As such, the regulation of content delivered through subscription-based multichannel video platforms draws separate treatment by the FCC and heightened First Amendment scrutiny by reviewing courts.
In 2001, the FCC drew analogies to its experience with direct broadcast satellite (DBS) service. In determining the applicability of the Communications Act's provisions on foreign ownership to SDARS during a hearing involving Sirius (then known as Satellite CD Radio), the FCC concluded:
We agree ... that the issues regarding foreign ownership for DBS and SDARS are virtually identical and thus we affirm the Bureau's determination that Section 310(b) of the Communications Act does not apply to subscription SDARS licenses because the service offered is neither broadcast, common carrier, aeronautical en route or aeronautical fixed service.85
Thus, the FCC made clear that XM and Sirius did not fit within the existing regulatory pigeonhole of radio broadcasting. Rather, SDARS is a distinctly different medium, warranting separate application of content-based regulation.
In 2004, Mt. Wilson FM Broadcasters asked the Commission to apply the indecency rules to SDARS. The Commission declined to do so, saying that, "[c]onsistent with existing case law, the Commission does not impose regulations regarding indecency on services lacking the indiscriminate access to children that characterizes broadcasting."86 Clearly, any extension of the regulations to subscriber-based radio would be highly vulnerable to constitutional challenge.87 In upholding the constitutionality of legislation permitting the regulation of indecent broadcast content, the Supreme Court has, since the Pacifica decision in 1978, focused on the following governmental interests: pervasiveness of the media, its unique accessibility to children, and the fact that unwilling listeners or viewers can happen upon indecent material while tuning their radios or televisions.88 Compared to terrestrial broadcast radio, satellite radio is less pervasive because it is a subscription-based service. Satellite radio also affords far more listener control than does terrestrial broadcast radio. In addition to requiring consumers to subscribe to the content, both XM and Sirius have measures in place that empower users to decide when they will encounter adult material. XM, for instance, denotes stations that frequently feature explicit language with an "XL" and allows users to block them.89 Sirius permits channel blocking and requires listeners to opt-in to receive Playboy Radio.90
Courts assessing the applicability of existing indecency statutes and regulations to SDARS would likely analogize the service to cable television. Unlike its First Amendment decisions concerning the broadcast media, the Supreme Court's decisions concerning the constitutionality of content-based cable regulations have applied strict scrutiny.91 If the Court recognizes voluntary channel blocking—offered by both Sirius and XM—as a less restrictive alternative to content restrictions, then the application of existing broadcast-based indecency regulation to SDARS would surely be held to be unconstitutional.
b. Would a Significant Portion of the Satellite Digital Audio Radio Content Be Deemed Indecent in a Broadcast Environment?
Surveys of XM's and Sirius's channel offerings show a selection of programming that frequently contains explicit discussion of sexual or excretory activities or organs, or extensive profanity. Table 1 lists the relevant channels. As Table 1 shows, a significant number of popular channels on both XM and Sirius contain indecent or profane material.
|
In contrast to XM and Sirius, terrestrial radio broadcasters have been self-censoring material that, before the FCC's the increase in indecency enforcement, would almost certainly have been aired unedited. For example, radio stations have pulled or edited Lou Reed's "Walk on the Wild Side" and Steve Miller's "Jet Airliner"—iconic rock songs that radio broadcasters had aired unedited for more than a generation.93 Other stations have instituted zero-tolerance policies for on-air talent, prompting some personalities to take out "indecency insurance."94 Only fourteen of 300 public television stations aired an unedited version of a documentary on the war in Iraq, in which soldiers swore while under fire.95 During the 2006 Super Bowl halftime show, the Rolling Stones were bleeped twice by the network, once during "Start Me Up" (a song previously played uncensored on broadcast radio since its release in 1981) and again during a new song, "Rough Justice."96
Strong evidence indicates that indecent content attracts a significant portion of the paying audience for SDARS. For instance, XM's CEO has identified the Opie & Anthony Show and XM's comedy channels as among its most popular.97 Playboy Radio, which requires subscribers to opt-in, reportedly drew more than 1 million customers to Sirius over three months.98 Equity analysts have documented the growth in Sirius subscriptions following the addition of Howard Stern to its lineup.99 Stern, who precipitated numerous FCC indecency enforcement actions in the past,100 left terrestrial radio in 2006 after signing a five-year, $500 million contract with Sirius.101 Stern specifically cited the freedom from indecency regulations on satellite radio as the reason for his decision to switch to a different distribution platform for his show.102
In the approximately two years since Stern announced that he would leave terrestrial radio, Sirius's subscriber base increased from less than 700,000 to more than 6 million.103 Analysts attribute between 1 and 2 million of these subscribers to Stern himself.104 Sirius paid Stern bonuses totaling $219 million in 2006 and $83 million in 2007 after Sirius exceeded the subscription targets specified in Stern's contract.105 In 2006, Sirius announced its acquisition of the rights to more than 23,000 hours of Stern programming, which it intends to air unedited.106 In its annual report filed in March 2006 with the Securities and Exchange Commission, XM specifically identified Stern as a possible competitive threat.107 In their SEC filings, both XM and Sirius identified their uncensored programs as marquee content.108
c. The Treatment of Advertisement-based Broadcast Services and Subscription-based Services by Other Federal Agencies and Courts
The FCC, the DOJ, and the federal courts have identified factors that implicitly or explicitly segment media programming product markets between advertisement-based broadcast and subscription-based services, and those factors apply equally to television and radio.109 In proposed mergers and acquisitions among broadcast radio station operators, the DOJ has regarded broadcast radio as a separate and relevant product market.110
As modern subscription-based programming evolved, Congress recognized its competitive implications, as evidenced in the Cable Television Consumer Protection and Competition Act of 1992.111 The Act's findings reflected Congress' position that cable television in general constituted a separate product when compared to broadcast television, so much so that cable's existence threatened that of broadcast: "As a result of the growth of cable television, there has been a marked shift in market share from broadcast television to cable television services."112 The Act recognized that the broadcast medium could not effectively compete with the emerging and increasingly popular multichannel subscription-based services, declaring that "without the presence of another multichannel video programming distributor, a cable system faces no local competition. The result is undue market power for the cable operator as compared to that of consumers and video programmers."113 The regulatory remedy that Congress created was a "must carry" provision that requires cable providers to devote channel capacity to local broadcast television stations.114 In effect, the must-carry provision creates a legislatively mandated duty to deal to preserve the existence of television broadcasters as suppliers of local content. Such an arrangement inherently involves two distinct product markets—one (the market for origination of local content) that Congress feared could not survive without being assured access to the other (the market for multichannel video program distribution). In this merger case, broadcast radio is analogous to broadcast television, and SDARS is analogous to cable television. The analogy does not imply that the two separate markets (terrestrial radio and SDARS) will interact in the same way that broadcast and cable television have, but merely that the separate markets exist for similar reasons, and a monopoly in either market threatens consumer welfare.
The FCC was prepared to make key distinctions that separated SDARS from terrestrial radio, drawing direct analogies from its experience with subscription-based television. In a 1987 proceeding "to determine what criteria may be used by the Commission to determine whether a communications service should be treated as ‘broadcasting’ under the Communications Act,"115 the FCC found that "the definition of ‘broadcasting’ ... was intended to differentiate between services intended to be received by an indiscriminate public and those intended only for specific receive points,"116 and that "transmissions designed to be available only to paying subscribers clearly demonstrate the intent of the licensee."117 Thus, the FCC found that subscription-based television service was not a form of broadcasting, and so the subscription-based service was not subject to existing regulations governing broadcast media: "[I]n all cases, the purveyor and its audience are engaged in a private contractual relationship. That relationship, enforced by the need for special equipment and/or decoders, obviates the need for the traditional broadcast type regulation that has been developed over the past 40 years."118
In 1997, the FCC authorized two licensees, Sirius and XM, "to launch and operate satellites to provide SDARS."119 From the beginning, the FCC treated SDARS differently from terrestrial radio broadcasting. In a portion of its 2001 notice for granting licenses to XM and Sirius, the FCC highlighted the exclusivity of the two companies that would occupy a reserved portion of the spectrum, making no reference to terrestrial radio: "There are only two SDARS providers authorized to provide service in the DARS spectrum band, XM Radio, Inc. and Sirius Satellite Radio, Inc."120 This exclusivity implies that entry by a third SDARS provider would be costly.
When Clear Channel proposed to merge with AMFM in 2000, the DOJ issued a competitive impact statement, declaring:
Clear Channel and AMFM are two of the three largest operators of broadcast radio stations in the United States. Clear Channel's and AMFM's radio stations compete head-to-head against one another for the business of local and national companies seeking to advertise on radio stations in many cities throughout the United States, including Allentown, Pennsylvania; Denver, Colorado; Harrisburg, Pennsylvania; Houston, Texas; and Pensacola, Florida.121
The DOJ specifically found the relevant product market to be radio-based advertising.122 Thus, even if SDARS had existed as a viable force at the time, it would not have been included in the relevant product market.
3. Do Satellite Digital Audio Radio Subscribers Perceive Other Forms of Audio Services to be Close Substitutes?
Some commentators to the XM–Sirius merger proceeding argued that the relevant product market for purposes of analyzing the merger should contain an array of services in addition to SDARS.123 However, those arguments are not pervasive as a matter of antitrust analysis. The weak substitution possibilities for current SDARS customers imply that a hypothetical monopoly provider of SDARS could profitably impose a SSNIP. Sirius's own website included a press release that emphasized that, from the consumer perspective, its SDARS product bears little resemblance to terrestrial radio:
Currently, SIRIUS utilizes its satellite broadcast technology to transmit 100 digital ‘streams’ of entertainment that include 60 streams of 100% commercial-free music, and 40 streams of news, sports, and entertainment for $12.95 per month. Unlike today's radio channels, these digital streams from SIRIUS can also carry video signals or other data.124
Similar claims could be found on XM's website.125 This press release emphasized the absence of commercials, ubiquity, and large number of channels as the characteristics that distinguish SDARS from terrestrial radio broadcasts. Through such statements, XM and Sirius manifested their own belief that consumers view SDARS as significantly different from terrestrial radio.
a. Advertiser-supported Terrestrial Radio
It is a mistake to think that SDARS subscribers would substitute to "free" terrestrial radio broadcasting in response to a SSNIP. Instead, the effective price for a given subscriber of advertiser-supported radio is the reduction in utility associated with having to endure commercials. Not surprisingly, evidence suggests that advertiser-supported terrestrial radio is able to compete only weakly with SDARS by reducing commercial time. For example, at the end of 2004, Clear Channel decided to cut its ad time and reduce the length of commercial spots from 60 to 30 seconds in an attempt to "win back listeners, boost ratings, and in turn lead to higher ad rates."126 According to Forrester Research, the success of SDARS partly reflects listeners' desire to avoid advertising.127
Even for SDARS subscribers who are willing to endure commercials, the number of terrestrially delivered radio stations available in any given geographic market is severely constrained relative to the number of channels available on SDARS. In 2000, there were only 47 terrestrial radio stations as listed by Arbitron broadcasting in New York City; in many metropolitan areas outside the largest 50 markets (such as Jacksonville, Louisville, and Oklahoma City) there are 30 or fewer terrestrial radio stations as listed by Arbitron.128 Bernstein Research believes that digital terrestrial radio "poses little threat to the growth in satellite radio subscriptions" because it "cannot address four key factors that drive consumer adoption of satellite radio: commercial-free music; a large range of channels in a variety of formats; exclusive programming; and satellite radio's distribution advantage as the auto OEMs [original equipment manufacturer]."129
Unlike SDARS, advertiser-supported terrestrial radio stations lack a ubiquitous footprint. XM's nationwide service can reach nearly 100 million listeners age 12 and over who are beyond the range of the largest 50 markets as measured by Arbitron.130 XM estimated that, of these 100 million listeners, 36 million live beyond the largest 276 Arbitron markets.131 XM also estimated that 22 million people age 12 and older receive five or fewer stations.132 A significant percentage of radio listeners, such as truckers (who numbered roughly 3 million in 2004),133 routinely travel through two or more Arbitron radio markets on a frequent basis.134 Those consumers clearly would not perceive terrestrial service to be a reasonable substitute for SDARS.
Much of the marquee content on SDARS would be considered indecent if delivered via broadcast radio. In other words, by regulatory constraint consumers cannot turn to terrestrial radio broadcast to receive such content. Regulation constrains demand substitutability between terrestrial radio and SDARS. Not only does satellite radio offer a much broader range of content, far fewer commercials, integration with other communications technology, often better quality sound, and national coverage, it also offers content that is unavailable on terrestrial radio—namely material that would invite indecency enforcement if aired over terrestrial broadcast radio outside the safe harbor period permitted by the FCC.
Finally, new survey data suggest that satellite radio subscribers do not perceive terrestrial radio to be a close substitute for satellite radio. In June 2007, Wilson Research Strategies conducted a survey of current satellite radio subscribers at the request of the NAB. The survey polled 501 current SDARS subscribers on a range of questions on their reasons for subscribing and their demographic characteristics.135 The survey results suggest that a significant number of satellite radio subscribers (1) are less likely to have a sufficient amount of terrestrial radio service by virtue of their geographic location, (2) value certain attributes of satellite radio that are not available on terrestrial radio, and (3) do not perceive MP3 players to be substitutes for satellite radio.
The survey data confirm that a majority of satellite radio subscribers reside in a small city, town, or rural area. Because local radio coverage declines with the size of the local population, this fact suggests that satellite radio subscribers reside in areas of below-average terrestrial radio coverage. The majority (58 percent) indicated that they lived away from a large city.136 This finding suggests that many XM and Sirius subscribers would be vulnerable to an increase in the price of satellite radio.
The survey data suggest that satellite subscribers value SDARS for qualities that are unavailable on terrestrial radio. These qualities include commercial-free music, uninterrupted signal, and greater number of channels. According to the survey, 87 percent of respondents listed commercial-free music as an "important" reason for subscribing; 77 percent of satellite subscribers cited "uninterrupted signal nationwide" as an "important" reason for subscribing; and another 77 percent identified "number of channels" as an "important" reason for subscribing.137 Because these features are not available on terrestrial radio, it is reasonable to infer that terrestrial radio does not constrain the price of satellite radio.
Finally, the survey shows that a majority (53 percent) of satellite subscribers own or use an MP3 player.138 Thus, most satellite subscribers are aware of MP3 players and do not perceive them as a substitute for satellite radio, because they continue to subscribe to XM or Sirius. Satellite subscribers more likely view MP3 players and satellite radio as distinct products used for different purposes.
b. HD Radio
HD radio is a technology that allows for digital transmission of AM and FM terrestrial broadcasts on the same frequencies on which they are currently broadcast.139 For several reasons, HD radio is not likely to constrain the pricing of SDARS. First, like analog radio, HD radio suffers from a limited national footprint. BusinessWeek has projected that only 2500 of the nation's 13,000 commercial radio stations will be digital by 2010.140 Because not all terrestrial stations have launched HD service, the footprint of HD signals is a subset of the footprint of terrestrial radio.
Second, HD radio currently lacks unique or compelling content.141 In its current form, it is merely a parallel broadcast of analog terrestrial radio signals. HD radio is also subject to the same indecency standards as conventional broadcast radio, which prevents HD radio from offering indecent content. Moreover, much of the marquee content available on SDARS is under exclusive contracts with XM or Sirius.
Third, HD radio requires high upfront costs for consumers. HD receivers currently cost at least $200.142 Thus, potential marginal SDARS customers would have to incur a nontrivial switching cost as a penalty for substituting to HD radio. High switching costs imply that a SSNIP by a hypothetical monopoly provider of SDARS is more likely to be profitable.
The opinion that SDARS are distinct from HD radio services is corroborated by industry analysts, who believe that HD radio is not a viable alternative to SDARS:
Seven terrestrial radio companies announced yesterday that they had formed a partnership to accelerate the rollout of digital radio (based on the "HD Radio" format developed by Ibiquity). Although we believe that this is a step in the right direction for digital radio, we continue to believe that digital terrestrial radio poses little threat to the growth in satellite radio subscriptions.143
Bernstein Research also explains that HD radio cannot compete effectively with SDARS due to satellite radio's distribution advantage with automobile manufacturers.144 Finally, Bernstein Research notes that the entry barriers for radio stations are significant, which should also limit substitution possibilities. In particular, the average HD conversion costs were $100,000 in 2005.145 As of December 2005, only 600 stations of a total of more than 13,000 radio stations (4.6 percent) had been upgraded to the HD radio transmission format.146 By the end of 2006, only 1300 stations (10 percent) were expected to have converted to digital.147
c. Podcasts Delivered over an iPod
Podcasts are broadcasts downloaded to an MP3 player for later use.148 Unlike SDARS, podcasts are not delivered in real time. SDARS are superior for consumers whose time is too scarce to load new songs onto an iPod and create new playlists. The programming on SDARS is constantly updated. Second, the docking technology for iPods in automobiles is cumbersome and prone to interference. In contrast, the SDARS device is built into the car or installed by a dealer. According to Bernstein Research, the "cross-price elasticity of demand between the two platforms [podcasts and SDARS] is likely overstated, and satellite radio has a number of advantages over iPods in cars. In our view, the two are likely to be more complementary."149 Former FCC Chief Economist Dr. Gerald Faulhaber explains the critical difference between an iPod and satellite radio:
With satellite radio, they do programming; they're real programmers. They offer a choice of formats. With iPod, you're picking your own music and that's fine but it's a different experience. They also do not have the personalities on iPod that they do on XM and Sirius radio.150
Based on those differences, Gerald Faulhaber concludes that "the iPod is a very different service than Satellite radio."151
d. Mobile Internet Radio
Mobile Internet radio provides for programming delivered over the Internet and to the end user through a mobile phone. Mobile Internet radio is not a close consumer substitute to SDARS for at least three reasons. First, Internet radio lacks the ubiquity of SDARS. Mobile Internet radio requires a connection to the Internet, most often through a cellular telephone network. Current cellular networks lack the ubiquity of SDARS for even the most basic voice services, let alone 3G data services.152 Indeed, analysts predict that wireless networks will never have service areas that are comparable to SDARS.153
Second, the quality of mobile Internet radio is significantly inferior to SDARS. A Harvard Business School case study concluded that mobile Internet radio had noticeably inferior audio quality.154 In an article in PC Magazine, Bill Machrone, vice president of technology at Ziff Davis Publishing, also questioned the quality of internet radio.155 In contrast, SDARS received high customer satisfaction levels.156 Accordingly, analysts have been skeptical of the near-term economic viability of mobile Internet radio.157
Third, mobile Internet radio is more expensive than SDARS because mobile Internet radio combines the direct cost of a subscription and, in most cases, the consumer's imputed time cost of listening to commercials. A network connection for in-car Internet radio is expensive. As of February 23, 2007, the cheapest monthly data connection capable of supporting Internet radio from Cingular Wireless was $44.99.158 In addition to the out-of-pocket costs of connecting to Internet radio, "free" Internet radio relies on advertisements for revenue.159 Moreover, all wireless operators limit the amount of downloading per month, even under their "unlimited" plans.160 As we explained above, advertisements impose a real cost on consumers and should not be viewed as costless.
An examination of a proposed Internet radio offering from Sprint-Nextel reveals the inferiority of mobile Internet radio to SDARS. In September 2005, Sprint-Nextel announced a joint venture with RealNetworks to offer six music channels including 1970s and Country (similar in format to SDARS) and at least one streaming radio station for $16.95 a month (equal to a $6.95 service fee with a minimum $10 data plan).161 In contrast, SDARS offer over 100 channels of music at $12.95 a month.
e. Music Services Available on Wireless Telephones
XM and Sirius recounted the latest offers by other mobile telephone providers to argue that wireless telephones should be included in the relevant product market.162 However, SDARS customers are not likely to perceive these offerings to be close substitutes to SDARS. To borrow one obvious example, it is not clear how a sports program downloaded onto a Verizon VCast mobile phone could be played over the speaker in one's car. XM and Sirius failed to link Verizon's VCast or any of these offerings to anticipated demand-side substitution among SDARS subscribers in response to price changes. Stated differently, XM and Sirius cannot reject the hypothesis that these offerings came about completely independently of how the wireless carriers perceive the demand response of SDARS customers. Instead, it seems far more likely that these carriers were motivated by a desire to capture ancillary revenues in the upstream content markets, primarily music downloads. In other words, Verizon's VCast is a closer substitute to the iPod than it is to SDARS.
Moreover, at the current prices sought by wireless carriers for audio content, it is highly doubtful that SDARS customers perceive these mobile telephone offerings to be close substitutes. A mobile voice subscriber to Sprint163 or AT&T164 must subscribe to an unlimited data package—priced between $20 and $50 per month, depending on the carrier—to avoid paying for data usage charges while listening to audio content over his mobile telephone. Setting aside the nontrivial incremental price for an unlimited data plan, Sprint offers ten commercial-free stations for $15 per month or 40 commercial-free stations for $20.165 By comparison, XM and Sirius each offer over 120 commercial-free stations for $12.99. A Sprint subscriber paying $20 per month for audio content would need to incur an additional $6.95 (a total of $26.95) to receive 20 Sirius channels. Because a Sirius subscriber already receives these channels and more in his satellite radio subscription, he would never be willing to substitute to Sprint's audio entertainment service for a higher price ($26.95 plus the unlimited data charge versus $12.95). Similarly, setting aside the price for an unlimited data plan, an AT&T customer must pay $8.99 per month to receive a small subset (25) of XM's channels. It is not credible that an XM customer would pay significantly more than $12.99 per month to forgo over 100 XM channels and the ability to listen to XM radio in his car.
Although Verizon offers MLB games (at $6.95 per month) and music downloads (at $1 per download) through its VCast service, Verizon does not offer a package of commercial-free stations for a monthly fee. Again, an XM subscriber, who receives MLB games under his current subscription for $12.99 per month, would not be willing to switch to Verizon VCast only to pay for an unlimited data plan ($50 per month) plus $6.95 per month for MLB and forgo over 120 channels. Table 2 summarizes these alternative audio entertainment offerings provided by mobile wireless operators.
|
As Table 2 shows, the monthly price differential between an SDARS subscription and any of the audio content offerings from mobile telephone operators is substantial, ranging from $4 (Alltel) to $24.98 (AT&T). It bears emphasis that two of the largest mobile telephone operators, Verizon and T-Mobile, do not even offer a base plan with a fixed number of audio channels. If these are the best options facing SDARS customers, then the unilateral price effects of the merger would be severe indeed.
Through their economists, XM and Sirius introduced a novel, theoretical concept called "dynamic demand" that would obscure market definition analysis and supposedly legitimate an unprecedented efficiency defense that is not cognizable under the Merger Guidelines.
XM and Sirius claimed that the standard SSNIP test used for market definition is inappropriate here because it ignores the long-term profitability considerations faced by SDARS providers:
We will explain why the "small but significant and nontransitory increase in price" (ssnip) test for market definition from the antitrust agencies' Horizontal Merger Guidelines must take into account the dynamic nature of demand and the important role of longer-term profit-maximization for Sirius and XM.166
Presumably, XM and Sirius would alter the standard SSNIP calculus—namely, a comparison of short-term profits before and after a price increase—by including additional terms for the hypothetical monopolist's long-term profits. XM and Sirius failed to cite any instance in which a court or an agency altered the SSNIP test in this way. Indeed, in the last six high-profile mergers reviewed by the Commission, the SSNIP test was applied without any alterations.167 XM and Sirius failed to provide an economic basis for its recommendation that the FCC deviate from the Merger Guidelines in such a fundamental way.
XM's and Sirius's novel "dynamic demand" analysis is wholly theoretical. Nowhere did the merger parties articulate the conditions that would have had to exist for the analysis to be applicable, let alone whether such conditions were in fact present. The "dynamic demand" concept provided no basis to claim that the postmerger dynamically optimal price will not be higher. There is no precedent for deviating from the Merger Guidelines by incorporating a concept that would vitiate the standard SSNIP test.168
2. The "Dynamic-Demand-Spillover" Problem
XM and Sirius further postulated that competition between the two satellite radio providers creates a significant impediment (a "dynamic demand spillover") to lower prices and better quality that would be eliminated by this merger. This "dynamic demand spillover" encourages free riding by XM and Sirius, which allegedly undermines each provider's incentive to engage in "demand-enhancing investments, such as mounting advertising campaigns, improving the quality of its products and services, and investing in low penetration prices."169 But the merger parties failed to provide an analysis of how many resources (if any) were being held back by XM and Sirius due to this hypothesized free-rider problem. They also failed to provide an analysis of how much consumers would benefit from continued rivalry between XM and Sirius.
Moreover, XM's and Sirius's "dynamic demand spillover" conjecture was inconsistent with its market definition position. It was not consistent for XM and Sirius to argue on the one hand that the other types of audio entertainment compete with SDARS, but on the other that the merger would solve the problem of "dynamic demand spillover."170 XM and Sirius have neglected to consider that, after the merger, the alternative audio entertainment devices that allegedly compete with SDARS will still be able to free ride on the "demand-enhancing" investments made by a combined XM–Sirius. Alternatively, if the "dynamic demand spillover" is truly specific to the two SDARS providers (such that there is no spillover to other audio entertainment services), then one must conclude that those alternatives are not in the same product market. Stated differently, if there is a newly created incentive after the merger to engage in penetration pricing and promotions, then it must be the case that iPods and HD radio do not compete with SDARS; otherwise the "demand-enhancing" investments that would occur after the merger would still generate demand for iPods and HD radio.
In summary, the "externality" that XM and Sirius invoked is properly described as product differentiation, and it is precisely the force that is constraining the price of SDARS today. The merger can be counted on to "solve" this "competition problem" between XM and Sirius. But the result will be higher SDARS prices (in the absence of a price-freeze concession). For that reason, the externality problem should have been ignored.
3. What Is the Proper Role for Novel Economic Theories of Antitrust Analysis?
In its supplemental report, XM and Sirius aimed to resuscitate its dynamic demand arguments.171 Yet XM and Sirius were not able to identify a single instance in which dynamic demand considerations had been recognized by an antitrust court or agency during a merger review. The closest that the merging parties came to satisfying this burden were citations to the "economics literature and the marketing literature."172 Although several abstract theories have been developed by economists over the years, none of them serves as a basis for deviating so radically from the Merger Guidelines.
Put simply, the relevant question for the Commission was whether it was ready to depart from recognized antitrust analysis in light of a novel theory that could have some bearing on merger analysis but had not yet been recognized by any antitrust authority. In an effort to build precedence for such a radical approach, XM and Sirius cited language from the Merger Guidelines, Merger Commentary, and the AMC report,173 each of which admittedly tolerates some "flexibility" in merger analysis.174 Indeed, the very quote provided by XM and Sirius admits exactly where the AMC is willing to entertain new economic theories:
In industries in which innovation, intellectual property, and technological change are central features, just as in other industries, antitrust enforcers should carefully consider market dynamics in assessing competitive effects and should ensure proper attention to economic and other characteristics of particular industries that may, depending on the facts at issue, have an im