June 29, 2018
On June 25, 2018, the U.S. Supreme Court, in a 5-4 decision by Justice Thomas, held that provisions in American Express Company’s (“American Express”) and its operating subsidiary’s contracts with merchants that restricted the ability of these merchants to steer customers to other credit or charge cards did not violate the Sherman Act. Ohio v. American Express Co., 585 U.S. __, slip op. at 1 (2018). The Court held that plaintiffs—the United States Department of Justice and the Attorneys General of several states—failed to satisfy their burden of proving anticompetitive effects in the relevant market under the rule of reason. Id. at 10. The ruling has important implications for antitrust analysis, not only for the credit card industry, but for other industries that operate in two-sided markets where firms must compete simultaneously for different groups of customers whose demands are distinct but deeply interrelated.
American Express contracts with merchants to accept charges on American Express cards as payment for goods and services in return for American Express’s agreement to reimburse merchants for those charges, minus a designated merchant discount fee that American Express retains as compensation for its services. American Express’s acceptance contracts with merchants typically contain some form of non-discrimination provisions (“NDPs”) in which the merchant agrees not to discriminate against American Express by, inter alia, indicating a preference for another credit card or attempting to dissuade cardholders from using the American Express card (collectively, “steering”). United States v. American Express Co., 838 F.3d 179, 191 (2d Cir. 2016) (describing American Express’s standard form NDPs).
The DOJ, along with 17 state Attorneys General, challenged these agreements under Section One of the Sherman Act, alleging that American Express’s NDPs unreasonably restrained competition in the alleged “network services market” because, by inhibiting merchants from steering consumers to use cards with a lower cost to the merchant, they reduced the incentive of credit card networks to reduce merchant fees because reducing merchant fees would not necessarily result in greater volume. Id. at 192. After a lengthy bench trial, the District Court agreed with the DOJ, finding that the NDPs were unreasonable restraints on competition in violation of the Sherman Act. United States v. American Express Co., 88 F.Supp. 3d 143 (E.D.N.Y. 2015). Based on this finding, the court entered a sweeping injunction requested by the plaintiffs that not only prohibited American Express from enforcing the contractual NDPs, but prohibited American Express from unilaterally treating merchants differently based on whether they steered or not (i.e., the injunction prohibited not only the challenged contractual restraints, but any unilateral action by American Express based on a merchant steering or not steering American Express cardholders). United States v. American Express Co., 2015 WL 1966362 (E.D.N.Y. Apr. 30, 2015). On appeal, the Court of Appeals reversed the District Court’s order and directed the District Court to enter judgment for American Express, rejecting the District Court’s findings on market definition, and holding that the plaintiffs had failed to prove the NDPs violated the Sherman Act.
The parties all agreed that the NDPs are vertical restraints subject to analysis under the rule of reason. Thus, the Supreme Court began its analysis with market definition. The District Court, relying largely on United States v. Visa USA, Inc., 344 F.3d 229 (2d Cir. 2003), had found that the relevant market was for the provision of “network services,” meaning the “core enabling functions provided by networks, which allow merchants to capture, authorize, and settle transactions for customers who elect to pay with their credit or charge card.” 88 F. Supp. at 171. In doing so, the District Court had expressly rejected the idea that the market should be evaluated as a “single platform-wide market for transactions” that encompassed both merchant and consumer interactions, finding that this analysis would have taken “the concept of two-sidedness too far.” Id. at 171-72.
In affirming the Second Circuit’s rejection of the District Court’s approach to market definition, Justice Thomas’s majority opinion (joined by Chief Justice Roberts and Justices Kennedy, Alito and Gorsuch) agreed that the credit card industry operates in a two-sided market, meaning that it simultaneously provides service to two different groups: cardholders and merchants. Id. at 2. In assessing how the NDPs affect competition, the Court focused on the highly interconnected nature of the respective demands on each side of the market—demand for merchants to accept a network’s card and the demand for consumers to use it—explaining that a credit card “is more valuable to cardholders when more merchants accept it, and is more valuable to merchants when more cardholders use it.” Id. The Court described the dynamic relationship between these two groups of consumers as an example of “indirect network effects,” a feature of two-sided platforms where the value of the platform to one group depends on how many members of another group participate. Id. at 3. Indirect network effects, the Court observed, are critical considerations for competition among two-sided platforms, which “must take these indirect network effects into account before making a change in price on either side,” or they risk creating “a feedback loop of declining demand.” Id.
The Court found that the government failed to take these two-sided considerations into account in defining the relevant market and that the plaintiffs’ focus on merchant fees improperly ignored competition on the other side of the market: the competition for cardholders. Id. at 14. Focusing specifically on the plaintiffs’ evidence of increased merchant fees, the Court stated that “[e]vidence of a price increase on one side of a two-sided transaction platform cannot by itself demonstrate an anticompetitive exercise of market power.” Id. at 15.
Applying these two-sided principles to the case at hand, the Court held that an increase in fees charged to merchants was not in and of itself anticompetitive, as “Amex uses its higher merchant fees to offer its cardholders a more robust rewards program, which is necessary to maintain cardholder loyalty and encourage the level of spending that makes Amex valuable to merchants.” Id. at 16. Thus, in the Court’s view, higher prices charged by Amex to merchants were not an anticompetitive effect of the NDPs in its merchant contracts, but rather evidence that competition for cardholders—in the form of better cardholder rewards funded by those increased merchant fees—was in fact flourishing. Id. at 16-19.
Justice Breyer, in an opinion joined by Justices Ginsburg, Sotomayor, and Kagan, dissented. The dissenting justices argued that the two-sided nature of the credit card industry did not require a “special market definition approach” that considers both merchants and consumers. Id. at 15. Noting that Amex raised prices to merchants over 20 times and suggesting that Amex did not increase benefits or decrease cardholder prices during that period, the dissent argued that price increases to merchants were sufficient proof of actual anticompetitive effects of the contractual NDPs to discharge plaintiffs’ initial rule-of-reason burden. Id. at 3.
The American Express case contains several lessons for antitrust analysis and for antitrust litigation, and these are not limited to two-sided markets. While, with the exception of the two-sided market analysis, these lessons are not particularly novel, they bear repeating because they found an unusual application in this case and because they occasionally seem to be forgotten, including by regulators and enforcers.
Antitrust analysis must consider the total market affected by a restraint, including two-sided platform markets where appropriate. In their case-in-chief, the American Express plaintiffs and their economic expert acknowledged that the credit card market was two-sided but then made minimal efforts to incorporate that fact in their analysis. The District Court followed the plaintiffs down this path, not wishing to take “the concept of two-sidedness too far.” 88 F.Supp. 3d at 172-73. The Supreme Court, however, devoted substantial discussion to the proper approach to two-sided analysis, including an examination of the academic literature on the subject. Slip op. at 3-5. In light of this analysis, it will be very difficult for a future litigant to ignore or merely pay lip service to two-sided issues when they are properly presented. This is important because the payments industry is not the only industry in which two-sided markets are present. Examples of industries in which two-sided or multi-sided market analysis may be important range from ridesharing services (drivers and passengers) to media companies (viewers and advertisers) and insurance companies (patients and providers). In all two-sided markets, the firm must balance the value offered on one side of its platform with the value offered on the other side. Often, the firm sets the price on one side of the platform as free or even negative in order to generate demand on the other side of the platform and thereby maximize its total revenue. Analysis of competition in these and other platform industries should consider actual and potential effects of the challenged conduct on both sides of the platform, including the kind of feedback and network effects that were central to the Court’s analysis here.
Price increases, without consideration of either the nature or cost of the product or market conditions, do not by themselves demonstrate market power. As discussed above, the Court rejected the District Court’s reliance on American Express’s increased merchant rates as proof of either market power or anticompetitive effects because this analysis did not consider prices on the other side of the platform. Slip op. at 16. Nor did the plaintiffs attempt to establish by any other means that American Express’s pricing or margins on transactions were supracompetitive. Id. In a competitive market, litigants cannot rely on price increases without more to prove their case.
Vertical restraints that protect differentiated product competition are important and valuable. The government’s theory and affirmative case of anticompetitive effects was both remarkably simple and remarkably narrow—merchant fees were higher than they would be absent the NDPs—completely ignoring the admitted innovation, competition and consumer benefits that characterized the industry in recent years. As the Supreme Court has recognized repeatedly, many firms in many industries use vertical restraints to promote and protect their differentiated product business models. See, e.g., Leegin Creative Leather Prods, Inc., v. PSKS, Inc., 551 U.S. 877 (2007); Business Electronics Corp. v. Sharp Electronics Corp., 485 U.S. 717 (1988); Continental T.V., Inc. v. GTE Sylvania Inc., 433 U.S. 36 (1977). Applying these principles to American Express’s NDPs, the Court found that they “promote interbrand competition” by protecting American Express’s investments in cardholder rewards, which in turn enable American Express to pursue a differentiated business model that “focuses on cardholder spending rather than cardholder lending.” Slip op. at 6, 19. While plaintiffs have an understandable desire for lower prices as a general matter, they should carefully consider the benefits of such differentiated product competition before challenging a restraint designed to promote such competition in hopes that the challenge will result in lower prices.
Output is key. Even with what they view as good price evidence, antitrust litigants ignore evidence relating to output at their peril. As the Supreme Court’s decision notes, the plaintiffs did not and could not attempt to prove that American Express’s NDPs reduced output. Slip. op. at 17. In fact, the evidence showed that competition for consumers was intense, manifesting itself in any number of ways, not the least of which were consumer rewards. Plaintiffs may wish to reconsider before bringing a case in which it is impossible to demonstrate that output has been, or is at least likely to be, restricted or quality reduced.