The U.S. Supreme Court ruled that courts must consider customers on both sides of credit card transactions—merchants and cardholders—when evaluating antitrust claims, from the outset. In so doing, the court affirmed the U.S. Court of Appeals for the Second Circuit's rejection of claims that anti-steering provisions in American Express's merchant agreements violated §1 of the Sherman Act. Ohio v. American Express Co., 138 S.Ct. 2274 (2018).

Proper adjudication of antitrust disputes requires careful study and exploration of the economic markets at issue. As such, developments in economic scholarship have often informed the evolution of antitrust jurisprudence. This decision continues in that tradition. The opinion is also significant for affirming the three-step burden-shifting framework lower courts devised for applying the rule of reason, describing how the rule applies to two-sided transaction platforms, and requiring plaintiffs to define a relevant market when challenging vertical conduct.

Background

The U.S. Department of Justice and 17 states sued American Express (Amex), Visa and MasterCard in 2010, claiming that anti-steering provisions in each company's merchant agreements reduced competition over fees charged to merchants. These rules—sometimes referred to as nondiscrimination provisions—prohibit merchants from steering their patrons to use a different card. For example, a contract with Amex prevents a grocer from encouraging shoppers to pay with a Visa card. Plaintiffs argued that preventing merchants from persuading customers to use a lower-cost card reduced card companies' incentives to offer merchants lower prices or better terms. Early in the case, MasterCard and Visa settled and agreed to cease using the disputed provisions. Plaintiffs continued to litigate their challenge to the vertical, non-price restraints in question against Amex, even though it accounts for only 26.4 percent of transaction volume in the credit card market—far less than market shares typically viewed as indicating market power.

Amex argued that the rules sharpen inter-brand competition for both merchants and cardholders. The rules allow Amex to compete for cardholders more vigorously by reducing friction at checkout, maximizing cardholder rewards and encouraging spending at merchants; absent the provisions, the company's business model would be threatened by increased free-riding, as merchants could lure Amex cardholders into their stores by advertising that they accept Amex, only to discourage them from using their preferred card at checkout.

Amex contended that plaintiffs could not demonstrate anticompetitive effects merely by showing the practice injured merchants. As a two-sided platform facilitating transactions between cardholders and merchants, Amex submitted, effects should be evaluated in the transaction services market as a whole, where cardholders and merchants jointly consume the service. The trial court disagreed, ultimately finding that the anti-steering rules violated antitrust law. Amex appealed. The Second Circuit reversed. The Department of Justice declined to seek further review, but the states persuaded the Supreme Court to take the case. The Supreme Court, in a five-to-four decision authored by Justice Thomas, affirmed the Second Circuit's ruling.

Economic Scholarship

The court's majority relied on the extensive economic research and literature on two-sided platforms, which began nearly two decades ago with papers by Jean-Charles Rochet and Nobel-laureate Jean Tirole. A two-sided platform offers a service to two groups that use it to interact with each other and whose demand is interdependent—each group's demand for the platform's service depends on the extent of the other's demand. Because demand is interdependent, the platform must determine what division of the service's cost between the groups will maximize the number of transactions.

The lesson that the interdependency of demand for a platform's two customer groups can have significant economic ramifications is now generally recognized in economics and has been studied and applied in the business world. The court cited repeatedly to Professors David S. Evans and Richard Schmalensee, leading experts in the field who submitted a brief as amici curiae (represented by the authors of this column). For an accessible discussion of two-sided economics cited by the court, see their recent book, Matchmakers: The New Economics of Multisided Platforms.

Burden-Shifting

The court decided that by neglecting to consider both sides of the platform, plaintiffs failed to meet their initial burden in a §1 restraint of trade claim under the rule of reason. The court has long held that while a few categories of conduct, like bid rigging, are per se illegal—because they “always or almost always tend to restrict competition and decrease output”—most practices are subject to the rule of reason, under which courts must decide if the challenged conduct unreasonably restrains competition by evaluating its effects on competition in a relevant market. For years, lower courts have used a three-step burden-shifting framework to apply the rule. In this case, the Supreme Court, for the first time, affirmed the framework's use. Plaintiffs, the court held, have the initial burden to prove the challenged practice has a substantial anticompetitive effect in a relevant market. If they succeed, the burden shifts to defendants to show the practice has procompetitive justifications. If defendants do so, the burden returns to plaintiffs to demonstrate that the procompetitive efficiencies could be reasonably achieved through less anticompetitive means.

Two-Sided Platforms

The court stated the product market at issue was transaction services for merchants and cardholders, not merchant services, and that assessing whether the anti-steering rules are anticompetitive required appraising their effects on both sides of the platform. The court noted that “it is not always necessary to consider both sides of a two-sided platform.” Whether that is necessary depends on the extent to which the two sides' pricing and demand are interconnected and will likely require a case-by-case analysis by lower courts. The court indicated that two-sided transaction platforms, such as Amex, that facilitate a single, simultaneous transaction between participants, require evaluation of both sides to accurately assess competition. These transaction platforms supply a single product —transactions—jointly consumed by two groups of customers and must find the right pricing balance that encourages the greatest number of transactions.

The court concluded the anti-steering provisions could be economically efficient and welfare enhancing even if they do raise merchants' costs. An increase in prices to merchants, the court stated, might increase the number or reduce the cost of transactions in the aggregate depending on how cardholders would respond. The court observed that evidence at trial suggested competition in the card payments industry had increased since the introduction of these provisions: transactions and cardholder benefits had both increased.

Two-sided analysis does not necessarily mean defendants always win. As Professors Evans and Schmalensee explained in their friend-of-the-court brief, looking at both sides can reveal anticompetitive harm that may go undetected under a one-sided inquiry, including, for instance, where the restraint is imposed on one side while the other side suffers competitive injury or when total prices may be relevant to the analysis.

Market Definition

The court also rejected plaintiffs' contention that it was unnecessary to define a relevant product market in this case. Plaintiffs argued that a party that presented direct evidence that a defendant's restraint had anticompetitive effects need not engage in the complex task of defining a relevant market and demonstrating that the defendant possessed market power. Plaintiffs claimed they had made a direct case, as the district court found they presented evidence that Amex had successfully and profitably raised merchant fees.

The court disagreed. Plaintiffs may not skip defining a relevant market when challenging vertical restraints like the anti-steering rules. The court noted that the cases relied upon by plaintiffs, where the court concluded it did not need to precisely define the relevant market, involved horizontal agreements between competitors not to compete. In contrast, vertical restraints are unlikely to threaten competition unless the implementing party has market power.

Dissent

In his dissent, Justice Breyer, joined by Justices Ginsburg, Sotomayor and Kagan, wrote the majority was incorrect to require parties challenging vertical restraints to define the relevant market in every case. As a defendant cannot cause anticompetitive effects without market power, Justice Breyer reasoned, it is unnecessary to require a plaintiff to prove market power if evidence exists that the company successfully raised prices without losing market share.

Justice Breyer also wrote the district court correctly defined the market as merchant services. The dissent stated that merchant and cardholder services should be analyzed as complements, like cameras and film, and added that the two-sided nature of Amex's platform did not merit merging the groups for market definition purposes. Because Amex's services to merchants cannot be substituted for its services to cardholders, Justice Breyer observed, the two services cannot be lumped together in the same relevant product market.

Finally, the dissent suggested that courts could consider the other side of the platform, only later in its analysis. The trial court may weigh procompetitive effects in the “shopper-related market” at steps two and three of the rule-of-reason inquiry, according to Justice Breyer, but not at step one.

Elai Katz is a partner of Cahill Gordon & Reindel. Andrew J. Fuller, an associate at the firm, assisted in the preparation of this article.