Supreme Court Rules on Two-Sided Platforms Antitrust Case

July 02, 2018

In Ohio v. American Express Co.,1 the United States Supreme Court held that American Express Co. (Amex) did not violate Section 1 of the Sherman Antitrust Act by including “antisteering” provisions in its agreements with merchants. The Amex decision establishes important precedent on the application of the antitrust rule of reason to two-sided platforms and to antidiscrimination clauses, both of which are common in e-commerce. The Court’s ruling establishes that, although not all two-sided platforms are to be treated the same, antitrust analysis of at least “transactional” two-sided platforms cannot be conducted in isolation and must consider the impact to users on both sides of the platform. Applying this framework, the Court held that it was improper to analyze the effects of the antidiscrimination provisions on merchants in isolation and that, when properly looking at overall effects on both merchants and cardholders together, plaintiffs failed to show that the provisions had anticompetitive effects. 

The Challenged Antisteering Provisions 

Amex offers a different business model than its rivals in the credit card industry. While its rivals focus on earning revenues from interest on cardholders that carry balances, Amex instead offers its cardholders generous rewards to encourage increased spending, which is valuable to merchants, and Amex in turn charges merchants higher fees. 

Amex’s investments in superior customer rewards would be undermined, however, if merchants were to steer Amex cardholders to use other lower fee / lower reward cards at the point of sale through means such as charging the consumer a surcharge or penalty for using Amex. Such steering could have a substantial impact on Amex because, according to the Court, the vast majority of Amex cardholders also carry Visa or MasterCard, but only a small percentage of Visa and MasterCard holders also carry Amex. 

Amex thus includes “antisteering provisions” in its merchant agreements. These provisions are analogous to most-favored-nation clauses (MFNs), price-parity clauses, or other antidiscrimination clauses commonly used by e-commerce platforms and other intermediaries that facilitate transactions but do not themselves set prices in those transactions.2 By preventing sellers from discriminating against the platform when setting prices, such provisions protect Amex’s investment in the benefits, user experience, or amenities that it uses to attract consumers, but make it difficult for sellers to reward consumers for using platforms that charge the seller a lower fee. 

The federal government and several states alleged that Amex’s antisteering provisions unreasonably restrained trade in violation of Section 1 of the Sherman Act by preventing merchants from steering consumers to lower-fee cards. The parties agreed that the claims do not allege a per se antitrust violation and therefore must be evaluated under the “rule of reason” and its three-step, burden-shifting rubric. The first step in that analysis—which formed the center of the dispute in this case—requires plaintiffs to prove that the defendant’s conduct, had a substantial anticompetitive effect on consumers in the relevant market.3 

The key question throughout each stage of the case was whether that competitive effect should be assessed with respect to the merchants alone or whether the courts could properly consider effects on cardholders as well. The answer to this question depends on how one defines the “relevant market.” The plaintiffs alleged that credit card transactions involve two distinct markets—one market between consumers and the credit card company and a separate market between the credit card company and the merchants. They argued that the antisteering provisions have anticompetitive effects within the latter market because they increase the fees that merchants must pay. Amex argued that the relevant market must include both sides of a credit card transaction because each sale necessarily requires both a consumer and a merchant and because there were feedback loops in which rewards to cardholders encouraged more spending to the benefit of the merchants, and in which higher fees to merchants could cause the merchant to refuse Amex, thus reducing utility to cardholders. According to Amex, when looking at the market as a single but two-sided market for “credit-card transactions” Amex’s antisteering provisions were not anticompetitive. 

The district court agreed with the plaintiffs, analyzed the provisions’ impact on merchants in isolation, and held that the restraints were unreasonable in violation of Section 1. But, on appeal, the Second Circuit Court of Appeals agreed with Amex’s view, and held that the district court erred in focusing only on the interests of merchants in seeking lower fees. When looking at both sides of the market, the Second Circuit held that the evidence was insufficient to meet plaintiffs’ burden to show net harm to the market as a whole. 

The Supreme Court’s Decision 

On June 25, 2018, the Supreme Court affirmed the Second Circuit, holding that the relevant market included both sides of the credit-card platform and that plaintiffs had not shown that Amex’s antisteering provisions caused anticompetitive effects within that overall market.4 Writing for the Court, Justice Thomas explained that credit cards are a two-sided platform because they have “indirect network effects,” in which the value of the platform to each set of participants increases as the number of participants on the other side of the platform increases. Thus, in order to “ensure sufficient participation, two-sided platforms must be sensitive to the prices that they charge each side.” 

The Court, however, cautioned that it is not always necessary to consider both sides of a two-sided platform, and that a market should be treated as one sided when the impacts of indirect network effects are minor. Using the example of newspapers, the Court explained that the value to advertisers increases with the number of subscribers, but the value to subscribers does not increase based on the amount of advertising. 

Justice Thomas explained that credit card networks are different from newspapers because they facilitate a single simultaneous transaction. Because these platforms cannot make a sale unless both sides of the platform simultaneously agree to use it, such “transaction platforms” have more pronounced indirect network effects and interconnected pricing and demand. The Court thus held that transaction platforms like credit cards are better understood as supplying only one product – the transactions. 

After defining the relevant market to include both sides of the platform, the Court concluded that the plaintiffs’ evidence was insufficient to demonstrate anticompetitive effects. “To demonstrate anticompetitive effects on the two-sided credit-card market as a whole,” the Court explained, “plaintiffs must prove that Amex’s antisteering provisions increased the cost of credit-card transactions above a competitive level, reduced the number of credit-card transactions, or otherwise stifled competition in the credit-card market.” Because plaintiffs did not prove that any of those effects occurred, the Court held that they failed to meet their burden under the rule of reason. 

In reaching its conclusion, the Court rejected plaintiffs’ argument that increasing merchant fees alone constituted an anticompetitive effect. The Court held that a price increase on one side of a two-sided platform cannot by itself demonstrate an anticompetitive exercise of market power and that plaintiffs were required to show that the price increases were above a competitive level, which plaintiffs failed to show. Contrary to the plaintiffs’ characterization of supracompetitive pricing, the Court found that Amex’s higher fees reflected the higher value of its cardholders to merchants and the higher costs of Amex’s transactions. 

The Court further rejected the argument that the difference between Amex’s pricing and its rivals’ pricing showed pricing above competitive levels. The fact that Amex’s rivals seek to recover costs through interest charges to cardholders while Amex seeks to recover costs through higher fees to merchants reflected a difference in the model and not a difference between competitive and supracompetitive pricing. The Court also noted evidence that increased merchant fees were not caused by the antisteering provisions. In particular, Visa and MasterCard’s fees were not lower at merchant locations where the merchant refused to accept Amex. The Court also emphasized that there was no evidence to suggest that the antisteering provisions dampened output or quality. On the contrary, credit card transactions increased during the relevant time period, credit became more readily available to lower-income customers, and Amex’s actions spurred rivals to offer premium cards with better rewards. Moreover, evidence showed that the antisteering provisions did not prevent rivals from gaining broader merchant acceptance than Amex by offering lower merchant fees, and that fees industry wide had decreased by more than half during the period in which the antisteering provisions have been in effect. Accordingly, the Court concluded that the plaintiffs had failed to show that the antisteering provisions reduced competition within the credit card industry. 

Given that plaintiffs failed to meet their required showing of harm, the Court did not need to reach the procompetitive benefits of the clauses. Nonetheless, the Court noted that antisteering clauses stem negative externalities that would undermine Amex’s investments in cardholder rewards, which would not only harm consumers but also the merchants that benefit when cardholders spend more to earn rewards. 

Practical Impact on Antitrust Doctrine 

The Court’s decision will significantly impact the manner in which lower courts assess antitrust claims against two-sided platforms common as a result of the growth in e-commerce. By defining the relevant market to include both sides of a transaction, the Court’s holding mandates a broad assessment of economic effects under the rule of reason for platforms that facilitate simultaneous transactions. Such transaction platforms can include travel booking and online shopping services, app stores, and perhaps healthcare insurance providers. 

The opinion also suggests that platforms that make revenues through advertising, such as search engines or other online tools or news, entertainment, information, or blogging sites will be less likely treated as two-sided. Such non-transactional platforms, however, will often be found to be in broader markets because they compete against a variety of other media for advertising dollars. 

The Court’s decision may also have broader implications for the treatment of Section 1 claims involving vertical arrangements. It reinforces the need to define the relevant market with accuracy and plaintiffs’ obligation to demonstrate both market power and a causal link between the challenged conduct and the alleged anticompetitive effects. Higher prices do not necessarily indicate market power or harm to competition. And it suggests that antitrust defendants will be able to highlight expanded output, innovation, and competitive responses to undermine a plaintiffs’ case at the first stage of the rule of reason analysis. 

Finally, the Court’s discussion of the benefits of antisteering provisions could provide analogous support in defense of other types of antidiscrimination provisions commonly used by e-commerce platforms, which often result in disputes between the sellers and the platform over transaction fees. The Amex case provides a useful roadmap to litigating such future disputes. 


1) Ohio v. American Express Co., No. 16-1454.
2) See, e.g., United States v. Apple, Inc., 791 F.3d 290 (2d Cir. 2015) (discussing MFNs preventing publishers from charging a higher price for e-books sold on Apple’s platform vs Amazon); in re Online Travel Company Hotel Booking Antitrust Litig., 3:12-cv-3515 (N.D. Tex. 2014) (discussing rate parity clauses that prevented hotels from charging higher room rates on online travel company sites).
3) As framed by the Court, if a plaintiff satisfies the first step by proving anticompetitive effects, then the burden shifts to the defendant in the second step to prove a procompetitive basis for the restraint. And if the defendant meets its burden, then the burden shifts back to the plaintiff in the third step to show that the defendant could have achieved the procompetitive benefits through less anticompetitive means.
4) Justice Breyer filed a dissent, joined by Justices Ginsburg, Sotomayor, and Kagan, arguing that the two-sided platform is actually two separate markets and that Amex’s higher merchant fees were anticompetitive.

Subscribe to Dechert Updates