The Supreme Court on Ohio

v. American Express


abe wickelgren teaches law at the University of Texas at Austin.

Published July 16, 2018


The Supreme Court’s decision in June that let American Express off the hook in a decade-old antitrust case may or may not be the right decision on the merits, but one thing is certainly clear: the court’s analysis does not focus on the important issues. The majority opinion (on behalf of the five conservatives justices) starts off correctly arguing that one cannot determine the competitive effects of behavior in a “two-sided” market — a market in which two parties interact using a common platform — without analyzing the impact on both sides of the market. It then proceeds to decide the case without any careful analysis of the competitive effects at all.

Anatomy of the Case

Let’s start with the facts not in dispute. American Express offers credit cards with higher transaction fees paid by merchants accepting the cards than most other card issuers (think Visa and Mastercard). But it does devote some of the extra revenue to attracting and keeping cardholders by sweetening cardholder benefits. Merchants are, of course, happy to sell to Amex cardholders, but would prefer them to use the lower-fee plastic in their pockets — or to pay in cash. So, in order to prevent merchants from encouraging the switch by offering discounts, American Express requires them to sign contracts that prohibit giving customers any incentive to use another card.

Merchants object to the anti-steering provisions since it forces them to forgo steering or give up their affiliation with Amex. The U.S. Department of Justice along with several states sued American Express, alleging that these contracts are anti-competitive and hence a violation of Section 1 of the Sherman Act.

The goal of the Sherman Act is to protect consumers by preventing (ideally, by deterring) firms from engaging in activities that limit competition in ways that raise prices or restrict consumer choice. But that’s easier said than done. Competitive strategies are often complex, making it hard to distinguish between pro-competitive behavior that hurts rivals but benefits consumers and behavior that hurts both competitors and consumers. 

As a result, antitrust law is guided by a series of tests that break up the diagnosis of market behavior into manageable steps. But in a great many cases, these guidelines do not actually help distinguish pro-competitive from anti-competitive behavior, thereby tempting courts to base decisions on proxies that miss the mark. In the American Express case, both sides were guilty of this casual analysis in making their arguments.

As both sides acknowledge, American Express’ anti-steering contract provisions (and the resulting higher transaction fees) enable the company to win over cardholders with extra goodies — everything from cash back to free air travel to extended product warranties. American Express’ anti-steering provisions violate antitrust law if this bundle of extra benefits and higher merchant costs ends up making consumers as a group worse off and would have been eliminated by market competition that the anti-steering provisions inhibit. This implies that the first step in a proper analysis is to explain why some consumers would use Amex cards if, on balance, it harmed consumers as a group. A hypothetical example shows why this might indeed be the case.    

Imagine that a store’s profit-maximizing price for, say, the latest fashion sneakers is $97 if they are purchased with a low-fee credit card (e.g., Visa or Mastercard). Now assume that American Express charges the store an extra $3 on the same sale, but also provides card users with a benefit worth an extra $3. (It is easiest to think of this as cash-back, but it could be returned in frequent flyer miles or, for that matter, credit toward dinner with a celebrity chef.) 

If no consumers use an American Express card, the profit-maximizing price for the sneakers remains $97. If all consumers use an American Express card, however, the store can achieve a nearly identical profit-maximizing outcome by raising its price to $100. In this latter case, the store still nets $97. And the Amex cardholders still pay a net price of $97 because they receive a rebate equivalent to $3 after paying $100. 

So, why would anyone use American Express if the effect on card users is a wash? If everyone else is using American Express, but I don’t (and, thanks to the anti-steering contract, the store can’t give me a discount for not doing so), I will still be stuck paying the full $100 — but I won’t receive any perks from Amex. Thus, in this case, the fact that everyone else is using American Express will make me worse off. 

Suppose (in a scenario closer to reality) one customer in three uses an American Express card and the rest use Visa or Mastercard. Then, merchants pay the extra $3 American Express fee in only one-third of the sneaker purchases, so the store’s profit-maximizing price will be roughly $98. Suppose further that American Express passed only $2 of its $3 surcharge back to Amex cardholders in the form of benefits. American Expressprofits; American Express cardholders gain — they receive $2 in extra benefits yet pay only $1 more ($98 rather than $97). But the two-thirds who use Visa or Mastercard now pay $98 (an extra $1) and get nothing in return. 

The point is not that American Express’s strategy necessarily hurts consumers, just that it could hurt consumers and still survive in the market in which American Express is allowed to impose anti-steering rules.

Of course, it is possible that American Express’s strategy, on balance, benefits consumers as a whole. Maybe it is able to deliver rewards (e.g., frequent flyer miles) that are worth more to consumers than it costs American Express to provide them. This could be the case, for example, if these miles are used to fill seats that otherwise might be empty. The point is not that American Express’s strategy necessarily hurts consumers, just that it could hurt consumers and still survive in the market in which American Express is allowed to impose anti-steering rules. (All this, it’s worth noting, depends on the American Express brand being sufficiently strong that a lot of stores can’t afford to refuse the use of the Amex card.)

Eyes on the Prize

The more important point is that neither the majority ruling in American Express’ favor nor the dissenters ever addressed this issue in their opinions. Instead, the bulk of the analysis in both opinions was directed at largely irrelevant formalistic disputes about market definition and market power. In fact, in footnote 7, Justice Thomas goes so far as to say that evidence of anti-competitive effects is insufficient to establish liability without defining a market or establishing market power, even though the only reason to do either is to determine competitive effects. (He says this, by the way, right after saying that market realities matter more than formalistic distinctions.) 

To the extent that the majority opinion is interpreted as actually requiring plaintiffs in two-sided market cases to analyze the net effects of contractual provisions on both sides of the market, it is hardly devastating to antitrust law. Indeed, that is exactly what plaintiffs, defendants and courts should be doing. But this interpretation gives the Supreme Court more credit than it deserves. The fact that the majority could not look beyond the plaintiffs’ doctrinal focus on just one side of the market to make a careful assessment of the competitive effects of the contractual provisions independent of how one defines the market is unfortunate — and, alas, not much of a break from the way both liberal and conservative courts have handled cases in the past.

main topic: Competition Policy