Epic failed to define the relevant market or present a less restrictive alternative.

On April 24, 2023, a three-judge panel from the Ninth Circuit largely affirmed a district court’s judgment following a bench trial that Epic Games, Inc., failed to prove its antitrust claims against Apple, Inc., challenging its rules:

• Restricting app distribution on iOS devices to Apple’s App Store.

• Requiring in-app purchases on iOS to use Apple’s in-app payment processor.

• Limiting the ability of app developers to communicate the availability of alternative payment options to iOS device users.

Despite finding several legal errors in the district court’s analysis, the Ninth Circuit upheld its ultimate conclusions that Epic failed to properly define the relevant market and provide a sufficiently specific less restrictive alternative (“LRA”) for Apple to accomplish its procompetitive justifications for its restrictions.

Unfortunately, the Ninth Circuit’s order contains clumsy analysis not founded in antitrust principles or precedent.

As a result, this order threatens to create case law that makes it unnecessarily arduous for plaintiffs to allege a relevant market comprised of a single-brand aftermarket and for antitrust law to reach the increasingly important platform aftermarkets like apps.

The Ninth Circuit Misapplies Eastman Kodak Co. v. Image Tech. Servs.

The Ninth Circuit conflated Kodak’s relevant market analysis with its market power analysis. As a result, the court required Epic to prove far more than is necessary, relevant, or desirable to prove that its proposed single-brand aftermarket constituted a relevant market.

To understand the Ninth Circuit’s error, we need to briefly summarize Kodak’s pertinent holdings. The plaintiffs in Kodak claimed the relevant market was a single-brand aftermarket comprised of parts and services for Kodak equipment. Kodak argued that a single-brand aftermarket could not serve as a relevant market. The Supreme Court sided with the plaintiffs, finding the relevant market was comprised of “the choices available to Kodak equipment owners” and that “[b]ecause service and parts for Kodak equipment are not interchangeable with other manufacturers’ service and parts, the relevant market from the Kodak equipment owner’s perspective is composed of only those companies that service Kodak machines.”

When analyzing market power in a separate section of the opinion, the Supreme Court found that competition in the equipment foremarket did not prevent Kodak from possessing market power in the parts and services aftermarket because consumers (i) lacked knowledge to effectively lifecycle price (i.e., understand the total price of equipment, service, and parts at time of purchase) and (ii) face high switching costs given the upfront cost of the equipment at issue. Notably, the Supreme Court did not find that inability to lifecycle price or high switching costs were required to prove market power in a single-brand aftermarket – only that it sufficed in that case. And, as highlighted above, it did not raise these issues when analyzing market definition.

Unfortunately, the Ninth Circuit blended Kodak’s relevant market and market power analyses. Using Kodak as authority, it rejected Epic’s proposed single-brand aftermarket consisting of iOS apps because Epic did not prove iPhone users lacked knowledge about Apple’s restraints at the time of purchase or faced high switching costs. But, as Kodak demonstrates, these issues have nothing to do with defining a relevant market, which involves the hypothetical monopolist test, cross price elasticity of demand, and, to the extent they pull in the same direction as the economics, the Brown Shoe criteria.

By requiring proof that consumers lack information to lifecycle price and face high switching costs, the Ninth Circuit elevates inter-platform competition above intra-platform competition, which is a normative conclusion not based in economically sound antitrust analysis. It also ignores the scale and depth of today’s behemoth digital platforms and aftermarkets. As of 2022, there were more than 200 million iPhone users in the US. And these people rely on the app aftermarket to do everything from banking, to ordering food, to shopping, to gaming, to videoconferencing, to streaming videos, to obtaining directions, etc. Good antitrust policy simply cannot assume that competition between the small number of cellphone manufacturers will discipline competition in the vast and growing app aftermarkets.

The Relevant Market Adopted by Ninth Circuit Makes No Sense

The Ninth Circuit’s misreading of Kodak also caused it to craft a relevant market that defies black-letter antitrust law. Relevant markets define the area of effective competition, meaning they include the product at issue and all reasonable substitutes. To determine whether products are reasonable substitutes, courts often analyze whether consumers would switch from one to the other in response to a small but significant non-transitory increase in price (“SSNIP”).

Here, the court determined the relevant market was mobile gaming transactions, which includes all gaming apps on all operating systems. But nearly all consumers use a single smartphone. Therefore, to switch from an iOS gaming app to a gaming app using a different operating system, most consumers would have to spend hundreds of dollars to purchase or activate a second smartphone. Consumers would not do this in response to a SSNIP in iOS gaming app prices, meaning mobile games on iOS and different operating systems are not reasonable substitutes and should not be included in the same relevant market. If the Ninth Circuit had applied this analysis (i.e., the market definition analysis from Kodak), then it likely would have blessed Epic’s proposed single-brand aftermarket.

Despite Mistakes, the Ninth Circuit Did Make Important Corrections

While some of its analysis was legally and/or factually incorrect, the Ninth Circuit did curb the amount of bad case law it created by correcting multiple errors by the district court.

• For one, the Ninth Circuit struck down the trial court’s categorical rule that relevant product markets can never be comprised of products that are neither licensed nor sold. Not only is this rule inconsistent with case law, but it elevates form over substance and ignores commercial realities, both of which antitrust law repeatedly warns against. In today’s economy, several of the largest companies were built by products that are neither licensed nor sold. For example, Google does not license or sell its search engine, Facebook does not license or sell its social media platform, and Amazon does not license or sell its retail platform. Yet these companies generate billions of dollars annually from these products by monetizing them through methods like collecting and selling consumer data or generating ad revenue.

• Second, the Ninth Circuit overruled the trial court when finding that contracts of adhesion can form the basis of Section 1 claims. This rule is consistent with Section 1’s plain language, which says the law pertains to “[e]very contract, combination . . . ., or conspiracy” that unreasonably restrains trade. Consistent with Section 1’s language, the Supreme Court, Ninth Circuit, and other circuits have consistently acknowledged that Section 1 can apply to agreements between a willing and unwilling party.

Epic Should Have Done More to Show Less Restrictive Alternative

Not all blame for Epic’s loss should be directed at the Ninth Circuit. Even using the Ninth Circuit’s erroneous market definition, Epic may have won the day had it provided a more detailed explanation of how Apple could achieve its procompetitive rationales for its restrictions through less restrictive alternatives.

Under the rule of reason test that applied to its antitrust claims, Epic was required first to prove Apple’s restraints had substantial anticompetitive effects that harmed consumers, which it did. Apple was then required to provide a non-pretextual, legally cognizable procompetitive rationale for its restraints, which it did, in part, by arguing its restrictions ensure consumer-demanded security and privacy and compensate Apple for its IP investment. Finally, Epic was required to show that Apple’s procompetitive rationale could be achieved through less restrictive alternatives. Both the district court and the Ninth Circuit determined Epic did not satisfy this requirement.

Regarding the distribution restriction, Epic largely hung its hat on the fact Apple currently employs a less restrictive alternative on its computers that permits iOS apps to be distributed outside of the App Store. However, Epic failed to provide important specifics concerning how this would be implemented on iPhones, including how Apple could eliminate the higher incidence of malware caused by the LRA and how Apple would be compensated for third party use of its IP. Regarding the app payment processing restriction, Epic proposed access to alternative payment processors as an LRA. But Epic again failed to explain how this LRA would achieve Apple’s IP-compensation rationale.

Although one can argue the trial court and Ninth Circuit erred when blessing Apple’s purported non-pretextual, legally cognizable procompetitive rationale for its restraints, Epic should have assumed they would be accepted and been better prepared to address them. If Epic had provided a detailed plan for implementation of their LRAs that directly and exhaustively explained how they would achieve Apple’s procompetitive rationales, then Epic would have satisfied the third prong of the rule of reason test and may have ultimately succeeded on its antitrust claims. This can serve as a learning opportunity for antitrust practitioners moving forward.