For those of us following the Epic v. Apple case, it was sometimes difficult to follow or immediately grasp the relevance of the meandering lines of questioning that covered a wide range of themes. Intermingled with all of the evidence that found its way into the record via testimony and discovery were references to key antitrust law concepts and case history. As we await a ruling, Judge Yvonne Rodriguez Rogers’s job is to consider the facts in this case against United States and California antitrust laws. We thought it would be helpful to explore some of those concepts to better understand how that analysis might look.
The claims Epic brought initially arise under the federal Sherman Act and the analogous statute in California, the Cartwright Act. In general, the Cartwright Act explicitly prohibits certain activities that federal courts have interpreted to be illegal under the much more general Sherman Act. So, Cartwright seems mostly to reinforce Sherman (or at least has a basis in Sherman concepts), but since state laws are usually brought in state courts instead of federal courts, they are interpreted by a completely separate set of judges with a separate appeals path—and that means there could be some significant, relevant differences between conclusions made under the Cartwright and Sherman laws. So, to keep things simpler, we will stick to the federal law:
Market Definition
At issue in the Epic case are Epic’s claims that Apple is liable for “unilateral” anticompetitive conduct, meaning that the case does not involve agreements or conspiracies with other companies, but is instead focused on Apple’s conduct all by itself. Generally, it is impossible to show that a single company is liable under federal or state antitrust laws for unilateral conduct unless that company has market power and that it used its market power to harm competition and consumers.
Market power refers to a company’s ability to raise prices or reduce output or quality, unchecked by the usual competitive forces. To test for market power, courts often use the “hypothetical monopolist test,” by asking—for example—whether a firm can impose a “small, but significant, and non-transitory increase in prices,” (also referred to as an SSNIP) for which other competitors are unable to punish the company by luring away their customers. If you define a market narrowly enough, every company has market power (for example, Ford has a monopoly on the market for new Ford automobiles). It’s no surprise, then, that Epic’s proposed market definitions are—you guessed it—the markets for iOS app distribution and iOS in-app payment processing(IAP) (see, e.g., pages 51 and 96 in the linked filing from Epic).
In general, whether this is a valid antitrust market, however, depends on whether market actors (consumers, developers, and other platforms) behave as though it is a separate market. One of the most pressing questions, therefore, is whether consumers and developers see iOS app distribution as a substitute for another platform’s app distribution services. If they are substitutes and consumers and developers actually make use of both or either, such that iOS app distribution is in direct competition with another platform’s app distribution, then Epic loses the market definition battle.
The fight over market definition explains why the witnesses spent a lot of time talking about what constitutes a substitute for an in-app purchase on iOS. Since Apple’s proposed market definition is the market for game app transactions, its attorneys and witnesses focused on showing that iOS game transactions are substitutes for game transactions on other platforms. This view of the market tracks much more closely with how ACT | The App Association’s members reach consumers—they usually develop software for distribution on multiple platforms, not just iOS. Complicating matters a little bit is the fact that game app transactions take place in a “two-sided market” under federal law: developers transact with the platform on one side of the market and consumers transact with the platform on the other side of the market. The service the platform provides to consumers and developers is the transaction itself. A consequence of this—again under U.S. federal law as clarified in the U.S. Supreme Court (SCOTUS) case Ohio v. American Express—is that the platform’s activity must harm both sides of the market for it to constitute illegal monopolization.
For precedential support for its claim that iOS apps are a market, Epic turns to Eastman Kodak Co. v. Image Technical Servs. In that case, SCOTUS ruled that aftermarket servicing of Kodak copiers was itself an entire market. The market for iOS apps, Epic argues, are likewise a single-brand aftermarket good. However, SCOTUS noted that a single brand name’s products or services almost never constitute a market, unless the very specific circumstances surrounding Kodak exist. Specifically, Kodak sold enough high-volume copiers to claim a substantial portion of that market and then (after buyers had purchased the items) imposed on the owners of those copiers and independent service operators (ISOs) a set of restrictions locking copier owners into a service contract with Kodak and preventing ISOs from servicing Kodak copiers. Those circumstances were especially important to SCOTUS’ analysis because Kodak had created unseen barriers to servicing its products to which that the market’s consumers never agreed.
In the case of Apple, on the other hand, consumers know the only app store on iOS is the one Apple provides, and that the “walled garden” is not only something consumers and developers know Apple insists on but is also a selling point for consumers and developers. They choose Apple’s walled garden over Google’s more open platform all the time precisely because it is a closed system. This competitive dynamic contrasts sharply with—and has the opposite evidentiary value of—the situation in Kodak where the defendant tricked consumers into exclusive contracts. Tricking consumers prevented them from having a choice in the matter and eliminated the competitive dynamics that would otherwise exist, which in turn would widen the definition of the market.
Antitrust professors are known for teaching their students that market definition is often “the whole ballgame.” Hopefully, this brief section above explains why that is the case and why it takes up so much of the witnesses’ time and discussion in Epic v. Apple.
Essential Facility
Apple’s Proposed Conclusions of Law filed with the court before the trial began indicates that Epic had “abandoned” its essential facility claim (page 120). However, despite Epic’s witnesses avoiding the issue, Epic has not formally dropped it as a claim. Antitrust attorneys seem to view the essential facilities doctrine as one that applies only in narrow circumstances. The case that created the doctrine, United States v. Terminal Railroad Association of St. Louis, involved a railroad collective that controlled all railway bridges and switching yards into and out of St. Louis. Later cases would flesh out the concept and it wasn’t until MCI v. AT&T in 1983 that federal courts developed a test to identify “essential facilities.”
In that case, the 7th Circuit found that something might be an essential facility if: 1) the monopolist controls access to an essential facility; 2) the facility cannot be duplicated practicably by a competitor; 3) the monopolist has denied access to the competitor; and 4) it was feasible for the monopolist to grant such access. Implied in this four-part test is that the plaintiff needs to have established that the defendant is a monopoly before getting to the question of whether it is an essential facility. It may be that Epic is not pressing this claim because its elements are even harder to prove for Epic than the rest of its claims. Specifically, for most of its claims, Epic only has to show that Apple has market power and that Apple’s use of its market power harms competition and consumers. With an essential facilities claim, Epic would have to show that it is unable to reach consumers of game transactions (or, if you take Epic’s market definition, consumers of iOS game transactions) unless it can access iOS, a higher threshold than market power and harm to competition and consumers. The claim is also plainly untrue, as iOS customers can access Epic’s offerings outside of iOS (on the internet or another platform). Apple is simply not the only railroad into St. Louis.
Anti-steering
The witnesses have also hotly debated Apple’s “anti-steering” terms of service. Coincidentally (or not), the AmEx case focused mainly on American Express’s anti-steering provisions. Apple prohibits app developers from using their iOS app to tell their customers that they could transact more cheaply on another platform. This is a common provision to avoid competitors from using each other’s resources to lure customers away (7-11 doesn’t allow other convenience stores to advertise inside 7-11s). A highlight of the testimony involved one of the Apple attorneys triangulating one of the elements of an anti-steering claim, which is that the prohibited steering must direct the consumer away from an offering that competes directly with the offering of the defendant. Unfortunately for Epic, the claim itself requires the presence of a direct competitor to iOS game app transactions, which is mutually exclusive with a market definition of “iOS game app transactions” and not only implies—but necessitates—the existence of a broader market.
Certain business practices that courts have encountered over and over again are sometimes analyzed in an antitrust context using a truncated approach where courts have ruled the practice to be per seanticompetitive. In narrow sets of circumstances, an antitrust target’s practice of using market power to force buyers to accept separate products that are “tied” together could fall under this per se category. The vast majority of antitrust claims are analyzed under the much more arduous “rule of reason” analysis and that is true of most “tying” claims as well, unless the various specific factors are present.
A fairly recent case where the court grappled with whether to apply the per se rule to a claim is Leegin Creative Leather Products v. PSKS, and ultimately it concluded that per se rules are inappropriate where the effects of a tying arrangement could have “either procompetitive or anticompetitive effects.” Not only that, but courts typically decline to apply per se analysis to cases involving new technologies and that lack an established precedent of analogous activity from which to conclude that the activity is probably anticompetitive. In this case, Apple argues that per se is inappropriate and if anything, the court has to use the rule of reason. But more importantly, Apple argues its IAP service is integrated with app distribution, so the two are not separable offerings (pp. 234-235). This seems more in line with reality because consumers and developers are not requesting or demanding that Apple provide in-app purchasing services as a standalone service outside of iOS. If the notion that the services are integrated is true, the tying claim has to fail because Epic has failed to show that the two services are completely separate (yet artificially tied together).
Antitrust law is necessarily a moving target that evolves along with the competitive harms that plague new kinds of markets. The applicability of old caselaw to new scenarios is always a difficult task for courts and makes the job of predicting outcomes especially tricky, but we hope this quick analysis of just a few of the main legal concepts is illuminating. If anything, the prospects seem dimmer for Epic and its allies and brighter for the rest of the app economy; but we will see in the coming months.