Amicus Curiae’s Identity, Interest, and Authority to File

The Information Technology and Innovation Foundation (ITIF) is an independent, nonprofit, and nonpartisan think tank.[1] ITIF’s mission is to formulate, evaluate, and promote policy solutions that accelerate innovation and boost productivity to spur growth. ITIF’s core focus lies at the intersection of technological innovation and public policy, including economic issues related to antitrust and competition policy and policy issues in the areas of information technology, broadband telecommunications, and mobile platforms.

ITIF adheres to the highest standards of research integrity and is guided by an internal code of ethics grounded in analytical rigor and independence from external direction or bias. The University of Pennsylvania has recognized ITIF as setting the global standard for excellence in science and technology policy and as one of the top think tanks in the United States.[2]

ITIF strives to provide policymakers around the world with high-quality information, analysis, and recommendations they can trust. ITIF engages in policy and legal debates, both directly and indirectly, by presenting policymakers, courts, and policy influencers with compelling data, analysis, and proposals to advance effective innovation policies—including amicus briefs before this Court in connection with the ongoing Epic Games, Inc. v. Apple Inc. litigation.[3]

ITIF’s Schumpeter Project on Competition Policy promotes an understanding of antitrust law and economics that is designed to maximize dynamic competition and innovation, with a particular focus on promoting sound antitrust enforcement in the digital economy and other high-tech industries. ITIF draws heavily from the work of the Austrian-American economist Joseph A. Schumpeter, who famously described how competition takes the form of “creative destruction” by large firms that have the incentive and ability to engage in the risk-taking and research and development that are necessary to drive innovation.

Summary of Argument

Apple’s iOS platform embodies a complex and symbiotic relationship between Apple, developers, and users. Apple has historically funded this ecosystem through a commission assessed only against certain app types and monetization strategies—most apps pay Apple nothing. And even if Apple were charging anticompetitively high commissions, the general effect on users is ambiguous. Some users may not use apps that are subject to Apple’s commission and thus would not be harmed by anticompetitive commissions. Other users may also be unharmed because the prices the users actually paid in the apps they use were not affected by Apple’s allegedly anticompetitive commission. Indeed, some developers may actually charge lower fees in the face of a higher Apple commission because it may be more profitable to maximize output and focus on revenue from other means (e.g., advertising). Users who may see increased prices for some apps may therefore see reduced prices for others, potentially offsetting any overall harm.

As such, it is not necessarily the case that any given iOS user who paid Apple for one or more in-app purchases or for subscription purchases would have been harmed even if Apple had set anticompetitively high commissions. To identify those potential class members who would be harmed, a properly defined class must identify which iOS users actually faced artificially high prices. Appellants failed to do that: they could not match individual transactions with the individual who made those transactions. And what’s more, Appellants’ strategy of simply exempting persons from the proposed class who paid $10 or less in total app purchases substitutes the problem of overbreadth with a problem of underinclusiveness by improperly excluding iOS users who would be truly harmed and should be included in the class.

Gerrymandered class actions are especially troubling in this context: unless the requirements of Rule 23(b)(3) are strictly applied, the Supreme Court’s decision in Apple Inc. v. Pepper, 587 U.S. 273 (2019), risks opening the floodgates to a deluge of class action litigation in platform markets across the economy. This could have a serious chilling effect on innovation. Further still, allowing Appellants to respond to their failure to defend a well-defined class by changing the nature of antitrust injury to a mere loss-of-choice theory would contravene the consumer welfare framework that has guided antitrust law for decades.

Argument

Appellants’ Class Fails to Satisfy Rule 23(b)(3)’s Predominance Requirement

Rule 23(b)(3) permits certification only when “questions of law or fact common to class members predominate over any questions affecting only individual members.” Fed. R. Civ. P. 23(b)(3). To prove there is a common question of law or fact in an antitrust class action, plaintiffs “must establish that essential elements . . . , such as the existence of an antitrust violation or antitrust impact, are capable of being established through a common body of evidence, applicable to the whole class.” Olean Wholesale Grocery Coop., Inc. v. Bumble Bee Foods LLC, 31 F.4th 651, 666 (9th Cir. 2022) (en banc) (quotation marks and citations omitted). Here, because Appellants offered no reliable classwide method of proving antitrust injury, individualized questions continued to predominate, and the district court did not abuse its discretion in decertifying the class.

Higher Commissions Do Not Establish That Any Particular Consumer Paid a Higher Price

Apple’s App Store launched in 2008 with 538 third-party applications; by 2023, it offered iOS users more than 1.6 million. See 2-ER-37. The apps transact with iOS users at various price tiers, typically at $0.99 and subsequent $1 intervals. Approximately 82 percent of these applications pay nothing to Apple for these transactions. See Apple’s Br. at 6 (citing 2-ER-43). Only a minority—namely, apps that cost a fee to download or that allow in-app purchases of digital content—pay Apple a commission, which is the mechanism Apple uses to invest in, monetize, and support its iOS platform. Apple does not charge commissions based on in-app advertising. See 4-ER-482.

If, as Appellants contend, Apple were charging a commission rate enabled by anticompetitive practices, many iOS users would be unaffected. First, some users may not make purchases that are subject to the iOS commission, and these users thus would be unaffected by a commission that deviates from the hypothetical, competitive “but-for” rate. Moreover, other users who make purchases in applications that do pay the purportedly anticompetitive commission rate may also not face higher prices. This is because some apps may decide to bear the cost of the anticompetitive surcharge in toto if they find that moving to a higher price tier would reduce their profitability, including through reduced advertising revenues due to a loss of users. See Apple’s Br. at 12 (citing 8-ER-1522, 9-SER-2139–57).

To be sure, some iOS users would pay higher prices for purchases in those apps that face and decide to pass on, either in whole or in part, high commission rates. But that does not mean that even those users are harmed, as they may pay lower rates for purchases in other apps. Indeed, some apps that face an especially high elasticity of demand, low marginal costs, or especially high advertising revenues may find it more profitable at the current commission rate to move to a lower price tier (or eliminate a download fee altogether) and increase output relative to output in the but-for world—where the developer retains a greater share of download and in-app purchase fees. As a result, some applications would necessarily charge higher prices in the but-for world. Accordingly, anticompetitively high commissions may result in some users paying lower prices for some apps and higher prices for other apps, and for any given user, the benefits from the former could outweigh the harm associated with the latter.

Appellants’ Class Does Not Account for iOS Dynamics

As the district court correctly observed, “[c]onsumer plaintiffs and their experts acknowledge that not all app developers respond in the same way to Apple’s commission rate” and that “some developers may choose to pass through some or all of Apple’s charged commission to their consumers, while others may not.” 1-ER-25. However, Appellants nonetheless define their class to include iOS users “who purchased one or more iOS applications or application licenses” or “who paid Apple for one or more in-app purchases, including, but not limited to, any subscription purchase.” 2-ER-109.

A “model supporting a plaintiff’s damages case must be consistent with its liability case, particularly with respect to the alleged anticompetitive effect of the violation.” Comcast Corp. v. Behrend, 569 U.S. 27, 35 (2013) (quotation marks and citation omitted). Here, Appellants have failed to reconcile the economic reality underlying their theory of liability with their class. That a potential class member has paid Apple for application downloads or in-app purchases does not at all imply that this person faced higher prices as a result of an app being charged anticompetitively high commissions—either because the commission was not passed on or because reduced prices from other apps offset price increases elsewhere. That is precisely why the district court required Appellants to model the but-for world—one in which some transactions would have cost more and others less—and offer a methodology for matching each transaction with a user. See 1-ER-25–26. Because, as the district court held, Appellants have failed to articulate a reliable methodology for matching transactions to class members, see 1-ER-12–21, their theory is not “capable of establishing antitrust impact on a class-wide basis,” Olean, 31 F.4th at 678.

In a failed attempt to save their class from legally fatal overbreadth, Appellants limit their class to “those persons who paid more than $10.00 in total to Apple during the Class Period for iOS application and in-app purchases from any one Apple ID account.” 2-ER-109. That threshold may have reduced the model’s estimated share of unharmed accounts, but it did not provide a legitimate mechanism for identifying which consumers actually paid a net overcharge—and thus suffered antitrust injury. What’s more, arbitrarily excluding a whole segment of potentially injured members from the class to eliminate false positives may give rise to other problems, including, for example, the potential for an additional class action brought by the excised members who have every bit as much standing to sue as those who were injured and paid over $10. Indeed, as the Fourth Circuit explained only last year, “excising such a large share of potential claimants from the proposed class . . . might expose defendants to a continued trickle of individual lawsuits and thereby impair the efficiency goal which is a key purpose of the class action mechanism.” Mr. Dee’s Inc. v. Inmar, Inc., 127 F.4th 925, 932 (4th Cir. 2025).

Certifying Gerrymandered Classes Is Dangerous in Platform Markets

Platform markets that share similar dynamics with iOS—namely, multi-sided markets that connect two groups of users and that exhibit indirect network effects whereby the more users there are on one side of the platform, the more valuable the platform becomes to users on the other side—are not isolated contexts. On the contrary, platform markets are ubiquitous in the modern economy, making judicial treatment of certification in this case hugely consequential.

Indeed, the widespread nature of platform markets and the billions in commerce they enable in the U.S. economy are all the more reason for lower courts to apply more—not less—scrutiny to consumer class definitions like Appellants’, so as to curtail both litigation costs and error costs in the form of overdeterrence. As Justice Gorsuch warned in his Apple v. Pepper dissent, determining antitrust injury will involve “difficult questions about apportionment of damages between app developers and their customers, along with the risk of duplicative damages awards.” 587 U.S. at 292 (Gorsuch, J., dissenting). Those “difficult questions” cannot be swept under the rug by a strategic class definition or an incomplete expert analysis.

Appellants’ attempt to skirt their burden of proof in the face of this complexity by claiming that their class of iOS users is harmed by a loss of choice again merely substitutes one problem for another—just as plaintiffs did in revising their class definition. See Appellants’ Opening Br. at 53–54. This Court has rejected attempts to equate a limitation on consumer choice with antitrust injury without a connection to harm to the competitive process and to the plaintiff. See, e.g., Somers v. Apple, Inc., 729 F.3d 953, 966–67 (9th Cir. 2013). It has likewise explained that reduced choice and higher prices do not, without more, establish injury to competition because those effects may arise from lawful or procompetitive conduct. See Brantley v. NBC Universal, Inc., 675 F.3d 1192, 1202 (9th Cir. 2012). Put simply, reduced choice may be relevant in an appropriate antitrust case, but it alone is not a substitute for proving causation and injury.

Conclusion

For all these reasons, this Court should affirm the district court’s decision to decertify Appellants’ class.

Endnotes