Apple Seeks Reversal of Class Certification in App Store Antitrust Case

The petition was filed Friday and docketed Tuesday. Robert Pepper, Stephen Schwartz, Edward Hayter and Edward Lawrence are the named plaintiffs in the case.

The antitrust suit alleged Apple unlawfully monopolizes the retail market for the sale of apps by charging developers hefty commissions that are passed on to consumers in the form of increased prices for apps or subscriptions. The complaint was dismissed on standing grounds in 2013, but the U.S. Supreme Court ruled in a 5-4 decision in May 2019 that iPhone owners who bought apps from Apple’s App Store are “direct purchasers" and have standing to sue Apple for alleged monopolization of the iPhone apps aftermarket.

In March 2022, U.S. District Judge Yvonne Gonzalez Rogers for Northern California in Oakland denied plaintiffs’ motion for class certification, granting in part and denying in part Apple’s Daubert motions to exclude expert testimony of Daniel McFadden, an economics professor at the University of California, Berkeley. While Rogers found the economist’s model and assumptions flawed, she anticipated the flaws could be addressed, and denied plaintiffs’ motion for class certification without prejudice.

On Feb. 2, Rogers signed an order (docket 4:11-cv-06714) in U.S. District Court for Northern California in Oakland saying plaintiffs had resolved the deficiencies in their motion for class certification, denying Apple’s Daubert motion to exclude the experts’ testimony and granting the plaintiffs’ renewed motion for class certification.

A recurring issue in antitrust class actions is determining whether Rule 23(b)(3)’s demanding predominance requirement is satisfied -- most significantly with respect to whether plaintiffs can prove that each class member suffered an injury, said Apple's petition.

Apple put forth in the petition questions presented by its proposed appeal. One, is the certification order “manifestly erroneous because the district court conditionally certified a class without rigorously analyzing whether individualized issues relating to each class member’s injury will predominate over common questions?" Also, Apple said, “Can a district court certify a class based on an injury model that masks individualized differences affecting whether millions of class members were injured at all?”

Plaintiffs are iPhone and iPad users who claim Apple unlawfully monopolized a single-brand market for the device’s apps and in-app purchases by requiring that all such apps be distributed through its App Store. Those restraints allowed Apple to charge “supracompetitive” commissions to developers who in turn set higher prices for consumers, the petition said.

In-mid 2021, the plaintiffs first sought certification of a class of millions of U.S. consumers with App Store accounts who paid for an app or made an in-app purchase since 2008, the petition said. The plaintiffs “relied solely on expert testimony to show classwide injury,” it said. McFadden first estimated the commission rate Apple would have charged without the challenged conduct, then used an “econometric model” to estimate prices developers would have set with the lower commission rate. He used the “but-for” prices to estimate how much each account would have paid for the apps without the restraints, the petition said.

McFadden’s model predicted prices for games, music and entertainment apps, out of 27 categories of apps the App Store sells, said the petition. The model also predicted injury only related to Apple ID accounts, not individual class members, it said. “Plaintiffs propose identifying harmed individuals in a post-trial claims-administration process,” it said.

In another antitrust case against Apple in the Northern District of California, brought by Epic Games, the court “found for Apple on all antitrust claims,” said the petition. The court determined Apple’s App Store conduct is “is procompetitive because it helps ‘improve device security and user privacy’ and ‘increases interbrand competition,’” it said.

When the district court initially denied certification without prejudice because plaintiffs’ model was inadmissible under Rule 702 for expert testimony, they then couldn’t meet their “predominance burden” under Rule 23(b)(3) because their common proof of impact “could not ‘reliably demonstrate which members, and how many, were injured,’” the petition said.

The plaintiffs sought certification again in late 2022, relying on the same methodology “with minor tweaks,” the petition said. Their model “still found that more than 30 million Apple ID accounts would have paid more for apps and in-app content -- and thus suffered no injury -- absent Apple’s challenged conduct,” it said. To "reduce the number of unharmed accounts," plaintiffs limited the class definition to consumers who spent more than $10 from a single account, it said. “Even with this narrowing, the model still leaves about 10.28 million unharmed accounts,” or 7.9% of class accounts, the petition said.

The court granted the plaintiffs' motion the second time, ordering the parties to report back “whether plaintiffs’ model can successfully ascertain the number of uninjured class members and limit them,” the petition said. It indicated it would consider “decertification” if the model “turned out to be defective,” it said.

Review is warranted when a class-certification decision is “manifestly erroneous” or “presents an unsettled and fundamental issue of law relating to class actions, important both to the specific litigation and generally, that is likely to evade end-of-the-case review,” said the petition, citing Chamberlan v. Ford Motor Co. That's the case here where the district court didn't “rigorously analyze whether individualized issues relating to injury predominate,” it said.

District courts have disagreed regarding when a model’s unsupported assumptions about individualized factors bearing on injury, such as developers’ marginal costs, made a model incapable of showing that class members were injured, said the petition. In In Re: Google Play Store Antitrust Litigation, “it was the plaintiffs’ expert’s ‘unproven assumptions’ that ‘fl[ew] in the face of the huge diversity of apps’ in the store that led to decertification,” it said. That the same district court took the opposite view “supports Rule 23(f) review” in Apple’s case, it said.