In re Google Play Consumer Antitrust Litigation

Background

This multidistrict antitrust litigation includes several groups of plaintiffs challenging Google’s Play Store practices. The order primarily concerns the consumer case, In re Google Play Consumer Antitrust Litigation, No. 20-cv-05761-JD. The consumer plaintiffs allege that Google illegally monopolized the Android app distribution market in violation of Section 2 of the Sherman Antitrust Act, causing overcharges and other injuries.

The consumers offered Dr. Hal J. Singer, an economist, to provide trial testimony about injury and damages. At the earlier class-certification stage, the court had denied Google’s request to exclude Dr. Singer’s opinions and had certified a consumer class. At the merits stage, Google challenged only the injury and damages portions of Dr. Singer’s work, not his qualifications, the relevance of his testimony, or all of his opinions. Google also sought to exclude merits opinions by Dr. Marc Rysman, but the court stated that motion would be addressed in a separate order.

Rule 702 and the Pass-Through Model

Federal Rule of Evidence 702 permits qualified experts to testify when their specialized knowledge will help the factfinder, their opinions rest on sufficient facts or data, their methods are reliable, and they reliably applied those methods to the case. Applying Rule 702 and the standards from Daubert v. Merrell Dow Pharmaceuticals, Inc., the court emphasized that the proponent of expert testimony bears the burden of establishing its reliability.

A central part of Dr. Singer’s damages analysis was a pass-through formula based on a logit demand model. The formula estimated an app developer’s pass-through rate as one minus the app’s share of its self-selected Play Store category. The formula was used in damages calculations for the Android App Distribution Market and the In-App Aftermarket.

The court found that the logit model depended on an “independence of irrelevant alternatives” assumption, meaning that the model treated apps in a category as substitutes in proportion to their market shares. The court concluded that this assumption was not supported for the Play Store’s broad categories, which included apps with very different functions. It used examples involving language-learning and plant-identification apps to illustrate why category share alone could not reliably predict substitution.

The court also found that Dr. Singer had not adequately shown that regression analysis validated the logit model for this market, had not compared the model’s fit with an alternative demand model, and had not supported the assumption that the apps were proportional substitutes. Because the pass-through formula was the “keystone” of his overcharge analysis, the court excluded the testimony and injury and damages opinions that depended on it.

Consumer-Subsidy Models

The court separately excluded Dr. Singer’s alternative consumer-subsidy models, which did not use the pass-through formula. For the Play Points model, the court found that Dr. Singer relied on unsupported assumptions, including that an expanded Play Points program would resemble the existing program and that Google would retain a 60 percent market share in a competitive “but-for” world.

The court also found inadequate support for Dr. Singer’s Amazon Appstore discount model. It concluded that the report did not sufficiently explain why Amazon’s discounts were a reasonable approximation or benchmark for damages in this case. The court determined that both subsidy models were connected to the available data by too large an analytical gap to be admitted at trial.

Disposition

The court granted Google’s motion for leave to file a supplemental brief in support of its Rule 702 motion. Judge James Donato granted Google’s motion to exclude Dr. Singer’s merits opinion testimony, docket number 487. The order did not rule on the underlying antitrust claims or on Dr. Rysman’s merits opinions.