Must Experts Duke It Out Before Class Certification?

In a prior column, we discussed the U.S. Court of Appeals for the Third Circuit’s influential opinion in In re Hydrogen Peroxide Litig., 552 F.3d 305 (3d Cir. 2008), and its potential impact on courts’ treatment of expert disputes at the class certification stage.¹ While the Supreme Court has not yet waded squarely into the issue, it did recently dip its toes in the water with some language in Wal-Mart Stores Inc. v. Dukes, 564 U.S.—, 131 S. Ct. 2541 (2011), suggesting an endorsement of the Third Circuit’s holding that expert disputes related to class certification must be resolved before certification can issue. This language from the nation’s highest court should be taken into consideration by lower courts addressing class certification motions. However, as illustrated by the U.S. Court of Appeals for the Eighth Circuit’s opinion less than one month later in In re Zurn Pex Plumbing Products Liability Litigation, 644 F.3d 604 (8th Cir. 2011), petition for cert. filed, No. 11-740 (Dec. 15, 2011), at least one court has not required a full and conclusive Daubert review at the class certification stage, a holding that drew a vigorous dissent and that is likely to herald continued disagreement by other circuits.

‘Hydrogen Peroxide’

Purchasers of hydrogen peroxide and related chemicals brought forward an expert in an antitrust class action to establish that the impact of an alleged conspiracy among manufacturers to set prices in the hydrogen peroxide industry was capable of proof at trial through evidence common to the class. Hydrogen Peroxide, 552 F.3d at 312-13. The defendants responded by bringing forth their own expert to show that the statistical methods employed by the plaintiffs’ expert could not feasibly establish commonality to satisfy Rule 23(a). Id. at 313-14. The district court declined to weigh the credibility of the competing experts and instead ignored the defendants’ expert, holding that plaintiffs need not show at the class certification stage that their proposed statistical methods would actually work, but instead need only “demonstrate their intention to prove a significant portion of their case through factual evidence and legal arguments common to all class members.” Id. at 321-22.