In securities cases, defendants often take the position that an investor class may not be certified because plaintiffs failed to satisfy the predominance requirement of Fed. R. Civ. P. 23. Under Rule 23(b)(3), a plaintiff must show that the case presents “questions of law or fact common to class members [that will] predominate over any questions affecting only individual members.” Fed. R. Civ. P. 23(b)(3). The issue of predominance figures prominently in many certification decisions for both securities and non-securities cases. The parties often rely on experts at class certification whose opinions they claim are relevant to assessing compliance with the predominance requirement. Defendants may also choose to attack the methodology employed by plaintiffs’ expert in the hopes of undermining any reliance on his or her opinions for purposes of the predominance inquiry.

The Court of Appeals for the District of Columbia Circuit recently issued an opinion that articulates a rigorous standard of review for expert testimony offered to prove predominance. See In re Rail Freight Fuel Surcharge Antitrust Litig., No. 12-7085, slip op. (D.C. Cir. Aug. 9, 2013). This recent opinion from an antitrust case is instructive for any type of putative class action case where statistical models or regression analysis may be utilized in an attempt to bolster plaintiffs’ claims of predominance.

The claims in the case involved an alleged price-fixing scheme related to railway fuel surcharges, which “set[] up a[] classic battle of the experts” at certification. Id. at 7. Before the district court, the parties’ experts debated inter alia whether plaintiffs could show through common evidence injury in fact to all class members from the defendants’ conduct and thereby satisfy the predominance requirement for class certification. Id. at 6-7.

The Circuit Court’s analysis of the predominance question on appeal focused on defendants’ criticisms of plaintiffs’ expert whose damages model they deemed to be “defective.” Id. at 13. According to defendants, the model employed by plaintiffs’ expert would yield “false positives” because it would “detect[] injury where none could exist”, i.e., as to shippers who were not charged allegedly inflated rates because they were subject to legacy contracts pre-dating any alleged conspiratorial behavior. Id. The district court viewed the damages model offered by plaintiffs’ expert as “essential” to its decision to certify a class, but failed to “address the defendants’ concern that the damages model yielded false positives with respect to legacy shippers.” Id.

The Court of Appeals held that the U.S. Supreme Court’s recent certification decision in Comcast Corp. v. Behrend, 133 S. Ct. 1426 (2013), informed its analysis in the case. Slip op. at 14. In Comcast, also an antitrust case, the Supreme Court held that it was error to certify a class based on a potentially defective damages model and the decision to do so was an indication of the failure to conduct the type of rigorous analysis demanded by Rule 23. Id. As the Circuit Court observed, the Supreme Court in Comcast specifically rejected the notion that “attacks on the merits of the [plaintiffs’] methodology . . . have no place in the class certification inquiry.” Id. (internal citations and quotation marks omitted).

The Court of Appeal concluded that, based on Comcast, “[i]t is now indisputably the role of the district court to scrutinize the evidence before granting certification, even when doing so ‘requires inquiry into the merits of the claim.’” Id. at 14-15 (citing Comcast, 133 S. Ct. at 1433). Thus, “[i]f the damages model [in this case] cannot withstand this scrutiny then, that is not just a merits issue.” Id. at 15. “No damages model” means “no predominance” which, in turn, means “no class certification.” Id.

Contrary to Comcast, the district court had certified a class based on its determination that the damages models offered by plaintiffs were merely “plausible” and without considering the defendants’ arguments related to false positives that undermined their accuracy. See id. at 17-18. This was error because “Rule 23 not only authorizes a hard look at the soundness of statistical models that purport to show predominance” but “the rule commands [such an inquiry].” Id. at 18. For this reason, the Court of Appeals vacated the prior certification order and remanded to the district court with instructions to consider the flaws in Plaintiffs’ models with the benefit of the guidance supplied by Comcast. Id.