In an August 9, 2013 opinion in In re Rail Freight Fuel Surcharge Antitrust Litigation, the US Court of Appeals for the DC Circuit overturned the district court's certification of an antitrust class action and remanded the case for reconsideration in light of the US Supreme Court decision in Comcast Corp. v. Behrend. The court applied Comcast to make clear that class certification under FRCP 23(b)(3) commands courts to rigorously scrutinize statistical models used to satisfy the predominance requirement. The court held that because the damages model advanced by the plaintiffs did not reliably prove class-wide injury, class certification was improper.

The plaintiffs, a group of shippers who paid inflated fuel surcharges to four major freight railroads, filed suit in the US District Court for the District of Columbia against the freight railroads for allegedly violating Section 1 of the Sherman Act by colluding to raise fuel surcharge rates. The plaintiffs sought class certification by offering two statistical models that in conjunction show predominance, including the damages model. The damages model purported to quantify the injury-in-fact to all class members caused by the defendants' collusive conduct. Before the decision in Comcast was issued, the district court accepted the plaintiffs' models as plausible and granted class certification.

On appeal, the court noted that before Comcast, courts were far more accommodating to class certification under FRCP 23(b)(3). In Comcast, Comcast cable television customers in the Philadelphia area filed suit against Comcast for violating antitrust laws in the provision of cable services. The Comcast court reasoned that because the required damages calculation could not be measured class-wide under the theory of liability advanced by the plaintiffs, class certification was improper.

The DC Circuit applied Comcast to vacate the class certification because it found plaintiffs' damages model defective. The court found that the damages model yielded false positives by detecting injury where none could exist. When applied to shippers who were subject to legacy contracts, meaning that they were bound by rates negotiated before any price fixing was alleged to have occurred, the damages model yielded similar results. The court reasoned that common questions of fact cannot predominate where there is no reliable means of proving class-wide injury. When a case turns on individualized proof of injury, separate trials are needed. The district court did not address the concern that the damages model yielded false positives for legacy shippers.

Practitioners are reminded that after Comcast, courts will undergo a more rigorous analysis of statistical models that purport to prove predominance in granting class certification under FRCP 23(b)(3).