On Wednesday, the United States Supreme Court handed down its opinion in another of this term’s major class action cases. Following on the heels of Standard Fire Insurance Co. v. Knowles, where the Court closed a loophole which had allowed plaintiffs to attempt to stipulate around the threat of removal to Federal court pursuant to the Class Action Fairness Act, the Court overturned class certification in Comcast Corp. v. Behrend, holding that plaintiffs had failed to establish that damages could be proven on a class-wide basis. There’s no doubt that Behrend was a big win for the defense bar, but the question which has divided attorneys in the days since it came down is – how big? Our initial post on Behrend, discussing the facts and lower court rulings in depth, is here.
Behrend involved a certified class of more than two million former and current subscribers to Comcast’s cable services in the Philadelphia metropolitan area. The plaintiffs alleged that Comcast had violated Sections 1 and 2 of the Sherman Act by deliberately buying up cable systems in areas where Comcast had a significant market presence, often in return for giving up systems in areas where Comcast was less concentrated. The plaintiffs argued four different theories of antitrust injury, but ultimately only one was certified – the proposition that Comcast’s alleged behavior deterred entry by “overbuilders” – companies who deliberately entered a market where another cable provider was already established.
Comcast’s cert petition presented the case as a slam-dunk violation of Wal-Mart Stores, Inc. v. Dukes, going so far as to seek summary reversal. Comcast argued that the Third Circuit had simply ignored the Supreme Court’s Dukes instruction to conduct a “rigorous analysis” in certification proceedings, whether or not the issues overlapped the merits. After a number of relists, the Supreme Court ultimately granted cert on a somewhat different question: whether a district court could certify a class without deciding whether the plaintiffs had introduced admissible evidence, including expert testimony, to show that damages could be awarded on a class-wide basis. In other words: are Daubert challenges appropriate at the certification stage?
When the opinion came down last week, at first glance it seemed that prospects for a major Daubert/class certification decision had struck an iceberg: the defendants had never raised a Daubert challenge to the plaintiff’s expert. Noting the arguable waiver, the majority reformulated the question presented slightly, announcing that the issue was whether “certification was improper because respondents had failed to establish that damages could be measured on a classwide basis.” But reading the majority opinion as a whole, it appears that the majority arguably answered the question they planned to answer all along.
The district court in Behrend had held that to be entitled to certification under Federal Rule 23(b)(3), the plaintiffs were required to show that “the damages resulting from [the injury] were measurable ‘on a class-wide basis’ through use of a ‘common methodology.’” The sole basis for the court to make that finding was the multiple regression model of cable pricing built by the plaintiffs’ experts. But there was a problem with the model: the expert’s regression didn’t separate out damages flowing from overbuilder deterrence from damages flowing from other issues, including the other three possible theories of antitrust injury proposed by plaintiffs and rejected by the district court (at least for classwide treatment). Instead, the model merely predicted what prices would have purportedly prevailed in a “but-for” world absent the defendant’s alleged anticompetitive activities. The defendants had pointed this out before the Third Circuit, but the court had refused to even consider the issue, labeling it a merits question.
In fact, Justice Scalia’s opinion for a five-Justice majority found, the Court’s recent precedents on Rule 23 not only permitted, but flatly required such an analysis. “The first step in a damages study is the translation of the legal theory of the harmful event into an analysis of the economic impact of that event,” the Court noted, quoting the Federal Judicial Center’s Reference Manual on Scientific Evidence. Since the plaintiffs had presented no classwide theory of damages flowing from their one certified theory of injury, Rule 23 was not satisfied and certification was improper. Justices Ginsburg and Breyer filed an unusual joint dissent, joined by Justices Sotomayor and Kagan. The dissenters argued that once the Court discovered the procedural problems with the Daubert issue, the Court should have dismissed certiorari as improvidently granted (what the Court calls a “DIG”).
As I noted above, the bar has been somewhat split in the days since Behrend about just how important a win it is. For example, some attorneys interviewed by Law 360 argued that the decision was a powerful new tool for defense counsel, while others suggested that given the expectations for a full-blown analysis applying Daubert to certification, Behrend had turned out to be much ado about relatively little. Global Regulatory Enforcement Law Blog comments that Behrend is “sure to give class action defendants ammunition to attack expert evidence of classwide damages.” Class Action Countermeasures calls the opinion “limited,” but agreed that it “provides some help” to defendants.
Several conclusions seem clear from the opinion. First, as the employment bar has already noted, cases such as wage and hour class actions, where damages tend to be inherently individualized, may prove significantly harder for plaintiffs to get certified. Second, as Insurance Class Actions Insiderhas correctly pointed out, Behrend will make it far more difficult for the plaintiffs’ bar to persuade lower courts that individual issues in damages calculations are irrelevant to the certification issue, since all of the statements found in Behrend trying to limit the breadth of the majority opinion on this issue are found in the dissent.
But there’s more here than that. The Supreme Court originally granted certiorari on the question of whether a Daubert
challenge – the proposition that the plaintiffs had no admissible model of class-wide damages – could block class certification. Although the majority seems to gently turn aside that issue, the majority then holds that the regression model built by plaintiffs’ expert is insufficient because it fails to distinguish between various sources of possible higher-than-competitive pricing. But this is simply another way of saying that had the defendants brought a Daubert
challenge, the expert’s model would have been excluded; it answers the wrong question and is therefore irrelevant testimony. Specifying the appropriate number of variables to separately quantify every plausible influence on price is hardly an unusual issue in properly constructing a multiple regression model – it can fairly be said that that is what such models are for. The statistical biases created by using too few variables are well understood in the literature. At the very least, defense counsel should raise every appropriate Daubert
challenge at the certification stage post-Behrend
in order to give the Court to opportunity to formally and expressly decide the issue they wanted to address here. But closely examining the opinion with an understanding of multiple regression theory suggests that in fact, a majority of the Court has already made up its mind about the Daubert