Argument recap: You say tomato . . .
on Nov 8, 2012 at 3:10 pm
There were moments during Monday’s oral arguments in Comcast Corp. v. Behrend when the Justices expressed real frustration with the parties, both of whom focused their arguments on the facts of the case rather than the legal rule the Court should adopt. But the parties did communicate some crucial differences in how courts should evaluate expert testimony on damages for class certification purposes.
Let’s call the whole thing off
Both parties focused a great deal on the damage model at issue in the case. Miguel Estrada, arguing for Comcast, emphasized two flaws. First, he noted that the district court only certified a class as to one theory of antitrust liability –that Comcast’s alleged monopolization of the Philadelphia area through clustering deterred competitors, or “overbuilders,” from competing. However, the record only showed one potential overbuilder, RCN, and the evidence showed that RCN only threatened to overbuild in 5 of the 649 franchise areas that comprise the Philadelphia area. Accordingly, Estrada argued that the damage model presumed that RCN was “the little engine that could” whose effects would “ripple” or “radiate” throughout the area. Second, he argued that it was “startling” that the damage model was not linked to the only theory certified by the court. He concluded that the model was “bunk,” and the district court should not have relied on it to certify the class.
For obvious reasons, Barry Barnett, arguing for the respondents, was more bullish on the damage model. He stressed that the model was “agnostic” as to the underlying antitrust theory of liability, and thus would apply regardless of which theory, or theories, the court certified. Moreover, Barnett argued enthusiastically that the damage model was far from flawed, stating that “[t]his is a fantastic model.” He even asserted that the respondents would “embrace whatever Daubert standard anybody wants to apply retroactively.”
The focus on the damage model seemed to annoy the Justices because the parties seemed to agree on the underlying standard that should apply. Justice Kagan pointed out to Estrada that, apart from “the ultimate determination” on the expert’s report, the district court seems to have done “exactly what you say they should have done.” Similarly, when Barnett suggested that the respondents would “embrace” any standard the Court adopts, Justice Sotomayor responded, “so you are not disagreeing with your adversary on a legal standard.” At one point an exasperated Justice Kagan exclaimed that “I am still in search of a legal question that anybody disagrees with here.”
A Daubert hearing by any other name?
A disagreement began to surface once the Justices focused on the respondents’ waiver argument. In their brief, the respondents argued that Comcast waived any challenge to the admissibility of the damage model by failing to move to exclude the damage model under Daubert until the eleventh hour. At first the Justices prodded Estrada on the issue. Justice Kagan expressed some skepticism about Estrada’s argument that the issue was not waived because one can infer that Comcast raised the issue by challenging the damage report, noting that “it seems like a remarkable proposition, honestly, especially with a client like yours that is well lawyered.”
Justice Scalia, however, threw a lifeline to support Estrada’s implied challenge argument. He noted that Comcast had originally framed the issue as whether a district court could consider the merits for class certification purposes, and it was the Court who reformulated the issue to whether a district court must consider the admissibility of the evidence under Daubert. Justice Scalia remarked that, in reframing the issue, “we must have thought” Comcast meant to challenge the admissibility of the damage model under Daubert, “or else we wouldn’t have reformulated the question this way, right?”
Prior to Justice Scalia’s lifeline, Justice Sotomayor attempted to move the conversation away from the waiver issue to what, exactly, the district court was required to do, Daubert label or not. Justice Sotomayor pushed Estrada to “get out of the Daubert language, okay?” and asked “what rule we [should] announce” so that courts know what to do when faced with challenges to expert reports on damages at the class certification stage. Barnett tried to resist this more functional approach to the issue by insisting that labels matter, and that he is not advocating a “magic words” approach: “if you’re in a trial court and somebody says Daubert or somebody says Rule 702 or somebody says I object to this expert’s testimony, that has profound significance.”
The Justices, following Justice Sotomayor, eventually focused on whether a Daubert analysis, no matter what it is called, should be conducted by the district court for purposes of class certification. Estrada argued, in response to the waiver issue, that a Daubert analysis includes a look at “fit” and “reliability,” and argued that Comcast has consistently argued that the damage model “is so unreliable that it is just not usable period, full stop.” Barnett argued for a standard in which the district court is persuaded that it is “[m]ore likely than not that” the damage model “will be admissible at trial, and it will meet the standard that’s required to get to a verdict.” In response, Justice Sotomayor seemed to take up Estrada’s parsing of Daubert, telling Barnett that “the bottom line is, can a district court ever say that it’s persuaded by unreliable or not probative evidence.” Justice Alito, noting that the dispute was “factual,” also asked “could this report be probative if it did not satisfy Daubert?”
As the Justices focused more functionally on what a district court should do in considering expert testimony for purposes of class certification, the parties eventually displayed some differences. Estrada stressed that, under Federal Rule of Civil Procedure 23, a court cannot certify a class unless it finds that common issues “predominate.” He then argued that, in making a finding of predominance, a district court cannot simply conclude that damage issues could be proven on a classwide basis eventually, “that providence will provide; we’ll think about it in the morning.” Although Barnett was very confident about the validity of the damage model, he also admittedly “hesitate[d]” in accepting the Chief Justice’s reformulation of the standard as “[m]ore likely than not whether it can be used at trial,” noting that usability is “something you don’t know.” Justice Scalia suggested an approach where a district court simply focuses on the reliability prong of Daubert and defers any “Daubert motion” for later, an approach that Barnett seemed to embrace: “Justice Scalia, I would say what you’re doing is what I suggest the Court ought to do.” Thus, unlike Estrada, Barnett expressed some comfort with a conditional finding of reliability for certification purposes, stating during his argument that a district court does not “have to decide whether this model is perfect.”
As to the issue of whether the model had to be perfect, Justice Kennedy had an interesting exchange with Barnett. Justice Kennedy questioned whether the full Daubert analysis was appropriate at the class certification stage when the judge plays no “gatekeeping” role in terms of excluding prejudicial evidence from the jury. He noted that “[t]he trial judge at the end of the day can hear the testimony, say: You know, I admitted this testimony, but it doesn’t make any sense. It doesn’t work.” Barnett disagreed with Kennedy’s position, arguing that the judge is, in fact, a “gatekeeper, and what he’s doing or she’s doing is projecting: What’s this trial going to look like based on the evidence in front of me.” At first glance Barnett’s push back was surprising because Justice Kennedy seemed to be agreeing with the conditional approach favored by the respondents. But the push back makes sense because Barnett probably was concerned that Justice Kennedy was analogizing class certification to a bench trial without taking into account the role of the certification inquiry – to ensure that a class action would be manageable at trial, not to determine the merits of the claims. To his credit, Justice Kennedy stated that “I’ll think about it, but that’s not my understanding.”
Reading the tea leaves
Given the murkiness of the parties’ positions, particularly on the waiver issue, the Chief Justice suggested a middle approach whereby the Court would define the standard of review and then “send [the case] back for the court to determine whether or not the parties adequately preserved that option or not.” Although it is difficult (and perhaps foolish) to divine the Justices’ positions from their questions, it appears that a healthy majority favor some form of finding of reliability of expert testimony on damages for class certification purposes without necessarily requiring a full-blown Daubert analysis.
Nevertheless, two outlier positions are worth highlighting. First, Justice Ginsburg, a former civil procedure professor, questioned whether any type of finding of reliability was necessary. She noted that in contexts like Title VII, “if the liability question can be adjudicated on a class basis, then the damages question may be adjudicated individually.”
Second, Justice Breyer, focusing on the substantive antitrust law, noted that a monopolization claim under Section 2 of the Sherman Act does not require the plaintiffs to “trace the damages to the exclusionary conduct.” Thus, Breyer pressed Estrada on whether the damages could be calculated simply by comparing actual prices to a competitive baseline, allowing the court to “[c]ertify the class and leave the damages for later.” Breyer, in fact, repeatedly suggested that a bifurcation approach is a matter of “hornbook law.”
As I noted in my preview, a bifurcation approach was in fact the accepted approach for antitrust class actions, an approach taken by Justice Sotomayor while she was on the Second Circuit, but would later retreat from somewhat in a later opinion. Based on the questioning, it appears unlikely that a bifurcation approach will remain “hornbook law.”
[Disclosure: Goldstein & Russell, P.C., whose attorneys work for or contribute to this blog in various capacities, is among the counsel to the respondents in this case. However, the author of this post is neither associated with the law firm nor involved in the case.]