It isn't complicated. A majority of the Supreme Court holds that, where an expert's model will not prove damages on a classwide basis, “[q]uestions of individual damage calculations will inevitably overwhelm questions common to the class,” and Rule 23(b)(3) predominance is not satisfied. As Justice Scalia explains, “[t]his case thus turns on the straightforward application of class-certification principles . . . .” There is nothing unclear about the Court's holding.
Yet no sooner does the Court issue its opinion than the debate begins, in the form of the dissent co-authored by Justices Ginsburg and Breyer. Despite the clarity of the majority decision, they write that it “should not be read to require, as a prerequisite to certification, that damages attributable to a classwide injury be measurable 'on a class-wide basis.'” In other words, it doesn't mean what it says. The dissent relies on the parties' agreement that the plaintiff needed to establish damages on a classwide basis, even though the majority opinion expressly rests, not on the parties' agreement, but on its interpretation of Rule 23 and “straightforward class action principles.” The dissent's effort to nullify the majority opinion requires disregard of its structure and wording. Other nullifiers try to limit the decision to antitrust cases (a strategem that was attempted and failed with the Supreme Court's decision in Twombly), or to cases in which the expert's methodology does not match the plaintiffs' liability theories.
Underlying the dissenting opinion is the widely held view that, “when adjudication of questions of liability common to the class will achieve economies of time and expense, the predominance standard is generally satisfied even if damages are not provable in the aggregate.” In support, they cite W. Rubenstein, Newberg on Class Actions (a treatise that courts often cite when granting class certification without questioning its objectivity) and several federal appellate decisions. According to these authorities, “a class may be certified for liability purposes only, leaving individual damages calculations to subsequent proceedings.” Since this view “that invidividual damages calculations do not preclude class certification under Rule 23(b)(3) is well nigh universal,” the dissent concludes, courts should not interpret the majority opinion as changing it.
So far, most federal appellate courts line up with the dissent. The Sixth Circuit, in In re: Whirlpool Corp. Front-Loading Washer Products Liability Litigation, and the Seventh Circuit, in Butler v. Sears, Roebuck & Co., both of which were remanded from the Supreme Court after Comcast was decided, have held that Comcast does not preclude certification of a liability-only class, leaving damages issues for later determination. The D.C. Circuit, on the other hand, vacated and remanded a district court's certification decision in In re: Rail Freight Fuel Surcharge Antitrust Litigation, holding that, after Comcast, it is clear “that Rule 23 not only authorizes a hard look at the soundness of statistical models that purport to show predominance–the rule commands it.”
Courts that follow the Comcast dissent invoke Rule 23(c)(4), which permits certification of issues rather than whole cases. Years ago, the Fifth Circuit, in Download Castano v American Tobacco Co, decried this application of Rule 23(c)(4), explaining in a footnote that (c)(4) is simply a housekeeping rule, and that for purposes of certification under Rule 23(b)(3), a case as a whole must satisfy predominance; otherwise, because virtually every case contains some common issues, the nimble use of (c)(4) would allow virtually any case to be certified, and (b)(3) would be eviscerated. The Comcast majority opinion seems closer to Castano than to the newer wave of cases.
I have always had problems with the notion that courts can certify liability-only classes in cases where damages are individualized. In my view, courts that do so fail to apply the “rigorous analysis” that Rule 23 requires, because they have no clear view at the time of certification how the damages issues will play out. Judge Posner, in Butler, relies on his belief that “[t]he parties probably would agree on a schedule of damages based on the cost of fixing or replacing class members' mold-contaminated washing machines,” and if they do, “the hearings would be brief; indeed the case would probably be quickly settled.” He and other judges also rely on the ability to certify subclasses to help resolve damages issues, without defining what types of subclasses will resolve the problem. Both the Butler and Whirlpool decisions also suggest that defendants “should welcome” certification of liability classes that could result in defense judgments binding on the class, a view that is out of touch with the realities of corporate risk management. As the Fifth Circuit explained in Castano, and some courts continue to recognize, the mere pendency of a class action can place disproportionate pressure on a corporate defendant to settle unmeritorious cases. While courts certainly have a responsibility to protect the rights of potential class members, they should be equally vigilant in preventing legalized extortion.
Perhaps the reason for this stark contrast between the Comcast majority's approach and the approach of the dissent is that we tend to view the choice through polarized lenses: either liability and damages can be bifurcated in all cases or they can be bifurcated in none. The first of these extremes is evident from Whirlpool, in which the Sixth Circuit relied on its earlier precedent that “'[n]o matter how individualized the issue of damages may be,' determination of damages 'may be reserved for individual treatment with the question of liability tried as a class action . . . .'” The other extreme would hold that Comcast has changed the application of Rule 23(c)(4) and that it no longer is appropriate to certify liability-only classes in any 23(b)(3) case.
A closer look at the Rule and the case law points to the possibility of a more balanced interpretation. Rule 23(c)(4) permits issue certification “[w]hen appropriate.” This language suggests that the propriety of bifurcation should be determined on a case-by-case basis, an approach that would be more consistent with the requirement of rigorous analysis. Whether bifurcation is appropriate could turn on the degree to which damages are individualized, and the degree to which individualized determinations of damages can be managed, consistent with due process, in the event of a classwide finding of liability. Some categories of cases may lend themselves to bifurcation more than others. For example, class members' damages may be readily determinable in a securities fraud case, where the only issue after liability is how much each class member paid for his or her shares. In contrast, wage and hour cases, in which each employee's claim may require fact-intensive investigations, would be too individualized for bifurcation; bifurcation would simply delay the resolution of individualized issues, not avoid them. Rather than simply assume that every case can be bifurcated, courts should consider the type of case before them, the evidence presented and the damages theories invoked to determine whether class members' damages will be capable of resolution through a manageable process if liability is found.
Also, to be true to the “rigorous analysis” mandate, courts should determine what that process will be at the time of the certification decision. Too often, courts simply recite the variety of mechanisms available for determining damages without determining how or even whether those mechanisms can work in the case before them. Courts also rely on the probability of settlement, which improperly assumes that the defendant will relinquish its defenses to damages, and its concomitant due process rights, if liability is established. While such approaches may seem responsible and pragmatic, they more closely resemble wishful thinking than rigorous analysis. They have no place in Rule 23 analysis, especially in this post-Comcast world.