Advance Sheet: Class Action Answers—and Questions

The protagonists in Dukes seemed right out of central casting for a morality play of this kind, the outcome of which would determine decisively which side the legal system was on. For class action activists, the plaintiffs were necessarily the “good guys” or, rather, a deserving class of women wage earners who claimed to have been victimized in hundreds, if not thousands, of decisions on promotions in a system dominated by a boys club of male supervisors. Their employer and supposed tormentor was Walmart, the contemporary bogeyman of activist labor policy. Surely, there could be no greater proof of the usefulness and validity of the class action.

From the standpoint of the detractors, however, the defendant was not a bad actor at all, but the epitome of civic virtue standing tall against the depredations of class action lawyers. Not only did Walmart have a staunch written policy against discrimination in any form, but as those who have represented Walmart in recent years (including this author) are keenly aware, that company has worked extra hard to bring true diversity into its hiring and promotions, launching initiative after initiative on women’s issues, among others. From the perspective of Walmart and its supporters, the class attack seemed atavistic and certainly more than a little unfair.

The Court’s Workmanlike Opinion

The decision disappointed those on both sides who demanded a moralistic conclusion to what really was a far more mundane legal controversy. Dukes was mostly about the practical elements of Federal Rule of Civil Procedure 23, not moral issues of epic proportions. And the Court rightly responded not ideologically, but in a sober, workmanlike fashion, achieving remarkable unanimity on most of the underlying questions.

Nor should this have surprised anyone. Whether viewed from a theoretical or a practical standpoint, there is some truth in both views of class actions, and complete truth in neither. If Rule 23 and its state court counterparts were devised, and have served, as a remedy for the disadvantaged and unrepresented, they have also too often fallen into the hands of those interested only in pelf. A novel procedure, and still a relatively new one from the perspective of our centuries-long experience with the common law, class actions should never have been expected to operate successfully in all cases. There were certain to be cases of abuse, particularly in the hands of shrewd lawyers pursuing a payday. But misuse has not been a feature of most cases, let alone every one. The question in recent times has been whether the correct balance of pluses and minuses has been upset—whether, through the wiles of lawyers and the laxity of the courts, the class action had lost its place as a true means of providing a remedy to the needy and had become a resource for greedy lawyers alone.

Likewise, in practical terms, the case itself presented less of a stark controversy than may first have appeared, notwithstanding the very substantial financial stakes involved. The women were not complaining so much that Walmart had intended to discriminate. Its alleged sin was to have given too much discretion to local managers, with the result that existing biases among male supervisors could operate unconsciously. And the issue was less whether Walmart’s anti-discrimination policies were right or proper than whether the best way to test and remedy any defects in them lay within the procedures available under Rule 23(b)(2).

For this reason, perhaps, the result had a greater effect on subsequent class actions than any grand statement of class action policies or purposes might have had. Take, for example, the Court’s reading of the differences between Rule 23(b)(2) and Rule 23(b)(3). Dukes was brought under the former, which provides for injunctive relief, but sought back pay more familiar under the latter. Was this permissible? Second, how far should the courts probe in assessing compliance with Rule 23’s standards? A third issue was the true meaning of the common question requirement of Rule 23(a).

On the first question (actually discussed last in the opinion), the entire Court agreed that back pay damages of the sort sought by the Walmart employees did not fit within a Rule 23(b)(2) class action. According to the Court, the slimmed-down procedures of Rule 23(b)(2), with its lack of concern about individualized notice and absence of an opt-out procedure, are incompatible with a back pay remedy. There is a risk, the Court said, that the damages question would simply overwhelm the proper use of Rule 23(b)(2), which is to stop unlawful conduct, not to seek compensation for it. And Walmart, it said, was entitled to raise individual defenses to back pay determinations, though it is unclear why those could not be raised in the course of the separate and later back pay hearings. In any event, from now on, a class action seeking back pay damages for the most part will be relegated to Rule 23(b)(3).

This is far from insignificant. A cursory view of the rule would suggest just this division of labor between Rule 23(b)(2) and Rule 23(b)(3), the former being for injunctive relief (where the individualized procedural demands would be less) and the latter for damages actions. Yet, everyone knows that such a division between equity and law is never so sharp in reality. It was exactly this fact that permitted clever plaintiffs’ lawyers to try to get the benefits of a (b)(3) claim in a (b)(2) action without the burdens associated with the former. The Court unanimously resurrected a bright-line division, though it remains to be seen whether it is a workable one. Indeed, it would seem that Rule 23(b)(2) will now be used only rarely, in those few cases where purely equitable relief is all that is wanted or needed.

The Court had a more difficult time with the procedures for evaluating a class action. Both the majority and the dissent seemed to believe that some probing was necessary. Their disagreement was less about whether it should be done than about what standard should be applied to the underlying evidence. The majority doubted that experts, for example, should be given less than full Daubert scrutiny in the context of a class certification motion. The dissent seemed less comfortable with such exacting analysis but was unwilling to give experts a free pass either.

Likely Short-Term Effect

This aspect of the decision in Dukes is likely to have the most immediate impact. To begin with, the Court’s rule would seem to apply to all actions brought under any of the parts of Rule 23(b). Heretofore, class certification petitions have not always, if ever, garnered exacting scrutiny from the courts. If a case appeared, mostly from the pleadings, to meet the relevant standards, then certification was likely. The invitation to the courts—and, more important, to defendants’ lawyers—to attack a certification motion on the merits is likely now to turn every class certification motion into a substantive battle, where fewer such motions will survive. If the point of Dukes was to reconsider the balance between plaintiffs and defendants, this is a serious shift in procedure toward the defense bar.

Dukes also will impose at least some period of re-assessment on the courts. Exactly what level of scrutiny applies is anybody’s guess. Doubtless the trial courts will have to figure that out, even as the class actions continue. Eventually, whether through immediate appeals under Rule 23(f) or appeals from final judgments, greater clarity should emerge.

The greatest degree of difference between the majority and the dissenters on the Court arose in the context of determining what constitutes a common question. The majority criticized the Ninth Circuit and the courts generally for approaching the common question requirement of Rule 23(a) from too much of a semantic standpoint. It noted that clever lawyers are always able to find common questions. As a simple example, they could ask, “Are the plaintiffs entitled to a finding of liability on discrimination?” Clearly, this is a question that pertains to everyone in the class and therefore appears common. But is this the kind of common question the rule has in mind?

The Court’s majority emphatically said “no.” But instead of investigating what kind of question is truly common, the majority changed the inquiry. The issue, it said, was not really common questions, but common answers. That is, if the common question is whether plaintiffs are entitled to a finding of liability on discrimination, it turns out not to be common to the class because the answer depends on a host of factors where each plaintiff will appear different. If each plaintiff is required to make an individualized showing in order for liability to attach, the question is not common at all.

There is a certain sense to this approach, but one wonders whether the Court’s new formulation is proper, let alone likely to be truly superior to the common-question inquiry that preceded it. The rule clearly speaks of common questions, not common answers, though the majority seemed to address this by saying it was referring to common questions that were susceptible to common answers. The dissenters, by contrast, argued that this was a subterfuge for introducing into Rule 23(a) the requirement that existed only in Rule 23(b): that common questions predominate.

They have a point. Isn’t the liability of a defendant to a plaintiff always dependent on individual determinations? Shouldn’t the question be whether the factors governing these individual determinations are common or different? Recall that the case was one in which Walmart was accused of having given its managers too much discretion. Necessarily, every decision was individual. But if those individual determinations followed a common pattern or derived from the same factors, was there not a common question and even a common answer? In Dukes, the real problem was that the plaintiffs were unable to provide a basis to conclude that there was a common pattern or factor at all, rendering the inquiry devoid of both common questions and common answers.

The Court’s Underlying Goal

In the end, however, what’s truly at stake here is what kind of common questions are appropriate and what kind are not. The common answer doctrine limned by the majority seems only a way of classifying which common questions qualify. It remains to be seen whether this new formulation really has the long-term effect of bringing the common question requirement back into a more tractable mode.

But there are reasons to doubt it will. Won’t clever lawyers now find ways to state the common questions so that they appear to have common answers? If so, whatever reform might have been achieved by Dukes may turn out to be short-lived. There is also the converse concern. Will defendants’ lawyers now be at too great an advantage, chopping down even meritorious class claims because the relevant questions can be stated or re-formulated so as not to appear to have common answers at all?

These inquiries are important. The goal of the Court in Dukes, among others, clearly was to reduce the number of class actions by making them tougher to bring. Restricting Rule 23(b)(2) claims to cases of pure injunctive relief will have that impact to some extent. Likewise, if the courts now give certification petitions greater factual scrutiny, there are sure to be fewer sustainable class actions. Undoubtedly, the common-question approach was intended to have a similar impact, reducing the number of class actions that will pass muster. The goal of moderating the impact of class actions will have been poorly served, however, if the Court has provided us with rules that do not, in the end, limit the scope and breadth of class cases in a legitimate way.

Is a reduction in the number of class actions enough? Don’t we want the right kind to survive and the wrong kind to wither away? Will these new rules have such an effect? Unclear. And there is a further and related question—whether the Court is currently going about the reform of litigation pleading and procedure in the right way. In recent years, there has been a strong sense among legal scholars and practitioners that the court system was out of kilter, at least in the increased volume of litigation. Has anyone noticed that the Court seems to have set out to fix the problem through a kind of incremental jurisprudence? The first stage in this process was the rewriting of the pleading rules in Bell Atlantic Corp. v. Twombly, 550 U.S. 544 (2007), and Ashcroft v. Iqbal, 129 S. Ct. 1937 (2009), which have fundamentally changed what is required to get a lawsuit started. Now the Court has taken on class actions, making changes that are designed to reduce the number, and the scope, of such proceedings.

The question is whether these changes are for the better. The answer does not lie simply in whether they will be effective in reducing the number of cases. Although the procedural simplicity of initiating an action is a factor in the proliferation of lawsuits, the origins of this crisis lie much deeper. There has been a noticeable change in society itself, where bringing lawsuits has become far more acceptable and common. Making slight adjustments in the procedural rules hardly seems like the correct prescription for the problem. Indeed, it has a little bit of the character of a band-aid solution for a much more substantial wound.