THE justification for class actions rests on two main grounds: compensating victims whose claims are too small to be brought individually and deterring wrongdoing by aggregating claims to facilitate private enforcement.1 These two rationales overlap and compete with one another, as does their application to class actions certified under different subdivisions of Federal Rule of Civil Procedure 23. Broadly speaking, class actions certified under subdivision (b)(3) focus on compensation to individual class members, with deterrence resulting only from the defendant’s exposure to liability for paying such compensation, while class actions certified under subdivision (b)(2) focus on injunctions that prevent or deter future wrongdoing, without regard to the relief awarded to individual class members.2 In the recent decisions in Wal-Mart Stores, Inc. v. Dukes3 and AT&T Mobility LLC v. Concepcion,4 the Supreme Court cast further doubt on the deterrent function of the class action. More precisely, it sacrificed deterrence when compensation could not be accurately given. Wal-Mart restricted the remedies available in (b)(2) class actions to exclude individual monetary relief, and it also restricted the conditions under which any class action could be certified. AT&T Mobility restricted the conditions under which plaintiffs could get to court to bring a class action in the face of contracts requiring individual arbitration. These decisions are all the more significant for being widely misunderstood.
The misunderstandings begin with the alignment of the Justices on the precise holdings in each case. The decisions are cast as exhibits in the conventional left-right divide of the Roberts Court, with the five Justices identified as conservatives supporting the interests of businesses and voting to restrict class action, and the four Justices identified as liberals taking the opposite view. This standard view is true as far as it goes, but it does not go very far. Both decisions accord with longstanding trends, in which Justices from across the political spectrum have acquiesced. Wal-Mart represents only the latest in a long series of decisions in which the Supreme Court has disapproved of massive class actions. In employment discrimination cases, as Wal-Mart itself was, this series goes back to East Texas Motor Freight v. Rodriguez5 and in the law of class actions generally to Eisen v. Carlisle & Jacquelin,6 both decisions from the 1970s. AT&T Mobility follows an equally long line of cases endorsing arbitration of a variety of different claims, from admiralty to age discrimination, dating back to a securities fraud claim in Scherk v. Alberto-Culver Co.7 The recent decisions therefore do not reflect new-found sympathy for business or new-found hostility to class actions.
The line-up of the Justices in each also fails to conform to a simple left-right divide. Wal-Mart was unanimous on denying certification under (b)(2) as a class action in which “final injunctive relief or corresponding declaratory relief is appropriate respecting the class as a whole.” It had to be certified, if at all, under (b)(3) and meet the requirement that “questions of law or fact common to class members predominate over any questions affecting only individual members.” The Court divided five-to-four on the distinct question whether the action met the requirement under subdivision (a)(2) that “there are questions of law or fact common to the class.” This “commonality” requirement was not met, according to the majority, because the particular form of sex discrimination alleged by the plaintiffs—subjective decisions by Wal-Mart’s supervisors and managers denying raises and promotions to women—was not supported by evidence that generally applied to the 1.5 million current and former employees in the class. The Justices who dissented on this issue strongly disagreed with this conclusion, choosing to defer instead to the district court’s finding that there were common issues.
The division among the Justices on commonality under (a)(2) has received the most attention, but the unanimous holding on certification under (b)(2) has important implications in its own right. In employment discrimination cases, it eliminates the formerly widespread practice of certifying mixed cases for injunctions and monetary relief under (b)(2). In other cases, it also eliminates the possibility of combining broad injunctive or declaratory relief with individual awards of damages or restitution, all in a single class action under (b)(2). The diminished prospects for combined relief also diminish the prospects for recovery of attorney’s fees, either through fee-shifting statutes, fees from a common fund awarded to the class, or contingent-fee contracts, all of which make the amount of fees awarded more or less proportional to the relief granted. With the decline in recoverable fees comes a decline in the incentives of plaintiffs’ attorneys to bring class actions, with an accompanying decrease in the deterrent effect on potential wrongdoing. Denying monetary relief in (b)(2) class actions sacrifices the deterrent function of the class action to the compensatory function by limiting it to the special procedures in (b)(3) class actions. The procedures in (b)(3) class actions seek to assure individual class members of their right to obtain individual relief, mainly by giving them individual notice and the right to opt out. The requirement of predominance under (b)(3) also makes a difference, but the cost of individual notice and the risk of class members opting out pose greater obstacles to maintaining (b)(3) class actions. By contrast, class actions under (b)(2) involve less demanding forms of notice and impose mandatory participation upon class members.
No one can object to giving priority to the compensatory function of the class action—so long as deserving class members eventually receive compensation. Yet the claims of absent class members might be so small that, even within a class action, the cost of notice and administering a settlement overwhelms the amount at stake. The costs of litigating the class issues might further deplete the resources available for compensating the class, by focusing the litigation on class issues and leaving fewer assets in the hands of the defendant for compensation. The well-known conflicts of interest between class attorneys and members of the class also diminish the prospect that genuine relief will eventually reach class members. Class attorneys often have succumbed to the temptation to augment their own fees at the expense of compensation to individual class members. These problems also appear in class actions under (b)(2)—and perhaps in more aggravated form because of the less elaborate procedures than under (b)(3)—but they are less disturbing from the perspective of deterrence rather than compensation. The deterrent function of class actions can be served by extracting money and other relief from wrongdoers, regardless of whether it eventually goes to the class. The compensatory function requires more elaborate mechanisms of distribution and therefore increases the cost of litigation.
The Court’s unanimous holding denying certification under (b)(2) heightens the paradox of the compensatory class action: that protection of the rights of class members to compensation often comes at the expense of any class action at all. The best—in the form of precise awards of individual compensation—becomes the enemy of the good—in the form of deterrence and approximate relief. The Court also disapproved of any attempt to average awards to individual class members, dismissing it as “Trial by Formula.”8 The pragmatic advantages of deterrence and approximate relief had to be subordinated to the existing structure of Rule 23—and perhaps to the defendant’s right to hearing on individual relief under the Due Process Clause. The rule requires enhanced procedural protections for class members in (b)(3) class actions, and it limits (b)(2) class actions to claims for injunctive and declaratory relief for the entire class. The lower courts have tried to work around this scheme by splicing awards of monetary relief onto (b)(2) class actions by characterizing them as “incidental” or, in the case of back pay, “equitable” relief.9 The Court in Wal-Mart disapproves of both approaches, as well it might in a class action in which monetary relief could easily reach into the billions of dollars. The underlying structural problem with the rule remains, however, with little prospect that it might be resolved by expanding the terms of (b)(2). The recent amendments to the rule have left (b)(2) unchanged, while they have elaborated on other provisions in the rule, and the relevant legislation, such as the Private Litigation Securities Reform Act10 and the Class Action Fairness Act,11 has mainly restricted class actions. All of these developments assure the continued ascendance of precise compensation at the expense of deterrence as the rationale for Rule 23.
The same holds true of the Court’s divided holding on the absence of common issues of law and fact under (a)(2), a requirement that must be met for all class actions. In support of this holding, the Court offers a long-overdue qualification to the opinion in Eisen, which disapproved of a preliminary inquiry into the merits as part of the certification process. In Wal-Mart, the Court corrects any implication from Eisen that the merits are irrelevant to certification. Quoting opinions since Eisen, the Court endorses the principle that the “class determination generally involves considerations that are enmeshed in the factual and legal issues comprising the plaintiff’s cause of action.”12 For the majority, that principle had sharply negative implications for the commonality of the plaintiffs’ claims: that the practice of delegating promotion and pay decisions to local officials had a disparate impact upon women and that the failure to correct this impact amounted to disparate treatment. Legal doctrine under Title VII of the Civil Rights Act of 1964 supports such claims of disparate impact and disparate treatment arising from subjective decision-making procedures.13 The question in Wal-Mart was whether those claims were common to the class members, which in turn depended on the common evidence for those claims.
That evidence was sorely lacking. The plaintiffs submitted affidavits from 120 female employees recounting instances of sex discrimination, but these represented only a minuscule fraction of the total number of class members and applied only to scattered stores across the country. The plaintiffs also submitted a regression analysis, finding a nationwide shortfall in the number of women promoted, but this suffered from the opposite defect as the affidavits: it was not connected to decisions in particular regions or stores. Lastly, the plaintiffs’ principal expert testified to the presence of “implicit bias” against women, but he could not quantify its effect on Wal-Mart’s actual employment decisions with any precision. According to his own testimony, it might have affected as few as 0.5% or as many as 95% of the decisions.14 As noted earlier, the Justices who dissented on commonality strongly disagreed with the majority’s interpretation of both the evidence and the standard for finding commonality.
Putting these disagreements to one side, their very existence raises the question whether Wal-Mart would have come out differently if the plaintiffs had submitted stronger evidence of common adverse effects across the class. Suppose the plaintiffs had submitted a better regression analysis that met the majority’s objections and that made up for the limitations of the anecdotal evidence and the expert testimony on implicit bias. Requiring such evidence, however, approaches ever more closely a full-fledged examination of the merits, which still remains barred under Eisen. Wholly apart from that conceptual problem, requiring such evidence raises the cost to the plaintiffs of obtaining a favorable ruling on certification. In sum, the holding on commonality in Wal-Mart diminishes the prospect of certification and in doing so, diminishes the likelihood that a class action will be brought. The net effect is to reduce the defendant’s exposure to class-wide liability and the deterrent effect of class actions generally.
That is the dismaying lesson of Wal-Mart. It might not foretell the death of class actions, which the Supreme Court has continued to endorse in other respects,15 but it does diminish the frequency of class actions. AT&T Mobility has gone in the same direction by a different means—through enforcement of contractual clauses requiring arbitration, and by implication, barring class actions. The precise issue in AT&T Mobility concerned the pre-emptive effect of the Federal Arbitration Act on state law governing the unconscionability of contractual arbitration clauses. The plaintiffs had agreed to arbitration of any claim of any dispute with AT&T Mobility but only in their “individual capacity, and not as a plaintiff or class member in any purported class or representative proceeding.”16 Nevertheless, the plaintiffs filed an action in federal court, which was consolidated with a class action against AT&T Mobility. The plaintiffs relied upon decisions under applicable California law that made the contractual prohibition against class actions unconscionable, rendering the entire arbitration clause unenforceable.
The lower federal courts allowed the plaintiffs’ case to go forward, also relying upon California law, but the Supreme Court reversed. The majority opinion by Justice Scalia held that California law was preempted by the FAA and that the plaintiffs’ only remedy was through individual arbitration of their claims. The plaintiffs could not bring their claims in court, and they were bound by the contractual prohibition against class arbitration. California law, to the extent that it required some form of class relief, discriminated against the arbitration clause as compared to other contract terms and impeded the enforcement of the clause according to its terms, contrary to the purpose of the FAA. Justice Thomas joined the opinion of the Court only “reluctantly,” because of his views of the limited preemptive effect of the FAA.17 He reasoned that California law was not confined to grounds that would support the revocation of “any contract,” which defined the permissible role of state law under the FAA, and so, like Justice Scalia, he concluded that California law was preempted. Justice Breyer dissented on behalf of the four Justices conventionally identified as liberals. He argued that California law was broadly based on principles of unconscionability and not focused narrowly on disfavoring arbitration. The stability of the holding in AT&T Mobility therefore depends upon Justice Thomas, who would not enforce the FAA at all in proceedings in state court.18 It remains to be seen whether a future case might see him allied with the dissenting Justices.
Whether and how such a case would arise remains a matter of doubt, but more immediate limitations on AT&T Mobility might be inferred from the arbitration clause in the case itself. Whatever else might be said of the clause, it apportioned the costs and burdens of arbitration fairly among the parties, for instance, in requiring the company to bear the costs of arbitrating any nonfrivolous claim. Companies might now simply imitate these contract terms, with the expectation of frustrating any claim of unconscionability and forcing their customers and employees into individual arbitration. If this is the response to the decision, it will succeed in augmenting the ability of individuals to recover compensation through arbitration, but again, as in Wal-Mart, at the expense of the deterrent function of the class action. As Justice Breyer pointed out in his dissent, no one would litigate (or perhaps even arbitrate) an individual claim for the $30.22 at stake in AT&T Mobility.19 The ostensible commitment to individualized compensation, implemented through arbitration, frustrates the ability of plaintiffs to have their claims aggregated mainly for purpose of deterrence and approximate compensation.
The ingenuity of class action plaintiffs may yet avoid or defeat the restrictive consequences of Wal-Mart and AT&T Mobility. They are far more likely to succeed at this task through the incremental process of litigation, and perhaps state legislation, than they are through changing the Supreme Court’s mind—without a change in membership—or in obtaining liberalizing amendments to Rule 23 through the rulemaking process or legislation. Nearly fifty years have elapsed since the 1966 amendments to the Federal Rules of Civil Procedure transformed Rule 23 into the principal vehicle for mass litigation in our legal system. The decisions in Wal-Mart and AT&T Mobility reveal continued ambivalence, if not outright hostility, to using class actions for deterrence alone, even when compensation remains a distant goal that can be achieved only by approximations, which now have also been thrown into doubt. Because of the cost of administration, the attempt to precisely measure awards of individual relief often frustrates the goal of actually compensating victims of wrongdoing. Much can be said against this paradox—denying deterrent class actions when compensation class actions have no realistic prospect of success—but it appears to be an enduring feature of class action practice. These recent decisions raise the question whether it has become an obstacle that can only be removed by a fundamental restructuring of the rules that govern this form of litigation.
Preferred citation: George Rutherglen, Wal-Mart, AT&T Mobility, and the Decline of the Deterrent Class Action , 98 Va. L. Rev. In Brief 24 (2012), http://www.virginialawreview.org/inbrief/2012/04/14/Rutherglen.pdf.
[Back to this issue]