Elizabeth Chamblee Burch is an Associate Professor at the University of Georgia School of Law, where she teaches Complex Litigation, Mass Torts, and Civil Procedure. She has written extensively about class actions and multidistrict litigation and is the author of the forthcoming second edition of The Law of Class Actions and Other Aggregate Litigation (with Richard A. Nagareda, Robert G. Bone, Charles Silver, and Patrick Woolley). She is also a co-editor of the Mass Tort Litigation Blog.
The Supreme Court's decisions this Term, particularly Wal-Mart Stores, Inc. v. Dukes and AT&T Mobility, LLC v. Concepcion, are the latest in a long line of both Congressional and court-based maneuvers that make certifying a class increasingly difficult and nonclass aggregation exceedingly likely. Since the mid-1990s, both Congress and the courts have taken steady measures to curtail the prevalence of class actions by providing federal courts with jurisdiction, making it harder to plead claims, and raising the threshold for class certification. Consider, for example, the changes wrought over the past fifteen years by the Private Securities Litigation Reform Act, the Securities Litigation Uniform Standards Act, Amchem Products, Inc. v. Windsor, Ortiz v. Fibreboard Corp., and the Class Action Fairness Act, to name but a few.
Although many of these changes made class actions more difficult to certify, Wal-Mart Stores, Inc. v. Dukes and AT&T Mobility, LLC v. Concepcion further threaten the already fragile class action and, as I discuss shortly, procedural legitimacy. The latter decision enables companies that contract with consumers "“ think phone, credit-card, Internet, mortgage, and cable companies "“ as well as those that contract with their employees (and what company doesn't?) to include class-action waivers and arbitration provisions within those contracts. This decision effectively shields companies against small claims that are not economically viable to litigate individually. Although several commentators have suggested arguments for distinguishing Concepcion, such as instances in which plaintiffs cannot effectively vindicate their statutory rights and thus have no access to justice or where a ban on class actions would safeguard companies from liability, the courts have only just begun to consider the parameters and merits of these arguments. So far, courts have split principally along state-federal lines, with state courts finding ways around the Concepcion opinion and federal courts dutifully adhering to it.
For claims that escape arbitration, Wal-Mart Stores, Inc. v. Dukes dramatically changed the relative ease with which plaintiffs' counsel could certify not only a Rule 23(b)(2) class, but classes across the board. The Dukes majority endowed defendants with three new forms of artillery: a formidable commonality standard under Rule 23(a), an impossibly high standard for recovering monetary remedies on a classwide basis under Rule 23(b)(2), and, lest a class squeak by under Rule 23(b)(3), the ability to raise individual defenses. To illustrate these barriers, imagine a scenario in which Employer with offices throughout the country discriminates against women in its pay and promotion policies. Of course, Employer learned long ago not to put any discriminatory policy on the books or to use biased testing procedures, so the discrimination manifests through supervisory discretion. So, to clear Rule 23(a)'s commonality barrier, plaintiffs will have to propose bite-sized classes that target a specific supervisor, store, or plant. Less sprawling class definitions, however, mean that plaintiffs will tend to be geographically concentrated and fewer in number. Thus, while targeted classes may clear the commonality hurdle, they may not meet Rule 23(a)'s numerosity requirements.
Nevertheless, assume for argument's sake that plaintiffs do clear Rule 23(a)'s hurdles and reach Rule 23(b)(2). The attorney has two options. First, she could request a hybrid class in which indivisible relief proceeds under Rule 23(b)(2) and divisible relief "“ such as back pay, compensatory damages, and punitive damages "“ proceeds under Rule 23(b)(3). Or, second, she could ask the court to certify the Rule 23(b)(2) class for indivisible relief as an issues class and then permit each class member to proceed with requests for divisible remedies individually. The former option, however, is riddled with questions about when compensatory and punitive damages cause individual issues to predominate over common ones, a question that Dukes did not address.
Courts considering whether to award punitive or compensatory damages on a classwide basis before Dukes split over the issue. On one hand, several courts, including the Ninth Circuit in Dukes, suggested that punitive damages might clear the predominance hurdle under Rule 23(b)(3) if liability was predicated on a classwide theory. Even though plaintiffs must demonstrate harm before receiving punitive damages, requests for back pay "“ even in the absence of compensatory damages "“ have satisfied this requirement. On the other hand, some lower courts thought that Title VII required a fact-specific inquiry into whether each class member would be entitled to divisible remedies, which kept common questions from predominating under Rule 23(b)(3). Though it did not reach the issue, the Supreme Court's rationale in Dukes lends support to this latter view. Hybrid certifications are thus far riskier propositions, which makes it more likely that plaintiffs must proceed individually with damages claims, if they can proceed at all.
For the sake of argument, however, pretend that our attorney has somehow successfully shepherded the class through commonality and demonstrated that divisible damages do not defeat the predominance requirement. Still, Dukes adds a final hurdle: individual defenses. As Justice Scalia explained, under Title VII, "if the employer can show that it took an adverse employment action against an employee for any reason other than discrimination" "“ showing up late, poor customer service, false information on an application, or insubordination "“ "the court cannot order the "hiring, reinstatement, or promotion of an individual as an employee, or the payment to him of any backpay.'" Contrary to pre-Dukes case law, it also means that a defendant can inject individuality into an otherwise cohesive class, which undermines the predominance of common questions and suggests that a class action is no longer superior to individual cases. And individual defenses are not unique to Title VII cases; one needn't think too creatively to name a few "“ waiver, statute of limitations, assumption of the risk, and contributory negligence come readily to mind.
The point is, a putative class faces an ever-increasing number of certification barriers; there are more barriers today than there were three months ago and more three months ago than there were fifteen years ago. The upshot of both Concepcion and Dukes is that would-be class actions may either cease to enter the litigation realm at all "“ thereby creating an access-to-justice problem for meritorious claims "“ or, if appropriately funded, will proceed as mass torts typically do, as nonclass aggregation. Nonclass aggregation, where attorneys and courts join claims through Rules 20 and 42 and aggregate them through multidistrict litigation, falls into a black hole somewhere in between class actions and individual lawsuits.
In individual lawsuits, we trust that a client can monitor her own case. But when a single law firm represents hundreds or thousands of claimants, information about one's individual case is hard to come by. In class actions, where attorneys represent thousands of absent class members, we are justifiably concerned that when a decision pits the attorney's best interests against the members' best interests, the attorney's interests will win out. Accordingly, the judge steps in to appoint the class attorneys, determine whether to certify the class, approve the settlement, and award attorneys' fees. Trouble is, these quintessential principal-agent problems exist in nonclass aggregation just as they do in class actions, but, absent class certification, judges lack clear-cut policing authority.
As I have argued in more detail elsewhere, arbitration and nonclass aggregation raise broader questions about procedural justice and institutional legitimacy. That the Supreme Court's recent decisions will likely expand the prevalence and scope of both makes these questions all the more urgent. For example, procedural-justice research indicates that litigants prefer: (1) an adversarial system before an impartial decision-maker with error-correcting mechanisms such as new trials and appeals; (2) either well-established court rules or the ability to participate in designing dispute-resolution procedures; and (3) an opportunity to take part and be heard in the adjudicatory or deliberation process. But arbitration and nonclass aggregation procedures are often at odds with these preferences. In the mass-tort context, compensation grids and confidential settlements have replaced traditional error-correction mechanisms. Participation and voice opportunities wane as the number of litigants increase and attorneys find it increasingly taxing to communicate with their clients. Finally, judicial impartiality may be compromised by self-interest, demands for systemic efficiency, and the lack of quality information. Ultimately, plaintiffs may view the process itself as illegitimate. To offset this risk, institutional designers need to think more about how to balance efficiency with both substantive and procedural fairness and how to monitor the agency relationship, and less about additional ways to stifle class certification.