A common criticism of modern academic legal writing is its lack of usefulness in the real world of practice. A common criticism of writing directed toward solving workaday legal problems is its lack of theoretical sophistication. Neither of these criticisms can be levied against the new paper by Francis McGovern, a Special Master in the National Prescription Opiate Litigation multidistrict litigation (MDL), and William Rubenstein, a consulting expert to the MDL court. The paper is dazzlingly conceived, and Judge Dan Polster, who is presiding over the Opiate MDL, has employed the proposal and its reasoning to find a solution to one of today’s most intractable mass torts.
The proposal harnesses Rule 23 to a new purpose: to create a class action designed specifically to negotiate a settlement. Today class actions are certified to litigate claims and, more controversially, to settle them. But this proposal is different. It forms a class around a specific settlement structure, in advance of negotiating the settlement itself; in effect, without knowing whether the defendants will settle or for how much, the class agrees about how the proceeds of a possible settlement will be distributed among class members. Class members have a right to opt out once the structure is known, so those dissatisfied with the proposed distribution plan may exit and the defendants may know in advance of negotiations what the binding effect of the settlement will be. Should a settlement be achieved, class members, who at this point will know what their share of the settlement will be, then vote to approve or reject the settlement, with approval requiring the favorable vote of a supermajority. Class members also retain rights to object to the settlement before the judge decides whether to approve it.
This new form of class action is designed specifically for what they coin “heterogeneous classes.” In a “homogeneous class,” all class members have small claims. No class member has an incentive to opt out and pursue separate litigation because individual litigation would cost more than the claim is worth; as a result, the defendant can be confident that, in negotiating a settlement, there will be no opt outs and no knock-on litigation. In a heterogeneous class, some individual claims are large enough to be worth pursuing. Hence, when a defendant negotiates a settlement with the class, the defendant must hold back some money or negotiate escape clauses to hedge against class members with valuable claims opting out. That dynamic prevents the class from harvesting the “peace premium:” the extra amount that a defendant would be willing to pay to make the entire litigation go away once and for all.
Enter the negotiation class action. Before commencing negotiations with the class, the defendant knows who is in the class and who will be bound if the court approves any ultimate settlement. As the authors argue, it behooves the putative class members and class counsel to establish a settlement structure that will keep class members with large-value claims from exercising their opt-out right. With those large-value claimants remaining in the class, the defendant can obtain global peace, while class counsel can negotiate from a position of strength, capture the peace premium, and divvy up the proceeds of any settlement according to the agreed-on distribution plan.
McGovern and Rubenstein bolster their proposal with insights from the academic class-action literature. Agency costs—the fear that the class’s agents (the representatives and counsel) will sell out the interests of the class to enrich themselves—represent one of the great concerns of class actions. The traditional responses to agency costs include enhancing the agent’s loyalty (by requiring adequate representation), giving the class members voice (by allowing them to intervene and object), and permitting them to exit (by providing an opt-out right). The authors suggest that the negotiation class action employs a novel, fourth approach to reduce agency costs: cooperation. At the same time, the negotiation class action relies on traditional mechanisms: exit (through the initial opt-out right) and voice (through the right to object to the settlement). The authors also offer a new voice mechanism, borrowed from the ALI’s Principles of Aggregate Litigation and from bankruptcy reorganization, of requiring supermajority class approval of the settlement.
The negotiation class action is both a brilliant and a brilliantly flawed proposal. Harnessing the ALI’s supermajority voting requirement to the class-action context, in which voting rights can be employed without doing violence to the principle of litigant autonomy that has otherwise doomed the ALI’s proposal, is a clever way to give class members more voice than the often-ineffective route of filing an objection to a settlement. At the same time, a voting-rights proposal generates more questions than it answers: What is the proper supermajority (66%? 75%? 90%?)? How are votes counted (one person, one vote? weighted based on claim value? weighted based on other considerations, such as a class member’s location or claim strength?)? And is the majority calculated on the basis of the votes cast or on the basis of the class size (an essential question when class members may not be easily found or when small-claim class members may not bother to return their ballots)? The devil of the proposal lies in such details, and the authors only advert to these issues without providing a framework for thinking about them.1
Some theoretical underpinnings of the proposal are also debatable. The existence of a peace premium has become an article of faith in some academic circles, despite only a modicum of evidence that it might exist in the real world of aggregate-settlement negotiation. Another academic difficulty is the argument that the concept of cooperation is a fourth strategy to deal with agency costs. It is not; cooperation assumes away agency costs or, at best, is a way to execute a loyalty strategy. A simple tell reveals the weakness of this cooperation strategy: to ensure that agency costs do not infect the initial intra-class negotiation on the proper settlement structure, the authors must rely on strategies of exit (opt-out rights) and voice (objections and supermajority voting).
These concerns are perhaps troubling only to an academic. More practically troubling is the idea of the heterogeneous class. Claims of class members may be heterogeneous along a number of vectors. One vector—on which the authors build their proposal—is differences in the amount of damages that a class member suffers. In this regard, the authors’ idea creates a new label (“heterogeneous claims”) for an old principle—existing case law recognizes that claims of different amounts do not necessarily create internal differences or conflicts sufficient to scuttle class certification.
But another vector is the comparable strength or weakness of claims, in light of differences in applicable substantive law, defenses, or evidence. Class certification of heterogeneous mass torts with these characteristics often founder against one or more of the elements necessary to attain certification (commonality, typicality, adequacy of representation, or predominance). What Richard Nagareda called the pre-existence principle, which seems to underlie the Court’s decisions in Amchem and Wal–Mart, provides that these elements cannot be satisfied by a common negotiation strategy or distribution plan. It becomes difficult to imagine that negotiation classes could be certified except in cases in which the only relevant differences among class members is the amount of damages they suffer. But these are the cases in which a novel negotiation class action is least needed, for certification of a litigation or settlement class action is already possible.
The advantage that the negotiation class may offer over litigation or settlement class actions is a means to keep large-scale claimants in the class. But these claimants will remain in the class, rather than opt out, only if they perceive that the distribution plan negotiated at the outset favors them—that they stand to gain more from the class action than from individual litigation. A negotiation class action might promise more gain for a number of reasons: because they share in a peace premium, because they can spread litigation costs among all class members, or because the settlement structure overcompensates them in relation to small claimants. The first possibility, on which the authors seem to hinge the proposal, is speculative—in no small part because, once the defendants know the distribution plan, they can game their negotiating strategy accordingly. Unable to use the possibility of large-claimant withdrawal as leverage to negotiate a sweeter deal, class counsel might be negotiating with one arm tied behind its back. The second possibility is not unique to negotiation class actions but is shared with litigation or settlement class actions. And the last possibility raises disqualifying conflicts of interest within the class—conflicts of the kind that cratered the settlement class actions in Amchem and Ortiz.
Given its conservatism on class actions, the Supreme Court would almost surely “disapprove th[e] novel project” of negotiation class actions, as it disapproved “trial by statistics” in Wal-Mart. At best, the device might find use in a narrow swath of class actions in which claims are heterogeneous only with respect to the amount of damages that class members suffered. Time will tell, and the early returns are coming soon—the order certifying the Opiate Litigation negotiation class action is on appeal.
It pains me to be skeptical. The need to create useful solutions to resolve the intractable problem of mass torts is palpable. And the authors’ proposal is one of the most promising and innovative developments to come along in a great while.
- In his opinion certifying the class of municipalities and counties seeking recovery from pharmaceutical companies for saturating their communities with opioids, Judge Polster contemplated that each governmental unit would cast one vote, which would then be weighted along three metrics on one scale and two on another, for a total of six voting classes. Approval for each voting class required a favorable vote from a 75% supermajority.