Emory Law Journal

Disaggregative Mechanisms: Mass Claims Resolution Without Class Actions
Jaime Dodge Assistant Professor of Law, University of Georgia. I am indebted to Richard Freer for his role in making this colloquy possible. I would like to thank Dianne Amman, Beth Burch, Dan Coenen, Michael Coenen, Glenn Cohen, Jim Grenier, John Goldberg, Matt Hall, Bert Huang, Jeremy Newman, Matthew Stephenson, and Rory Van Loo for their encouragement and immensely helpful comments. I am also grateful to participants in colloquia at Emory University School of Law and Peking University’s School of Transnational Law, as well as the Harvard/Stanford/Yale Junior Faculty Forum for their generous feedback on earlier versions of this paper. Rory Weeks and Alan Rozenshtein provided assistance transcending my expectations.

Abstract

Aggregation has long been viewed as the primary, if not sole, vehicle for mass claims resolution. For a half century, scholars have consistently viewed the consolidated litigation of similar claims as the only mechanism for efficiently resolving mass claims. In this Article, I challenge that long-standing and fundamental assumption. This Article seeks to reconceptualize our understanding of mass claims resolution, arguing that we are witnessing the birth of a second, unexplored branch of mass claims resolution mechanisms—which I term “disaggregative” dispute resolution systems because they lack the traditional judicial aggregation of victims that has been the hallmark of mass claims litigation.

Disaggregation returns to a focus on the individual akin to that of the single-plaintiff system, but uses either procedural or substantive streamlining, or a shift of costs to the defendant, to ameliorate the asymmetries that prompted the creation of class actions. Many of our most innovative claims structures—from the BP Gulf Coast Claims Fund and the fund created in the wake of the Costa Concordia disaster, to the common single-plaintiff arbitration clauses in consumer and employment agreements—use this new, bottom-up model of disaggregative mass claims resolution instead of the familiar top-down aggregative model.

These next-generation systems have been heralded as a significant advancement in mass claims resolution, capable of awarding greater compensation to claimants more quickly and at a lower cost than aggregate litigation. But like the single-plaintiff and aggregate litigation systems that preceded it, disaggregation has its flaws. Because the defendant typically designs these systems, they often give rise to questions about legitimacy and the accuracy of compensation. Yet, because these systems are the product of contract, attempts to restrict them have largely failed. This Article tees up not only the problem these privately ordered processes are creating for the enforcement of traditional public mechanisms of aggregation, but also the problems with the public system that drove the private demand for disaggregation. The Responses to this Article published in the remainder of this colloquy begin to explore the consequences of this new approach to mass claims and the array of potential public mechanisms for bringing aggregation and disaggregation into balance.

Introduction

Aggregation was arguably the single most important procedural innovation of the past century. 1See David L. Shapiro, Class Actions: The Class as Party and Client, 73 Notre Dame L. Rev. 913, 913 (1998) (“Perhaps the most dramatic development in civil procedure in recent decades has been the growth of interest in the class action . . . .”); accord John C. Coffee, Jr., Class Wars: The Dilemma of the Mass Tort Class Action, 95 Colum. L. Rev. 1343, 1344–45, 1350 (1995); Arthur R. Miller, Of Frankenstein Monsters and Shining Knights: Myth, Reality, and the “Class Action Problem, 92 Harv. L. Rev. 664, 664 (1979). It held the promise of remedying previously irremediable harms; 2Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (noting that the emergence of aggregate litigation through Rule 23 was largely an attempt to vindicate “the rights of groups of people who individually would be without effective strength to bring their opponents into court at all” (quoting Benjamin Kaplan, A Prefatory Note, 10 B.C. Indus. & Com. L. Rev. 497, 497 (1969)) (internal quotation marks omitted)). it was the device to which we entrusted our most important cases, from catastrophic mass torts like Agent Orange 3In re “Agent Orange” Prod. Liab. Litig., 996 F.2d 1425 (2d Cir. 1993). and asbestos, 4Amchem Prods., 521 U.S. 591. to the school desegregation of Brown v. Board of Education 5347 U.S. 483 (1954). and the quest for workplace equality. 6Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541 (2011). Yet, if aggregation was the innovation of the past century, what could be the innovation of this century?

In recent years, we have increasingly sought to resolve our nation’s greatest tragedies and harms without aggregate litigation. In the aftermath of the terrorist attacks, Congress created the September 11th Victim Compensation Fund for the express purpose of removing cases from the traditional litigation system—insulating the airlines from the costs of defense and streamlining the costs of public funds dissemination through the creation of a non–Article III claims tribunal. 7Air Transportation Safety and System Stabilization Act, 49 U.S.C. § 40101 (2006). For similar programs, see Smallpox Emergency Personnel Protection Act of 2003, 42 U.S.C. §§ 239–239h (2006) (Smallpox Compensation Program); Public Readiness and Emergency Preparedness Act, id. §§ 247d-6d to -6e (Countermeasures Injury Compensation Program); Radiation Exposure Compensation Act, id. § 2210 (Radiation Exposure Compensation Program); Public Safety Officers’ Benefits Act of 1976, id. §§ 3796–3796c-2 (Supp. 2008) (Public Safety Officers’ Benefits Program); Civil Liberties Act of 1988, 50 U.S.C. app. § 1989b-3 (2006). Private defendants have likewise eschewed aggregate litigation, as exemplified most prominently by the BP Gulf Coast Claims Fund (GCCF) created in the wake of Deepwater Horizon’s explosion. 8For the official website, see http://www.gulfcoastclaimsfacility.com. Although BP would ultimately reach a class action settlement approved in 2012, it resolved more than two-thirds of its claims—more than $7.8 billion—through a private claims resolution facility, as required by the Oil Pollution Act of 1990. See 33 U.S.C. § 2705(a) (2012). In the wake of the class action settlement, the GCCF website was closed, directing participants to the class settlement website—a transition discussed in more detail in Parts II and III of this Article.

But companies do not only bypass aggregation in addressing national tragedies and mass disasters. Increasingly, companies seek to avoid aggregation before and after more ordinary disputes arise as well. With the Supreme Court’s blessing, companies have begun to include arbitration clauses as a predispute mechanism for precluding aggregation, ensuring individual processing of employment 9See, e.g., Nitro-Lift Techs., L.L.C. v. Howard, 133 S. Ct. 500 (2012) (remanding noncompete claims to arbitration); Rent-A-Center, W., Inc. v. Jackson, 561 U.S. 63 (2010) (enforcing a provision giving an arbitrator authority to determine all challenges to enforceability of an arbitration agreement between employee and employer, in the context of a race discrimination claim). and small-value consumer claims. 10See, e.g., CompuCredit Corp. v. Greenwood, 132 S. Ct. 665 (2012) (holding a dispute over $257 in credit card fees was properly subject to arbitration because the statute created a right to proceed in class action litigation but was silent as to whether claims could proceed in arbitration, and thus did not override the Federal Arbitration Act’s presumption of arbitrability); AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011) (holding a $30 dispute was properly subject to a single-plaintiff arbitration agreement). Putative defendants have also adopted mass compensation schemes to resolve contract and mass tort cases, ranging from disputes over seats at the Super Bowl to the running aground of the Costa Concordia cruise ship, in a postdispute effort to prevent or weaken postdispute aggregation. 11For discussion, see infra Part II.B.

Yet our scholarship continues to conceive of mass claims resolution through the lens of aggregate litigation—class actions and, in recent years, multidistrict litigation and “quasi–class action” settlements. 12When federal civil actions pending in different districts have common questions of fact, multidistrict litigation (MDL) is utilized to combine and transfer the actions to any district for consolidated pretrial proceedings. 28 U.S.C. § 1407 (2012). The new trend within the scholarship has been toward the recognition of quasi–class action settlement structures through MDL as contrasted with traditional class actions; yet both mechanisms fall within the traditional conception of aggregation. See, e.g., Elizabeth Chamblee Burch, Financiers as Monitors in Aggregate Litigation, 87 N.Y.U. L. Rev. 1273, 1274–75 (2012); John C. Coffee, Jr., Class Action Accountability: Reconciling Exit, Voice, and Loyalty in Representative Litigation, 100 Colum. L. Rev. 370, 371 (2000); Troy A. McKenzie, Toward a Bankruptcy Model for Nonclass Aggregate Litigation, 87 N.Y.U. L. Rev. 960, 962–63 (2012); Charles Silver & Geoffrey P. Miller, The Quasi–Class Action Method of Managing Multi-District Litigations: Problems and a Proposal, 63 Vand. L. Rev. 107, 110–11 (2010). As a consequence, many of these innovations are dismissed as ad hoc responses to extraordinary situations. 13See, e.g., Tracy A. Thomas, Introduction to Symposium, Remedies for Big Disasters: The BP Gulf Oil Spill and the Quest for Complete Justice, 45 Akron L. Rev. 567, 570 (2012) (summarizing the conclusion of symposium participants that the GCCF should not serve as a template for other disasters or compensation programs); Adam S. Zimmerman, Funding Irrationality, 59 Duke L.J. 1105, 1114, 1118 (2010) (attributing the lack of scholarship on public funds to the conception that they are “sui generis” products of “special legislation”).For studies of particular tribunals, see Kenneth R. Feinberg, What Is Life Worth?: The Unprecedented Effort to Compensate the Victims of 9/11 (2005); Robert M. Ackerman, The September 11th Victim Compensation Fund: An Effective Administrative Response to National Tragedy, 10 Harv. Negot. L. Rev. 135 (2005); Robert L. Rabin, Indeterminate Future Harm in the Context of September 11, 88 Va. L. Rev. 1831 (2002); Byron G. Stier, The Gulf Coast Claims Facility as Quasi-Public Fund: Transparency and Independence in Claim Administrator Compensation, 30 Miss. C. L. Rev. 255 (2011).

I argue that a far more fundamental transition is underway: We are witnessing the birth of an entirely new branch of nonaggregative mass claims resolution systems, of which these mechanisms are mere exemplars.

These emerging systems are based upon a wholly different dispute resolution methodology than those described in the existing literature—one that bypasses the determination of common questions, at the core of aggregate mechanisms, entirely. 14Leading scholars are now focused upon doctrinal trends within aggregation toward smaller classes and multidistrict litigation—what I suggest here is a far more fundamental shift in mass claims resolution. See, e.g., Elizabeth Chamblee Burch, Disaggregating, 90 Wash. U. L. Rev. 667 (2013) (discussing the trend toward smaller classes); Alexandra D. Lahav, The Case for “Trial By Formula, 90 Tex. L. Rev. 571 (2012). Instead, these mechanisms use individualized claims determination as the vehicle for mass claims resolution. 15See Francis E. McGovern, The What and Why of Claims Resolution Facilities, 57 Stan. L. Rev. 1361, 1380–81 (2005) (“Except for the option of individual trials, all of the currently available litigation procedures for an endgame in disputes involving large numbers of claims contemplate a claims resolution facility. [Rule 23 federal class actions], state class actions, bankruptcy, multidistrict litigation, and mass settlements all reach closure with a claims resolution facility.”). I therefore term this new, second, unexplored branch of mass claims resolution mechanisms “disaggregative” dispute resolution systems. While “disaggregation” has typically referred to the breaking down of an aggregate class into smaller subclasses, 16See, e.g., Judith Resnik, Fairness in Numbers: A Comment on AT&T v. Concepcion, Wal-Mart v. Dukes, and Turner v. Rogers, 125 Harv. L. Rev. 78, 93 (2011) (using the term “disaggregation” to reference the Dukes Court’s insistence on smaller classes and Concepcion’s upholding of predispute arbitration provisions that effectively barred aggregation, noting that both opinions “insisted on disaggregation, devolution, and privatization”). this Article posits a far more radical form of disaggregation, in which the would-be class is broken into individual claims. The consequences of this pure disaggregation are far more extreme, but also create a very different set of normative concerns than the more restricted form of disaggregation occurring within the public aggregation mechanisms. 17See infra Part II.

The public class action device contributed to “a sea change in our understanding of both substantive and procedural law,” 18Shapiro, supra note 1, at 914. generating an entirely new field of legal scholarship. 19Id. at 914–16 & n.2 (querying whether any law review existed that had not published an article on class actions and providing an excellent index of the leading works within the field). The rise of private disaggregation has the potential to create an equally dramatic shift in the legal landscape, as this new approach to dispute resolution is driving many of the most innovative claims structures emerging today. 20For a traditional discussion of disaggregation as a mechanism for effectuating global class settlement, see Richard A. Nagareda, Autonomy, Peace, and Put Options in the Mass Tort Class Action, 115 Harv. L. Rev. 747, 751 (2002) (discussing the asbestos, silicone gel breast implant, and fen-phen class settlements, which he noted “all aspire to create some form of private administrative system that . . . promises more efficient compensation for plaintiffs, long-term peace for defendants, and a reduced litigation burden for the courts”).

In the mass claims context, individual plaintiffs often lack the incentive to bring suit, and where they do bring suit, they often rationally underinvest in litigation compared to the defendant. 21See Randy J. Kozel & David Rosenberg, Solving the Nuisance-Value Settlement Problem: Mandatory Summary Judgment, 90 Va. L. Rev. 1849, 1859–60 (2004). This resource asymmetry led to the creation of aggregation mechanisms, but created a new set of undesirable results, from nuisance suits to sweetheart deals. 22See infra notes 98–103 and accompanying text. See generally Richard A. Nagareda, Mass Torts in a World of Settlement 11–54 (2007) (discussing the dynamics of mass tort litigation and dysfunctions that arise in this context). Disaggregative mechanisms are thus a next-generation attempt to correct the resource asymmetries of the single-plaintiff era, without generating the familiar problems of aggregation.

Disaggregative systems often streamline procedure or substantive inquiries, or shift the costs of litigation to the defendants, allowing the pursuit of claims that could not rationally be pursued in the pre–Rule 23 world because the litigation costs exceeded the potential recovery. 23See infra Part III.A. Effective in 1966, Rule 23 of the Federal Rules of Civil Procedure set the prerequisites for class certification, as well as the types of classes permitted and their respective requirements. 1 Alba Conte & Herbert B. Newberg, Newberg on Class Actions § 1:8, at 28–29 (4th ed. 2002). The accompanying advisory committee notes explained the rationale for this innovation, focusing largely on the risk of inconsistent judgments and the need for classwide relief in discrimination cases. Fed. R. Civ. P. 23 advisory committee’s note, reprinted in 39 F.R.D. 69, 98–102 (1966).But, today’s class action practice focuses far more on these cost asymmetries, which can prevent either the pursuit of a claim or the successful pursuit of a claim. See David Rosenberg, Mandatory-Litigation Class Action: The Only Option for Mass Tort Cases, 115 Harv. L. Rev. 831, 852–53 (2002). For an excellent introduction for the unfamiliar reader to the phases of class litigation and the settlement pressures and “dysfunctions” apparent at each phase of litigation, see generally Nagareda, supra note , at 11–54. As a result, there is no need for the aggregation of claims; each individual can autonomously decide whether to pursue relief or not. This structure has the consequence of removing the agency relationship inherent to class litigation 24See Burch, supra note 12, at 1274–81 (discussing agency problems in both class actions and MDL and the possibility of third-party financiers acting as monitors in the context of nonclass aggregative litigation to help manage principal–agent problems); Jonathan R. Macey & Geoffrey P. Miller, The Plaintiffs’ Attorney’s Role in Class Action and Derivative Litigation: Economic Analysis and Recommendations for Reform, 58 U. Chi. L. Rev. 1, 7–9 (1991) (discussing agency in the context of Rule 23 class litigation and the agency costs unique to aggregate litigation). See generally Samuel Issacharoff, Litigation Funding and the Problem of Agency Cost in Representative Actions, 63 DePaul L. Rev. 561 (2014) (discussing the agency costs inherent in representative actions and the potential impact of alternative litigation financing). —and with it the substantial costs aggregation entails, 25See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 598 (1997) (noting that the transaction costs exceeded recovery in the asbestos cases by a two-to-one ratio). allowing the settlement of cases that could not be certified for class treatment consistent with due process. 26See id. at 597–98 (holding the class settlement proposing global resolution of current and future claims could not be certified consistent with due process, but noting that lack of certification meant that “future claimants may lose altogether” given the specter of exhaustion of assets).

As a result, these features of disaggregation create the possibility for claimants not only to receive compensation more quickly but also to receive more compensation than would have been available in aggregation. 27See infra Part II.B. Yet, even where this supercompensation is offered, defendants may see a decrease in overall costs, as litigation costs and free riding decline. 28See infra Part II.C. Thus, disaggregation holds the laudable promise of a simultaneous, superior, win-win outcome for both parties by eliminating the systemic costs and delays inherent to aggregate litigation—suggesting that these mechanisms may become a more common feature in our legal system in the years to come. 29Cf. John H. Beisner et al., Class Action “Cops”: Public Servants or Private Entrepreneurs?, 57 Stan. L. Rev. 1441, 1444–57 (2005) (discussing distrust of class actions by the public and the impact upon the selection and settlement of lawsuits).

But it is already foreseeable that disaggregation will also generate a new set of deviations from our preferred outcomes, like the single-plaintiff and aggregate litigation structures that preceded it. 30See Nagareda, supra note 22, at 11–54. While aggregation is often described as one of the most controversial features of our litigation system, attracting more scholarly attention than almost any other, 31See, e.g., Margaret H. Lemos, Aggregate Litigation Goes Public: Representative Suits by State Attorneys General, 126 Harv. L. Rev. 486, 488 (2012) (“Few aspects of contemporary civil litigation have attracted as much scholarly attention as the damages class action. Commentators have criticized class actions as either too powerful or not powerful enough . . . .”(footnote omitted)); Richard A. Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L. Rev. 97, 97 (2009) (“Few pretrial motions in our civil justice system elicit as much controversy as those for the certification of class actions.”); Sergio J. Campos, Class Actions All the Way Down, 113 Colum. L. Rev. Sidebar 20, 20 (2013), http://www.columbialawreview.org/wp-content/uploads/2013/02/20_Campos.pdf (“[T]he class action [is] arguably the most controversial procedure in civil litigation.”). disaggregation may become even more controversial. 32See, e.g., Myriam Gilles, Opting Out of Liability: The Forthcoming, Near-Total Demise of the Modern Class Action, 104 Mich. L. Rev. 373, 373–79 (2005) (predicting with disappointment the end of the class action procedure as a result of “the demise of mass tort class actions and the rise of contractual class action waivers”); Linda S. Mullenix, Prometheus Unbound: The Gulf Coast Claims Facility as a Means for Resolving Mass Tort Claims—A Fund Too Far, 71 La. L. Rev. 819, 825 (2011) (criticizing one such “defendant-created and directed” disaggregative device as “a model essentially unconstrained by law”); Resnik, supra note 16, at 161–68 (critiquing recent Supreme Court decisions that subsequently imposed greater entry barriers to federal courts through disaggregation).

Disaggregation has the potential to substantially limit the experiment with aggregation that has occurred over the past half century—giving defendants no less than the option to avoid class actions altogether. The Rehnquist and Roberts Courts, in both their procedural 33See, e.g., Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2560–61 (2011); Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 619–22 (1997). and substantive 34See, e.g., Philip Morris USA v. Williams, 549 U.S. 346, 353–63 (2007); State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408, 416 (2003). rulings, have already focused on the due process limitations on aggregation, chipping away at the potential expanse of aggregation. 35See, e.g., Burch, supra note 14, at 667; Resnik supra note 16, at 93. But now, disaggregation gives private parties the power to chip away at aggregation from the opposite end of the spectrum: Pre-dispute, contracting parties can use arbitration provisions to waive aggregation—preventing the pursuit of any future class action. 36See, e.g., AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1746–53 (2011); Stolt-Nielsen S.A. v. AnimalFeeds Int’l Corp., 559 U.S. 662, 681–89 (2010). Post-dispute, defendants in contract and tort actions alike can now use disaggregative settlements not just to deter the filing of a class action, but to prevent a class action altogether. 37See, e.g., Winzler v. Toyota Motor Sales U.S.A., Inc., 681 F.3d 1208, 1210–15 (10th Cir. 2012); In re Aqua Dots Prods. Liab. Litig., 654 F.3d 748, 751–53 (7th Cir. 2011).

The primary carveout to this newly emerging doctrine exists where the damages incurred are subject to debate, such that the court can find that the corporation has not offered full compensation in its disaggregative process. 38See, e.g., In re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on April 20, 2010, 295 F.R.D. 112, 161 (E.D. La. 2013) (certifying class and approving class settlement). Yet, in many cases the same individualized inquiries that interfere with damages may also drive variation in liability or manageability questions sufficient to preclude class certification. 39See Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2550–52, 2557 (2011) (reversing a decision to certify a class for a lack of commonality among the 1.5 million members, and finding their claims for backpay to have been improperly certified under Rule 23(b)(2)). Thus, private disaggregative mechanisms may have the greatest power in shielding defendants in the cases that were the last stand of the class action system. The confluence of this early precedent then suggests that, unless we change the direction of our doctrine, if the defendant prefers to avoid a class action it can readily use disaggregative mechanisms to do so. The practical impact of this innovation is the radical notion that we are on the precipice of a world in which many class actions may be maintained only with the defendant’s blessing.

In the shadows of this broader systemic shift lurk concerns with the operation of these private mechanisms. In the shadow of aggregation, the particular outcomes obtained have the potential to either substantially undercompensate or overcompensate claimants. 40See infra Part III.A. These consequences flow in part from the bypassing of merits determinations that many disaggregative processes formally employ or informally incentivize. 41For an excellent discussion of the normative consequences of a privately ordered shift toward minimal process in the single-plaintiff context, see Nora Freeman Engstrom, Sunlight and Settlement Mills, 86 N.Y.U. L. Rev. 805 (2011). But these outcomes are also the result of a process that is typically defendant driven, with the defendant acting as the systems designer, not only crafting the rules of the game but also selecting the umpire who will make the calls—giving rise to substantial legitimacy questions. 42See, e.g., Jean R. Sternlight, Creeping Mandatory Arbitration: Is It Just?, 57 Stan. L. Rev. 1631, 1631–35 (2005).

The existing literature typically frames this dynamic as a conflict between party autonomy and private ordering on the one hand, 43See, e.g., Albert Choi & George Triantis, Strategic Vagueness in Contract Design: The Case of Corporate Acquisitions, 119 Yale L.J. 848, 922–23 (2010); Howard M. Erichson, The Problem of Settlement Class Actions, 82 Geo. Wash. L. Rev. (forthcoming 2014) (manuscript at 5, 13–15), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2243155; Deborah E. Greenspan & Matthew A. Neuburger, Settle or Sue? The Use and Structure of Alternative Compensation Programs in the Mass Claims Context, 17 Roger Williams U. L. Rev. 97, 129–36 (2012); Keith N. Hylton, Agreements to Waive or to Arbitrate Legal Claims: An Economic Analysis, 8 Sup. Ct. Econ. Rev. 209, 248 (2000). and enforcement of substantive law, social justice, and the larger public interest on the other. 44See, e.g., Cynthia L. Estlund, Between Rights and Contract: Arbitration Agreements and Non-compete Covenants as a Hybrid Form of Employment Law, 155 U. Pa. L. Rev. 379, 379–80, 383 (2006); Mullenix, supra note 32, at 825; Resnik, supra note 16, at 161–68; Adriaan Lanni, Note, Protecting Public Rights in Private Arbitration, 107 Yale L.J. 1157, 1161 (1998). Yet those seeking to restrict disaggregation have struggled to find a legal basis to do so, and, more broadly, to reconcile such a restriction with the default rule permitting parties to settle their claims without court approval or intervention. 45See, e.g., Myriam Gilles & Gary Friedman, After Class: Aggregate Litigation in the Wake of AT&T Mobility v. Concepcion, 79 U. Chi. L. Rev. 623 (2012); David Horton, Mass Arbitration and Democratic Legitimacy, 84 U. Colo. L. Rev. (forthcoming 2014) (reviewing Margaret Jane Radin, Boilerplate: The Fine Print, Vanishing Rights, and the Rule of Law (2013)), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2235901.

This Article reconceptualizes this conflict, arguing that the privatized nature of the existing disaggregation system is the source of many of the problems identified. 46See infra Part III. Thus, while parties elect private disaggregative mechanisms because they yield better outcomes than the default litigation system, in many disputes a public disaggregative mechanism would offer a superior option. Yet no public mechanism exists.

This Article proceeds in three parts. Part I examines the systemic costs and failures of public aggregation mechanisms that gave rise to disaggregation. Part II introduces the concept of private disaggregation, and then sets forth a typology of existing disaggregative systems. Part III explores the consequences of disaggregation for the twin goals of compensation and deterrence, as well as the broader public interest in the availability of legitimate and transparent processes. In contrast to the existing literature, this Part argues that a web of interrelated factors—including the size, similarity, and sophistication of the population seeking relief—can create a bias favoring not only defendants but also claimants. This analysis suggests that current limits on disaggregation, while well intentioned, do not obtain optimal results. Finally, this Article offers a few concluding observations on the impact the innovation of private disaggregation has upon our ability to rely upon private rights of action as a mechanism for enforcement of law.

I. The Traditional Conception of Disaggregation as a Handmaiden of Aggregation

A. The Necessity of Disaggregation in Mass Claims Compensation

Scholars have long debated the optimal dispute resolution structure for processing mass claims. The inherent challenge stems from the realization that single-plaintiff litigation results in suboptimal levels of compensation and deterrence where wrongs are perpetrated on a large scale. 47See Nagareda, supra note 20, at 751, 754, 770, 821 (“[T]he existence of a credible mass tort class settlement removes the possibility of a deterrence gap by virtually eliminating the prospect that too few compensation claims will be filed.” (internal quotation marks omitted)). The nature of these claims can vary widely as the problem of vindication is quintessentially one of the ratio of damages to litigation costs, such that even high-value claims can fall within this category. See id. at 750. Mass tort claims are an exemplar, as the alleged damages may include substantial claims for personal injury or wrongful death, but establishing causation may require the costly development of epidemiological evidence, expert testimony, and voluminous document review. See id. at 750, 827. The emergence of aggregate litigation through Rule 23 was thus largely an attempt to vindicate “the rights of groups of people who individually would be without effective strength to bring their opponents into court at all.” 48Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (quoting Kaplan, supra note 2, at 497). Through aggregation, the asymmetrical investment incentives that would otherwise exist could be corrected, encouraging both parties to make an optimal investment in the litigation process. 49See Rosenberg, supra note 23, at 852. Aggregation held the promise of not only improved enforcement and deterrence, but also more systematic treatment of wrongs and greater consistency across similarly situated victims, 50See Judith Resnik et al., Individuals Within the Aggregate: Relationships, Representation, and Fees, 71 N.Y.U. L. Rev. 296, 308, 382 (1996). with more efficiency than individualized litigation. 51See Jack B. Weinstein, Individual Justice in Mass Tort Litigation 135–36 (1995) (describing the efficiency benefits of aggregation).

Despite its promise, aggregation could only resolve the common questions of litigation; it could not determine individual-specific elements of a claim, defenses, or damages. 52See Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2560–61 (2011) (holding that the defendant has a due process right to present individualized defenses in a class action proceeding). Thus, disaggregation became a crucial feature of aggregation-based mechanisms for dispute resolution. 53See McGovern, supra note 15, at 1380–81. The dynamics of class actions—and aggregation generally—almost inevitably lead to a high probability of settlement prior to trial, given the stakes for both sides in the all-or-nothing verdict at trial and the intensive resource allocation class litigation requires. 54See In re Rhone-Poulenc Rorer Inc., 51 F.3d 1293, 1299 (7th Cir. 1995) (Posner, C.J.) (describing certification as “forcing these defendants to stake their companies on the outcome of a single jury trial” or “to settle even if they have no legal liability” out of “fear of the risk of bankruptcy”). But see Stephen B. Burbank & Tobias Barrington Wolff, Redeeming the Missed Opportunities of Shady Grove, 159 U. Pa. L. Rev. 17, 59–61 (2010) (“What is striking . . . is Judge Posner’s lack of attention in the first part of his analysis to the policy differences that States might have regarding the ‘mature tort’ problem and the relative merits of decentralized adjudication, which offers the benefit of accreted wisdom over time but may produce results that lack uniformity and appear arbitrary, versus a high-stakes industry-wide trial, which creates greater risks of inaccurate or unreliable results but also provides greater parity and fairness among claimants.”). Disaggregation has therefore become the mechanism for distributing the proceeds of a settlement to individual class members. 55See Adam S. Zimmerman, Distributing Justice, 86 N.Y.U. L. Rev. 500, 507–12 (2011).

Modern class settlements looked to the public law model of disaggregation, creating eligibility and compensation frameworks in the settlement agreement akin to those embedded in the congressional authorizations of public tribunals. 56See, e.g., Richard A. Nagareda, Turning from Tort to Administration, 94 Mich. L. Rev. 899, 921 (1996). Claims settlement facilities are then employed to distribute individual compensatory awards pursuant to those frameworks. 57See McGovern, supra note 15, at 1365. Because claims facilities have received little scholarly attention, a brief introduction is useful in understanding the operation of the existing aggregate system, the impetus it creates toward disaggregation, and the ways in which disaggregation can enlarge the pool of funds available to the parties—allowing both sides to simultaneously receive a superior net outcome as contrasted with aggregation. 58See Deborah R. Hensler, Alternative Courts? Litigation-Induced Claims Resolution Facilities, 57 Stan. L. Rev. 1429, 1429 (2005) (noting that claims resolution facilities have “largely escaped the scrutiny of legal scholars”).

B. The Class Action Settlement Process

Once the court has approved a class settlement, as required by Rule 23, an administrator is typically engaged to provide class notices and manage the claims fund. The small cadre of leading settlement administration firms has developed best practices for managing the complex operational details that frequently emerge during the notice and distribution processes. There are generally two roles incorporated within this function: (1) determining whether the individual is properly entitled to participation in the settlement, and (2) determining the proper amount of compensation and overseeing its provision.

The predominant form of aggregate settlement provides for an allocation-only function, in which the claims administrator allocates a defined fund among claimants. 59See Zimmerman, supra note 55, at 511–12. Typical allocation structures include flat rate, in which all claimants receive the same relief; flat-rate election, in which all claimants are given the choice among a small number of remedies; and formula or grid structures, in which certain predefined criteria are used to calculate the compensation due. 60See Francis E. McGovern, Distribution of Funds in Class Actions—Claims Administration, 35 J. Corp. L. 123, 126–33 (2009) (describing various settlement structures as case studies, without creating a typology). While one could envision a system in which the administrator is given the discretion to make an individualized determination, both parties typically favor providing clear guidance to the administrator. Moreover, this clarity is useful in demonstrating to the court that the absent class members have enough information about their likely recovery to make informed decisions about participation. 61Cf. McGovern, supra note 15, at 1376–77 (recognizing the tensions inherent in requiring opt-out before awards are calculated). As a result, most settlements adopt a clear payout formula.

1. Barriers to Compensation

In recent years, claims forms have become a common feature in settlements, whether because of difficulties in determining the identity or location of class members to send payment or because of potential variations in the harm suffered by the class. 62See Judges’ Class Action Notice and Claims Process Checklist and Plain Language Guide, Fed. Jud. Ctr. 6 (2010), http://www.fjc.gov/public/pdf.nsf/lookup/NotCheck.pdf/$file/NotCheck.pdf (treating claim forms as a common practice). While a class action binds all members of the class, only the fraction who submit claims forms receive compensation where such a form is required. A recent study by Rust Consulting—one of the preeminent settlement administration firms—found that the claims form completion rate is approximately 2%–20% in consumer cases, while higher rates are obtained in securities cases (20%–35%) and employment cases (20%–85%). 63See Tiffany Allen, Anticipating Claims Filing Rates in Class Action Settlements, Class Action Persp. (Rust Consulting, Inc., Minneapolis, Minn.), Nov. 2008, available at http://www.rustconsulting.com/Portals/0/pdf/Monograph_ClaimsFilingRates.pdf. As a matter of best practices, there is pressure for courts to ensure that claims forms are actually necessary in any particular case—either to identify the individual or determine the amount of compensation due. 64See Judges’ Class Action Notice and Claims Process Checklist and Plain Language Guide, supra note 62, at 6. Nevertheless, this leads to the astonishing notion that many class actions are only providing compensation to a small fraction of the harmed individuals, while preclusion operates to bar the remaining individuals’ claims. 65See, e.g., Am. Law Inst., Principles of the Law: Aggregate Litigation § 3.07 (2010) (recommending judges limit such payments “to circumstances in which direct distribution to individual class members is not economically feasible, or where funds remain after class members are given a full opportunity to make a claim” (internal quotation marks omitted)).This observation does not account for the practice of settling class claims without even attempting to make distribution to the class—a phenomenon that has troubled class action scholars who view the practice as an abandonment of one’s duty to the class in favor of a charitable group comprised of individuals who are not clients (but are instead, for example, political allies). Yet, lawyers have argued that in small-value cases the transaction costs of payment can exceed damages. See Lemos, supra note 31, at 528–29 (describing objections to cy pres distributions and documenting examples of politically directed cy pres recoveries).

Equally notable, the structure of class actions requires that absent class members be given the choice not only to file a claim form, but also to opt out of the settlement and retain their right to sue individually, or object to the settlement altogether. 66See Fed. R. Civ. P. 23(b)(3), (e)(4). While these options promote due process and autonomy, the practical result is often confusion among absent class members who, confronted with various forms, complete the wrong paperwork—for example, opting out of the settlement when they intend to accept payment. 67McGovern, supra note 60, at 127. In one notable case, 10,000 claims forms were filed, but 20,000 requests for exclusion were filed. 68Id. at 126–27.

2. Fixed-Fund Versus Claims-Made Settlements

In these structures there remains a question of precisely what compensation will be determined to be due, and successfully paid, given the high probability of absent class members who fail to file claims forms or even cash checks mailed to their last known addresses. 69Five to twenty-five percent of checks issued in class actions remain uncashed. See Class Notice & Settlement Administration: Avoiding the Pitfalls—Part I, Class Action Persp. (Rust Consulting, Inc., Minneapolis, Minn.), Mar. 2007, available at http://www.rustconsulting.com/Portals/0/pdf/Monograph_Pitfalls_I.pdf. Therefore, the establishment of a total fund in the settlement agreement is common. 70With improved technology, some funds have been able to make automatic payments to all class members. However, these payment structures require substantial information about claimants, which is frequently unavailable. See, e.g., Zimmerman, supra note 13, at 1167. Innovative multitiered payment schedules can be utilized to minimize the residual, but typically the balance will ultimately be donated to charity through a cy pres distribution. 71A cy pres distribution occurs when class damages cannot be feasibly distributed or when a balance remains; under this system, the funds are to be distributed to the next best compensation use, typically a charitable donation. See 4 Conte & Newberg, supra note 23, § 11:20. For an excellent discussion of the cy pres system, see William B. Rubenstein, On What a “Private Attorney General” Is—And Why It Matters, 57 Vand. L. Rev. 2129, 2159–71 (2004). A total settlement fund benefits both the defendant and the class counsel. For the defendant, this provides certainty to financial markets about the extent of liability, a step commonly regarded as advantageous to stock price. 72Francis E. McGovern, A Model State Mass Tort Settlement Statute, 80 Tul. L. Rev. 1809, 1811–12 (2006). For the class counsel, this provides a ready benchmark by which the court can measure the counsel’s success, a key component in the determination of attorneys’ fees. 73See John C. Coffee, Jr., Understanding the Plaintiff’s Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions, 86 Colum. L. Rev. 669, 677–84 (1986) (describing fee structures and incentives for class counsel).

Yet notable outliers exist. For example, Congress gave the special master administering the September 11th Compensation Fund the ability to determine the overall size of the claims fund. 74For varied analysis, see Symposium, After Disaster: The September 11th Compensation Fund and the Future of Civil Justice, 53 DePaul L. Rev. 205 (2003). Similarly, the silicone gel breast implant settlement is one of the most prominent examples of a private fund utilizing this model, as private defendants typically demand certainty and global peace in exchange for their agreement to settle. 75See McGovern, supra note 15, at 1363–64 (discussing the September 11th Compensation Fund and silicone gel breast implant settlements as exemplars of funds that do not establish overall fund limits but instead “the total amount of damages would be the sum of the subsequently determined individual amounts”). In that case, defendants in an MDL proceeding agreed to form a “claims-made” fund, in which those individuals submitting claims and meeting the established criteria would receive payments according to a defined schedule. 76Settlement Facility and Fund Distribution Agreement Between Dow Corning Corporation and the Claimants’ Advisory Committee at 2–3, In re Dow Corning Corp., 244 B.R. 705 (Bankr. E.D. Mich. 1999) (No. 95-20512). The unlimited fund feature of these programs can be a useful bridge in negotiations, particularly where counsel cannot agree upon the likely value and number of claims—and thus cannot reach agreement on the derivative total settlement figure. 77See McGovern, supra note 60, at 1389.

But, arbitraging this asymmetry in expectations does include risk. If the damages are greater than expected, there is no cap or pro rata cramdown as is provided for in any potentially undercapitalized fixed-fund settlement. 78For an excellent discussion of undercapitalization and the resulting need for a second opt-out opportunity, see Laurens Walker & John Monahan, Scientific Authority: The Breast Implant Litigation and Beyond, 86 Va. L. Rev. 801, 803–07 (2000). If the liability is too great, it may risk bankrupting the involved defendants—as occurred in the asbestos and breast implant litigation contexts. 79Anthony J. Sebok, What Do We Talk About When We Talk About Mass Torts?, 106 Mich. L. Rev. 1213, 1213 (2008) (reviewing Nagareda, supra note 22). While bankruptcy does not bar recovery, it may require presentment of claims and set cutoff dates that are impossible for those with latent claims to meet, resulting in diminished or permissive payout structures—the reality faced by many breast implant claimants. 80See Agreed Order Allowing Certain Late Claimants Limited Rights to Participate in the Plan’s Settlement Facility at 3, In re Dow Corning Corp., No. CV-00005-DPH (E.D. Mich. Dec. 12, 2007), available at http://www.mied.uscourts.gov/Information/Dow/pdf/orderpdf/agreed%20Order.pdf. For these reasons, claims-made funds have generally been disfavored by counsel to both parties. Claims-made funds thus remain a tool most commonly used by public institutions seeking to effectuate broader instrumental goals, transcending the immediate resolution of claims.

3. The Hidden Costs of Fixed Funds

The prevalence of fixed-fund structures then creates an allocation problem in all but the simplest cases. In order to reach agreement upon an aggregate total, either the amount of each absent class member’s claim must be determined in advance, or the administrator must be given power to ascertain these awards. Determining a binding claim value in advance requires counsel to ascertain the relevant variables to compensation and reach agreement on the details of compensation at a micro level. But, often parties are instead incentivized to simply agree to an overall settlement figure first, then later reach agreement on how that fund should be allocated. As a result, rather than having a bottom-up settlement that reflects the aggregate value of the individual claims, the settlement is a top-down settlement in which claimants effectively compete for allocation of a fixed-fund. Moreover, in high-profile settlements in which objections are anticipated from outside plaintiffs’ firms, errors or omissions in defining these variables may lead to denial of the motion to approve the class settlement. 81See William B. Rubenstein, The Fairness Hearing: Adversarial and Regulatory Approaches, 53 UCLA L. Rev. 1435, 1459 (2006) (concluding that “the track record of professional objectors to date . . . has been less than stellar”).

In contrast, if the parties seek to reach agreement on an overall figure without providing a specific formula for allocation of compensation using already-known variables, undercapitalization is possible. Indeed, the initial silicone gel breast implant settlement suffered from precisely this defect, resulting in a cramdown, or ratcheting, of the absent class members’ awards and triggering a second opt-out opportunity for the class members. 82Walker & Monahan, supra note 78, at 803–07. Because of the potential for undercapitalization, it is at times a “best practice” to withhold payment of class members’ claims until all payments have been calculated in order to ensure that sufficient monies exist to satisfy all awards at the same pro rata payout percentage.

Withholding in turn creates another set of tradeoffs: delaying payments in full requires claimants to wait until the conclusion of the proceedings to receive monies that may be needed immediately for resulting medical treatment and other expenses. However, creating a multi-tranche payout scheme increases administrative costs, limiting the pool of resources available for the settlement. These dynamics make clear that in fixed-fund cases, there may be real conflict among class members who are, de facto, competing for a share of the fund.

4. Settlement Motivations and Agency Costs

Given this complexity, why is it that disaggregation has been viewed as a mere ministerial detail? 83See Hensler, supra note 58, at 1429, 1439 (lamenting the lack of scholarship on claims administration). For defendants, as deterrence theory suggests, the dominant consideration in settlements is the total payout and the degree of closure they will receive in exchange for that payment. 84See Nagareda, supra note 56, at 914 (“From the defendants’ standpoint . . . settlement is attractive only to the extent that it does not merely resolve pending cases but, in addition, provides assurance that such action will not result in a deluge of marginal cases.”). For class counsel, fees are calculated based upon the overall aggregate recovery. 85See Charles Silver & Lynn Baker, I Cut, You Choose: The Role of Plaintiffs’ Counsel in Allocating Settlement Proceeds, 84 Va. L. Rev. 1465, 1527 (1998). Thus, for both sets of counsel, so long as the allocation is not so unfair as to prompt a high opt-out rate or later collateral attack, there is no structural incentive to invest additional resources in improved allocation mechanisms. 86For a discussion of the agency dynamics that drive class settlement allocations between class members, see generally Samuel Issacharoff et al., Bargaining Impediments and Settlement Behavior, in Dispute Resolution: Bridging the Settlement Gap (David A. Anderson ed., 1996); Marcel Kahan & Linda Silberman, The Inadequate Search for “Adequacy” in Class Actions: A Critique of Epstein v. MCA, Inc., 73 N.Y.U. L. Rev. 765 (1998); Silver & Baker, supra note 85.

The incentive is therefore toward creating the broadest possible class in settlement, which provides greater closure to the defendant and enhances the attorneys’ fees award to class counsel—a structural incentive checked only by increasingly stringent court review. However, the structure of settlement approval hearings necessarily operates so that counsel for both sides are unified in their desire to obtain approval, thus minimizing any issues appearing before the court. 87See Rubenstein, supra note 81, at 1445.

Moreover, absent class members often lack not only the incentive, but also the discovery necessary to uncover these differences, given the negative-value considerations that prompted the initial creation of aggregation. Thus, unless a competing plaintiffs’ firm seeks to appear as counsel for an objecting party as a means of preventing settlement or gaining a seat at the negotiating table—a dynamic that typically occurs in only the largest settlements, but is unlikely in most class actions—the Rule 23 structure does not result in any actor with both the incentive and ability to raise these conflicts to the court. 88See generally Richard A. Nagareda, Administering Adequacy in Class Representation, 82 Tex. L. Rev. 287, 370–71 (2003) (describing the lack of meaningful checks imposed by objections upon disloyal class counsel).

C. The Impetus for Privatization

In the existing public litigation system, disaggregation is simply a mechanism for implementing an aggregate settlement. While the aggregate settlement figure may reflect a proper deterrent value, the individual settlements that result in its shadow may fail to provide an accurate level of compensation because of the emphasis on collective treatment and minimization of administrative costs. Yet, for an absent class member, opting out of the settlement will mean litigating the claim individually or with the handful of other opt-outs—frequently a negative-value proposition. 89See John C. Coffee, Jr., Litigation Governance: Taking Accountability Seriously, 110 Colum. L. Rev. 288, 304–05 (2010). The rational choice for an absent class member is then to remain in the aggregate settlement, even if it is substantially undercompensating his or her losses. Yet this dynamic suggests the possibility from the claimant’s perspective of a more optimal compensation system focused upon disaggregated compensation, rather than aggregate deterrence. 90See David Rosenberg, Decoupling Deterrence and Compensation Functions in Mass Tort Class Actions for Future Loss, 88 Va. L. Rev. 1871, 1871–79 (2002).

II. The New Generation of Disaggregative Structures

Aggregation promised to ameliorate the irremediability of certain types of harms, to create improved deterrence for widespread but negative-value wrongs, and to correct the asymmetries inherent to the single-plaintiff litigation system that emerged in the context of mass claims. Yet the solution of aggregation created its own well-documented costs resulting from the additional procedural protections aggregation entailed, a new set of dislocations 91For every claim, there is an optimal point of both compensation and deterrence associated with the value of the claim. Determining this optimal point within the law is, in the first instance, a matter for the legislature in crafting the available remedies, but then requires substantial individualized discovery and assessment to determine the precise value of any particular legal right in a particular case. In addition, scholars will often argue that particular legal frameworks are not obtaining the optimal levels of compensation or deterrence given particular normative goals.This Article argues that there are a number of factors that can shift settlement values from these optimal points, wherever set. The term “dislocation” is used throughout my analysis to show that these shifts are bilateral, in contrast with the traditional terminology that often suggests a merely one-directional shift from this optimal value point. in settlement values, and ultimately dysfunction harmful to plaintiffs and defendants alike. In the intervening decades, scholars and lawmakers have proposed countless modifications, expansions, and alterations to the aggregate litigation system designed to ameliorate these dislocations, but the underlying dysfunction remains.

In this Part, I argue that disaggregation is emerging as a new approach to mass claims resolution, creating a new set of dispute resolution mechanisms distinct from those of aggregation. This Part outlines the sources of parties’ dissatisfaction with the traditional aggregative process, which created the conditions that fostered the birth of these new private mechanisms. The analysis then turns to the ways in which procedural and substantive streamlining operate, identifying the situations in which parties should elect these mechanisms as superior to the public litigation and aggregation systems.

Disaggregation is already the mechanism by which every award of compensation to a mass claims victim is made. Disaggregation is essential even in aggregative proceedings because an individual’s status as a victim eligible for compensation must, by definition, be determined on an individual basis. 92The use of disaggregation, even in this traditional setting, has “largely escaped the scrutiny of legal scholars.” Hensler, supra note 58, at 1429, 1439 (“It is time to get serious about understanding the role claims resolution facilities are playing in the delivery of civil justice in the United States.”). But see Georgene Vairo, Why Me? The Role of Private Trustees in Complex Claims Resolution, 57 Stan. L. Rev. 1391, 1392–93 (2005) (advocating for the use of administrative trusts run by private persons). Disaggregation has therefore traditionally been employed as a handmaiden of aggregation. But I argue that today parties are engaging in a trend of “responsive opt-out,” in which they contractually bypass aggregation altogether, employing disaggregation not as a mere handmaiden but as the mechanism for reaching settlement.

In recent years, controversy has swirled around predispute arbitration agreements and postdispute settlement offers made by defendants to undermine class actions, as exemplified most prominently by the BP Gulf Coast Claims Fund. 93See supra note 8 and accompanying text. I argue that these mechanisms are each instantiations of a new trend toward disaggregative systems design, as defendants have increasingly realized that reliance upon disaggregation as a substitute for aggregative mechanisms can yield vastly superior outcomes in contrast with the default public litigation system. This reconceptualization reveals that disaggregation is not limited to its traditional role as a handmaiden to aggregation, but is instead a new, second, and co-equal, but heretofore unexplored, branch of mass claims systems design—rivaling the field of aggregation.

In this new world, parties are no longer merely selecting among the available mechanisms for aggregation within our public litigation system. This emerging innovation in the pre- and postdispute use of disaggregation reflects a grassroots realization that in certain cases the costs of aggregation to the parties, in both outlays to third parties and the costs derivative of uncertainty and a delay in settlement, will at times be greater than the initial difference in claims valuation between the parties. The result is such that seriatim, individual settlement is superior to undertaking the costs of aggregation, particularly given that disaggregation remains the endgame of aggregate settlements.

These new disaggregative mechanisms are thus able to remove not only the high procedural costs that aggregation entails for the protection of both absent class members and defendants, but also to unwind the transformation of the bargaining process that aggregation yields. Moreover, for defendants, postdispute disaggregative settlements may offer substantial public relations benefits as they attempt to “make right” a wrong, while simultaneously reassuring shareholders about the company’s ability to quickly resolve the pending liability. These second-generation opt-out structures have generated substantial controversy, framed alternatively as incredibly innovative and efficient mechanisms superior in their results for both parties, or as unconscionable procedural mechanisms created by corporations to exploit victims and externalize the costs of their tortious or otherwise illegal behavior.

To be clear, my claim is not that aggregation has no value. Rather, that from a systems design perspective, it yields suboptimal outcomes for the parties with respect to particular types of claims. For this reason, I argue, we see a trend toward disaggregation not only in private party disputes but also in public mass claims tribunals. While the motivations of the parties and, in turn, the necessity of external limitations are inherently very different in the public and private contexts, I reserve this issue for the later Parts of this Article. In this Part, I only seek to establish the existence of this branch of disaggregative mass claims resolution mechanisms and begin to identify the situations in which parties may turn to disaggregative tribunals.

The consequences of this investigation are significant. This transition toward private ordering creates a new, second branch of dispute resolution mechanisms and scholarship. But, it also has substantial implications for our aggregate litigation scholarship. In the design of public aggregate litigation mechanisms we must no longer focus exclusively upon the ability of the rules to create optimal outcomes in the trial, or settlement, of mass claims. Rather, parties are increasingly treating these mechanisms as mere default rules—as a starting point in negotiation for a disaggregative process. Yet the question of whether aggregation rules are optimal, not as mandatory rules but instead as default rules, has been largely overlooked in the literature.

A. Basic Typology

As the preceding discussion suggests, despite aggregate litigation’s promise of correcting asymmetries and promoting efficiency, the reality was far different. Aggregation surely provided a benefit and helpful innovation over the single-plaintiff and joinder systems that preceded it. Yet, these first-generation mass claims aggregation mechanisms entailed substantial costs, 94For an excellent discussion of agency in Rule 23 classes and MDL, see Macey & Miller, supra note 24, at 7–8. which could easily swamp even high-value litigation recoveries. 95See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997). At the same time, limitations on the benefits of aggregation became apparent, as protecting the due process rights of absent class members limited the scope and content of settlements—and in turn the degree of closure provided to defendants. 96See Howard M. Erichson & Benjamin C. Zipursky, Consent Versus Closure, 96 Cornell L. Rev. 265, 270–74 (2011). In imposing these restrictions, the Supreme Court recognized the intractable problems this presented in mass claims resolution: In certain types of cases, global aggregate settlements could not be approved consistent with due process, while continued litigation was not only value destructive, but would create a new set of dislocations as claimants were forced to rush to the courthouse before the remaining funds were dissipated. 97See Amchem, 521 U.S. at 598. The Court expressly invited congressional intervention, recognizing the inability of aggregation mechanisms to solve the problem posed by the massive liability created by asbestos use. Id. at 598–99; Ortiz v. Fibreboard Corp., 527 U.S. 815, 865 (1999) (Rehnquist, C.J., concurring). Yet, in practice, this solution too has proven untenable. Georgene Vairo, Mass Tort Bankruptcies: The Who, The Why and The How, 78 Am. Bankr. L.J. 93, 95 (2004) (“The Supreme Court’s decision in Amchem and its later decision in Ortiz have made it more difficult for companies seeking global peace in resolving a mass tort to use Rule 23 settlement classes. Moreover, Congress appears to have failed in its attempt to resolve legislatively the protracted asbestos litigation.” (footnotes omitted)).

As sophistication with the tools of aggregation increased, sources of structural dysfunction became apparent at all phases of the claims maturation process. Settlements drifted away from their optimal value—a trend exemplified by the nuisance litigation, 98See, e.g., Kozel & Rosenberg, supra note 21, at 1879–90; James J. Park, Rules, Principles, and the Competition to Enforce the Securities Laws, 100 Calif. L. Rev. 115, 161 (2012) (discussing the incentive of entrepreneurial enforcers to bring low probability claims). blackmail settlements, 99See, e.g., Bruce Hay & David Rosenberg, “Sweetheart” and “Blackmail” Settlements in Class Actions: Reality and Remedy, 75 Notre Dame L. Rev. 1377, 1378 (2000). and sweetheart deals 100See, e.g., Richard A. Nagareda, The Preexistence Principle and the Structure of the Class Action, 103 Colum. L. Rev. 149, 163 (2003). that evolved in the shadow of aggregation. 101See, e.g., Jonathan T. Molot, Litigation Finance: A Market Solution to a Procedural Problem, 99 Geo. L.J. 65 (2010). The settlements occurring in the shadow of aggregate litigation predictably were contaminated by this dysfunction, triggering a widespread critique of aggregation as providing suboptimal levels of both compensation and deterrence. 102See, e.g., Nagareda, supra note 22, at 11–54. Sadly, little has changed in the nearly two decades since John Coffee offered his summary of class litigation: “All that is certain about mass tort litigation is that it places a heavy burden on the federal courts, while producing often modest and delayed benefits to plaintiffs.” 103Coffee, supra note 1, at 1346.

Private disaggregative mass claims resolution mechanisms are substantially responsive to these sources of dysfunction in aggregate litigation, which in turn responded to the failures of single-plaintiff and joinder models of litigation. These emerging second-generation mass claims structures focus upon addressing the underlying failures of the single-plaintiff model through mechanisms other than aggregation, in an effort to avoid the dysfunction aggregation is perceived to entail. Broadly characterized, one set of mechanisms bars aggregation in favor of individualized hearings, but typically substitutes streamlined procedures aimed at reducing or shifting the litigation costs that necessitated aggregation. 104See infra Part II.A.2. A second set of mechanisms removes the common questions prerequisite to aggregation, allowing victims to receive compensation based upon simply satisfying the individualized proof requirements set forth. 105See infra Part II.A.1. Typically, the first set of mechanisms has been utilized pre-dispute, while the second set is more common post-dispute—but this results from pragmatic factors, which are discussed in this Part, rather than from any absolute requirement. Indeed, many structures combine a certain degree of both of these features, creating a broad array of potential systems designs.

At a deeper level, both mechanisms for responsive opt-out can be reconceptualized as insights about the limits of the default rules of aggregation. By identifying these false assumptions, corporations can craft mechanisms that manipulate these features of the system, creating alternatives that are superior for both participants. This Part turns to providing a basic typology of these two methods, with a particular focus on the systems design choices inherent to both structures, highlighting the similarities and contrasts between existing public and private disaggregative systems as they operate pre- and post-dispute.

B. Postdispute Disaggregative Mechanisms

The proliferation of early individualized settlement offers contravenes the established wisdom that defendants will assert classwide defenses and insist upon individualized proof as a means of threatening to increase the cost of litigation and thus decrease settlement values. 106See Nagareda, supra note 22, at 11–54. At the outset, then, one must wonder why defendants would create disaggregative structures, surrendering the bargaining power derived from delay. 107See, e.g., Feinberg, supra note 13 (arguing that private defendants would not have an incentive to enter into early, postdispute settlement regimes). The impetus for the these funds lays in the recognition that even in cases with substantial common questions of general liability, the costs of aggregation may overwhelm the benefits to the parties—making opt-out a mutually value-enhancing option.

In the wake of the GCCF, a variety of defendants have followed the same model, offering various forms of immediate, prelitigation settlement that require participants to forgo the possibility of aggregate litigation. 108See, e.g., Edward F. Sherman, The BP Oil Spill Litigation and Evolving Supervision of Multidistrict Litigation Judges, 30 Miss C. L. Rev. 237, 243 (2011). While BP’s GCCF was formed months after the Deepwater Horizon explosion, the delay in formation is quickly shortening: the Costa Concordia settlement grid was negotiated within weeks, while the Chevron refinery explosion fund was created in only days. 109See Paul Elias, Thousands File Claims After Chevron Refinery Fire, AP (Aug. 11, 2012, 4:39 AM), http://bigstory.ap.org/article/thousands-file-claims-after-chevron-refinery-fire. Just as defendants are becoming more comfortable with quickly instituting these funds, a shortened delay in victim acceptance of these funds also seems to be occurring—the Chevron explosion injured upwards of 4,000 people, of whom 3,800 sought settlement offers through its disaggregated claims procedure by the end of the very week the explosion occurred. 110See id. Chevron’s fund ultimately ballooned to 23,900 claims filed after 15,000 people sought medical treatment, yielding a total payout of approximately $10 million through its claims settlement fund. See Chevron Has Paid Out $10 Million for Refinery Fire, ABC7 News (Jan. 30, 2013), http://abclocal.go.com/kgo/story?section=news/local/east_bay&id=8973226.

Skeptics have suggested that these tribunals are simply mechanisms for exploiting victims, using legal sleight of hand to offer less compensation than available by law and to avoid transparency. 111See, e.g., Mullenix, supra note 32, at 825 (reviewing challenges to the legitimacy of the September 11th Fund and arguing that “[f]or those concerned with the rule of law, equity, and fundamental fairness, the GCCF [created by BP] ought to be a cause for concern”). There are certainly risks and disadvantages associated with these structures, which are detailed in the next Part. But the commonality between public and private opt-out systems suggests that there are legitimate structural motivations for these innovations. 112Cf. Albert Choi & George Triantis, Completing Contracts in the Shadow of Costly Verification, 37 J. Legal Stud. 503, 504–09, 523–25 (2008) (characterizing contractual predispute substantive provisions as value-adding and clarifying underlying obligations).

The trend toward private disaggregative mechanisms represents a return to first principles and creates a degree of unity between public and private approaches to mass claims resolution. From nuclear weapons testing to war crimes, the government has often been placed in a situation analogous to that of corporate defendants, confronted by numerous, similarly situated individuals seeking compensation for wrongs or harms. 113This preference dates from the earliest days of our nation. See Michele L. Landis, “Let Me Next Time Be ‘Tried By Fire’”: Disaster Relief and the Origins of the American Welfare State 1789–1874, 92 Nw. U. L. Rev. 967, 983 (1998) (describing the transition from “individual to group eligibility criteria in relief legislation” as based in the growth of the administrative state). For recent examples, see statutes listed supra note 7. Surely, sovereign immunity accords Congress greater latitude than private defendants possess in deciding how to respond to these claims in the first instance. 114See, e.g., United States v. Mitchell, 463 U.S. 206, 212–13 (1983) (describing the origins of sovereign immunity waivers); cf. Gregory C. Sisk, The Tapestry Unravels: Statutory Waivers of Sovereign Immunity and Money Claims Against the United States, 71 Geo. Wash. L. Rev. 602, 611–15 (2003) (noting that Congress often waives sovereign immunity conditioned upon the plaintiff’s consent to participate in a non–Article III system of adjudication as to the claim). Yet despite Congress’s constitutional role in creating Article III courts, it has often looked not to Article III courts or their aggregation mechanisms, but instead has preferred non–Article III tribunals vested with the authority to award compensation pursuant to a defined statutory scheme. 115For an excellent discussion of the constitutionality of these shifts, see Richard H. Fallon, Jr., Jurisdiction-Stripping Reconsidered, 96 Va. L. Rev. 1043, 1050–51 (2010); and Caleb Nelson, Adjudication in the Political Branches, 107 Colum. L. Rev. 559, 583–84 (2007).

Rather than simply waiving sovereign immunity and permitting adjudication of general liability and damages as defined by background law, Congress typically substitutes an eligibility and compensation structure. The same waiver of questions of general liability—in favor of a pure focus upon the determination of individualized issues of eligibility and compensation—typifies emerging early settlement offers and funds by private defendants.

For both public and private funds, the structure of payments is a complex decision, embedded with substantial symbolic value. In a purely retrospective system, the key stakeholders will typically have a sense of their rights in the default system and any asymmetries between their individualized situation and the typical harms of others within the system. The result is that as design decisions are formulated and established, the stakeholders will frequently be aware of the impact it will have upon their particular recovery. In both public and private disaggregation, three general compensation structures predominate: flat-rate offers, grid/formula structures, and ad hoc determinations by the tribunal. 116See McGovern, supra note 60, at 126–33. Both flat-rate offers and grid/formula structures can be categorized as “substantive offers,” while ad hoc determinations are more aptly characterized as “process offers.”

1. Substantive Offers: Flat-Rate Offers, Grids, and Formulas

Flat-rate payouts, as with the reparations made to Japanese-Americans interned during World War II, 117See Jamal Greene, The Anticanon, 125 Harv. L. Rev. 379, 400 (2011) (discussing reparations in the context of Korematsu v. United States, 323 U.S. 214 (1944)). are the most streamlined payment structure—the administrator need only determine that the claimant suffered the identified harm, without making any determination of particularized damages. Such a payout system can have an important symbolic effect in valuing each life lost or harm suffered as equal in value—a critique special master Ken Feinberg offered in reflecting on his design of the September 11th Fund, which provided differential payouts based upon a formulaic calculation with some discretion for modification in exceptional circumstances. 118See Feinberg supra note 13, at 177–79.

But flat-rate payment structures inherently fail to distinguish among levels of harm. For example, the Japanese-American citizen who was interned, and the citizen whose internment resulted in the additional loss of his home and business, would receive identical compensation payments despite clear differences in their actual harms. Moreover, although these compensation systems are by definition extrinsic to the baseline tort system, participants’ conceptions of fairness are often shaped by a tort-law conception of damages—an expectation exemplified by September 11th widows dissatisfied with the earnings caps imposed by the Fund upon high-income earners killed in the tragedy. 119See Elizabeth M. Schneider, Grief, Procedure, and Justice: The September 11th Victim Compensation Fund, 53 DePaul L. Rev. 457, 477 (2003).

Private tribunals traditionally operate as litigation substitutes, so the efficiency and viability of flat-rate offers are typically correlated with commonality in harm suffered. While this type of commonality can exist in contractual claims, many regimes need to incorporate either a grid or formula component in order to distinguish among levels of harm and thus better approximate the legal value of the claim. For example, after the Costa Concordia cruise ship ran aground, the cruise line offered all passengers who had not suffered substantial physical injuries or death a flat-rate payment of around $14,500, in addition to reimbursement of their tickets, costs of flights home, and other documented expenses incurred in Italy in the days following the wreck. 120See Gaia Pianigiani, Italy: Cruise Line Offers Cash to Shipwreck Passengers, N.Y. Times, Jan. 28, 2012, at A11.

Substantive offers may not only incorporate separate payment tiers to account for the objective differences among plaintiffs, but also incorporate an election of remedies, allowing plaintiffs to select among objectively similar options based upon individual, subjective preferences. The NFL’s settlement offer resulting from its shuffling of a number of ticketholders into less desirable seats or standing room only areas in the Dallas Cowboys’ stadium during Super Bowl XLV exemplifies this opportunity to create added value. 121See Jeremy Stahl, Gameday Theory, Slate (Feb. 11, 2011), http://www.slate.com/articles/sports/sports_nut/2011/02/gameday_theory.html. The NFL offered a tiered compensation structure, which each ticketholder could choose to accept in lieu of litigation. 122See id. The 2,800 ticketholders who were delayed or relocated could choose (a) the face value of their ticket or (b) a ticket to a future Super Bowl. 123See id. The 475 ticketholders who did not receive any seat were given three options to select among: (a) $2,400 plus a ticket to the 2012 Super Bowl, which they could use or sell in the secondary market, (b) a ticket to a future Super Bowl, including airfare and a four-night hotel stay, or (c) a payout of $5,000 or a check for more than $5,000 with documentation supporting their expenditures on the actual ticket price paid, travel, lodging, and meal expenses. 124See id.

Although the attorney who had filed a class action suit on behalf of the ticketholders urged them not to take the settlement, third-party pundits advised the putative class through the media that the deal provided a greater recovery than parties would receive in court—particularly after one considered the risk of loss given that the case for general liability “seems very hard to make.” 125Id. Moreover, while the court would be limited to the contractual damages available, the offer of tickets to a future Super Bowl of the ticketholder’s choosing created an offer objectively superior to anything potentially available in court. 126Id. Not only did the NFL offer a subjective value for plaintiffs given the difficulty in obtaining tickets for a particular Super Bowl, but it also offered an enhanced objective value given the secondary market premium on tickets. 127See id.

The NFL example highlights the benefits of disaggregation relative to aggregation. The ticketholders were returned to a position of autonomy, as the structure operated on an opt-in rather than opt-out basis. Those who failed to act would retain their right to sue, in contrast to a class action that would likely fail to provide relief to an absent class member who failed to file a claim form while simultaneously precluding future suit. Those who chose to settle were given greater compensation than available in the courts, without the reduction in net compensation for attorneys’ fees and costs. And, this relief was available immediately, avoiding the years of delay common to class action litigation. Thus, for the ticketholders, the offer was objectively superior to the best possible outcome in litigation.

The NFL was likewise able to decrease its net payout relative to litigation. While it paid supercompensation, it saved a substantial sum in attorneys’ fees and costs and was able to spread the payout over many years rather than making an immediate lump sum payout, to the extent that ticketholders chose tickets to future Super Bowls. In addition to these direct monetary benefits, the NFL also obtained the public relations benefit of appearing to “do the right thing” in taking responsibility for the mistake and compensating the victims quickly and fully. Given that the defendant decreased its net payments while each claimant received more than he would in litigation, without the delay of aggregation, one might posit that the only losing party in the arrangement was the putative class counsel.

It thus appears that the emerging popularity of disaggregation results precisely from its potential for mutually superior outcomes for both parties—as juxtaposed with the dysfunction and delay inherent to aggregate litigation.

2. Process Offers: Ad Hoc Determinations

While substantive offers are efficient in reducing administrative cost, they typically eliminate both the need for a hearing—at which the victim can speak about the loss suffered—and the resulting benefits of closure and a sense of fairness that may accompany such an opportunity to be heard. Moreover, they necessitate the creation of an objective mechanism for payment and compensation, which may entail substantial costs if, despite the similarity in cause, the particular harms suffered are highly variable. For this reason, Congress has granted some public tribunals the ability to make individualized damages determinations. For example, the Marshall Islands Nuclear Claims Tribunal could make reparations without particular direction or guidelines on how to allocate the allotted funds, 128See Agreement Between the Government of the United States and the Government of the Marshall Islands for the Implementation of Section 177 of the Compact of Free Association, arts. X, XII, available at http://www.nuclearclaimstribunal.com/177text.htm. and the special master in the September 11th Fund was empowered to create the compensation structure for victims. 129See September 11th Victim Compensation Fund of 2001, 28 C.F.R. § 104 (2002).

From a systems design standpoint, this allows for greater accuracy in the determination of compensation where the cause of the injury may be common but the harms are highly individualized. These systems are typically more efficient, as questions and variations in harm need not be anticipated in advance, but can be addressed on an as-raised basis—mirroring our system of common law precedent. Over time, these systems typically develop their own interpretations of legal questions common to the relief, taking on a higher degree of standardization akin to substantive offers and enhancing the legitimacy of the tribunal’s determination. 130See Kenneth R. Feinberg, Unconventional Responses to Unique Catastrophes, 45 Akron L. Rev. 575, 578 (2012).

Consider the Gulf Coast Compensation Fund created by BP in the wake of the Deepwater Horizon explosion. 131Initially, BP created a claims facility as required by the Oil Pollution Act; however, two months later this was replaced with the GCCF pursuant to BP’s negotiations with President Obama. For discussion, see Mullenix, supra note 32, at 833–35. BP framed the GCCF as a full compensation fund: it would pay out at a 100% rate for all damages recognized by the Fund’s administrators. 132See id. at 834 & n.69. Full-payout settlements are relatively uncommon in aggregate and single-plaintiff litigation, as defendants commonly negotiate for a risk discount for their surrender of remaining defenses, a reduction to net present-value prices, and even a reduction in litigation costs. 133See McGovern, supra note 15, at 1380. But disaggregation does not allow for these negotiations with an agent—in contrast to the discounts negotiated with class counsel in aggregate settlements.

Instead, BP needed to entice each claimant to affirmatively opt into its fund. Its selection of a full-payout rate provided a preliminary assurance to potential claimants that, so long as the administrator properly determines the claims payable, they should receive the greatest possible compensation for their losses—free of the allocational error, diminution for attorneys’ fees and costs, and delay inherent to the class action system. 134See id. This substantive promise of full compensation combined with the procedural structure and selection of Ken Feinberg—the September 11th Fund administrator—enhanced the legitimacy and perception of fairness for some claimants. 135See Mullenix, supra note 32, at 819–21.

Although the GCCF was formally structured as a fixed fund, its structure created very different incentives and efficiencies than Rule 23 fixed funds. With some exceptions, parties in class settlements typically first negotiate the size of the fund, only secondarily turning to questions of allocation. 136See supra Part I. This negotiating structure flows naturally from the defendant’s concern with limiting its total expenditure on the one hand, and the class counsel’s concern with payment structure, which is keyed to the aggregate amount recovered, on the other hand. But it is equally a result of the tradeoffs that can be made in the negotiation process—for example, expanding or contracting the class definition or scope of claims covered by the settlement to obtain a balance between the scope of peace granted to the defendant and the price paid.

Although technology has improved the distribution process and costs to decrease the residual, 137For example, in some recent settlements, monies have been directly credited to absent class members’ accounts, yielding substantial cost savings that allow payments to be made to victims suffering far smaller wrongs than was traditionally possible. See, e.g., Settlement Agreement and Release ¶¶ 81, 87, In re Checking Account Overdraft Litigation, No. MDL 2036 (S.D. Fla. May 6, 2011), available at http://www.bofaoverdraftsettlement.com/LinkClick.aspx?fileticket=rxwM1SXyxBc%3d&tabid=67&mid=415 (stating that the settlement agreement provided that awards for improper overdraft fees would be paid by crediting each account holder’s account, only allowing a physical check to be mailed if the credit is unfeasible or unreasonable; the settlement also stipulated a cy pres distribution for the residual). there inevitably remains a balance in the settlement fund. Courts expect as a matter of course that the residual will be donated through a cy pres charitable distribution. The return of the residual balance to the defendant is not permitted—as courts expect that the settlement value reflected the proper level of damages—and parties, having reached the end of the settlement process, are reluctant to risk rejection of the settlement over the allocation of the cy pres. The traditional class settlement thus requires extensive discovery with respect to the likely scope and magnitude of damages prerequisite to settlement.

In contrast, the GCCF’s structure did not require this investment of resources prior to settlement. The Fund was given substantial resources that currently appear to far exceed the actual damages it is meant to secure, and, were the Fund to become insolvent, uncompensated claimants could avail themselves of the default litigation system. 138See Mullenix, supra note 32, at 834. This largesse was possible because the residual of the Fund reverts to BP, and because the payments to the Fund are structured longitudinally, allowing BP to fund the GCCF with future revenues. 139See id. at 834 n.69, 835–36. This allowed the Fund to be created while the oil spill was still ongoing—a result that would be highly improbable under the traditional fixed-fund class settlement, given the fiduciary obligations of class counsel. Thus, although disapproved under the traditional Rule 23 structure, an opt-in early settlement fund can utilize a BP-style fund structure to deliver compensation much more quickly than traditional settlement funds. 140Press Release, Dep’t of Justice, Audit of Gulf Coast Claims Facility Results in $64 Million in Additional Payments (Apr. 19, 2012), http://www.justice.gov/opa/pr/2012/April/12-asg-500.html (describing fund as “a significant advance in disaster response”).

From this perspective, the waiver of substantive liability questions can generate a substantial joint benefit sufficient to create a zone of agreement without litigation or discovery, given that aggregation costs can equal—and in some cases even exceed—the alleged liability.

C. Predispute Creation of Disaggregative Mechanisms

Prospective waivers of aggregation are premised upon a simple insight: The aggregate litigation system results in high costs, diminishing net value to both parties. Eliminating these costs—both the procedural checks aggregation entails to ensure due process and the free-riding by passive absent class members—can create joint gains.

Publicly created prospective tribunals present a somewhat different design challenge than purely retrospective tribunals, because they are intended to create a streamlined framework for continuing or future harms. Typically, these systems are designed to promote the efficient resolution of claims against the government as an employer, as with the Public Safety Officers’ Benefits Program, Radiation Exposure Compensation Program, Countermeasures Injury Compensation Program, and Smallpox Compensation Program, which operate as specialized, public workers’ compensation systems. 141See statutes cited supra note 7; see also Peter H. Meyers, Fixing the Flaws in the Federal Vaccine Injury Compensation Program, 63 Admin. L. Rev. 785, 816–37 (2011) (analyzing the structural features of public workers’ compensation as well as general public compensation programs). Thus, these systems usually employ strict liability structures in which the substantive burden upon claimants is reduced and compensation is awarded based upon a grid—one that often awards less compensation than the tort system would for the same harm.

Prospective agreements between private parties have likewise modified the compensation framework, but they have taken a very different form. Private defendants have typically offered supercompensation in excess of that available in the default legal system—in contrast to the decrease in compensation common in the public model. Both models provide for procedural streamlining, reducing the litigation costs to the would-be defendant and, typically, providing greater compensation to claimants. Viewed from the perspective of any claimant, the offers of private defendants are actually superior to not only the default litigation system, but even the prospective systems created by the government as a defendant.

Why then have these private aggregation waivers engendered so much opposition? 142See, e.g., Aaron-Andrew P. Bruhl, The Unconscionability Game: Strategic Judging and the Evolution of Federal Arbitration Law, 83 N.Y.U. L. Rev. 1420, 1457 (2008); Gilles, supra note 32, at 373; Jean R. Sternlight, As Mandatory Binding Arbitration Meets the Class Action, Will the Class Action Survive?, 42 Wm. & Mary L. Rev. 1, 5 (2000). The simple answer is that, in contrast to the high-value claims common to the public system, these corporate provisions have the greatest value where the expected claims rate will be particularly low. This observation implicates difficult questions about optimal deterrence and the role of class action litigation.

One of the predominant uses of class action litigation in recent decades has been the pursuit of small-value wrongs against large groups, correcting the underenforcement of claims; consumer claims provide a familiar example. But the high costs of litigating pressure even defendants that expect to prevail on the merits to consent to an early, nuisance settlement of a claim. This decision to settle will occur even in cases the defendant believes are so frivolous that few if any plaintiffs would have pursued it in single-plaintiff litigation—not because of the costs, but because of the weak merits of the case. In recent years, with the increasing expanse of hold orders in the new era of e-discovery, the costs of avoiding claims of spoliation have only increased this pressure. For defendants, aggregation imposes substantial costs—whether correctly imposed for small-value wrongs, or inefficiently imposed through nuisance litigation—that they would rationally prefer to avoid.

Streamlined procedural provisions are thus commonly used where a defendant anticipates claims that few absent class members are likely to bring individually. Most often, these explicit or implicit aggregation waivers are embedded in a dispute resolution provision providing for streamlined individual hearings or arbitration. However, the same result may be substantially achieved by contracting for forums that fracture the would-be class.

To better frame our discussion, consider the provision at issue in AT&T Mobility LLC v. Concepcion. 143131 S. Ct. 1740 (2011). The plaintiffs sued alleging that when AT&T offered new subscribers a free cell phone, it should not only have provided the cell phone free of charge but also paid the state sales tax on the phone. 144See id. at 1744. AT&T’s form contract provided that the parties agree to resolve disputes valued at under $10,000, at the customer’s option, through arbitration or small-claims court, with AT&T waiving its rights to attorneys’ fees, agreeing to pay double plaintiff’s attorneys fees, and finally agreeing to pay a minimum of $7,500 to the claimant if the amount awarded to the customer in arbitration exceeds the last written settlement offer of AT&T. 145See id. The company sought to provide more efficient and cost-effective remedies for consumers than the default system—in exchange for a waiver of the consumer’s ability to bring claims on behalf of third-party consumers. 146See id. The Supreme Court upheld AT&T’s provision. 147See id. at 1748–49.

In its opinion, the Court noted that AT&T’s system permitted the effective vindication of consumers’ rights in ways that many consumers would reasonably view as superior to the default system. 148See id. at 1745. By limiting the scope of the claim to those who choose to vindicate their claims, the risk of loss for the corporation is decreased, allowing it to not only streamline procedure and offer supercompensation, but also make quick settlement offers. 149See id. at 1752–53.

Contextualizing these provisions within the spectrum of disaggregative mechanisms, procedural streamlining offers plaintiffs a weaker benefit relative to other responsive opt-out mechanisms. Pure streamlining provisions reduce the costs of litigation, converting claims that would be negative value under the default rules governing single-plaintiff litigation into viable claims. 150See id. at 1749. However, no provision is made for either waiving common questions or aggregating the costs entailed in their proof. The consequence is to return these contracting parties to the redundancies and investment asymmetries that existed prior to Rule 23’s enactment.

However, in these mechanisms, publicly available aggregation mechanisms function as a default rule, against which the company bargains for waiver. The result is typically an upward adjustment of compensation and streamlined procedure for the handful of claimants who seek to enforce their rights on an individual basis—creating a subsidy by the defendant of the costs entailed in individual, rather than aggregated, claims resolution. 151Common provisions include corporate agreements to pay the costs of arbitration, reasonable attorneys’ fees, or an incentive payment substantially in excess of the alleged damages. See id. at 1744. In this paradigm, defendants contract to offer supercompensation to the handful of claimants who choose to bring claims in a disaggregative system, while still reducing overall costs to the corporation. As the Supreme Court noted in Concepcion, these provisions are thus superior for those claimants who choose to pursue relief—providing a larger recovery far sooner than aggregation would likely yield. 152See id. at 1752–53. Removing both the costs of aggregation and its de facto mandatory class generates this zone of shared gain.

D. Common Features Across Structures

This Part has posited that the array of unilateral settlement offers and private claims funds are not isolated one-offs, but instead are part of a far broader phenomenon. Across the spectrum of claims types, from small-value claims for defective children’s toys to high-value claims for mass torts, defendants are increasingly making unilateral settlement offers. 153See Samuel Issacharoff & Robert H. Klonoff, The Public Value of Settlement 78 Fordham L. Rev. 1177, 1182–83 (2009) (providing examples of recent settlements). While prominent scholars like Samuel Issacharoff have noted that the lack of closure will substantially limit these funds, this Article posits a far more complex and nuanced picture. 154See, e.g., Samuel Issacharoff & D. Theodore Rave, The BP Oil Spill Settlement and the Paradox of Public Litigation, 74 La. L. Rev. 397 (2014).

Understanding when a defendant will choose to propose a disaggregative structure requires not only situating these funds within the genus of postdispute settlement offers, but also the broader family of disaggregation. By recognizing that these forms of disaggregation share certain common motivations and purposes, we can, for the first time, begin to develop a more nuanced analysis and typology of these structures. In turn, this allows for a more careful consideration of the benefits and potential risks of these structures.

As the defendant typically proposes disaggregation, we should expect to see these proposals made with higher frequency where one of three conditions exists. First, where the defendant expects it can obtain a high participation rate, substantially achieving closure through its ex post creation of a tribunal. Second, where the defendant expects that its settlement offer will prevent the successful certification of a class—whether by dissuading the filing of a class action or by preventing certification under Rule 23. And third, the defendant may propose a predispute aggregation waiver where it expects that its largest costs will come from either small-value or nuisance suits, premised on alleged violations that few of those with whom it contracts would pursue.

The potential combination of features and structural configurations of disaggregative tribunals creates an almost limitless set of design possibilities and structures. 155See Hensler, supra note 58, at 1443 (referring to these as “a dizzying array of potential claims resolution facility designs”). Yet, while disaggregative systems have been widely utilized by Congress, given their superiority in efficiently administering compensation, congressional structures are inherently public in their creation and operation, subject to the indirect checks of democracy. In contrast, the nature of privatized disaggregation raises questions about both the existence of dysfunction in the disaggregative process and the extent to which these mechanisms further or undermine the obtainment of the public purposes of the litigation system. It is to those questions that the next Parts turn.

III. The Specter of Dysfunction in Disaggregation

Disaggregation stands as a new dispute systems design framework, a next-generation system built upon the foundations and dysfunctions of the aggregate and single-plaintiff litigation systems that preceded it. Inevitably—and indeed, intentionally—this new privatized system substantially alters the nature of both compensation and deterrence, rendering our existing analyses in the single-plaintiff and aggregate contexts inapplicable. This experimentation with disaggregation thus prompts us to embark upon the creation of a new branch of scholarship focused upon the assessment of these distinctly unique strategic behaviors and their fit with our normative goals.

It is already foreseeable that disaggregation will generate a new set of dislocations and dysfunctions, like the aggregate and single-plaintiff litigation structures that preceded it. Just as aggregate litigation could not be fully explicated in a single article—but instead has led to tens of thousands of scholarly articles and books 156Shapiro, supra note 1, at 914–16 & n.2 (querying whether any law review existed that had not published an article on class actions). —so too the questions posed by disaggregation are far too broad and complex for any single article to address comprehensively. This Part therefore seeks only to commence the dialogue about the instrumental consequences of this new generation of private, disaggregative dispute resolution mechanisms.

Implicit within this discussion is a recognition of the potential for the rise of disaggregative mechanisms to radically reshape the role of aggregation mechanisms in two key ways. First, it reshapes the central inquiry of our study of aggregation. Complex litigation scholarship has long focused upon developing a set of mechanisms for aggregate litigation that provides an optimal balance of autonomy and collective interest in the context of mass claims, furthering both compensation and deterrence. 157See Nagareda, supra note 20, at 750–51, 754. The use of disaggregation now creates a second line of inquiry, focused upon whether aggregation mechanisms operate to generate optimal results when employed as default rules against which parties bargain—rather than as mere ends in themselves.

Second, parties are crafting systems designed to expand the pool of resources available for private division between the parties by removing those systemic features designed to further the public’s interest. The question of which of disaggregation’s alterations are normatively desirable thus implicates fundamental questions about the dual public-private role of private rights of action as mechanisms for the enforcement of public interests.

A. The Compensatory Function of Law

While the savings inherent to disaggregation are a universal driver in its adoption, the legal and market dynamics differ between pre- and postdispute mechanisms, creating different structures and potential dysfunctions. In theory, disaggregative systems can offer near-perfect compensation and deterrence, through streamlined, individualized determinations of damages and the elimination of litigation costs that distort both compensation and deterrence. Yet, in the shadow of aggregation, defendants are not paying the net expected liability to each plaintiff. Contrary to the existing literature on these mechanisms, 158See, e.g., Issacharoff, supra note 24; Issacharoff & Rave, supra note 154. I argue that bidirectional dislocations exist and identify these competing pressures’ impact in shifting compensation away from the value of the expected legal recovery.

1. Pure Procedural Streamlining

Predispute disaggregation provisions are most common where the defendant anticipates aggregate litigation brought by a handful of plaintiffs, with whom the company has a contractual relationship, on a claim that the company believes few class members would choose to pursue individually. For these defendants, supercompensation can be offered given the anticipated cost savings generated by avoiding a de facto mandatory class. These anticipated savings are typically so great that the company is willing to streamline procedure, realizing that even if the resulting award is erroneous it will be limited to just a single plaintiff. 159See AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1752 (2011) (“Informal [arbitration] procedures do of course have a cost: The absence of multilayered review makes it more likely that errors will go uncorrected. Defendants are willing to accept the costs of these errors in arbitration, since their impact is limited to the size of individual disputes, and presumably outweighed by savings from avoiding the courts.”). This observation is particularly important as it reverses the unique asymmetric settlement pressures of aggregation upon defendants. 160See id. (“Faced with even a small chance of a devastating loss, defendants will be pressured into settling questionable claims.”). Equally important, plaintiffs are substantially returned to the world of single-plaintiff litigation, bearing the obligation to establish all elements of their stated claim and damages. These new dynamics give rise to the potential for bidirectional deviations from optimal compensation.

In the predispute disaggregation context, terms have often been more favorable to claimants than required by law, as exemplified by the combination of supercompensation and fee shifting with procedural streamlining incorporated in the provision in Concepcion, profiled in Part II. 161See supra notes 143–52. But, as precedent develops in support of these provisions and the minimum requirements for enforcement become clear, the provisions’ contours should rationally be determined by a combination of market functioning and cost. 162For a discussion of current doctrinal trends in lower courts, see Myriam Gilles, Killing Them with Kindness: Examining “Consumer-Friendly” Arbitration Clauses After AT&T Mobility v. Concepcion, 88 Notre Dame L. Rev. 825 (2012).

As I have argued elsewhere, predispute procedural terms are uniquely susceptible to market failure, as contrasted with substantive terms. 163See Jaime Dodge, The Limits of Procedural Private Ordering, 97 Va. L. Rev. 723, 755–64 (2011). This dynamic results from the rational decision by recipients of form terms to not bear information costs. 164Id. at 758–59. This dynamic occurs because the cost of studying the term would exceed the potential benefit or cost of these contract terms as adjusted for probability and magnitude of expected risk. 165See id. at 759. The full analysis is complex, but the underlying logic is familiar: if it would cost $250 to retain an attorney to advise on whether a procedural clause accompanying a $20 printer cartridge is fair, the consumer will not seek legal counsel. 166See id. As a result, individuals cannot properly price procedural terms. 167See id. at 757. For a discussion of the frequency of these form contracts, see Russell Korobkin, Bounded Rationality, Standard Form Contracts, and Unconscionability, 70 U. Chi. L. Rev. 1203, 1203 (2003) (noting that “nearly all commercial and consumer sales contracts are form driven”); and W. David Slawson, Standard Form Contracts and Democratic Control of Lawmaking Power, 84 Harv. L. Rev. 529, 529 (1971) (estimating that more than ninety-nine percent of contracts are based upon form terms). This rational risk-taking and resulting market failure should trigger a race to the bottom for corporations with respect to these terms—decreasing the extent to which compensation or terms offered favor claimants—although, as I discuss below, there are limits to this race. 168See Jean R. Sternlight & Elizabeth J. Jensen, Using Arbitration to Eliminate Consumer Class Actions: Efficient Business Practice or Unconscionable Abuse?, Law & Contemp. Probs., Winter/Spring 2004, at 75, 96–99.

In the shadow of this market failure, private forums have a structural incentive to favor the repeat-player drafting party. 169See, e.g., Developments in the Law—Access to Courts, 122 Harv. L. Rev. 1151, 1174–75 (2009) (detailing concern with repeat-player bias). The risk of institutional capture encompasses both the crafting of procedural rules and the selection of particular arbitrators. 170See Dodge, supra note 163, at 742 (discussing forms of institutional capture in arbitration, litigation, and the crafting of substantive law); Gilles, supra note 32, at 410–12 (discussing repeat-player leverage in the context of institutions’ class arbitration policies). The public identification of the institutions that favor corporate repeat players over those whose business model focuses upon scrupulous neutrality often takes a number of years, as win-loss statistics do not distinguish institutional bias from a selection bias in the underlying cases—a phenomenon exaggerated and further contaminated by the likelihood of settlement. 171See Resnik, supra note 16, at 109–11 (discussing the delay in identifying the National Arbitration Forum as a potentially biased forum, in which repeat players “won virtually all the cases”). To the extent that structural biases promote some degree of bias toward the drafting party, this may impact the likelihood of a finding of liability as well as the amount of compensation awarded. 172See Miles B. Farmer, Note, Mandatory and Fair? A Better System of Mandatory Arbitration, 121 Yale L.J. 2346, 2356–57 (2012) (discussing systemic biases in favor of repeat players and resulting impacts upon substantive outcomes). Thus, while the aggregate determination of compensation introduced allocational error, 173See supra Part II.A. privatized disaggregation risks a different type of error—bias in the underlying determination of the case.

To be clear, having served as both counsel in arbitration and as an arbitrator, I do not believe that all or even many arbitrators are consciously or unconsciously biased, nor are all arbitral institutions. Instead, my claim is that the presence of a minority of bad actors creates pressure upon those that prefer to operate in a fair system, particularly to the extent that the selection of an arbitral forum is a low-salience or even nonsalient term for consumers. 174While a robust literature discusses this market failure, the best evidence of how widespread this phenomenon is may well be its appearance at the center of a South Park episode’s plotline. Compare Dodge, supra note 163, at 755–64, 767 (reviewing literature supporting a conception of “ex ante exceptionalism” with respect to market failure for procedural terms in form contracts), with South Park: HUMANCENTiPAD (Comedy Central television broadcast Apr. 27, 2011), available at http://www.southparkstudios.com/full-episodes/s15e01-humancentipad (focusing upon the unintended consequence of Kyle Broflovski’s agreement to Apple’s terms and conditions without first reading the terms). Corporations that take advantage of procedural market failure are able to use these gains to provide more generous salient terms, or improve profits—creating pressure upon other corporations that incentivizes a race to the bottom. 175See Dodge, supra note 163, at 760–61.

Thus, one set of pressures suggests the potential for undercompensation, resulting from the increased probability of market failure. This assertion is not to say that all terms facially beneficial to the nondrafting party will be phased out with the clarification of precedent and transition to a more pure market system. To the contrary, mutually beneficial terms are likely to continue. For example, the streamlining of procedure is a benefit to not only the plaintiff, but the defendant as well. Likewise, offers of supercompensation are typically conditioned upon events that will only occur in cases of substantial malpractice by corporate counsel. Corporations are therefore likely to continue to offer these terms, as they suggest fairness and legitimacy to those consumers who review the policy, yet are cost neutral or even favorable to the corporation.

A second set of pressures weighs in the opposite situation, pressing the defendant to agree to provide supercompensation—compensation in excess of that which would be objectively determined to be due. For many small-value claims, removing the specter of aggregation has the impact of making the claim one in which the costs of defense exceed the alleged liability. The cost-shifting provisions required to demonstrate that the procedural alterations have not prevented the vindication of the underlying substantive law exacerbate this dynamic. A company should rationally opt to settle these cases rather than bear the costs of defense—even for claims that it believes are meritless. 176See, e.g., Hay & Rosenberg, supra note 99, at 1378.

There is an outer limit imposed on this settlement principle—where the corporation expects that offering quick settlement will cause an increase in the number of claims sufficient to outweigh the cost savings. Yet this will frequently remain an outer limit and not the ordinary corporate response, because the small amount in dispute not only incentivizes the company to rationally settle but also will disincentivize many potential claimants from filing claims given the opportunity cost even streamlined litigation entails. 177See AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1761 (2011) (Breyer, J., dissenting).

Taken together, these competing tensions result in a substantial shift in the dynamics of compensation, distinctly different from those of either single-plaintiff or aggregate litigation. In contrast to aggregation, predispute waivers of aggregation typically yield full compensation for those who file claims—in many ways harkening back to the compensation available in the single-plaintiff system. Yet, in order to survive judicial scrutiny, these provisions typically decrease the direct costs of pursuing these claims, converting claims whose direct costs would outweigh the relief available in the single-plaintiff litigation system into viable, compensable claims. Thus, the net compensation available for the claimant in disaggregation is theoretically superior to that of either of the existing public litigation mechanisms.

However, streamlining and cost shifting have the consequence of discouraging the meaningful assessment of the merits of the claim, as the game is frequently not worth the candle for the defendants. As a result, corporations should often agree to full payment of the alleged compensation due, rather than bear the greater costs of even streamlined merits determination. Moreover, in cases of sufficient monetary value, where contesting the claim is worthwhile, the question is whether market failure will encourage the defendant’s selection of a forum susceptible to repeat-player bias.

This also suggests the limits of disaggregation via streamlining. Well-drafted predispute provisions should carve out anticipated claims where the corporation would prefer to reduce the risk of inconsistent judgments or seriatim lawsuits—for example, antitrust objections to mergers are often better addressed in aggregate litigation. 178See Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 594 (2012) (granting certiorari to review the Second Circuit’s invalidation of a class arbitration waiver in the context of federal antitrust allegations).

Similarly, the value of disaggregative streamlining provisions is decreased with respect to claims that most victims will choose to bring even in a single-plaintiff format—for example, wrongful death claims. The disaggregation provision would in these cases yield a relativistic benefit for the corporation by returning to the investment asymmetries of the pre–Rule 23 world. However, the expansion of contractual aggregation, joint representation agreements, and MDL mitigate the benefit the corporation is likely to receive—and in turn reduce the joint gains available for division that are the foundation of private disaggregative mechanisms.

2. Substantive Streamlining

Postdispute disaggregation typically occurs where the costs of aggregate litigation, in both delay and added procedural checks, exceed the differential in aggregate settlement values, such that the defendant can propose a settlement superior for all parties as contrasted with the default of aggregation. 179See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 598 (1997) (“‘The most objectionable aspects of asbestos litigation can be briefly summarized: dockets in both federal and state courts continue to grow; long delays are routine; trials are too long; the same issues are litigated over and over; transaction costs exceed the victims’ recovery by nearly two to one; exhaustion of assets threatens and distorts the process; and future claimants may lose altogether.’” (quoting Report of the Judicial Conference Ad Hoc Committee on Asbestos Litigation 2–3 (1991))); see also David Rosenberg & Steven Shavell, A Simple Proposal to Halve Litigation Costs, 91 Va. L. Rev. 1721, 1727 (2005) (noting that, even in settlement, “on average, it costs approximately one dollar in legal expenses for the legal system to transfer one dollar from a defendant to a plaintiff”). But, given this win-win potential, why do some settlement offers succeed where others fail? And what does this suggest for the accuracy of compensation in these systems?

When confronted with angry Super Bowl ticketholders denied seats at the game, the NFL unilaterally offered ticketholders the choice of (a) cash plus a ticket to the next Super Bowl, (b) a ticket to a future Super Bowl plus airfare and a hotel, or (c) a pure cash payout of either $5,000 as a lump sum or a greater amount with documentation of expenses. 180See Stahl, supra note 121.

In the wake of the Concordia shipwreck, the cruise line made a similar offer to surviving passengers: a flat rate payment of around $14,500, in addition to reimbursement of their tickets, cost of flights home, and other documented expenses incurred in Italy in the days following the wreck. 181See Pianigiani, supra note 120. The amount of the settlement was carefully negotiated at arm’s length with prominent consumer groups, in contrast to the NFL offer’s unilaterally set terms. 182See Barbie Latza Nadeau, Why Survivors Aren’t Buying Costa Concordia’s Compensation Offer, Daily Beast (Jan. 27, 2012), http://www.thedailybeast.com/articles/2012/01/27/why-survivors-aren-t-buying-costa-concordia-s-compensation-offer.html.

Given the enhanced procedural legitimacy in the Concordia offer, what factors prompted the ticketholders to accept the NFL settlement 183See Gary Mihoces, Judge: Cowboys Off the Hook in Super Bowl XLV Seat Fiasco, USA Today (July 19, 2012, 9:35 PM), http://usatoday30.usatoday.com/sports/football/nfl/story/2012-07-19/dallas-super-bowl-lawsuit/56347366/1 (noting that most fans accepted the NFL’s settlement offer rather than participate in the class action litigation). while hundreds of passengers have filed suit rather than accept the arm’s length Concordia settlement? 184See Nadeau, supra note 182 (describing survivors’ reasons for hiring counsel for a class action lawsuit rather than accept the settlement offer). And equally important, what does this suggest about the viability and limits of postdispute disaggregative offers as a compensation mechanism?

Disaggregative settlement offers typically seek to generate joint gains through elimination of contested litigation and, in turn, the necessity of plaintiffs’ counsel. Yet, without legal counsel to educate claimants about their legal rights, the legal inquiry into liability and damages is transformed into an individualized assessment of the acceptability of the settlement offer, largely defined by subjective perceptions of fairness as defined by each victim.

This transformation suggests three limitations upon the types of disputes that are likely to generate significant participation in a disaggregative regime.

First, disaggregative offers may yield low participation rates where potential claimants, without the aid of counsel, cannot accurately ascertain the legally available compensation. In the NFL settlement, individual parties could easily determine their damages—indeed, they possessed superior information on this point as contrasted with either the defendant or the putative class’s counsel. 185See Stahl, supra note 121.

But in other circumstances, victims may not be able to accurately assess the settlement’s compensation as against legal damages. Like the Super Bowl ticketholders, the shipwreck victims had superior knowledge as to their out-of-pocket expenses related to the wreck. 186See Nadeau, supra note 182. But, while they knew of the terror they suffered during the shipwreck and the psychological damage many reported in the weeks and months that followed, most were unlikely to know the legal value of these harms. 187See id. (“These people have no idea what their agonizing experience is worth yet. It could take years to understand the full impact of the experience.” (internal quotation marks omitted)). Thus, while in many cases, claimants will have superior knowledge of the facts underlying their particular claims, they may lack the expertise to understand the value of these claims.

Second, the lack of plaintiffs’ counsel may prevent the effective education of potential claimants as to legal barriers to their recovery. Because the victims do not know how to calculate their damages, they may overestimate their likely recovery. But, it may also be the case that the victims do not properly understand the procedural hurdles or substantive problems with the merits of their case. As a result, even in cases where the parties can both easily agree on the value of damages, if the defendant seeks to reduce the compensation to adjust for the likelihood it would succeed in its defense, victims may believe the offer is unfair and refuse to participate.

As a result, defendants may not receive substantial participation where they seek a risk premium or deduction for statutory or contractual limits upon damages. Unlike the NFL offer, which was made at the top of the range of available damages without regard for uncertainty surrounding the liability phase, 188See Stahl, supra note 121. the Concordia settlement was made in the shadow of legal limitations on recovery. The cruise line passengers had agreed to a forum selection clause providing for litigation in the Italian courts, which would impose additional burdens, likely award less compensation than American courts, and would require the posting of a judiciary tax. 189See Costa Concordia Survivors Describe “Goliath” Fight Against Cruise Industry, NBC News (July 12, 2012, 8:22 AM), http://rockcenter.nbcnews.com/_news/2012/07/12/12701697-costa-concordia-survivors-describe-goliath-fight-against-cruise-industry?lite.

In addition, the cruise company alleged that indemnification limits made the proposed settlement more generous than the relief available in court, even though it was a substantial diminution as against the typical American judgment for such a claim. 190See Nadeau, supra note 182 (“This compensation package is higher than the current indemnification limits provided for in international conventions and laws currently in force . . . .” (internal quotation marks omitted)). In announcing the claim, the company pointed out that the offer exceeded the indemnification limits—although it did not state what those amounts were, nor whether these were contractual or statutory limits. 191See id. The company has made public that the wrongful death liability will be capped at $75,000, in contrast to the $2–$5 million typical in airplane fatalities. 192See Costa Concordia Survivors Describe “Goliath” Fight Against Cruise Industry, supra note 189. The would-be defendant struggled to gain credibility with the passengers and, in turn, to educate them as to the merits of the deal. Thus, not having class counsel created a substantial void in terms of both legitimacy and, in turn, persuading the victims to participate in the settlement.

Third, disaggregative mechanisms are likely to be most effective for mid-sized claims: small-value claims will have payout offers so small that few participants are willing to apply for compensation, while large-value claims may incentivize would-be plaintiffs’ counsel to take a public stance in opposition to the deal, suggesting that they can obtain superior compensation. 193In high-profile cases, plaintiffs’ counsel may publicly urge plaintiffs not to take the settlement but instead to retain their services. This dynamic is likely to have an asymmetric bias, as only counsel claiming the settlement is unfair will have an incentive to come forward. Indeed, in the Concordia settlement, the plaintiffs’ lawyers urging a class action were publicly asserting that all of the contractual forum selection and liability-limiting provisions would be invalidated. 194See Nadeau, supra note 182.

Implicit in this discussion is the recognition that while these structural features can act as barriers to settlement, they can also shift the substantive content of the settlement offer. In a traditional litigation paradigm, settlement occurs where the differential between the parties’ expected outcome at trial is smaller than the remaining projected litigation costs, such that the costs of continuing litigation exceed the projected gain for the parties. 195See Russell Korobkin, Aspirations and Settlement, 88 Cornell L. Rev. 1, 5–9 (2002); cf. Dodge, supra note 163, at 778. See generally Issacharoff et al., supra note 86. Put another way, litigation costs are rationally borne by parties only to the extent that they generate a convergence in expected outcomes. This convergence is essential to settlement, as it is necessary to the creation of a zone of potential agreement—the settlement range, in which the plaintiff’s projected net outcome at trial becomes less than the defendant’s projected net outcome at trial. 196See generally Roger Fisher et al., Getting to Yes (3d ed. 2011) (discussing negotiation theory); Korobkin, supra note 195, at 5–9 (discussing “reservation prices” in settlement negotiation).

Postdispute disaggregative devices are designed to decrease litigation costs, often through a defendant conceding liability on a particular substantive issue. Such a concession is rational where the defendant believes it is likely to be found guilty, such that the continuation of litigation is unlikely to yield a decrease in liability, but will certainly generate additional costs. But it can also result where the substantive issue is one whose determination is likely to cost more than the liability at stake—making it a negative-value defense. The addition of benefits from a public relations perspective of taking responsibility or the appearance of doing the right thing, and the provision of closure to the market, can add further value to early settlement—further increasing the amount of compensation that can be paid before exceeding the projected net expense of proceeding through trial.

For plaintiffs, the decrease in litigation costs and increased present value of a settlement should likewise result in a decrease in the compensation demanded, creating a bidirectional expansion of the zone of potential agreement between the parties based upon these net savings. However, a number of factors counterbalance this diminution in settlement value.

While claimants may have superior information about their individual damages, in many cases the defendant will have superior information about its liability for the underlying event at issue. Typically this information asymmetry is resolved through discovery, but most disaggregative mechanisms forgo discovery as part of the litigation cost savings. 197See, e.g., September 11th Victim Compensation Fund of 2001, 67 Fed. Reg. 11,233, 11,237 (Mar. 13, 2002) (codified at 28 C.F.R. pt. 104) (“[U]nder the tort system, while many claimants receive extremely large awards, many others walk away empty-handed due to the requirement that plaintiffs prove fault. In contrast, the Fund is a no-fault alternative to civil litigation designed to provide fair compensation in a matter of months.”); Gulf Coast Claims Facility: Frequently Asked Questions, Gulf Coast Claims Facility (Aug. 11, 2010), http://www.restorethegulf.gov/sites/default/files/imported_pdfs/library/assets/gccf-faqs.pdf (detailing that any individuals or businesses in regions affected by the oil spill could file claims with or without an attorney and describing the minimal documentation required as proof).

But given this information asymmetry, the defendant’s offer of settlement suggests an increased likelihood of guilt and thus does not trigger a risk premium in the form of decreased compensation. Put another way, the cost savings generated through avoidance of litigation costs should rationally be distributed between the two parties on the basis of bargaining power. Yet without counsel to educate victims as to the uncertainty with respect to the merits of a case and the legal valuation of their damages, defendants typically offer full compensation in order to obtain claimants’ participation. 198See notes 131–35 and accompanying text. But see Ackerman, supra note 13, at 161–65; Mullenix, supra note 32, at 854–59. The lack of knowledge and the opt-in nature of postdispute disaggregation essentially shift bargaining power in favor of the claimants.

For defendants to rationally agree to pay full compensation, the costs of litigation must exceed the expected value of the defenses they would raise, discounted for probability of success of those defenses. For defendants whose internal investigation has revealed guilt, disaggregative settlement offers clear benefits. But, as exemplified by the NFL settlement, even those defendants whose guilt is far from certain are offering settlements that do not just match but exceed that available by law. 199See Stahl, supra note 121.

This trend toward overcompensation is remarkable, insofar as it reveals that aggregation essentially serves as a tax upon defendants. Confronted with a class action, defendants’ litigation costs and indirect losses from the uncertainty mass claims generate are so great that even where substantial questions exist on the merits, the defendant may still find it cheaper to pay 100% (or more) of damages rather than litigate.

Because disaggregative settlements are contract based, each claimant must affirmatively sign a settlement agreement. To the extent that only a fraction of the victims are expected to file a claim, these monies can be redeployed toward offering enhanced compensation to those that do file. This creates a second basis for funding offers of supercompensation, which lies not in the reduction of litigation costs, but in a redistribution of compensation from those who do not decide to file claims to those who do.

But dislocations can also operate in the opposite direction. Although early settlement offers have thus far included this supercompensation premium to plaintiffs, as these processes gain increased acceptance and legitimacy, this premium may decrease. Indeed, both the September 11th Fund and the GCCF obtained the participation of individuals who received awards less than those available under tort law. 200See BDO Consulting, BDO USA, Independent Evaluation of the Gulf Coast Claims Facility: Report of Findings and Observations to the U.S. Department of Justice 10, 68, 76 (2012), available at http://www.justice.gov/opa/documents/gccf-rpt-find-obs.pdf (DOJ-commissioned report); Ackerman, supra note 13, at 161–65; Mullenix, supra note 32, at 854–59; David F. Partlett & Russell L. Weaver, BP Oil Spill: Compensation, Agency Costs, and Restitution, 68 Wash. & Lee L. Rev. 1341, 1359–61 (2011). Put another way, despite the promise of disaggregation, the efficiencies gained can be utilized to reduce the payout of compensatory damages, shifting the division of the joint gains in favor of the defendant. For example, an offer of only 80% of damages is still objectively superior to an award contested in litigation, given the attorneys’ fees, related costs, and time value of money—even before one accounts for the opportunity cost of delay for those plaintiffs facing unemployment, medical bills, foreclosure, tuition payments for children in college, or any of the other economic realities that might cause a personal preference for immediate payment.

To the extent that individuals rationally accept these offers, it risks underdeterrence by not requiring corporations to bear the externalities of their actions. It also reduces the effectiveness of aggregation for others and may even prevent formation of meaningful aggregation, such that parties effectively revert to the pre–Rule 23 asymmetries and irremediable claims if they refuse the disaggregative offer.

While in a theoretical sense plaintiffs acting in a unified manner could refuse the offer, this bargaining sophistication is somewhat atypical, particularly among a diffuse group of plaintiffs acting independently. As plaintiffs begin to accept the offer as superior to the litigation alternative, the aggregation benefits for those who do not opt into the fund begin to dissipate, increasing the pro-rata cost of litigation and thus decreasing the benefit to remaining in the class action. This strategy allows the defendant to capture the zone of potential agreement created by the avoided litigation costs. This may seem a minor criticism until one considers that the ratio between the administrative costs and recoveries of mass tort cases hovers around 2:3, 201See Robert L. Rabin, Tort System on Trial: The Burden of Mass Toxics Litigation, 98 Yale L.J. 813, 820–21 (1989) (collecting studies). and in some cases legal fees actually exceed compensation 202See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 598 (1997) (quoting Report of the Judicial Conference Ad Hoc Committee on Asbestos Litigation 2–3 (Mar. 1991)) (noting that the transaction costs exceeded the recovery in the asbestos cases by a 2-to-1 ratio). —allowing a substantial downward departure in compensation from the actual damages suffered.

Likewise, complexity generates a lack of transparency, providing opportunities for the defendant to capture gain. In the NFL and Concordia settlements, the determinations of eligibility and amount of damages were rather straightforward. Claims administrators could relatively easily determine who held an affected ticket or had been on the ship, and then review the receipts submitted to determine the total amount owed. But where these determinations are less clear—for example, in BP’s GCCF, the calculation of projected damages for businesses in the midst of a recession—then the design of the dispute resolution system and selection of a neutral have a far greater effect on the ultimate value of the payout in terms of both legitimacy and likely participation.

The nature of privatization typically places defendants in the role of initiators of disaggregation and in turn systems designers. With defendants unilaterally selecting the third-party neutral, there is a natural suspicion of bias, as exemplified by the challenges to the GCCF, notwithstanding the selection of Ken Feinberg, the special master of the September 11th Fund. 203See Mullenix, supra note 32, at 819–21. But for potential claimants, the information costs associated with investigating the neutrality of the proposed neutral will likely exceed the value of the information—particularly given the lack of transparency in arbitration, notwithstanding reporting requirements. 204See Resnik, supra note 16, at 108 (“The constitutional obligations of ‘open courts’ have produced a wealth of data on judges’ salaries, court budgets, case proceedings, and outcomes. In contrast, private dispute resolvers are left to do as they wish, subject only in a few jurisdictions, such as California, to requirements that arbitration providers ‘collect, publish . . . , and make available to the public’ information about parties, categories of disputes, time to disposition, and outcomes.”); Sternlight, supra note 42, at 1662–64; Roger J. Perlstadt, Comment, Timing of Institutional Bias Challenges to Arbitration, 69 U. Chi. L. Rev. 1983, 1985–87 (2002). Indeed, at present there is a lack of any such reporting requirements on these very new tribunals, and it is difficult to conceive of how such a report would be formulated. By definition, the question is not the binary one of liability, but one of relative compensation—thus requiring a third party to independently assess the claims to determine whether the values awarded were “correct” in the view of the third party.

Moreover, because these funds are structured as bilateral contracts between the defendant and each claimant, they are relatively insulated from public checks. Absent a showing of fraud, unconscionability, or other basis for invalidation, the agreements will be enforced. Claimants would therefore rationally operate upon the expectation of bias in assessing the proposed terms, potentially triggering a race to the bottom—as purportedly occurred in the context of consumer arbitration with the National Arbitration Forum. 205See George Padis, Note, Arbitration Under Siege: Reforming Consumer and Employment Arbitration and Class Actions, 91 Tex. L. Rev. 665, 681 n.102 (2013).

This suspicion may be correct even where the selected claims fund administrator is neutral, given the asymmetries discovered in the awards of the BP GCCF. As Linda Mullenix described it,

The serious challenges that scholars have raised with regard to the legitimacy of the September 11th Victim Compensation Fund have even more powerful resonance in relation to the Gulf Coast Claims Facility. . . . For those concerned with the rule of law, equity, and fundamental fairness, the GCCF ought to be a cause for concern. 206Mullenix, supra note 32, at 825.

Since the publication of Mullenix’s critique, the Department of Justice has confirmed her intuitions of systemic bias, finding in its audit of the tribunal’s awards that the average claimant received $8,800 less than the independent reviewers determined to be due, triggering aggregate payments of more than $64 million. 207See Press Release, Dep’t of Justice, supra note 140 (noting total fund payout of over $6.2 billion to more than 220,000 claimants). Moreover, the waivers and releases that GCCF claimants must sign cover not only BP but also any other potentially liable party—a breadth not provided for in the Oil Pollution Act of 1990 but drafted by Special Master Feinberg, to the advantage of BP and its potential codefendants, raising questions for the MDL judge about Feinberg’s impartiality. 208Mullenix, supra note 32, at 843–44 & n.115, 872, 888.

Disaggregation awards compensation on an individualized basis, rather than backing into the compensation due from a classwide settlement agreement subject to pressures toward allocational error. Moreover, the joint gains from the avoided litigation costs and delay make possible full compensation in a way not possible in either the aggregate or single-plaintiff litigation systems.

Yet, the dynamics inherent to a settlement in which claimants decide whether to participate without the benefit of discovery or perhaps independent counsel create a shift from legal entitlement under the law to subjective willingness to accept the deal. The inherently privatized nature of this disaggregation provides a further basis for shifts from the optimal level of compensation, resulting from the rational strategic choices the parties should make in the shadow of aggregation. While myriad factors will determine the direction of this dislocation in any particular settlement, it is clear that disaggregation creates a new set of dysfunctions in compensation that have broader normative consequences for the functions of law.

B. Optimal Enforcement and Deterrence

The innovation of disaggregative mechanisms has been the bypassing of the procedural costs and delay of an aggregate determination on the classwide merits of the case, allowing both parties to obtain better outcomes than traditional aggregation. Yet, even if bypassing the merits determination is rational for the parties, it creates a lack of precision in determining actual wrongdoing: Both innocent and guilty parties may find it cheaper to pay full damages to those who file claims in a disaggregative proceeding than to proceed through class certification and summary judgment.

Yet, unlike in class actions where the innocent defendant can negotiate a smaller settlement than a guilty defendant, in disaggregation there is no single agent to negotiate with on behalf of the plaintiff; therefore, guilty and innocent parties may pay substantially equal amounts to obtain plaintiffs’ participation. While this is a rational decision for the parties, it closes the differential between the two competing states of compliance and noncompliance with the law. As this differential creates the deterrent value of the law, the lack of merits-based determinations in disaggregative settlements may weaken the goal of deterrence. But this realization has deeper theoretical implications, illuminated through two concepts I term “fractional participation” and “reverse preference enforcement.”

1. Fractional Participation

Private rights have traditionally served not only the private interest in obtaining compensation for wrongs, but often have also functioned as a mechanism for the effectuation of the public’s interest in the enforcement of a particular right. Aggregative mechanisms sought to adjust this relationship, providing enhanced deterrence against a backdrop of perceived underenforcement of certain types of claims marked by low net compensation. In this aggregative paradigm, compensation became derivative of the overall settlement and subject to allocational error, resulting in “rough justice” for class members. 209See supra Part I.

Disaggregation upends this traditional conception. Parties are not just putting compensation first and allowing deterrence to follow derivatively; rather, they are trading enhanced compensation for decreased deterrence. Disaggregative mechanisms focus on improved individual compensation relative to both the single-plaintiff and aggregative models, but this autonomy often generates fractional enforcement as only a subset of victims choose to seek relief. The significance of this transformation cannot be overstated. Whereas aggregation focused on correctly assessing the total damages owed by the defendant but potentially erred in compensation, 210See supra notes 40–42 and accompanying text. disaggregation focuses on the individual damages owed to each plaintiff but potentially errs in the determination of total damages. 211See supra note 89–90 and accompanying text.

Disaggregation raises fundamental questions about how to define optimal enforcement and deterrence. Some will likely view this structure as obtaining the correct level of deterrence, providing compensation to those aggrieved enough to file a claim in a broader context of streamlining that generates joint gains for both parties—while preventing the frivolous suits based upon alleged wrongs recognized by only a few class members. 212Cf. AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1751–53 (2011). Others will likely view this structure as effectively precluding enforcement for small-value claims, as there does remain some opportunity cost or burden even in the most streamlined system. Claimants may fail to file claims even for guaranteed money if “claiming the $30.22 were to involve filling out many forms that require technical legal knowledge or waiting at great length while a call is placed on hold.” 213Id. at 1761 (Breyer, J., dissenting).

Disaggregation’s instrumental impact is thus best understood by bifurcating the private from the public interest. The small-value harms frequently the subjects of predispute disaggregation are the types of slights ordinarily borne in society without resort to the litigation system—indeed, “only a lunatic or a fanatic sues for $30.” 214Id. (“The realistic alternative to a class action is not 17 million individual suits, but zero individual suits, as only a lunatic or a fanatic sues for $30.” (quoting Carnegie v. Household Int’l, Inc., 376 F.3d 656, 661 (7th Cir. 2004))). Beyond making available a small-claims-court forum, our legal system does not aggressively facilitate the pursuit of these claims, recognizing the public resource investment as not justified by the benefit.

In contrast, where these small harms occur on a mass-claims basis, the legal system does recognize the necessity of creating an enforcement mechanism to correct the incentive for wrongdoing. But the harm to the individual plaintiff remains unchanged. Aggregate settlements of these small-value claims historically have not provided individual compensation, as the costs of administration would exceed the payment. 215See Six (6) Mexican Workers v. Ariz. Citrus Growers, 904 F.2d 1301, 1305 (9th Cir. 1990) (“When a class action involves a large number of class members but only a small individual recovery, the cost of separately proving and distributing each class member’s damages may so outweigh the potential recovery that the class action becomes unfeasible.”); Notice of Class Action, Proposed Settlement, and Hearing at 4, Settlement Recovery Ctr. v. ValueClick, Inc., No. CV-02638-FMC-(CTx) (C.D. Cal. May 29, 2008), available at http://www.affiliatefairplay.com/service/images/post/agreement_of_settlement.pdf (“Recognizing that administration costs for low-value payments can greatly exceed the value of these payments, no payment will be made if the Settlement Class Member’s pro rata share is calculated to be less than $1.00.”). Instead, the monies recovered are often donated to a charitable purpose through a cy pres distribution. 216See Lemos, supra note 31, at 529. Taken together, this suggests that in these low-value cases the true aim for both the class counsel and the court approving the settlement is deterrence. Compensation is simply a positive externality of the process, where it is possible.

In disaggregation’s shadow, legislatures can no longer presume that the creation of a private right of action will be a substantial mechanism for effectuating enforcement of the underlying substantive law. Instead, the costs of aggregation can be leveraged to create a dynamic in which those plaintiffs willing to seek compensation are overcompensated, while the defendant is able to benefit from fractional enforcement—making all parties better off as contrasted with aggregation, but not fulfilling the public interest. 217Some may even posit that nonfiling individuals are better off. To the extent they also would not have filed a claim form or would have received no compensation because the harm was so small as to obtain only cy pres relief, they receive no less compensation and may benefit from better terms as the company can now deploy the savings to better pricing or substantive product features. See Dodge, supra note 163, at 755–64, for a discussion of both sides of the market functioning argument. If the company is unsuccessful in obtaining a predispute aggregation waiver, then the case remains in the default litigation system with the potential for aggregate litigation, with its high transaction costs and de facto mandatory classes. For the legislator, the spectrum of potential dispute resolution mechanisms and their consequence for the level of deterrence obtained makes it impossible to precisely calculate the deterrent effect a private right of action will obtain.

2. Reverse Preference Enforcement

As with predispute waivers of aggregation, postdispute offers of settlement are likely to obtain only fractional participation. In contrast to predispute waivers that completely prohibit formal aggregation, postdispute settlement operates in the shadow of aggregation. Successful settlement offers will effectively preclude aggregation, but low participation rates allow aggregation to continue. Rejection of the settlement may reflect a belief that the offer was too small for the harm caused. 218See, e.g., Pianigiani, supra note 120 (quoting a representative of a consumer group who described the settlement offered by Costa Cruises as “ridiculous and disproportionate for the damage suffered” (internal quotation mark omitted)). But it might also reflect a belief that the purported wrong was not a wrong at all or not worth pursuing, particularly in cases of alleged mislabeling or other types of false advertising. To the extent that low participation rates are a populist rejection of the pending litigation’s substantive claim, it creates the paradox of “reverse preference enforcement.” The strongest claims de facto preclude aggregation as most claimants participate in the settlement, while the weakest claims allow the pursuit of aggregation and create a de facto mandatory class.

The court’s response to attempts to aggregate in the shadow of a postdispute disaggregative settlement offer can be expected to further alter the settlement dynamics. Facially pro-plaintiff decisions with respect to class action procedure have at times harmed the long-term recovery of victims in unrelated cases, as defendants reduce their settlement offers in response to the perceived degree of closure they receive. For example, where courts have allowed plaintiffs to reopen deals later determined to be disadvantageous to plaintiffs, but not defendants, this creates a one-way ratchet effect that should decrease the settlement values. 219See Richard A. Nagareda, The Law of Class Actions and Other Aggregate Litigation 129–31 (2009) (discussing the original Agent Orange class action settlement, the court’s decision to grant a collateral attack on the settlement resulting in a new round of payouts by the defendant, as undermining the degree of closure received by defendants and thus reducing the amount rational defendants should be willing to pay in future settlements as a result of this shift in the risk premium).

The courts’ response to fractional participation may operate in much the same way. If plaintiffs’ attorneys are able to file class action lawsuits on behalf of nonparticipants, the defendant must anticipate the potential for a low participation rate and the costs attending such litigation. Rather than allocating the cost savings of avoided litigation entirely to the settlement, the defendant should rationally make offers reflective only of the marginal cost of an additional absent class member in the suit—a nominal cost, to be sure. Thus, the creation of a seemingly pro-plaintiff doctrine permitting suit on behalf of those who do not participate in the private settlement may have the impact of reducing settlement offers to those willing to seek compensation.

Having identified the shadow aggregation may cast over disaggregation, the next section turns to exploring the problems this creates for the parties’ private ordering and the second-generation dysfunctions created by attempts of the courts to correct those problems.

C. Process Function and Legitimacy Values

1. First-Level Dysfunction: Defendants as Systems Designers

No disaggregative mechanism is available within our default civil procedure system. As a result, if the parties prefer disaggregation, they must privately create the system. Typically the corporate defendant has served as the systems designer. Pre-dispute, the corporation includes an arbitration or dispute resolution clause in its form contracts. 220See, e.g., AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1744 (2011). Post-dispute, the corporation will typically propose a settlement or resolution process, in which interested claimants may participate. 221See, e.g., supra note 8 and accompanying text; Nathan Bomey, Analyst: GM May Offer $2B-plus to Settle Recall Claims, USA Today (Mar. 26, 2014, 6:40 PM), http://www.usatoday.com/story/money/cars/2014/03/26/gm-recall-claims-settlement-prediction/6930339/. As systems designer, the corporation is then able to select the type of dispute resolution system, craft the procedural rules, decide what tradeoff to make as between accuracy and cost in light of the anticipated magnitude and complexity of claims, and select the neutral or administering body. The role of systems designer is thus a powerful one, capable of dramatically shifting the bargaining power of the parties and the ability to exit in cases of perceived unfairness. 222See generally Nancy H. Rogers et al., Designing Systems and Processes for Managing Disputes (2013) (explaining, through case studies and other examples, how to design and implement an effective alternative dispute resolution system). Moreover, by shifting evidentiary rules and in some cases even altering the underlying substantive law, the designer can substantially impact the degree of deterrence the private right of action will provide. 223See Dodge, supra note 163, at 744.

This power can often be used to design systems that are optimal for all parties involved, enhancing compliance with the underlying law, and even achieving transformative or cathartic ends that the litigation system itself may not obtain. The new mechanisms, procedural innovation, and, more broadly, dispute systems design techniques resulting from the grassroots experimentation with ADR have been lauded for their improved results for parties across a broad spectrum of methodological approaches—from law and economics extolling the ability to streamline costs and improve compliance with substantive terms 224See, e.g., Robert E. Scott & George G. Triantis, Anticipating Litigation in Contract Design, 115 Yale L.J. 814, 856–79 (2006) (exploring the impact of contract design, and in particular the selection of the procedural rules governing a contract, on the economics of litigation). to those focused on the increased party satisfaction and transformative effects alternative processes can generate. 225See, e.g., Fisher et al., supra note 196 (introducing the idea of principled negotiation on the merits that looks for mutual gain based on fair standards); Issacharoff & Klonoff, supra note 153, at 1179–84 (emphasizing the importance of the development of settlements in class actions and private aggregations of cases in the past twenty-five years); Carrie Menkel-Meadow, Peace and Justice: Notes on the Evolution and Purposes of Legal Processes, 94 Geo. L.J. 553 (2006) (commenting on the roles of different legal processes as a part of the justice system, including arbitration and mediation as alternatives to adjudication); Patricia M. Wald, Bureaucracy and the Courts, 92 Yale L.J. 1478, 1483 (1983) (calling for increased accuracy, clarity, and uniformity in the court system to enhance public confidence and satisfaction in the courts).

But, the public courts serve an essential function in the democratic system, from ensuring the public accountability of the courts to guaranteeing that the necessary procedural protections are in place for the obtainment of remedies, exemplified most prominently by the Civil Gideon movement of the past century. 226See, e.g., Harry T. Edwards, Alternative Dispute Resolution: Panacea or Anathema?, 99 Harv. L. Rev. 668, 671 (1986) (“[W]e must determine whether ADR will result in an abandonment of our constitutional system in which the ‘rule of law’ is created and principally enforced by legitimate branches of government . . . .”); Owen M. Fiss, Against Settlement, 93 Yale L.J. 1073, 1085–87 (1984) (arguing that settlement is not an adequate substitute for adjudication); Resnik, supra note 16, at 134–35 (noting that “[a]ggregation can . . . be a form of Civil Gideon for plaintiffs”). The transformation of procedure from public rules set to ensure a fair day in court to mere default rules subject to private ordering thus risks thinning “the substance of procedural due process.” 227Resnik, supra note 16, at 93. Indeed, to the extent that corporations are able to set the rules of procedure, they can shift the deterrent value of the law from that existing under the default rules of procedure.

2. Second-Level Dysfunction: Public Checks upon Disaggregation

Because disaggregative mechanisms are inherently contractual, the existing doctrine has generally placed minimal checks upon their usage. By looking at these mechanisms as a unified system, rather than isolated examples, a number of observations become apparent that have thus far been lurking within the shadows of our scholarship and doctrine.

As a general matter, disaggregative mechanisms are subject to the basic rules of contract and the defenses appurtenant thereto. With respect to predispute procedural modifications, the Supreme Court has created an additional protection, given the ability of alterations to undermine deterrence: procedural modification cannot thwart the substantive enforcement of law. 228See AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1746–48 (2011); Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 636–37 (1985) (“Where the parties have agreed that the arbitral body is to decide a defined set of claims . . . the tribunal therefore should be bound to decide that dispute in accord with the . . . law giving rise to the claim. And so long as the prospective litigant effectively may vindicate its statutory cause of action in the arbitral forum, the statute will continue to serve both its remedial and deterrent function.” (footnote omitted) (citation omitted)). In Stolt-Nielsen S.A. v. AnimalFeeds Int’l Corp., 559 U.S. 662 (2010); Shady Grove Orthopedic Assocs. v. Allstate Ins. Co., 559 U.S. 393; and E.E.O.C. v. Waffle House, Inc., 534 U.S. 279 (2002), the Supreme Court had also tried to limit the extent to which private modification court thwart the public interest. Of course, in practice, the majority has focused on whether a person could bring a claim, if so inclined, seemingly disregarding the opportunity cost of low-net-value claims, to the dismay of the dissenters. 229See, e.g., Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 2304 (2013). The true constraints imposed upon corporations’ use of disaggregation are then not the direct prohibitions of the doctrine, but the pragmatic constraints upon closure that operate in its shadow. 230See generally Erichson & Zipursky, supra note 96 (discussing the difficulty of achieving closure in mass tort settlements despite the demand for it by both plaintiffs and defendants); Nagareda, supra note 20 (exploring the conflict between autonomy for individual plaintiffs and global peace for defendants that is inherent in mass torts).

Predispute disaggregation is available only to contracting parties. Defendants lacking a contractual relationship with their accusers, as is the case with many tort claims, cannot take advantage of predispute disaggregation agreements.

This naturally arising dichotomy has some intuitive appeal: where parties are entering into a contractual relationship, they have the opportunity to assess the other’s reputation and their own interests and decide upon the level of assurance they prefer for their obligations. In contrast, where the parties are brought together by the liability-inducing event, ostensibly without the opportunity for due diligence, a stronger need for state protection is apparent. 231See Dodge, supra note 163, at 776–77.

Postdispute disaggregation now appears to be on the precipice of a similar bifurcation. In recent years, corporations have begun routinely recalling products, creating refund mechanisms, or otherwise offering the relief that would potentially be—or, perhaps more concerning, is already being—sought by a class action. In the last year, both the Seventh Circuit and Tenth Circuit have held that this prevents the pursuit of the class action. 232See, e.g., Winzler v. Toyota Motor Sales USA, Inc., 681 F.3d 1208 (10th Cir. 2012); In re Aqua Dots Prods. Liab. Litig., 654 F.3d 748 (7th Cir. 2011). In Winzler v. Toyota Motor Sales, the plaintiff brought a state law claim against Toyota alleging that the Toyota Corolla and Toyota Corolla Matrix had “defective ‘Engine Control Modules,’” which caused the cars to stall with no warning. 233Winzler, 681 F.3d at 1209. The court held that the voluntary recall filed with the applicable government agency rendered moot a class action because the recall had already provided the entire remedy to which the class members would be entitled. 234Id. at 1211.

In Aqua Dots, the defendants manufactured a toy consisting of small, brightly colored beads; several children ingested the beads and became severely ill. 235Aqua Dots, 654 F.3d at 749–50. The court held that the class could not be certified following a voluntary recall because the Rule 23(a) requirement of adequacy could not be met. “A representative who proposes that high transaction costs (notice and attorneys’ fees) be incurred at the class members’ expense to obtain a refund that is already on offer is not adequately protecting the class members’ interests,” notwithstanding the additional request of punitive damages as these damages would be based upon unmanageable questions of state law. 236Id. at 752. As Judge Easterbrook noted, the “substantial costs of the legal process” offered little value as “plaintiffs could have had refunds—and still can have them today.” 237 Id. at 751. Moreover, there was little reason to believe the court’s notice campaign would be more effective than that of the corporation—most consumers had returned the product, there were no complaints by the governing agency, and it was likely that many of the unreturned products had been used prior to the recall. 238Id.

In a world of fractional enforcement, where a large portion of the alleged victims may never take any action either in support of or opposition to any settlement offer, this substantially shifts the overall cost of a violation to the corporation. The income level of consumers, amount of the purchase, and extent of targeted notice will be among the myriad factors impacting the participation rate in such a program. By their nature, these programs are typically quietly administered and participation information is rarely disclosed, but the range of ratios of claims form return rates, spanning from 1:5 to 1:50, in consumer class actions suggests the magnitude of the shift in enforcement these programs may yield. 239See supra Part I. Thus, the emerging trend within the doctrine has the effect of allowing the defendant to avoid a class action not only before, but also after the dispute arises.

Of course, embedded within this emerging doctrine are also its limits. But to understand these limits, we must understand the preclusive effect of the disaggregative settlement on participating and nonparticipating victims. As lawyers, it seems self-evident that a company offering compensation would demand a waiver in exchange. In these cases, the settlement agreement is like any other single-plaintiff settlement, and can be revisited only upon narrow grounds such as fraud.

But at times, corporations offer compensation without demanding closure—perhaps most commonly to comply with a statutory requirement, to reduce management costs with very low value claims, to increase the legitimacy of the offer and rate of participation in the program, or for a public relations benefit. 240See, e.g., John Hoeffel, PG&E Announces $100-Million Relief Fund for San Bruno Gas Explosion Victims, L.A. Times (Sept. 13, 2010, 1:38 PM), http://latimesblogs.latimes.com/lanow/2010/09/pge-announces-100-million-relief-fund-for-san-bruno-gas-explosion-victims.html. In these situations, the company risks a subsequent suit alleging insufficiency of compensation and seeking the gap between the compensation paid and the total amount available by law. Because these voluntary payments will offset against any award, if the compensation offer is relatively accurate, even with a number of participants, the differential typically will not be sufficient to justify litigation costs. 241The notable exception is where the claims are high value and the litigation seeks punitive damages. See, e.g., Aaron Sankin, PG&E San Bruno Blast Source Revealed: Company Says 1956 Test Caused Deadly Explosion, Huffington Post (June 27, 2012, 3:26 PM), http://www.huffingtonpost.com/2012/06/27/pge-san-bruno-blast_n_1631550.html (noting that a class action for punitive damages had been filed despite PG&E’s disaggregative settlement fund and second quasi-governmental victims’ fund).

The closure problem then is not typically focused on participants, but on nonparticipants. Under the emerging precedent, if full compensation has been offered, then the class action is not seeking anything not already available to the nonparticipants. In such a case, individuals are of course not precluded from seeking individual adjudication or binding together to request an MDL. But, these functionally remain opt-in mechanisms requiring the individual to file a claim—in contrast to the class action, where all absent class members are presumptively included and must affirmatively opt out of the class. Thus, to obtain the aggregation benefits of a class action, successful plaintiffs’ counsel will need to argue that the disaggregative mechanism is overlooking some element of damages or is undervaluing those damages.

D. Constitutional Actors and Disaggregative Mechanisms

Public attempts to restrict disaggregative mechanisms have ebbed and flowed among the constitutional branches over time. With respect to arbitration, a vast literature has arisen addressing the shift from congressional support and judicial hostility at the time of the Federal Arbitration Act. 242See, e.g., Resnik, supra note 16, at 112–18 (summarizing the evolution of the Court’s treatment of the FAA over time). More recently, arbitration fairness acts, seeking to restrict predispute arbitration provisions in consumer and employment agreements, have become perennial proposals in the federal legislature. 243See, e.g., Arbitration Fairness Act of 2013, S. 878, 113th Cong. (2013); Arbitration Fairness Act of 2011, S. 987, 112th Cong. (2011); Arbitration Fairness Act of 2009, H.R. 1020, 111th Cong. (2009). But the judiciary has become more supportive of procedural private ordering, including arbitration, even in cases in which it undermines the enforcement of substantive law. 244See, e.g., Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 2304 (2013); Nitro-Lift Techs., L.L.C. v. Howard, 133 S. Ct. 500 (2012); AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011).

My goal in this section is not to recapitulate this literature. Instead, it focuses upon the undertheorized problem of public checks upon postdispute disaggregation. While Congress could restrict the ability of the parties to enter into bilateral settlements, these provisions are not widespread. 245See Dodge, supra note 163, at 772–76. The primary form of check has therefore been judicial. Again, because the settlements are constructed as individual settlements, the judiciary has few tools to directly intervene in most of these settlements. Instead, the power of the judiciary stems from its willingness or refusal to certify a class action notwithstanding the existence of a settlement offer. 246Compare In re Aqua Dots Prods. Liab. Litig., 654 F.3d 748, 752 (7th Cir. 2011) (denying class certification on the basis that a class action would not provide absent class members any relief not already offered by defendant), with In Re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on Apr. 20, 2010, 910 F. Supp. 2d 891, 920 (E.D. La. 2012) (granting class certification and ordering shutdown of the GCCF in favor of a new class action), aff’d sub nom. In re Deepwater Horizon, 739 F.3d 790 (5th Cir. 2014).

But what does this transition look like when a private fund is replaced with a judicially approved class settlement? The most prominent example—and one that exemplifies the problems with both private disaggregative tribunals and the oversight provided by the backstop of aggregation—is the response to the Deepwater Horizon rig’s explosion. 247Mullenix, supra note 32, at 912 (“The GCCF . . . illustrates an extreme and seemingly lawless expansion of the fund approach to resolving mass claims.”).

On April 20, 2010, the Deepwater Horizon rig exploded, transforming the Gulf Coast economy. 248See In re Deepwater Horizon, 739 F.3d at 803. The resulting oil spill continued for three months, releasing millions of gallons of oil, which forced the prohibition of fishing through large swaths of the Gulf of Mexico, while coastal tourism collapsed. 249See id. The secondary effects of these economic losses threatened even greater dislocations and potential economic collapse.

As the “responsible party” for the oil spill, BP had a statutory duty under the Oil Pollution Act of 1990 to “establish a procedure for the payment or settlement of claims for interim, short-term damages.” 25033 U.S.C. § 2705(a) (2012). Initially, this obligation was satisfied through an ad hoc claims administration process. 251See Mullenix, supra note 32, at 834. But based upon discussions with congressional officials and the President, BP replaced its ad hoc system with an “Independent Claims Facility,” later named the Gulf Coast Claims Facility. 252See id. at 833 n.62, 834 nn.65 & 69.

The GCCF had the potential to generate value for all of the stakeholders. For BP, the Fund’s structure permitted it to make payments in tranches over the course of years, allowing for longitudinal cost spreading. 253See id. at 834–35. While the first $3 billion payment came from BP’s cash reserves, the remaining $17 billion would be drawn from future Gulf Coast revenues—providing an incentive for the President to quietly rescind his suspension of Gulf Coast drilling. 254See id. The Fund also served as the centerpiece of BP’s public relations campaign in the years that followed, a sign—it proclaimed—of its corporate responsibility.

For claimants, the GCCF promised a quick payment to stem any consequential losses, particularly for those who had the most substantial harms. The Fund gave claimants the ability to select the most advantageous system as among a number of options, including the choice of flat-rate payment ($5,000 for an individual or $25,000 for a company), individualized interim quarterly payments based upon proof of ongoing losses, or an individualized one-time payout of all damages. 255See id. at 857. These rapid payments were beneficial to other stakeholder groups, from governments to businesses, concerned with the secondary effects of the direct job losses and property damage created by the spill.

Yet, despite the clear ability for a superior payout structure that avoided the costs and delay of aggregation, benefiting all parties, the privatized nature of the GCCF raised strong questions about its legitimacy. While many agree that Feinberg was selected as the administrator because of his unparalleled experience, including as the special master of the September 11th Fund, 256See id. at 820. many raised questions about whether his judgments favored BP. Critics pointed to the lack of public notice and comment period on the GCCF structure, 257See id. at 841–42 (comparing the relatively lax promulgation of GCCF standards and regulations with the more formal processes associated with the September 11th Victim Compensation Fund). inconsistent awards and the failure to create transparent payment-calculation frameworks, 258See id. at 913. and attempts to impose geographic restrictions and collateral source rules that deviated from the available legal remedies. 259See id. at 861. Also subject to criticism was the ultimate determination that the GCCF had systematically underpaid claimants relative to the payouts the Department of Justice determined to be properly owing—albeit by only one percent. 260See supra Part III.A.

As these questions swirled, many chose to participate in the GCCF nevertheless. 261See Mullenix, supra note 32, at 856 (noting that in the initial “emergency payment” phase of the GCCF, the facility paid out “$3.3 billion to approximately 251,000 claimants”); Margaret S. Thomas, Morphing Case Boundaries in Multidistrict Litigation Settlements, 63 Emory L.J. 1339, 1342–46 (2014). Even with these questions at the margins, it offered immediate payout of approximately the amount owing, without the need to bear attorneys’ fees or other expenses, and thus offered a superior net payout. Others openly opposed the regime as a private, unilateral offer lacking transparency and legitimacy and filed class action litigation against BP, which was brought into the MDL. 262See Mullenix, supra note 32, at 871. Many took no action at all, for reasons that we will likely never discern with certainty beyond scattered anecdotes.

On December 21, 2012, the presiding judge in the Deepwater Horizon MDL, U.S. District Court Judge Carl Barbier, granted final approval to a class settlement addressing medical and economic damages claims. 263In re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on Apr. 20, 2010, 910 F. Supp. 2d 891, 900 (E.D. La. 2012), aff’d sub nom. In re Deepwater Horizon, 739 F.3d 790 (5th Cir. 2014). The court’s order resulted in the immediate freezing of payouts by the GCCF, even to individuals who were negotiating individual settlements with the aid of counsel. 264See Derek Hawkins, BP Spill Fund Freezes Payments After Atty Fees Order, Law360 (Jan. 4, 2012, 1:47 PM), http://www.law360.com/articles/297159/bp-spill-fund-freezes-payments-after-atty-fees-order. This freezing resulted in part from the court’s order that claimants who had reached direct settlements with the GCCF, without the assistance of class counsel and in some cases with their own lawyers, would nevertheless be required to pay six percent of their awards to the MDL plaintiffs’ steering committee—underscoring the expense of aggregation relative to the disaggregative process. 265Id. Judge Barbier’s decision to approve the class settlement for economic and property damage was appealed to the 5th Circuit. The appellants argued that due to the potential inclusion of fictitious claimants, the class settlement did not meet the requirements of Rule 23 and the class settlement lacked Article III standing. In a decision filed January 10, 2014, the 5th Circuit affirmed Judge Barbier’s decision. In re Deepwater Horizon, 739 F.3d 790.The court held that class certification is not precluded if the class may include people who suffered no harm as a result of the defendant’s conduct. Id. at 803–04. The court further held that seeking recovery for an alleged harm suffered is sufficient to satisfy the requirements for Article III standing. Id.

What was the value added from the aggregation? The biggest benefit thus far seems to have been enhanced legitimacy, as the court approved the settlement’s terms and the selection of a claims administrator. Yet, beyond this shift at the top, the class settlement substantially paralleled the GCCF. The disaggregative Fund’s administrator had paid more than $6.2 billion to those who had submitted claims, 266BDO Consulting, BDO USA, Independent Evaluation of the Gulf Coast Claims Facility: Executive Summary 1 (2012), available at http://www.justice.gov/iso/opa/resources/697201241917226179477.pdf (DOJ-commissioned report). working with Garden City Group and PricewaterhouseCoopers as subcontractors in the claims administration process. 267BDO Consulting, supra note 200, at 15 fig.1. The new class settlement operates on a claims-made basis but is expected to pay out $7.8 billion to those individuals who chose not to file with the GCCF but now choose to file claims, working with Garden City Group and PricewaterhouseCoopers as subcontractors in the claims administration process. 268John Schwartz, Accord Reached Settling Lawsuit Over BP Oil Spill, N.Y. Times, Mar. 2, 2012, at A1. The programs are thus substantially similar on their face, offering full compensation for those who affirmatively file claims—although the aggregate settlement will preclude future litigation for those who missed the January 22, 2013 claims deadline, in contrast to the disaggregative system, which allowed nonparticipants to retain their claims.

At what cost was this legitimacy obtained? Despite BP’s demonstrated willingness to settle claims through the GCCF, the fees resulting from the aggregate proceeding were massive: the court approved an interim payment of $75 million to class counsel, with an overall cap of $600 million in fees, costs, and expenses. 269Cf. Plaintiffs’ Steering Comm.’s and BP Defendants’ Memorandum in Support of Joint Motion for (1) Preliminary Approval of Class Action Settlement, (2) Scheduling a Fairness Hearing, (3) Approving and Issuing Proposed Class Action Settlement Notice, and (4) BP’s Motion for Adjourning the Limitation of Liab. Trial at 7 n.3, In re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on Apr. 20, 2010, MDL No. 2179 (E.D. La. Apr. 18, 2012), 2012 WL 1322258. Moreover, as noted above, Judge Barbier sought to charge GCCF participants for the costs of this representation—despite objections that these disaggregative settlement participants received no benefit from the class action, as their settlement terms were those of the GCCF and not modified by the work of class counsel or the plaintiff’s steering committee in the MDL. 270In re Deepwater Horizon, 739 F.3d 790.

The lack of legitimacy and transparency is not just harming the parties or society, from the perspective of social justice. Rather, it is having trickle-down effects, creating dysfunction in the compensation paid in disaggregative systems. Yet aggregation’s delays and costs are so great that the over half a million claimants who filed claims with the GCCF and BP both were willing to bear dislocations in compensation, rather than litigate. 271See BDO Consulting, supra note 267, at 59–61. This suggests that both sides concluded that the risk of improper compensation was less than the benefit obtained from a disaggregative dispute resolution mechanism.

Against this backdrop, it seems there must be a better way to obtain the benefits of disaggregation without the legitimacy challenges of privatization nor the costs imposed by the current backstop of aggregation. The rest of this colloquy begins to address this question, exploring the potential for the use of alternative public mechanisms to allow parties to obtain the benefits of disaggregation, while ensuring that the public interest in deterrence and law enforcement is satisfied.

Conclusion

This Article documents the birth of an entirely new branch of mass claims resolution mechanisms—radically upending the traditional view that aggregation was the only way to resolve mass claims. Just as class actions were once in their infancy, so now is disaggregation. Disaggregation is an emerging trend that is gaining momentum, and for good reason—both claimants and defendants can receive better outcomes and far faster than in aggregate litigation. Yet, as this Article has detailed, purely privatized disaggregation is imperfect, creating a risk of suboptimal outcomes for not only the parties, but for our broader, normative goals.

This Article therefore offers a call to a new line of scholarship exploring the ways in which these new private ordering mechanisms are changing not only our conceptualization and use of aggregate procedure but also the enforcement of substantive law and the nature of the private right of action. In the wake of this transformation, what is the right approach? Should we move toward increasing use of quasi-administrative actions? 272See, e.g., Dana A. Remus & Adam S. Zimmerman, Aggregate Litigation Goes Private, 63 Emory L.J. 1317 (2014). Should we draft laws with increased qui tam rights, or create enhanced private penalties mirroring those of Shady Grove Orthopedic Associates v. Allstate Insurance Co.? 273559 U.S. 393 (2010). Or is it time to reform our private mass compensation schemes? 274See, e.g., Thomas, supra note 261.

As I indicated at the outset, my goal is not to answer these questions. Rather it is to identify this new branch of disaggregative mechanisms, beginning the conversation about how their existence has informed the viability of aggregation and, in turn, how the availability of aggregate mechanisms has shaped the adoption of disaggregative mechanisms. In beginning this discussion, I argue we must stop seeing aggregation as a mandatory mechanism and instead recognize the complex feedback loop operating between the default world of public aggregation mechanisms and the strategic choices it incentivizes as parties take their first steps into this new world of private disaggregative mechanisms.

Footnotes

Assistant Professor of Law, University of Georgia. I am indebted to Richard Freer for his role in making this colloquy possible. I would like to thank Dianne Amman, Beth Burch, Dan Coenen, Michael Coenen, Glenn Cohen, Jim Grenier, John Goldberg, Matt Hall, Bert Huang, Jeremy Newman, Matthew Stephenson, and Rory Van Loo for their encouragement and immensely helpful comments. I am also grateful to participants in colloquia at Emory University School of Law and Peking University’s School of Transnational Law, as well as the Harvard/Stanford/Yale Junior Faculty Forum for their generous feedback on earlier versions of this paper. Rory Weeks and Alan Rozenshtein provided assistance transcending my expectations.

1See David L. Shapiro, Class Actions: The Class as Party and Client, 73 Notre Dame L. Rev. 913, 913 (1998) (“Perhaps the most dramatic development in civil procedure in recent decades has been the growth of interest in the class action . . . .”); accord John C. Coffee, Jr., Class Wars: The Dilemma of the Mass Tort Class Action, 95 Colum. L. Rev. 1343, 1344–45, 1350 (1995); Arthur R. Miller, Of Frankenstein Monsters and Shining Knights: Myth, Reality, and the “Class Action Problem, 92 Harv. L. Rev. 664, 664 (1979).

2Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (noting that the emergence of aggregate litigation through Rule 23 was largely an attempt to vindicate “the rights of groups of people who individually would be without effective strength to bring their opponents into court at all” (quoting Benjamin Kaplan, A Prefatory Note, 10 B.C. Indus. & Com. L. Rev. 497, 497 (1969)) (internal quotation marks omitted)).

3In re “Agent Orange” Prod. Liab. Litig., 996 F.2d 1425 (2d Cir. 1993).

4Amchem Prods., 521 U.S. 591.

5347 U.S. 483 (1954).

6Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541 (2011).

7Air Transportation Safety and System Stabilization Act, 49 U.S.C. § 40101 (2006). For similar programs, see Smallpox Emergency Personnel Protection Act of 2003, 42 U.S.C. §§ 239–239h (2006) (Smallpox Compensation Program); Public Readiness and Emergency Preparedness Act, id. §§ 247d-6d to -6e (Countermeasures Injury Compensation Program); Radiation Exposure Compensation Act, id. § 2210 (Radiation Exposure Compensation Program); Public Safety Officers’ Benefits Act of 1976, id. §§ 3796–3796c-2 (Supp. 2008) (Public Safety Officers’ Benefits Program); Civil Liberties Act of 1988, 50 U.S.C. app. § 1989b-3 (2006).

8For the official website, see http://www.gulfcoastclaimsfacility.com. Although BP would ultimately reach a class action settlement approved in 2012, it resolved more than two-thirds of its claims—more than $7.8 billion—through a private claims resolution facility, as required by the Oil Pollution Act of 1990. See 33 U.S.C. § 2705(a) (2012). In the wake of the class action settlement, the GCCF website was closed, directing participants to the class settlement website—a transition discussed in more detail in Parts II and III of this Article.

9See, e.g., Nitro-Lift Techs., L.L.C. v. Howard, 133 S. Ct. 500 (2012) (remanding noncompete claims to arbitration); Rent-A-Center, W., Inc. v. Jackson, 561 U.S. 63 (2010) (enforcing a provision giving an arbitrator authority to determine all challenges to enforceability of an arbitration agreement between employee and employer, in the context of a race discrimination claim).

10See, e.g., CompuCredit Corp. v. Greenwood, 132 S. Ct. 665 (2012) (holding a dispute over $257 in credit card fees was properly subject to arbitration because the statute created a right to proceed in class action litigation but was silent as to whether claims could proceed in arbitration, and thus did not override the Federal Arbitration Act’s presumption of arbitrability); AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011) (holding a $30 dispute was properly subject to a single-plaintiff arbitration agreement).

11For discussion, see infra Part II.B.

12When federal civil actions pending in different districts have common questions of fact, multidistrict litigation (MDL) is utilized to combine and transfer the actions to any district for consolidated pretrial proceedings. 28 U.S.C. § 1407 (2012). The new trend within the scholarship has been toward the recognition of quasi–class action settlement structures through MDL as contrasted with traditional class actions; yet both mechanisms fall within the traditional conception of aggregation. See, e.g., Elizabeth Chamblee Burch, Financiers as Monitors in Aggregate Litigation, 87 N.Y.U. L. Rev. 1273, 1274–75 (2012); John C. Coffee, Jr., Class Action Accountability: Reconciling Exit, Voice, and Loyalty in Representative Litigation, 100 Colum. L. Rev. 370, 371 (2000); Troy A. McKenzie, Toward a Bankruptcy Model for Nonclass Aggregate Litigation, 87 N.Y.U. L. Rev. 960, 962–63 (2012); Charles Silver & Geoffrey P. Miller, The Quasi–Class Action Method of Managing Multi-District Litigations: Problems and a Proposal, 63 Vand. L. Rev. 107, 110–11 (2010).

13See, e.g., Tracy A. Thomas, Introduction to Symposium, Remedies for Big Disasters: The BP Gulf Oil Spill and the Quest for Complete Justice, 45 Akron L. Rev. 567, 570 (2012) (summarizing the conclusion of symposium participants that the GCCF should not serve as a template for other disasters or compensation programs); Adam S. Zimmerman, Funding Irrationality, 59 Duke L.J. 1105, 1114, 1118 (2010) (attributing the lack of scholarship on public funds to the conception that they are “sui generis” products of “special legislation”).For studies of particular tribunals, see Kenneth R. Feinberg, What Is Life Worth?: The Unprecedented Effort to Compensate the Victims of 9/11 (2005); Robert M. Ackerman, The September 11th Victim Compensation Fund: An Effective Administrative Response to National Tragedy, 10 Harv. Negot. L. Rev. 135 (2005); Robert L. Rabin, Indeterminate Future Harm in the Context of September 11, 88 Va. L. Rev. 1831 (2002); Byron G. Stier, The Gulf Coast Claims Facility as Quasi-Public Fund: Transparency and Independence in Claim Administrator Compensation, 30 Miss. C. L. Rev. 255 (2011).

14Leading scholars are now focused upon doctrinal trends within aggregation toward smaller classes and multidistrict litigation—what I suggest here is a far more fundamental shift in mass claims resolution. See, e.g., Elizabeth Chamblee Burch, Disaggregating, 90 Wash. U. L. Rev. 667 (2013) (discussing the trend toward smaller classes); Alexandra D. Lahav, The Case for “Trial By Formula, 90 Tex. L. Rev. 571 (2012).

15See Francis E. McGovern, The What and Why of Claims Resolution Facilities, 57 Stan. L. Rev. 1361, 1380–81 (2005) (“Except for the option of individual trials, all of the currently available litigation procedures for an endgame in disputes involving large numbers of claims contemplate a claims resolution facility. [Rule 23 federal class actions], state class actions, bankruptcy, multidistrict litigation, and mass settlements all reach closure with a claims resolution facility.”).

16See, e.g., Judith Resnik, Fairness in Numbers: A Comment on AT&T v. Concepcion, Wal-Mart v. Dukes, and Turner v. Rogers, 125 Harv. L. Rev. 78, 93 (2011) (using the term “disaggregation” to reference the Dukes Court’s insistence on smaller classes and Concepcion’s upholding of predispute arbitration provisions that effectively barred aggregation, noting that both opinions “insisted on disaggregation, devolution, and privatization”).

17See infra Part II.

18Shapiro, supra note 1, at 914.

19Id. at 914–16 & n.2 (querying whether any law review existed that had not published an article on class actions and providing an excellent index of the leading works within the field).

20For a traditional discussion of disaggregation as a mechanism for effectuating global class settlement, see Richard A. Nagareda, Autonomy, Peace, and Put Options in the Mass Tort Class Action, 115 Harv. L. Rev. 747, 751 (2002) (discussing the asbestos, silicone gel breast implant, and fen-phen class settlements, which he noted “all aspire to create some form of private administrative system that . . . promises more efficient compensation for plaintiffs, long-term peace for defendants, and a reduced litigation burden for the courts”).

21See Randy J. Kozel & David Rosenberg, Solving the Nuisance-Value Settlement Problem: Mandatory Summary Judgment, 90 Va. L. Rev. 1849, 1859–60 (2004).

22See infra notes 98–103 and accompanying text. See generally Richard A. Nagareda, Mass Torts in a World of Settlement 11–54 (2007) (discussing the dynamics of mass tort litigation and dysfunctions that arise in this context).

23See infra Part III.A. Effective in 1966, Rule 23 of the Federal Rules of Civil Procedure set the prerequisites for class certification, as well as the types of classes permitted and their respective requirements. 1 Alba Conte & Herbert B. Newberg, Newberg on Class Actions § 1:8, at 28–29 (4th ed. 2002). The accompanying advisory committee notes explained the rationale for this innovation, focusing largely on the risk of inconsistent judgments and the need for classwide relief in discrimination cases. Fed. R. Civ. P. 23 advisory committee’s note, reprinted in 39 F.R.D. 69, 98–102 (1966).But, today’s class action practice focuses far more on these cost asymmetries, which can prevent either the pursuit of a claim or the successful pursuit of a claim. See David Rosenberg, Mandatory-Litigation Class Action: The Only Option for Mass Tort Cases, 115 Harv. L. Rev. 831, 852–53 (2002). For an excellent introduction for the unfamiliar reader to the phases of class litigation and the settlement pressures and “dysfunctions” apparent at each phase of litigation, see generally Nagareda, supra note , at 11–54.

24See Burch, supra note 12, at 1274–81 (discussing agency problems in both class actions and MDL and the possibility of third-party financiers acting as monitors in the context of nonclass aggregative litigation to help manage principal–agent problems); Jonathan R. Macey & Geoffrey P. Miller, The Plaintiffs’ Attorney’s Role in Class Action and Derivative Litigation: Economic Analysis and Recommendations for Reform, 58 U. Chi. L. Rev. 1, 7–9 (1991) (discussing agency in the context of Rule 23 class litigation and the agency costs unique to aggregate litigation). See generally Samuel Issacharoff, Litigation Funding and the Problem of Agency Cost in Representative Actions, 63 DePaul L. Rev. 561 (2014) (discussing the agency costs inherent in representative actions and the potential impact of alternative litigation financing).

25See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 598 (1997) (noting that the transaction costs exceeded recovery in the asbestos cases by a two-to-one ratio).

26See id. at 597–98 (holding the class settlement proposing global resolution of current and future claims could not be certified consistent with due process, but noting that lack of certification meant that “future claimants may lose altogether” given the specter of exhaustion of assets).

27See infra Part II.B.

28See infra Part II.C.

29Cf. John H. Beisner et al., Class Action “Cops”: Public Servants or Private Entrepreneurs?, 57 Stan. L. Rev. 1441, 1444–57 (2005) (discussing distrust of class actions by the public and the impact upon the selection and settlement of lawsuits).

30See Nagareda, supra note 22, at 11–54.

31See, e.g., Margaret H. Lemos, Aggregate Litigation Goes Public: Representative Suits by State Attorneys General, 126 Harv. L. Rev. 486, 488 (2012) (“Few aspects of contemporary civil litigation have attracted as much scholarly attention as the damages class action. Commentators have criticized class actions as either too powerful or not powerful enough . . . .”(footnote omitted)); Richard A. Nagareda, Class Certification in the Age of Aggregate Proof, 84 N.Y.U. L. Rev. 97, 97 (2009) (“Few pretrial motions in our civil justice system elicit as much controversy as those for the certification of class actions.”); Sergio J. Campos, Class Actions All the Way Down, 113 Colum. L. Rev. Sidebar 20, 20 (2013), http://www.columbialawreview.org/wp-content/uploads/2013/02/20_Campos.pdf (“[T]he class action [is] arguably the most controversial procedure in civil litigation.”).

32See, e.g., Myriam Gilles, Opting Out of Liability: The Forthcoming, Near-Total Demise of the Modern Class Action, 104 Mich. L. Rev. 373, 373–79 (2005) (predicting with disappointment the end of the class action procedure as a result of “the demise of mass tort class actions and the rise of contractual class action waivers”); Linda S. Mullenix, Prometheus Unbound: The Gulf Coast Claims Facility as a Means for Resolving Mass Tort Claims—A Fund Too Far, 71 La. L. Rev. 819, 825 (2011) (criticizing one such “defendant-created and directed” disaggregative device as “a model essentially unconstrained by law”); Resnik, supra note 16, at 161–68 (critiquing recent Supreme Court decisions that subsequently imposed greater entry barriers to federal courts through disaggregation).

33See, e.g., Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2560–61 (2011); Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 619–22 (1997).

34See, e.g., Philip Morris USA v. Williams, 549 U.S. 346, 353–63 (2007); State Farm Mut. Auto. Ins. Co. v. Campbell, 538 U.S. 408, 416 (2003).

35See, e.g., Burch, supra note 14, at 667; Resnik supra note 16, at 93.

36See, e.g., AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1746–53 (2011); Stolt-Nielsen S.A. v. AnimalFeeds Int’l Corp., 559 U.S. 662, 681–89 (2010).

37See, e.g., Winzler v. Toyota Motor Sales U.S.A., Inc., 681 F.3d 1208, 1210–15 (10th Cir. 2012); In re Aqua Dots Prods. Liab. Litig., 654 F.3d 748, 751–53 (7th Cir. 2011).

38See, e.g., In re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on April 20, 2010, 295 F.R.D. 112, 161 (E.D. La. 2013) (certifying class and approving class settlement).

39See Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2550–52, 2557 (2011) (reversing a decision to certify a class for a lack of commonality among the 1.5 million members, and finding their claims for backpay to have been improperly certified under Rule 23(b)(2)).

40See infra Part III.A.

41For an excellent discussion of the normative consequences of a privately ordered shift toward minimal process in the single-plaintiff context, see Nora Freeman Engstrom, Sunlight and Settlement Mills, 86 N.Y.U. L. Rev. 805 (2011).

42See, e.g., Jean R. Sternlight, Creeping Mandatory Arbitration: Is It Just?, 57 Stan. L. Rev. 1631, 1631–35 (2005).

43See, e.g., Albert Choi & George Triantis, Strategic Vagueness in Contract Design: The Case of Corporate Acquisitions, 119 Yale L.J. 848, 922–23 (2010); Howard M. Erichson, The Problem of Settlement Class Actions, 82 Geo. Wash. L. Rev. (forthcoming 2014) (manuscript at 5, 13–15), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2243155; Deborah E. Greenspan & Matthew A. Neuburger, Settle or Sue? The Use and Structure of Alternative Compensation Programs in the Mass Claims Context, 17 Roger Williams U. L. Rev. 97, 129–36 (2012); Keith N. Hylton, Agreements to Waive or to Arbitrate Legal Claims: An Economic Analysis, 8 Sup. Ct. Econ. Rev. 209, 248 (2000).

44See, e.g., Cynthia L. Estlund, Between Rights and Contract: Arbitration Agreements and Non-compete Covenants as a Hybrid Form of Employment Law, 155 U. Pa. L. Rev. 379, 379–80, 383 (2006); Mullenix, supra note 32, at 825; Resnik, supra note 16, at 161–68; Adriaan Lanni, Note, Protecting Public Rights in Private Arbitration, 107 Yale L.J. 1157, 1161 (1998).

45See, e.g., Myriam Gilles & Gary Friedman, After Class: Aggregate Litigation in the Wake of AT&T Mobility v. Concepcion, 79 U. Chi. L. Rev. 623 (2012); David Horton, Mass Arbitration and Democratic Legitimacy, 84 U. Colo. L. Rev. (forthcoming 2014) (reviewing Margaret Jane Radin, Boilerplate: The Fine Print, Vanishing Rights, and the Rule of Law (2013)), available at http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2235901.

46See infra Part III.

47See Nagareda, supra note 20, at 751, 754, 770, 821 (“[T]he existence of a credible mass tort class settlement removes the possibility of a deterrence gap by virtually eliminating the prospect that too few compensation claims will be filed.” (internal quotation marks omitted)). The nature of these claims can vary widely as the problem of vindication is quintessentially one of the ratio of damages to litigation costs, such that even high-value claims can fall within this category. See id. at 750. Mass tort claims are an exemplar, as the alleged damages may include substantial claims for personal injury or wrongful death, but establishing causation may require the costly development of epidemiological evidence, expert testimony, and voluminous document review. See id. at 750, 827.

48Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997) (quoting Kaplan, supra note 2, at 497).

49See Rosenberg, supra note 23, at 852.

50See Judith Resnik et al., Individuals Within the Aggregate: Relationships, Representation, and Fees, 71 N.Y.U. L. Rev. 296, 308, 382 (1996).

51See Jack B. Weinstein, Individual Justice in Mass Tort Litigation 135–36 (1995) (describing the efficiency benefits of aggregation).

52See Wal-Mart Stores, Inc. v. Dukes, 131 S. Ct. 2541, 2560–61 (2011) (holding that the defendant has a due process right to present individualized defenses in a class action proceeding).

53See McGovern, supra note 15, at 1380–81.

54See In re Rhone-Poulenc Rorer Inc., 51 F.3d 1293, 1299 (7th Cir. 1995) (Posner, C.J.) (describing certification as “forcing these defendants to stake their companies on the outcome of a single jury trial” or “to settle even if they have no legal liability” out of “fear of the risk of bankruptcy”). But see Stephen B. Burbank & Tobias Barrington Wolff, Redeeming the Missed Opportunities of Shady Grove, 159 U. Pa. L. Rev. 17, 59–61 (2010) (“What is striking . . . is Judge Posner’s lack of attention in the first part of his analysis to the policy differences that States might have regarding the ‘mature tort’ problem and the relative merits of decentralized adjudication, which offers the benefit of accreted wisdom over time but may produce results that lack uniformity and appear arbitrary, versus a high-stakes industry-wide trial, which creates greater risks of inaccurate or unreliable results but also provides greater parity and fairness among claimants.”).

55See Adam S. Zimmerman, Distributing Justice, 86 N.Y.U. L. Rev. 500, 507–12 (2011).

56See, e.g., Richard A. Nagareda, Turning from Tort to Administration, 94 Mich. L. Rev. 899, 921 (1996).

57See McGovern, supra note 15, at 1365.

58See Deborah R. Hensler, Alternative Courts? Litigation-Induced Claims Resolution Facilities, 57 Stan. L. Rev. 1429, 1429 (2005) (noting that claims resolution facilities have “largely escaped the scrutiny of legal scholars”).

59See Zimmerman, supra note 55, at 511–12.

60See Francis E. McGovern, Distribution of Funds in Class Actions—Claims Administration, 35 J. Corp. L. 123, 126–33 (2009) (describing various settlement structures as case studies, without creating a typology).

61Cf. McGovern, supra note 15, at 1376–77 (recognizing the tensions inherent in requiring opt-out before awards are calculated).

62See Judges’ Class Action Notice and Claims Process Checklist and Plain Language Guide, Fed. Jud. Ctr. 6 (2010), http://www.fjc.gov/public/pdf.nsf/lookup/NotCheck.pdf/$file/NotCheck.pdf (treating claim forms as a common practice).

63See Tiffany Allen, Anticipating Claims Filing Rates in Class Action Settlements, Class Action Persp. (Rust Consulting, Inc., Minneapolis, Minn.), Nov. 2008, available at http://www.rustconsulting.com/Portals/0/pdf/Monograph_ClaimsFilingRates.pdf.

64See Judges’ Class Action Notice and Claims Process Checklist and Plain Language Guide, supra note 62, at 6.

65See, e.g., Am. Law Inst., Principles of the Law: Aggregate Litigation § 3.07 (2010) (recommending judges limit such payments “to circumstances in which direct distribution to individual class members is not economically feasible, or where funds remain after class members are given a full opportunity to make a claim” (internal quotation marks omitted)).This observation does not account for the practice of settling class claims without even attempting to make distribution to the class—a phenomenon that has troubled class action scholars who view the practice as an abandonment of one’s duty to the class in favor of a charitable group comprised of individuals who are not clients (but are instead, for example, political allies). Yet, lawyers have argued that in small-value cases the transaction costs of payment can exceed damages. See Lemos, supra note 31, at 528–29 (describing objections to cy pres distributions and documenting examples of politically directed cy pres recoveries).

66See Fed. R. Civ. P. 23(b)(3), (e)(4).

67McGovern, supra note 60, at 127.

68Id. at 126–27.

69Five to twenty-five percent of checks issued in class actions remain uncashed. See Class Notice & Settlement Administration: Avoiding the Pitfalls—Part I, Class Action Persp. (Rust Consulting, Inc., Minneapolis, Minn.), Mar. 2007, available at http://www.rustconsulting.com/Portals/0/pdf/Monograph_Pitfalls_I.pdf.

70With improved technology, some funds have been able to make automatic payments to all class members. However, these payment structures require substantial information about claimants, which is frequently unavailable. See, e.g., Zimmerman, supra note 13, at 1167.

71A cy pres distribution occurs when class damages cannot be feasibly distributed or when a balance remains; under this system, the funds are to be distributed to the next best compensation use, typically a charitable donation. See 4 Conte & Newberg, supra note 23, § 11:20. For an excellent discussion of the cy pres system, see William B. Rubenstein, On What a “Private Attorney General” Is—And Why It Matters, 57 Vand. L. Rev. 2129, 2159–71 (2004).

72Francis E. McGovern, A Model State Mass Tort Settlement Statute, 80 Tul. L. Rev. 1809, 1811–12 (2006).

73See John C. Coffee, Jr., Understanding the Plaintiff’s Attorney: The Implications of Economic Theory for Private Enforcement of Law Through Class and Derivative Actions, 86 Colum. L. Rev. 669, 677–84 (1986) (describing fee structures and incentives for class counsel).

74For varied analysis, see Symposium, After Disaster: The September 11th Compensation Fund and the Future of Civil Justice, 53 DePaul L. Rev. 205 (2003).

75See McGovern, supra note 15, at 1363–64 (discussing the September 11th Compensation Fund and silicone gel breast implant settlements as exemplars of funds that do not establish overall fund limits but instead “the total amount of damages would be the sum of the subsequently determined individual amounts”).

76Settlement Facility and Fund Distribution Agreement Between Dow Corning Corporation and the Claimants’ Advisory Committee at 2–3, In re Dow Corning Corp., 244 B.R. 705 (Bankr. E.D. Mich. 1999) (No. 95-20512).

77See McGovern, supra note 60, at 1389.

78For an excellent discussion of undercapitalization and the resulting need for a second opt-out opportunity, see Laurens Walker & John Monahan, Scientific Authority: The Breast Implant Litigation and Beyond, 86 Va. L. Rev. 801, 803–07 (2000).

79Anthony J. Sebok, What Do We Talk About When We Talk About Mass Torts?, 106 Mich. L. Rev. 1213, 1213 (2008) (reviewing Nagareda, supra note 22).

80See Agreed Order Allowing Certain Late Claimants Limited Rights to Participate in the Plan’s Settlement Facility at 3, In re Dow Corning Corp., No. CV-00005-DPH (E.D. Mich. Dec. 12, 2007), available at http://www.mied.uscourts.gov/Information/Dow/pdf/orderpdf/agreed%20Order.pdf.

81See William B. Rubenstein, The Fairness Hearing: Adversarial and Regulatory Approaches, 53 UCLA L. Rev. 1435, 1459 (2006) (concluding that “the track record of professional objectors to date . . . has been less than stellar”).

82Walker & Monahan, supra note 78, at 803–07.

83See Hensler, supra note 58, at 1429, 1439 (lamenting the lack of scholarship on claims administration).

84See Nagareda, supra note 56, at 914 (“From the defendants’ standpoint . . . settlement is attractive only to the extent that it does not merely resolve pending cases but, in addition, provides assurance that such action will not result in a deluge of marginal cases.”).

85See Charles Silver & Lynn Baker, I Cut, You Choose: The Role of Plaintiffs’ Counsel in Allocating Settlement Proceeds, 84 Va. L. Rev. 1465, 1527 (1998).

86For a discussion of the agency dynamics that drive class settlement allocations between class members, see generally Samuel Issacharoff et al., Bargaining Impediments and Settlement Behavior, in Dispute Resolution: Bridging the Settlement Gap (David A. Anderson ed., 1996); Marcel Kahan & Linda Silberman, The Inadequate Search for “Adequacy” in Class Actions: A Critique of Epstein v. MCA, Inc., 73 N.Y.U. L. Rev. 765 (1998); Silver & Baker, supra note 85.

87See Rubenstein, supra note 81, at 1445.

88See generally Richard A. Nagareda, Administering Adequacy in Class Representation, 82 Tex. L. Rev. 287, 370–71 (2003) (describing the lack of meaningful checks imposed by objections upon disloyal class counsel).

89See John C. Coffee, Jr., Litigation Governance: Taking Accountability Seriously, 110 Colum. L. Rev. 288, 304–05 (2010).

90See David Rosenberg, Decoupling Deterrence and Compensation Functions in Mass Tort Class Actions for Future Loss, 88 Va. L. Rev. 1871, 1871–79 (2002).

91For every claim, there is an optimal point of both compensation and deterrence associated with the value of the claim. Determining this optimal point within the law is, in the first instance, a matter for the legislature in crafting the available remedies, but then requires substantial individualized discovery and assessment to determine the precise value of any particular legal right in a particular case. In addition, scholars will often argue that particular legal frameworks are not obtaining the optimal levels of compensation or deterrence given particular normative goals.This Article argues that there are a number of factors that can shift settlement values from these optimal points, wherever set. The term “dislocation” is used throughout my analysis to show that these shifts are bilateral, in contrast with the traditional terminology that often suggests a merely one-directional shift from this optimal value point.

92The use of disaggregation, even in this traditional setting, has “largely escaped the scrutiny of legal scholars.” Hensler, supra note 58, at 1429, 1439 (“It is time to get serious about understanding the role claims resolution facilities are playing in the delivery of civil justice in the United States.”). But see Georgene Vairo, Why Me? The Role of Private Trustees in Complex Claims Resolution, 57 Stan. L. Rev. 1391, 1392–93 (2005) (advocating for the use of administrative trusts run by private persons).

93See supra note 8 and accompanying text.

94For an excellent discussion of agency in Rule 23 classes and MDL, see Macey & Miller, supra note 24, at 7–8.

95See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 617 (1997).

96See Howard M. Erichson & Benjamin C. Zipursky, Consent Versus Closure, 96 Cornell L. Rev. 265, 270–74 (2011).

97See Amchem, 521 U.S. at 598. The Court expressly invited congressional intervention, recognizing the inability of aggregation mechanisms to solve the problem posed by the massive liability created by asbestos use. Id. at 598–99; Ortiz v. Fibreboard Corp., 527 U.S. 815, 865 (1999) (Rehnquist, C.J., concurring). Yet, in practice, this solution too has proven untenable. Georgene Vairo, Mass Tort Bankruptcies: The Who, The Why and The How, 78 Am. Bankr. L.J. 93, 95 (2004) (“The Supreme Court’s decision in Amchem and its later decision in Ortiz have made it more difficult for companies seeking global peace in resolving a mass tort to use Rule 23 settlement classes. Moreover, Congress appears to have failed in its attempt to resolve legislatively the protracted asbestos litigation.” (footnotes omitted)).

98See, e.g., Kozel & Rosenberg, supra note 21, at 1879–90; James J. Park, Rules, Principles, and the Competition to Enforce the Securities Laws, 100 Calif. L. Rev. 115, 161 (2012) (discussing the incentive of entrepreneurial enforcers to bring low probability claims).

99See, e.g., Bruce Hay & David Rosenberg, “Sweetheart” and “Blackmail” Settlements in Class Actions: Reality and Remedy, 75 Notre Dame L. Rev. 1377, 1378 (2000).

100See, e.g., Richard A. Nagareda, The Preexistence Principle and the Structure of the Class Action, 103 Colum. L. Rev. 149, 163 (2003).

101See, e.g., Jonathan T. Molot, Litigation Finance: A Market Solution to a Procedural Problem, 99 Geo. L.J. 65 (2010).

102See, e.g., Nagareda, supra note 22, at 11–54.

103Coffee, supra note 1, at 1346.

104See infra Part II.A.2.

105See infra Part II.A.1.

106See Nagareda, supra note 22, at 11–54.

107See, e.g., Feinberg, supra note 13 (arguing that private defendants would not have an incentive to enter into early, postdispute settlement regimes).

108See, e.g., Edward F. Sherman, The BP Oil Spill Litigation and Evolving Supervision of Multidistrict Litigation Judges, 30 Miss C. L. Rev. 237, 243 (2011).

109See Paul Elias, Thousands File Claims After Chevron Refinery Fire, AP (Aug. 11, 2012, 4:39 AM), http://bigstory.ap.org/article/thousands-file-claims-after-chevron-refinery-fire.

110See id. Chevron’s fund ultimately ballooned to 23,900 claims filed after 15,000 people sought medical treatment, yielding a total payout of approximately $10 million through its claims settlement fund. See Chevron Has Paid Out $10 Million for Refinery Fire, ABC7 News (Jan. 30, 2013), http://abclocal.go.com/kgo/story?section=news/local/east_bay&id=8973226.

111See, e.g., Mullenix, supra note 32, at 825 (reviewing challenges to the legitimacy of the September 11th Fund and arguing that “[f]or those concerned with the rule of law, equity, and fundamental fairness, the GCCF [created by BP] ought to be a cause for concern”).

112Cf. Albert Choi & George Triantis, Completing Contracts in the Shadow of Costly Verification, 37 J. Legal Stud. 503, 504–09, 523–25 (2008) (characterizing contractual predispute substantive provisions as value-adding and clarifying underlying obligations).

113This preference dates from the earliest days of our nation. See Michele L. Landis, “Let Me Next Time Be ‘Tried By Fire’”: Disaster Relief and the Origins of the American Welfare State 1789–1874, 92 Nw. U. L. Rev. 967, 983 (1998) (describing the transition from “individual to group eligibility criteria in relief legislation” as based in the growth of the administrative state). For recent examples, see statutes listed supra note 7.

114See, e.g., United States v. Mitchell, 463 U.S. 206, 212–13 (1983) (describing the origins of sovereign immunity waivers); cf. Gregory C. Sisk, The Tapestry Unravels: Statutory Waivers of Sovereign Immunity and Money Claims Against the United States, 71 Geo. Wash. L. Rev. 602, 611–15 (2003) (noting that Congress often waives sovereign immunity conditioned upon the plaintiff’s consent to participate in a non–Article III system of adjudication as to the claim).

115For an excellent discussion of the constitutionality of these shifts, see Richard H. Fallon, Jr., Jurisdiction-Stripping Reconsidered, 96 Va. L. Rev. 1043, 1050–51 (2010); and Caleb Nelson, Adjudication in the Political Branches, 107 Colum. L. Rev. 559, 583–84 (2007).

116See McGovern, supra note 60, at 126–33.

117See Jamal Greene, The Anticanon, 125 Harv. L. Rev. 379, 400 (2011) (discussing reparations in the context of Korematsu v. United States, 323 U.S. 214 (1944)).

118See Feinberg supra note 13, at 177–79.

119See Elizabeth M. Schneider, Grief, Procedure, and Justice: The September 11th Victim Compensation Fund, 53 DePaul L. Rev. 457, 477 (2003).

120See Gaia Pianigiani, Italy: Cruise Line Offers Cash to Shipwreck Passengers, N.Y. Times, Jan. 28, 2012, at A11.

121See Jeremy Stahl, Gameday Theory, Slate (Feb. 11, 2011), http://www.slate.com/articles/sports/sports_nut/2011/02/gameday_theory.html.

122See id.

123See id.

124See id.

125Id.

126Id.

127See id.

128See Agreement Between the Government of the United States and the Government of the Marshall Islands for the Implementation of Section 177 of the Compact of Free Association, arts. X, XII, available at http://www.nuclearclaimstribunal.com/177text.htm.

129See September 11th Victim Compensation Fund of 2001, 28 C.F.R. § 104 (2002).

130See Kenneth R. Feinberg, Unconventional Responses to Unique Catastrophes, 45 Akron L. Rev. 575, 578 (2012).

131Initially, BP created a claims facility as required by the Oil Pollution Act; however, two months later this was replaced with the GCCF pursuant to BP’s negotiations with President Obama. For discussion, see Mullenix, supra note 32, at 833–35.

132See id. at 834 & n.69.

133See McGovern, supra note 15, at 1380.

134See id.

135See Mullenix, supra note 32, at 819–21.

136See supra Part I.

137For example, in some recent settlements, monies have been directly credited to absent class members’ accounts, yielding substantial cost savings that allow payments to be made to victims suffering far smaller wrongs than was traditionally possible. See, e.g., Settlement Agreement and Release ¶¶ 81, 87, In re Checking Account Overdraft Litigation, No. MDL 2036 (S.D. Fla. May 6, 2011), available at http://www.bofaoverdraftsettlement.com/LinkClick.aspx?fileticket=rxwM1SXyxBc%3d&tabid=67&mid=415 (stating that the settlement agreement provided that awards for improper overdraft fees would be paid by crediting each account holder’s account, only allowing a physical check to be mailed if the credit is unfeasible or unreasonable; the settlement also stipulated a cy pres distribution for the residual).

138See Mullenix, supra note 32, at 834.

139See id. at 834 n.69, 835–36.

140Press Release, Dep’t of Justice, Audit of Gulf Coast Claims Facility Results in $64 Million in Additional Payments (Apr. 19, 2012), http://www.justice.gov/opa/pr/2012/April/12-asg-500.html (describing fund as “a significant advance in disaster response”).

141See statutes cited supra note 7; see also Peter H. Meyers, Fixing the Flaws in the Federal Vaccine Injury Compensation Program, 63 Admin. L. Rev. 785, 816–37 (2011) (analyzing the structural features of public workers’ compensation as well as general public compensation programs).

142See, e.g., Aaron-Andrew P. Bruhl, The Unconscionability Game: Strategic Judging and the Evolution of Federal Arbitration Law, 83 N.Y.U. L. Rev. 1420, 1457 (2008); Gilles, supra note 32, at 373; Jean R. Sternlight, As Mandatory Binding Arbitration Meets the Class Action, Will the Class Action Survive?, 42 Wm. & Mary L. Rev. 1, 5 (2000).

143131 S. Ct. 1740 (2011).

144See id. at 1744.

145See id.

146See id.

147See id. at 1748–49.

148See id. at 1745.

149See id. at 1752–53.

150See id. at 1749.

151Common provisions include corporate agreements to pay the costs of arbitration, reasonable attorneys’ fees, or an incentive payment substantially in excess of the alleged damages. See id. at 1744.

152See id. at 1752–53.

153See Samuel Issacharoff & Robert H. Klonoff, The Public Value of Settlement 78 Fordham L. Rev. 1177, 1182–83 (2009) (providing examples of recent settlements).

154See, e.g., Samuel Issacharoff & D. Theodore Rave, The BP Oil Spill Settlement and the Paradox of Public Litigation, 74 La. L. Rev. 397 (2014).

155See Hensler, supra note 58, at 1443 (referring to these as “a dizzying array of potential claims resolution facility designs”).

156Shapiro, supra note 1, at 914–16 & n.2 (querying whether any law review existed that had not published an article on class actions).

157See Nagareda, supra note 20, at 750–51, 754.

158See, e.g., Issacharoff, supra note 24; Issacharoff & Rave, supra note 154.

159See AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1752 (2011) (“Informal [arbitration] procedures do of course have a cost: The absence of multilayered review makes it more likely that errors will go uncorrected. Defendants are willing to accept the costs of these errors in arbitration, since their impact is limited to the size of individual disputes, and presumably outweighed by savings from avoiding the courts.”).

160See id. (“Faced with even a small chance of a devastating loss, defendants will be pressured into settling questionable claims.”).

161See supra notes 143–52.

162For a discussion of current doctrinal trends in lower courts, see Myriam Gilles, Killing Them with Kindness: Examining “Consumer-Friendly” Arbitration Clauses After AT&T Mobility v. Concepcion, 88 Notre Dame L. Rev. 825 (2012).

163See Jaime Dodge, The Limits of Procedural Private Ordering, 97 Va. L. Rev. 723, 755–64 (2011).

164Id. at 758–59.

165See id. at 759.

166See id.

167See id. at 757. For a discussion of the frequency of these form contracts, see Russell Korobkin, Bounded Rationality, Standard Form Contracts, and Unconscionability, 70 U. Chi. L. Rev. 1203, 1203 (2003) (noting that “nearly all commercial and consumer sales contracts are form driven”); and W. David Slawson, Standard Form Contracts and Democratic Control of Lawmaking Power, 84 Harv. L. Rev. 529, 529 (1971) (estimating that more than ninety-nine percent of contracts are based upon form terms).

168See Jean R. Sternlight & Elizabeth J. Jensen, Using Arbitration to Eliminate Consumer Class Actions: Efficient Business Practice or Unconscionable Abuse?, Law & Contemp. Probs., Winter/Spring 2004, at 75, 96–99.

169See, e.g., Developments in the Law—Access to Courts, 122 Harv. L. Rev. 1151, 1174–75 (2009) (detailing concern with repeat-player bias).

170See Dodge, supra note 163, at 742 (discussing forms of institutional capture in arbitration, litigation, and the crafting of substantive law); Gilles, supra note 32, at 410–12 (discussing repeat-player leverage in the context of institutions’ class arbitration policies).

171See Resnik, supra note 16, at 109–11 (discussing the delay in identifying the National Arbitration Forum as a potentially biased forum, in which repeat players “won virtually all the cases”).

172See Miles B. Farmer, Note, Mandatory and Fair? A Better System of Mandatory Arbitration, 121 Yale L.J. 2346, 2356–57 (2012) (discussing systemic biases in favor of repeat players and resulting impacts upon substantive outcomes).

173See supra Part II.A.

174While a robust literature discusses this market failure, the best evidence of how widespread this phenomenon is may well be its appearance at the center of a South Park episode’s plotline. Compare Dodge, supra note 163, at 755–64, 767 (reviewing literature supporting a conception of “ex ante exceptionalism” with respect to market failure for procedural terms in form contracts), with South Park: HUMANCENTiPAD (Comedy Central television broadcast Apr. 27, 2011), available at http://www.southparkstudios.com/full-episodes/s15e01-humancentipad (focusing upon the unintended consequence of Kyle Broflovski’s agreement to Apple’s terms and conditions without first reading the terms).

175See Dodge, supra note 163, at 760–61.

176See, e.g., Hay & Rosenberg, supra note 99, at 1378.

177See AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1761 (2011) (Breyer, J., dissenting).

178See Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 594 (2012) (granting certiorari to review the Second Circuit’s invalidation of a class arbitration waiver in the context of federal antitrust allegations).

179See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 598 (1997) (“‘The most objectionable aspects of asbestos litigation can be briefly summarized: dockets in both federal and state courts continue to grow; long delays are routine; trials are too long; the same issues are litigated over and over; transaction costs exceed the victims’ recovery by nearly two to one; exhaustion of assets threatens and distorts the process; and future claimants may lose altogether.’” (quoting Report of the Judicial Conference Ad Hoc Committee on Asbestos Litigation 2–3 (1991))); see also David Rosenberg & Steven Shavell, A Simple Proposal to Halve Litigation Costs, 91 Va. L. Rev. 1721, 1727 (2005) (noting that, even in settlement, “on average, it costs approximately one dollar in legal expenses for the legal system to transfer one dollar from a defendant to a plaintiff”).

180See Stahl, supra note 121.

181See Pianigiani, supra note 120.

182See Barbie Latza Nadeau, Why Survivors Aren’t Buying Costa Concordia’s Compensation Offer, Daily Beast (Jan. 27, 2012), http://www.thedailybeast.com/articles/2012/01/27/why-survivors-aren-t-buying-costa-concordia-s-compensation-offer.html.

183See Gary Mihoces, Judge: Cowboys Off the Hook in Super Bowl XLV Seat Fiasco, USA Today (July 19, 2012, 9:35 PM), http://usatoday30.usatoday.com/sports/football/nfl/story/2012-07-19/dallas-super-bowl-lawsuit/56347366/1 (noting that most fans accepted the NFL’s settlement offer rather than participate in the class action litigation).

184See Nadeau, supra note 182 (describing survivors’ reasons for hiring counsel for a class action lawsuit rather than accept the settlement offer).

185See Stahl, supra note 121.

186See Nadeau, supra note 182.

187See id. (“These people have no idea what their agonizing experience is worth yet. It could take years to understand the full impact of the experience.” (internal quotation marks omitted)).

188See Stahl, supra note 121.

189See Costa Concordia Survivors Describe “Goliath” Fight Against Cruise Industry, NBC News (July 12, 2012, 8:22 AM), http://rockcenter.nbcnews.com/_news/2012/07/12/12701697-costa-concordia-survivors-describe-goliath-fight-against-cruise-industry?lite.

190See Nadeau, supra note 182 (“This compensation package is higher than the current indemnification limits provided for in international conventions and laws currently in force . . . .” (internal quotation marks omitted)).

191See id.

192See Costa Concordia Survivors Describe “Goliath” Fight Against Cruise Industry, supra note 189.

193In high-profile cases, plaintiffs’ counsel may publicly urge plaintiffs not to take the settlement but instead to retain their services. This dynamic is likely to have an asymmetric bias, as only counsel claiming the settlement is unfair will have an incentive to come forward.

194See Nadeau, supra note 182.

195See Russell Korobkin, Aspirations and Settlement, 88 Cornell L. Rev. 1, 5–9 (2002); cf. Dodge, supra note 163, at 778. See generally Issacharoff et al., supra note 86.

196See generally Roger Fisher et al., Getting to Yes (3d ed. 2011) (discussing negotiation theory); Korobkin, supra note 195, at 5–9 (discussing “reservation prices” in settlement negotiation).

197See, e.g., September 11th Victim Compensation Fund of 2001, 67 Fed. Reg. 11,233, 11,237 (Mar. 13, 2002) (codified at 28 C.F.R. pt. 104) (“[U]nder the tort system, while many claimants receive extremely large awards, many others walk away empty-handed due to the requirement that plaintiffs prove fault. In contrast, the Fund is a no-fault alternative to civil litigation designed to provide fair compensation in a matter of months.”); Gulf Coast Claims Facility: Frequently Asked Questions, Gulf Coast Claims Facility (Aug. 11, 2010), http://www.restorethegulf.gov/sites/default/files/imported_pdfs/library/assets/gccf-faqs.pdf (detailing that any individuals or businesses in regions affected by the oil spill could file claims with or without an attorney and describing the minimal documentation required as proof).

198See notes 131–35 and accompanying text. But see Ackerman, supra note 13, at 161–65; Mullenix, supra note 32, at 854–59.

199See Stahl, supra note 121.

200See BDO Consulting, BDO USA, Independent Evaluation of the Gulf Coast Claims Facility: Report of Findings and Observations to the U.S. Department of Justice 10, 68, 76 (2012), available at http://www.justice.gov/opa/documents/gccf-rpt-find-obs.pdf (DOJ-commissioned report); Ackerman, supra note 13, at 161–65; Mullenix, supra note 32, at 854–59; David F. Partlett & Russell L. Weaver, BP Oil Spill: Compensation, Agency Costs, and Restitution, 68 Wash. & Lee L. Rev. 1341, 1359–61 (2011).

201See Robert L. Rabin, Tort System on Trial: The Burden of Mass Toxics Litigation, 98 Yale L.J. 813, 820–21 (1989) (collecting studies).

202See Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 598 (1997) (quoting Report of the Judicial Conference Ad Hoc Committee on Asbestos Litigation 2–3 (Mar. 1991)) (noting that the transaction costs exceeded the recovery in the asbestos cases by a 2-to-1 ratio).

203See Mullenix, supra note 32, at 819–21.

204See Resnik, supra note 16, at 108 (“The constitutional obligations of ‘open courts’ have produced a wealth of data on judges’ salaries, court budgets, case proceedings, and outcomes. In contrast, private dispute resolvers are left to do as they wish, subject only in a few jurisdictions, such as California, to requirements that arbitration providers ‘collect, publish . . . , and make available to the public’ information about parties, categories of disputes, time to disposition, and outcomes.”); Sternlight, supra note 42, at 1662–64; Roger J. Perlstadt, Comment, Timing of Institutional Bias Challenges to Arbitration, 69 U. Chi. L. Rev. 1983, 1985–87 (2002).

205See George Padis, Note, Arbitration Under Siege: Reforming Consumer and Employment Arbitration and Class Actions, 91 Tex. L. Rev. 665, 681 n.102 (2013).

206Mullenix, supra note 32, at 825.

207See Press Release, Dep’t of Justice, supra note 140 (noting total fund payout of over $6.2 billion to more than 220,000 claimants).

208Mullenix, supra note 32, at 843–44 & n.115, 872, 888.

209See supra Part I.

210See supra notes 40–42 and accompanying text.

211See supra note 89–90 and accompanying text.

212Cf. AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1751–53 (2011).

213Id. at 1761 (Breyer, J., dissenting).

214Id. (“The realistic alternative to a class action is not 17 million individual suits, but zero individual suits, as only a lunatic or a fanatic sues for $30.” (quoting Carnegie v. Household Int’l, Inc., 376 F.3d 656, 661 (7th Cir. 2004))).

215See Six (6) Mexican Workers v. Ariz. Citrus Growers, 904 F.2d 1301, 1305 (9th Cir. 1990) (“When a class action involves a large number of class members but only a small individual recovery, the cost of separately proving and distributing each class member’s damages may so outweigh the potential recovery that the class action becomes unfeasible.”); Notice of Class Action, Proposed Settlement, and Hearing at 4, Settlement Recovery Ctr. v. ValueClick, Inc., No. CV-02638-FMC-(CTx) (C.D. Cal. May 29, 2008), available at http://www.affiliatefairplay.com/service/images/post/agreement_of_settlement.pdf (“Recognizing that administration costs for low-value payments can greatly exceed the value of these payments, no payment will be made if the Settlement Class Member’s pro rata share is calculated to be less than $1.00.”).

216See Lemos, supra note 31, at 529.

217Some may even posit that nonfiling individuals are better off. To the extent they also would not have filed a claim form or would have received no compensation because the harm was so small as to obtain only cy pres relief, they receive no less compensation and may benefit from better terms as the company can now deploy the savings to better pricing or substantive product features. See Dodge, supra note 163, at 755–64, for a discussion of both sides of the market functioning argument.

218See, e.g., Pianigiani, supra note 120 (quoting a representative of a consumer group who described the settlement offered by Costa Cruises as “ridiculous and disproportionate for the damage suffered” (internal quotation mark omitted)).

219See Richard A. Nagareda, The Law of Class Actions and Other Aggregate Litigation 129–31 (2009) (discussing the original Agent Orange class action settlement, the court’s decision to grant a collateral attack on the settlement resulting in a new round of payouts by the defendant, as undermining the degree of closure received by defendants and thus reducing the amount rational defendants should be willing to pay in future settlements as a result of this shift in the risk premium).

220See, e.g., AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1744 (2011).

221See, e.g., supra note 8 and accompanying text; Nathan Bomey, Analyst: GM May Offer $2B-plus to Settle Recall Claims, USA Today (Mar. 26, 2014, 6:40 PM), http://www.usatoday.com/story/money/cars/2014/03/26/gm-recall-claims-settlement-prediction/6930339/.

222See generally Nancy H. Rogers et al., Designing Systems and Processes for Managing Disputes (2013) (explaining, through case studies and other examples, how to design and implement an effective alternative dispute resolution system).

223See Dodge, supra note 163, at 744.

224See, e.g., Robert E. Scott & George G. Triantis, Anticipating Litigation in Contract Design, 115 Yale L.J. 814, 856–79 (2006) (exploring the impact of contract design, and in particular the selection of the procedural rules governing a contract, on the economics of litigation).

225See, e.g., Fisher et al., supra note 196 (introducing the idea of principled negotiation on the merits that looks for mutual gain based on fair standards); Issacharoff & Klonoff, supra note 153, at 1179–84 (emphasizing the importance of the development of settlements in class actions and private aggregations of cases in the past twenty-five years); Carrie Menkel-Meadow, Peace and Justice: Notes on the Evolution and Purposes of Legal Processes, 94 Geo. L.J. 553 (2006) (commenting on the roles of different legal processes as a part of the justice system, including arbitration and mediation as alternatives to adjudication); Patricia M. Wald, Bureaucracy and the Courts, 92 Yale L.J. 1478, 1483 (1983) (calling for increased accuracy, clarity, and uniformity in the court system to enhance public confidence and satisfaction in the courts).

226See, e.g., Harry T. Edwards, Alternative Dispute Resolution: Panacea or Anathema?, 99 Harv. L. Rev. 668, 671 (1986) (“[W]e must determine whether ADR will result in an abandonment of our constitutional system in which the ‘rule of law’ is created and principally enforced by legitimate branches of government . . . .”); Owen M. Fiss, Against Settlement, 93 Yale L.J. 1073, 1085–87 (1984) (arguing that settlement is not an adequate substitute for adjudication); Resnik, supra note 16, at 134–35 (noting that “[a]ggregation can . . . be a form of Civil Gideon for plaintiffs”).

227Resnik, supra note 16, at 93.

228See AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740, 1746–48 (2011); Mitsubishi Motors Corp. v. Soler Chrysler-Plymouth, Inc., 473 U.S. 614, 636–37 (1985) (“Where the parties have agreed that the arbitral body is to decide a defined set of claims . . . the tribunal therefore should be bound to decide that dispute in accord with the . . . law giving rise to the claim. And so long as the prospective litigant effectively may vindicate its statutory cause of action in the arbitral forum, the statute will continue to serve both its remedial and deterrent function.” (footnote omitted) (citation omitted)). In Stolt-Nielsen S.A. v. AnimalFeeds Int’l Corp., 559 U.S. 662 (2010); Shady Grove Orthopedic Assocs. v. Allstate Ins. Co., 559 U.S. 393; and E.E.O.C. v. Waffle House, Inc., 534 U.S. 279 (2002), the Supreme Court had also tried to limit the extent to which private modification court thwart the public interest.

229See, e.g., Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 2304 (2013).

230See generally Erichson & Zipursky, supra note 96 (discussing the difficulty of achieving closure in mass tort settlements despite the demand for it by both plaintiffs and defendants); Nagareda, supra note 20 (exploring the conflict between autonomy for individual plaintiffs and global peace for defendants that is inherent in mass torts).

231See Dodge, supra note 163, at 776–77.

232See, e.g., Winzler v. Toyota Motor Sales USA, Inc., 681 F.3d 1208 (10th Cir. 2012); In re Aqua Dots Prods. Liab. Litig., 654 F.3d 748 (7th Cir. 2011).

233Winzler, 681 F.3d at 1209.

234Id. at 1211.

235Aqua Dots, 654 F.3d at 749–50.

236Id. at 752.

237 Id. at 751.

238Id.

239See supra Part I.

240See, e.g., John Hoeffel, PG&E Announces $100-Million Relief Fund for San Bruno Gas Explosion Victims, L.A. Times (Sept. 13, 2010, 1:38 PM), http://latimesblogs.latimes.com/lanow/2010/09/pge-announces-100-million-relief-fund-for-san-bruno-gas-explosion-victims.html.

241The notable exception is where the claims are high value and the litigation seeks punitive damages. See, e.g., Aaron Sankin, PG&E San Bruno Blast Source Revealed: Company Says 1956 Test Caused Deadly Explosion, Huffington Post (June 27, 2012, 3:26 PM), http://www.huffingtonpost.com/2012/06/27/pge-san-bruno-blast_n_1631550.html (noting that a class action for punitive damages had been filed despite PG&E’s disaggregative settlement fund and second quasi-governmental victims’ fund).

242See, e.g., Resnik, supra note 16, at 112–18 (summarizing the evolution of the Court’s treatment of the FAA over time).

243See, e.g., Arbitration Fairness Act of 2013, S. 878, 113th Cong. (2013); Arbitration Fairness Act of 2011, S. 987, 112th Cong. (2011); Arbitration Fairness Act of 2009, H.R. 1020, 111th Cong. (2009).

244See, e.g., Am. Express Co. v. Italian Colors Rest., 133 S. Ct. 2304 (2013); Nitro-Lift Techs., L.L.C. v. Howard, 133 S. Ct. 500 (2012); AT&T Mobility LLC v. Concepcion, 131 S. Ct. 1740 (2011).

245See Dodge, supra note 163, at 772–76.

246Compare In re Aqua Dots Prods. Liab. Litig., 654 F.3d 748, 752 (7th Cir. 2011) (denying class certification on the basis that a class action would not provide absent class members any relief not already offered by defendant), with In Re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on Apr. 20, 2010, 910 F. Supp. 2d 891, 920 (E.D. La. 2012) (granting class certification and ordering shutdown of the GCCF in favor of a new class action), aff’d sub nom. In re Deepwater Horizon, 739 F.3d 790 (5th Cir. 2014).

247Mullenix, supra note 32, at 912 (“The GCCF . . . illustrates an extreme and seemingly lawless expansion of the fund approach to resolving mass claims.”).

248See In re Deepwater Horizon, 739 F.3d at 803.

249See id.

25033 U.S.C. § 2705(a) (2012).

251See Mullenix, supra note 32, at 834.

252See id. at 833 n.62, 834 nn.65 & 69.

253See id. at 834–35.

254See id.

255See id. at 857.

256See id. at 820.

257See id. at 841–42 (comparing the relatively lax promulgation of GCCF standards and regulations with the more formal processes associated with the September 11th Victim Compensation Fund).

258See id. at 913.

259See id. at 861.

260See supra Part III.A.

261See Mullenix, supra note 32, at 856 (noting that in the initial “emergency payment” phase of the GCCF, the facility paid out “$3.3 billion to approximately 251,000 claimants”); Margaret S. Thomas, Morphing Case Boundaries in Multidistrict Litigation Settlements, 63 Emory L.J. 1339, 1342–46 (2014).

262See Mullenix, supra note 32, at 871.

263In re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on Apr. 20, 2010, 910 F. Supp. 2d 891, 900 (E.D. La. 2012), aff’d sub nom. In re Deepwater Horizon, 739 F.3d 790 (5th Cir. 2014).

264See Derek Hawkins, BP Spill Fund Freezes Payments After Atty Fees Order, Law360 (Jan. 4, 2012, 1:47 PM), http://www.law360.com/articles/297159/bp-spill-fund-freezes-payments-after-atty-fees-order.

265Id. Judge Barbier’s decision to approve the class settlement for economic and property damage was appealed to the 5th Circuit. The appellants argued that due to the potential inclusion of fictitious claimants, the class settlement did not meet the requirements of Rule 23 and the class settlement lacked Article III standing. In a decision filed January 10, 2014, the 5th Circuit affirmed Judge Barbier’s decision. In re Deepwater Horizon, 739 F.3d 790.The court held that class certification is not precluded if the class may include people who suffered no harm as a result of the defendant’s conduct. Id. at 803–04. The court further held that seeking recovery for an alleged harm suffered is sufficient to satisfy the requirements for Article III standing. Id.

266BDO Consulting, BDO USA, Independent Evaluation of the Gulf Coast Claims Facility: Executive Summary 1 (2012), available at http://www.justice.gov/iso/opa/resources/697201241917226179477.pdf (DOJ-commissioned report).

267BDO Consulting, supra note 200, at 15 fig.1.

268John Schwartz, Accord Reached Settling Lawsuit Over BP Oil Spill, N.Y. Times, Mar. 2, 2012, at A1.

269Cf. Plaintiffs’ Steering Comm.’s and BP Defendants’ Memorandum in Support of Joint Motion for (1) Preliminary Approval of Class Action Settlement, (2) Scheduling a Fairness Hearing, (3) Approving and Issuing Proposed Class Action Settlement Notice, and (4) BP’s Motion for Adjourning the Limitation of Liab. Trial at 7 n.3, In re Oil Spill by the Oil Rig “Deepwater Horizon” in the Gulf of Mex., on Apr. 20, 2010, MDL No. 2179 (E.D. La. Apr. 18, 2012), 2012 WL 1322258.

270In re Deepwater Horizon, 739 F.3d 790.

271See BDO Consulting, supra note 267, at 59–61.

272See, e.g., Dana A. Remus & Adam S. Zimmerman, Aggregate Litigation Goes Private, 63 Emory L.J. 1317 (2014).

273559 U.S. 393 (2010).

274See, e.g., Thomas, supra note 261.