Bankruptcy v. Multidistrict Litigation for Mass Torts

A recent spate of major mass tort bankruptcies has renewed interest in Chapter 11 as a tool for resolving mass torts. Mass tort bankruptcy proponents argue that even highly solvent defendants should be able to take advantage of bankruptcy because it is a better procedural system for resolving mass tort controversies than multidistrict litigation (MDL). Bankruptcy has tools for delivering closure that MDL lacks, and closure in mass tort litigation creates value. Of course, mass tort defendants have no right to closure, but if the resulting surplus is shared, it can leave defendants and plaintiffs better off. This Article accepts the premise that bankruptcy can be an appropriate vehicle for resolving mass torts so long as it leaves tort victims no worse off than they would have been outside of bankruptcy. But it shows that when there is enough money to go around, tort claimants are unlikely to do better in bankruptcy than in MDL for two reasons. First, contrary to some claims, MDL has proven highly successful at resolving mass tort controversies. That is, MDL is not nearly as bad as mass tort bankruptcy proponents portray it, and this Article corrects their misperception by painting a more realistic picture of how MDL operates. Second, if tort claimants are going to share in any surplus created by the move to bankruptcy, they need leverage to bargain for it. And, as compared to MDL, many features of bankruptcy systematically shift bargaining leverage from mass tort plaintiffs to defendants. Most fundamentally, control over the terms of aggregation is a source of leverage. MDL enables tort claimants and their lawyers to aggregate on their own terms, while bankruptcy empowers the defendant to impose aggregate resolution unilaterally. As a result, bankruptcy is not an appropriate forum for resolving mass torts when the defendant is not in financial distress.

Table of Contents Show

Introduction

Faced with tens of thousands of lawsuits by women alleging that its iconic Baby Powder was contaminated with asbestos and gave them cancer, Johnson & Johnson—a company with nearly half a trillion dollars in market capitalization and a better credit rating than the United States—decided that bankruptcy was the answer to its mass tort problem. In a move known as the “Texas Two-Step,” J&J split its consumer products subsidiary into two new companies: a good company with all of the assets and a bad company with all of the tort liability. The bad company then filed for bankruptcy, with an agreement from the good company to cover the costs up to the value of the legacy company. The idea was to channel all of the present and future tort claims into the bankruptcy, while the good company continued to operate normally outside of bankruptcy, free and clear of mass tort litigation.[1]

Needless to say, this move was controversial. The bankruptcy was thrown out twice as a bad-faith filing, first by the Third Circuit and then again, after J&J refiled, by the Bankruptcy Court for the District of New Jersey. The courts reasoned that the company had ample assets to meet its tort liability and thus was not in “financial distress.”[2] Undeterred, J&J reincorporated its bad company in Texas and launched a third bankruptcy there, hoping the Texas Two-Step would find a friendlier reception in the Fifth Circuit, as it has in other courts.[3] Without reaching the propriety of the Texas Two-Step, the bankruptcy judge in Texas dismissed the case a third time after finding major flaws in the voting process and insurmountable obstacles to confirming a plan of reorganization.[4]

J&J is not the only corporate giant to see bankruptcy as a potential escape hatch from the tort litigation system generally, and federal multidistrict litigation (MDL), in particular. 3M took a subsidiary into bankruptcy in an unsuccessful attempt to escape what had become the largest MDL in history, the 3M Combat Arms Earplug Litigation.[5] In doing so, 3M argued that the MDL system was “broken beyond repair,” and that bankruptcy provided the only option for efficiently resolving its mass tort liability.[6] Other defendants, ranging from opioid manufacturers[7] to the Boy Scouts,[8] have also turned to bankruptcy to address their mass tort exposure. This recent spate of major filings, along with the Supreme Court’s decision in Harrington v. Purdue Pharma L.P.,[9] has renewed interest in Chapter 11 as a tool for resolving mass torts.

Mass tort bankruptcies are nothing new. For decades, distressed companies have turned to Chapter 11 in the face of overwhelming liability from products ranging from asbestos to breast implants.[10] Where tort liability exceeds a company’s capacity to pay, bankruptcy’s tools become essential to the orderly and equitable resolution of victims’ claims.[11] But the recent turn to bankruptcy in cases like J&J and 3M where there is enough money to go around is not premised on the need for equitable distribution. Rather, these defendants rely on the idea that bankruptcy is a better procedural system for resolving mass torts than federal multidistrict litigation, where most mass tort claims are currently pending. This Article pushes back against that idea.

Bankruptcy and multidistrict litigation have a lot in common, at least when it comes to mass torts.[12] Both centralize litigation in a single federal court. Both shift control from individual tort claimants to lawyers acting on behalf of large groups. And both typically end in some form of settlement that adopts an administrative process for handing out compensation in the place of trials. But there are several key differences. Bankruptcy has powerful tools for delivering closure that MDL lacks. These include the ability to compel all claimants to participate in a single proceeding, to cut off future claims for recovery, and to bind nonconsenting claimants to a resolution.[13] These tools make bankruptcy particularly attractive to mass tort defendants, even if they can pay their bills on time. As defendant sentiment has turned against MDL, some defendants have started to view bankruptcy as an escape hatch out of an MDL that is not going their way.[14]

Many critics see this strategy as an abuse of the bankruptcy process.[15] They argue that healthy companies in no financial distress lack a valid bankruptcy purpose for filing.[16] And without the equitable necessity to divide up a limited fund, binding tort claimants to a resolution over their objections would put their constitutional rights to due process and a jury trial at risk.[17]

But in an important and influential article, Professors Anthony Casey and Joshua Macey make the case for using bankruptcy for mass torts even when there is enough money to go around.[18] Casey and Macey argue that mass torts present the kind of collective action problem that bankruptcy was designed to solve. They say that bankruptcy’s efficient procedures and ability to compel participation and bind holdouts to a global resolution can create value for debtors and creditors alike. In their view, the Third Circuit was wrong in the J&J case to limit bankruptcy to situations where the debtor is in financial distress.[19] They argue that if bankruptcy’s tools can create value, then we should embrace bankruptcy as a means to resolve mass torts so long as tort victims share in the gains.[20] Casey and Macey’s argument has gained traction among bankruptcy courts and practitioners and was cited extensively in Justice Kavanaugh’s dissent in Harrington.[21]

As noted above, many would reject Casey and Macey’s premise that value creation can justify bankruptcy absent necessity. But for the purposes of this Article, I want to accept it. I am not against bankruptcy for mass torts.[22] I have little doubt that closure in mass tort litigation creates value and that bankruptcy has powerful tools to disable holdouts and deliver closure that are unavailable in MDL.[23] Of course, mass tort defendants have no right to closure just because it creates value, especially when delivering it requires disrupting plaintiffs’ preexisting rights.[24] But if the resulting surplus—the “peace premium”—is shared, it can leave both debtors and creditors (defendants and plaintiffs) better off.[25] This focus on tort claimant welfare is motivated by the creditors’ bargain at the heart of modern bankruptcy theory: Bankruptcy exists for the benefit of creditors, not the debtor.[26] I thus take as my starting point Casey and Macey’s premise that bankruptcy proceedings can be an “appropriate means of resolving mass torts as long as they leave tort victims no worse off than they would have otherwise been.”[27] I am, however, deeply skeptical that bankruptcy benefits mass tort claimants more than multidistrict litigation when there is enough money to go around. This Article explains why.

Judicial and academic proponents of mass tort bankruptcies often start from an implicit (or sometimes explicit) premise that the tort litigation system—and MDL in particular—is broken. They present a caricature of MDL, drawn from some of its sharpest critics, and compare that to an idealized version of bankruptcy.[28] My first goal in this Article, therefore, is to correct these misperceptions about how MDL works. I attempt to present a more realistic comparison of bankruptcy and MDL, which are fundamentally just different forms of aggregate dispute resolution.[29] Neither is perfect, neither is static, and neither is as bad as its critics make it out to be. Both typically end in a mass negotiated resolution shaped by the tools available in each system and the bargaining environments they create.

With a more realistic view of MDL in hand, my second goal in this Article is to analyze how the move from MDL to bankruptcy alters the bargaining dynamics in favor of defendants. For tort victims to share in any surplus value created by bankruptcy’s ability to resolve all outstanding claims, they need the leverage to bargain for it. But many features of bankruptcy—especially two-step bankruptcies—shift leverage from plaintiffs to defendants.[30] This Article will analyze how bankruptcy and MDL’s different approaches to things like stays, voting, future claims, valuation, and forum shopping affect bargaining dynamics.

At a fundamental level, control over the terms of aggregation in mass tort litigation is a source of power. When the plaintiffs can bundle their claims together for settlement, they can extract a premium from the defendant in exchange.[31] When the defendant can impose aggregation over the plaintiffs’ objections, the defendant can keep the resulting premium. Many features of MDL guarantee that plaintiffs and their lawyers control the terms of aggregation. Many features of bankruptcy enable defendants to impose aggregation unilaterally.

Even if bankruptcy makes the pie bigger, if defendants gain the leverage to demand a bigger slice, plaintiffs may be left worse off. When the defendant cannot afford to pay everyone that may be the best plaintiffs can do. However, when the realistic alternative for tort victims is neither a destructive race to exhaust a limited fund nor endless waiting in a trial queue, but rather a negotiated resolution in MDL, bankruptcy may not be an appropriate means for resolving mass torts.

Part I of this Article explains why mass tort defendants value closure and find bankruptcy an attractive means to achieve it. Part II corrects the caricature of MDL that mass tort bankruptcy proponents adopt. Contrary to some assertions, MDL has tools to fairly and efficiently achieve comprehensive resolution in mass tort disputes. Part III shows how the move from MDL to bankruptcy systematically shifts leverage from plaintiffs to defendants. Even if bankruptcy creates value, mass tort bankruptcy proponents like Casey and Macey are incorrect to think that tort claimants are likely to share in that value when there is enough money to go around. I conclude with some reflections on the Third Circuit’s “financial distress” requirement for companies to access bankruptcy. The requirement has been subject to criticism.[32] But it plays a key gatekeeping role in preventing mass tort defendants from abandoning MDL for bankruptcy simply to increase their bargaining leverage and helps limit bankruptcy to situations where mass tort victims are likely to share in the benefit.

I. Closure and the Allure of Bankruptcy in Mass Torts

A. Bankruptcy and Closure

Defendants in mass litigation value closure. The ability to resolve all outstanding claims in a single transaction yields efficiency and certainty. A comprehensive resolution avoids the risk of adverse selection.[33] Defendants understandably do not want to overpay to settle relatively weaker claims only to be left facing the strongest at trial. Nor do they wish to fund a war against themselves as plaintiffs’ lawyers invest proceeds from early settlements into recruiting new clients. Comprehensive resolution allows defendants to put the dispute behind them, thereby ending the uncertainty, negative publicity, unwanted regulatory scrutiny, and concomitant drag on stock prices that often accompany mass tort litigation.[34]

Because closure creates value, defendants are often willing to pay a premium to get it. Thus, closure can benefit plaintiffs too. If the plaintiffs can bundle all of their claims together and offer to settle them in a single transaction, they can bargain for a share of the resulting surplus—the peace premium.[35] But because plaintiffs in mass litigation retain individual control over their claims, delivering closure can be difficult.[36] Coordinating a large group is always challenging, and individuals may be tempted to forgo a group settlement and hold out for a better deal. In short, plaintiffs face a collective action problem.

As a result, dealmakers in mass tort litigation are constantly searching for procedures that can deliver closure.[37] But peacemaking in mass torts is difficult because individual plaintiffs’ control over their claims is constitutionally protected.[38] Attempts to resolve those claims without plaintiffs’ consent can raise serious due process and Seventh Amendment concerns.

Indeed, concerns about the legitimacy of overriding individual control led the Supreme Court to reject two attempts to resolve the “elephantine mass”[39] of asbestos litigation through class action settlements. In Amchem Products, Inc. v. Windsor, the Court rejected an opt-out class action settlement because the legions of known and unknown asbestos victims did not form a “sufficiently cohesive” unit to “warrant adjudication by representation” and the class lacked “structural assurances of fair and adequate representation” for future claimants.[40] In Ortiz v. Fibreboard, the Court rejected a mandatory class action that sought to resolve all of a manufacturer’s present and future asbestos liability because the parties failed to show that the defendant’s assets and insurance coverage constituted a limited fund insufficient to fully pay all claims.[41]

After Amchem and Ortiz appeared to hobble the mass tort class action, asbestos lawyers turned to bankruptcy.[42] And, it is no surprise that mass tort dealmakers find bankruptcy attractive. Bankruptcy has powerful tools for delivering closure that other procedures like class actions and MDL lack.[43] These include the ability to compel all claimants to participate in a single proceeding, cut off future claims for recovery, and bind nonconsenting claimants to a resolution.[44]

Bankruptcy gets around many of the due process and Seventh Amendment objections that sank the mass tort class action because the necessity of an equitable distribution can justify overriding creditors’ preexisting rights.[45] This is easiest to understand when the debtor is insolvent. When there is not enough money to go around, everyone needs to take a haircut. To make that work, all claimants need to be at the table. You cannot divide up the pie until you know how many people are entitled to a slice.[46] Nor can you do so if the first person to the table can eat the whole thing.

This necessity is what justifies the mandatory features of bankruptcy—and what was missing in Ortiz.[47] Without bankruptcy, massive verdicts in favor of early plaintiffs can destroy the defendant’s value as a going concern, leaving nothing for later plaintiffs, especially those whose injuries may not manifest for years to come. As Judge Whitley has explained: “While a party has a jury trial right and a property interest in his tort claim, those rights are necessarily limited in bankruptcy cases because of the limited pool of assets available to pay creditors.”[48] The due process problems inherent in mass torts (problems of consent, conflicts of interest, notice, etc.)[49] do not go away in bankruptcy. We tolerate them out of necessity—or by looking the other way.[50]

Of course, a debtor need not be insolvent to invoke the protections of the Bankruptcy Code, and bankruptcy’s powerful closure tools remain attractive to mass tort defendants who can pay their bills on time.[51] Indeed, there are numerous examples of large and profitable companies turning to Chapter 11 to resolve their mass tort liabilities (e.g., Johns Manville,[52] Dow Corning,[53] PG&E[54]). But, for the most part, mass tort bankruptcies have remained self-limiting because of the pain associated with bankruptcy.[55]

Bankruptcy can be a painful process for corporate managers and shareholders. It requires extensive disclosures and submission to judicial supervision.[56] It can be disruptive of operations and demoralizing to employees.[57] It inevitably involves expensive lawyers and other professionals and can drag on for years.[58] And it sends a terrible signal to the market. In short, bankruptcy is not something that most corporate managers are eager to try, except as a last resort. As a result, solvent companies historically have not been quick to invoke bankruptcy to address their mass tort liabilities. And the ones that have, often use so-called “prepacks” to get in and out of bankruptcy as fast as humanly possible.[59]

A recent innovation known as the Texas Two-Step, however, can dramatically reduce the pain of bankruptcy and make it attractive to solvent mass tort defendants.[60] In the first step of a Texas Two-Step, a corporate defendant divides itself into two new companies using Texas’s divisional merger statute (or some comparable law in another state).[61] One company (GoodCo) holds all of the assets and operations, and the other company (BadCo) holds all of the tort liability plus a funding agreement from GoodCo. The second step is to take BadCo into bankruptcy while GoodCo continues operating normally outside of bankruptcy.[62]

If done correctly with a proper funding agreement, a Texas Two-Step should not reduce the total amount of money available to pay tort claimants.[63] A proper funding agreement will require GoodCo to reimburse BadCo for whatever amount BadCo is determined to be liable to tort claimants, at least up to the value of the legacy company. (If it did reduce the money available, the divisional merger would likely be a fraudulent transfer.)[64]

But the Texas Two-Step does eliminate the need for speed. Because the debtor, BadCo, has no business operations, it can stay in bankruptcy indefinitely. GoodCo, meanwhile, can continue with business-as-usual outside of bankruptcy. This ability to partition off the tort liability and the pain of bankruptcy from the company’s business operations makes bankruptcy—with its powerful closure tools—much more attractive to mass tort defendants, even when they are not facing any sort of financial distress.[65]

That is where the Texas Two-Step has generated controversy—when healthy corporate giants like J&J or 3M invoke bankruptcy to deal with their mass tort liability not because they need to, but because they think it is more efficient. And it tees up the question of whether efficiency alone can justify resorting to bankruptcy, even when there is enough money to go around. Can we justify overriding tort claimants’ preexisting rights to control their claims not out of equitable necessity, but simply because doing so creates value?

B. Who Benefits from Closure?

I have little doubt that closure can create value in mass tort litigation and that bankruptcy has powerful tools for delivering closure that MDL and other parts of the tort litigation system lack. Indeed, dealing with collective action problems, like the ones that make closure elusive in mass torts, is bankruptcy’s forte.[66] But mass tort defendants have no right to closure just because achieving it creates value.[67] Nor do they have any right to keep the resulting peace premium for themselves. If we are going to use bankruptcy to alter or override tort claimants’ rights in order to generate a surplus, we need to be sure that the tort claimants share in the gains.

This prioritization of creditors’ rights is built into the fabric of bankruptcy. Bankruptcy is described as a “creditors’ bargain”—the set of tools that creditors would agree on to preserve value for themselves, not for the debtor.[68] Under bankruptcy’s absolute priority rule, creditors’ interests—including tort claimants’—come before shareholders’ interests.[69] And under the “best interests of the creditors” standard, every creditor—again including all tort claimants—must do at least as well in reorganization as they would have if the debtor were liquidated.[70] I am setting aside here any independent concern for the debtor’s shareholders, employees, community, or the like. Providing the debtor with a “fresh start” is not the motivating policy behind Chapter 11 corporate reorganizations,[71] and it is even less important in a Two-Step bankruptcy where the debtor has no operations to speak of.

So, for the purposes of this Article, I accept Casey and Macey’s premise that “bankruptcy proceedings are appropriate means of resolving mass torts as long as they leave tort victims no worse off than they would otherwise have been.”[72] But, do bankruptcy proceedings benefit mass tort victims more than the most likely alternative, multidistrict litigation?

Revealed preferences suggest that they do not. In the Texas Two-Steps to date, defendants have unilaterally chosen bankruptcy in an attempt to escape from MDLs and the tort litigation system more broadly.[73] Large groups of plaintiffs have resisted these efforts and sought to have the bankruptcy proceedings dismissed. But this pattern of behavior could be explained by agency costs. Plaintiffs’ lawyers might think they can obtain greater fees in MDL and thus resist bankruptcy, even if their clients would do worse in MDL. Or perhaps current plaintiffs prefer to litigate in the MDL where they believe they can achieve greater recoveries, even if that comes at the expense of future plaintiffs—the proverbial race to the courthouse.[74] So, we cannot fully trust revealed preferences and need to evaluate the two systems side by side.

Where bankruptcy solves a collective action problem, it may very well benefit tort claimants more than MDL. The clearest case is when there is not enough money to go around. If bankruptcy prevents a destructive race to the courthouse and preserves the defendant’s value as a going concern, tort claimants, as a group, are likely better off in bankruptcy than litigating in MDL. Some plaintiffs—the ones who win the race to the courthouse—might prefer the MDL to taking a haircut in bankruptcy. But basic principles of equity allow overriding their preferences in order to preserve the rights of similarly situated claimants.[75] I have no quarrel with mass tort bankruptcy when the defendant is in genuine financial distress, even if that means overriding claimants’ preexisting rights. That may be the best the tort victims can do.

Many would stop there. Bankruptcy’s infringement on individual rights is justified when needed to prevent a race to the courthouse. But it cannot be justified when it is not necessary to preserve the defendant’s going-concern value. Absent financial distress, they maintain, resolving mass tort liability more efficiently is not a valid bankruptcy purpose.[76] But at least for the sake of argument, I want to go further. Under a creditors’ bargain theory, bankruptcy could be justified, even without financial distress, if it leaves the creditors—here, the tort claimants—better off than they would be outside of bankruptcy.

Because defendants value closure, bankruptcy may create value even when there is enough money to go around. If bankruptcy’s closure tools allow tort claimants to overcome the collective-action problem they face in efficiently bundling their claims together to offer the defendant peace, they may be able to command a premium. I am setting aside here concerns about plaintiffs’ ability to participate, have their voices heard, or have their day in court[77]—not because such concerns are unimportant, but because I want to consider bankruptcy on its own terms. If the closure available in bankruptcy creates a peace premium, that could work to tort claimants’ material benefit.

But, in order to benefit from any value bankruptcy might create, tort claimants need leverage to bargain for a share of the surplus. Outside of bankruptcy, if mass tort defendants want peace, they must purchase it from tort victims through settlements that are generous enough to make participation more attractive than continued litigation.[78] In short, defendants must share the surplus with plaintiffs to get them to agree to settle. Bankruptcy, by contrast, includes tools that can override claimants’ consent in certain situations.[79] That can shift leverage in defendants’ favor. If the defendant can unilaterally impose closure on tort claimants without their consent, the defendant can capture all of the resulting surplus—or more. Bankruptcy does not give defendants unilateral control, of course, but it affects the sources of plaintiffs’ and defendants’ leverage in important ways.

So, in the end, whether mass tort bankruptcy benefits tort victims depends on two things: (1) How bad is the alternative system (i.e., MDL)? And (2) what does the shift from MDL to bankruptcy do to the bargaining dynamics? The next two Sections will consider those questions in turn.

II. Correcting the Caricature of MDL

Part of bankruptcy’s appeal for mass torts seems to stem from an implicit assumption that MDL, and the tort litigation system more broadly, are broken beyond repair. But proponents of mass tort bankruptcy, both on the bench and in the academy, often adopt a distorted view of MDL drawn from some of its sharpest critics. Proponents then compare this caricature to an idealized version of bankruptcy. This Section attempts to paint a more realistic picture of how MDL works in mass tort cases and highlight some of its strengths as well as its weaknesses.

A. The “Broken Beyond Repair” Myth

Many of mass tort bankruptcy’s defenders portray MDL as a failure. Indeed, even some mass tort bankruptcy critics think that bankruptcy is more efficient.[80] The narrative of MDL’s shortcomings seems to alternate between two poles. At one end, MDL is characterized by chaotic and inefficient duplication of procedures leading to endless trial queues. At the other, MDL is an inescapable vortex populated by greedy repeat-player plaintiffs’ lawyers who extort defendants with hordes of baseless claims while coercing their captive clients into unfair settlements. Neither description is accurate.

Some critics portray MDL as insufficiently comprehensive, making it slow, costly, and inefficient.[81] Without bankruptcy’s ability to compel all claimants to participate in a single proceeding, they say, MDL is plagued by overlapping litigation in multiple jurisdictions, redundant discovery, duplicative litigation of issues, and inconsistent application of rules.[82] MDL will, they assume, lead to endless trial queues where cases must be tried one by one, incurring astronomical litigation costs and lengthy delays for both plaintiffs and defendants.[83] Moreover, these critics bemoan the “lottery-like” effects of jury trials where some plaintiffs hit the jackpot while many are denied any recovery at all.[84] And they worry that MDL is unfair to future claimants who may be left out of settlements negotiated in the MDL only to find that the defendant has nothing left to compensate them when the claimants’ injuries manifest years later.[85] On this view, without the ability to bind all claimants—present and future—to a settlement, global resolution is impossible.[86]

Indeed, the dissenting Justices in Harrington picked up on many of these themes. Justice Kavanaugh wrote:

In many cases, there is no workable alternative other than bankruptcy for achieving fair and equitable recovery for mass-tort victims. “Outside of bankruptcy,” victims face “significant administrative costs” of multi-district litigation, “which has limited coordination mechanisms and no tools for binding future claimants.” And multi-district litigation cannot “solve the collective action problem because dissenting claimants can opt out of settlements even when super majorities favor them.”[87]

Another set of mass tort bankruptcy proponents portrays MDL as almost too powerful. Instead of being overly preservative of tort claimants’ autonomy, they say MDL forms “a captive settlement negotiation” from which neither plaintiffs nor defendants can escape or obtain a trial date.[88] In this picture: “Victims lack control in an MDL. Agency principles break down in the MDL process because the agents—plaintiffs’ attorneys—are invariably immune from the instructions and wishes of the principals, the victims.”[89] And “MDL judges are extremely invested in these cases and have exhibited a propensity to compel settlements that are coercive to individual plaintiffs.”[90] This results in “[c]ompelled settlements [that] rarely consider culpability, heightening the possibility of extortive litigation.”[91] These settlements “live in the shadows” because they “do not need court approval, and confidentiality agreements invariably prevent publication or assessment of the details.”[92]

While no one thinks MDL is perfect, these descriptions are drawn from some of MDL’s most prominent critics,[93] and they paint a one-sided picture that fails to recognize many of MDL’s strengths.[94] This distorted view of MDL is then compared to an idealized version of bankruptcy—generally one that incorporates mass tort bankruptcy proponents’ preferred reforms.[95] These reforms range from simple tweaks, like having the U.S. Trustee nominate candidates for future claims representatives,[96] limiting interim stays,[97] or improving disclosures about third-party releases,[98] to seemingly radical changes like eliminating the debtor-in-possession’s exclusive right to propose a plan of reorganization[99] or giving tort creditors priority over secured creditors.[100] I have no special insight into how feasible any of these reforms are (though some explicitly call for substantial new legislation).[101] But in comparing bankruptcy and MDL as vehicles for resolving mass torts, it seems important to consider both systems as they actually operate today, not as we wish they would operate.

B. The Realities of Multidistrict Litigation

In reality, MDL offers a highly effective set of tools for resolving mass tort controversies. I do not wish to describe MDL through rose-colored glasses. MDL has its weaknesses; I have written about some of them in the past.[102] My aim here is simply to dispel some misconceptions about how MDL works on the ground.[103] While it lacks bankruptcy’s complete aggregation and ability to bind dissenters, on the whole, MDL is well adapted to handling the challenges of mass torts in a decentralized federal system that allows states to make different choices about substantive tort law and affords constitutional protections to individual plaintiffs’ control over their tort claims. Even in the face of such challenges, MDL’s flexibility has allowed parties to negotiate efficient resolutions to some of the nation’s largest mass tort controversies.[104]

1.   MDL’s Split Personality

When many cases sharing a common question of fact are pending in federal district courts all over the country, the MDL statute authorizes the Judicial Panel on Multidistrict Litigation (JPML)—a panel of seven federal judges chosen by the Chief Justice—to transfer those cases to a single federal district judge for coordinated pretrial proceedings.[105] So, one judge—the MDL judge—will oversee discovery, motion practice, and other pretrial proceedings for all of the related federal cases, with the goals of conserving resources, ensuring consistent rulings, and potentially laying the groundwork for global resolution. At the conclusion of pretrial proceedings, in theory at least, transferred cases will be remanded back to the districts where they were originally filed for trial.[106] But in reality, remands are rare. Around 97 percent of cases are resolved in the MDL district.[107]

Created in 1968, MDL was intended to be the primary forum for the resolution of mass tort litigation. The small group of federal judges who drafted the MDL statute and shepherded it through Congress anticipated the coming explosion of product liability litigation and expected it to proceed in MDL.[108] They gave both the JPML and the MDL judge the procedural tools and flexibility necessary to manage such massive litigation towards resolution.[109] Although MDLs involve many types of cases (e.g., antitrust, sales practices, intellectual property, securities), in recent years mass tort cases have accounted for more than 90 percent of actions pending in MDLs (and more than half of the entire federal civil docket).[110]

MDL is more than just a tool for coordinated discovery. The statute gives the MDL judge all the powers of a federal district judge during the pretrial phase of litigation, including the powers to rule on dispositive motions like summary judgment, determine the admissibility of expert testimony, and manage the litigation towards settlement.[111] MDL judges are hand-picked by the JPML for their skill in case management and willingness to take on major mass tort litigation (no judge is forced to accept an MDL assignment).[112] Over many years and many thousands of cases, experienced MDL judges and practitioners have developed a set of tools to efficiently manage complex mass tort litigation.[113]

Mass torts typically involve claims under varying state laws by thousands of individuals from across the country. The plaintiffs’ alleged injuries can vary widely from minor or not-yet-manifest impairments to serious illnesses like cancer. And plaintiffs may face different challenges in proving individual causation depending on factors like length of exposure or underlying health problems. The plaintiffs will have retained their own lawyers, and with serious injuries and substantial amounts of money on the line, will often have a strong interest in their individual cases. Plus, claims will be governed by a multitude of different states’ substantive laws. Forcing such disparate plaintiffs into a single, undifferentiated class to litigate in the aggregate raises serious due process and federalism concerns.[114]

One reason MDL has proven so successful at resolving mass torts where the class action has failed is that MDL was actually designed to handle them. Multidistrict litigation simultaneously enables aggregation and respects the individual character of each case. In that sense, as Andrew Bradt and I have explained, MDL has a sort of split personality.[115] By formally preserving the individual nature of each case, MDL avoids many of the due process stumbling blocks of mass tort class actions and accommodates our federal system. Because it is only a temporary consolidation of individual claims, for example, we do not have to worry about whether the MDL court would have personal jurisdiction over all of the parties’ claims.[116] Nor does MDL have the class action’s requirement that common issues predominate over individual ones.[117] MDL is thus well suited to handle claims where different states’ laws will apply.[118] At the same time, MDL functions as a powerful aggregation tool. It forces litigants from around the country into a single forum where they must find a way to work together. Plaintiffs cannot simply opt out of an MDL; there is no opportunity for exit until the MDL judge decides that pretrial proceedings have concluded.[119] MDL can thus achieve substantial aggregation and set the stage for collective resolution, even when mass tort claims are too disparate for class treatment.

2.   Extensive, but Not Complete, Coordination

While aggregation in MDL is not as complete as bankruptcy, centralizing the federal cases in front of a single judge allows for extensive coordination. Doing so not only promotes efficient case management but also helps the plaintiffs work together, enables the maturation of the dispute, and preserves an external check through state courts.

a.       Leadership Appointments and Collective Action

Centralization in an MDL helps plaintiffs overcome the collective-action problem they face in maximizing their negotiating leverage against the defendant. Mass tort defendants have a structural advantage in many-against-one litigation because they can rationally spread costs and make tradeoffs across cases.[120] Individual plaintiffs, by contrast, care primarily about their own cases, not how they might impact the entire pool of cases; most would happily free ride on the efforts of others.[121] MDL can help balance the scales by forcing plaintiffs and their lawyers from all over the country to work together in a single forum.

The primary way that MDL facilitates plaintiff coordination is through the appointment of lead lawyers. An MDL judge will typically appoint a committee of experienced and well-resourced attorneys to coordinate the litigation, communicate with the court, manage common discovery, brief common legal issues, and generally handle “common benefit work” on behalf of all the plaintiffs.[122] Lawyers who are not appointed to leadership may be directed not to propound discovery requests or file anything with the court without working through the lead lawyers first.[123]

To properly incentivize lead lawyers to act on behalf of all the plaintiffs, MDL judges will typically award them “common benefit fees.” These fees help compensate lead lawyers for undertaking expensive and time-consuming tasks such as motion practice, common discovery, retaining expert witnesses, and other work on common issues that benefits all plaintiffs.[124] Common benefit fees are only awarded if the cases are successfully resolved, and they are typically funded by “taxing” nonlead lawyers’ fees.[125]

The appointment and payment of lead lawyers facilitates collective action. It forces potential rivals among the plaintiffs’ bar to work together. It concentrates control over the litigation in the hands of experienced attorneys who know how to make a deal and can credibly negotiate with the defendant.[126] And it provides those lawyers with the incentive to invest in steps in the litigation that will benefit all plaintiffs, not just their own clients.

b.       Maturation and Valuation

MDLs also create a centralized forum for the exchange of information in discovery.[127] Doing so avoids the massive costs and duplication of effort that would result if discovery proceeded independently in each case. Document repositories and discovery databases can be created to house common information. Instead of deposing an expert witness or the defendant’s CEO a thousand times, the witness can be deposed once. Inconsistent rulings on discovery motions can be avoided. This entire process will be overseen by the MDL judge, who may, for example, divide discovery into phases by staying all individual discovery until discovery into common issues is complete. The MDL judge can thus consciously sequence the matters to be decided, placing some issues on the back burner while the court and parties focus on others. In other words, MDLs “rationalize—and ration—the process of adjudication” to make sure that the litigation proceeds equitably and efficiently.[128]

In addition to creating efficiencies, centralized discovery plays an important role in enabling settlement.[129] A central challenge in resolving mass torts in any forum is the valuation of claims. Plaintiffs will say their claims are worth millions of dollars, while the defendant says they are worth nothing at all. Unless the parties can converge in their valuations, settlement will remain impossible. And we cannot just presume that all of the plaintiffs’ claims have the same value. Even if they have suffered the same injury or manifested the same disease, plaintiffs’ claims can have widely divergent values depending on the applicable substantive law and whether they can prove liability and individual causation.[130]

In order to figure out what each claim is worth, the parties need to test the underlying merits of those claims. In other words, as Professor Francis McGovern explained, the mass tort needs time to “mature” through the exchange of information in discovery and testing of theories of liability in litigation, to the point where the parties know enough to reach agreement on approximately how much different types of claims are worth.[131] They need not value claims precisely, of course, but they need enough information to identify the relevant variables that affect claim value and start constructing a framework for equitable settlement payouts.

This process of maturation inevitably takes time.[132] But MDL judges have employed many tools to help facilitate the exchange of information and speed up the process. Deciding substantive motions on common issues like preemption, general causation, and the admissibility of expert testimony under Daubert[133] can help inform all parties about the viability of different claims and legal theories. Indeed, such motions can sometimes resolve the entire litigation in one fell swoop if, for example, the MDL court grants summary judgment for the defendant on a common issue like general causation.[134]

MDL judges routinely require plaintiffs to fill out “fact sheets” that provide information about plaintiff characteristics, product use, exposure, and injury, often with supporting documentation and medical records.[135] These fact sheets operate like standardized interrogatories that probe specific causation and injury—and screen out meritless claims—even while formal discovery into individualized issues is stayed. In the 3M Combat Arms Earplug MDL, for example, simple fact sheets culled out tens of thousands of claims.[136] Plaintiff fact sheets also allow the parties and the court to build a database of pending claims and their relevant characteristics to more fully understand the universe of claims and identify the key issues that will have to be decided to drive the litigation towards resolution.[137]

Bellwether trials are another tool to generate information about the merits and help the tort mature. MDL judges will frequently select a handful of test cases for full individualized discovery and trial in the MDL court. Although they generally cannot try cases transferred under § 1407 after the Supreme Court’s decision in Lexecon v. Milberg, Weiss, Bershad, Haynes & Lerach,[138] MDL judges can try cases if plaintiffs file them directly in the MDL district or if the parties consent to a “Lexecon waiver.”[139] MDL judges often encourage the parties to enter “direct filing stipulations” that allow plaintiffs to skip the transfer process and file cases directly in the MDL district, further increasing the bellwether pool.[140] MDL judges can even (with the proper permission) follow cases back to their transferor districts and sit there by designation for bellwether trials, as Judge Fallon did in the Vioxx litigation.[141] Finally, as discussed below, MDL judges can coordinate with other judges to try bellwether cases on remand or in state court.

Such bellwether trials give the parties critical information about how legal and factual disputes play out in front of real live juries. MDL judges consciously construct bellwether trial pools (often drawn from the database populated with information from plaintiff fact sheets) to yield cases that will be representative of the larger universe of cases or that will most effectively tee up critical issues that need to be decided.[142] Even if some bellwether cases are settled or dismissed before verdicts are reached, those outcomes still yield valuable information about the strengths and weaknesses of different legal theories and evidence and help the parties place a dollar value on different types of claims.[143]

These tools need not be used in any particular order. The MDL statute and the Federal Rules of Civil Procedure give MDL judges tremendous flexibility to manage litigation in the manner they think will lead to the most efficient resolution.[144] Indeed, MDL judges can deploy many of these tools in tandem. A judge could, for example, issue a case census and direct plaintiffs to begin filling out fact sheets, while simultaneously moving forward with expert discovery into common issues like general causation. The judge might also appoint lawyers on both sides to settlement committees and direct them to mediation so that settlement discussions can take place contemporaneously with other activities in the case. Alternatively, a judge who thinks that ruling on a common issue, such as preemption or general causation, might resolve all the claims at once could stay all individualized discovery and focus on the potentially dispositive issue. No single approach is right for every MDL, and case management decisions are left to the discretion of the MDL judge. All of these tools help generate the kind of information that is critical to placing a valuation on the plaintiffs’ tort claims and enabling resolution.

c.       Coordination and Competition Across Jurisdictions

Unlike bankruptcy, MDL does not completely aggregate all claims against the defendant. Aggregation in an MDL is limited to claims that have been filed in or removed to federal court.[145] In most mass torts, there will be satellite state court litigation (often consolidated through state-level MDL-like procedures).[146] But that does not lead inevitably to wasteful duplication of effort, redundant discovery, and inconsistent rulings, as critics fear. MDLs work well at coordinating massive disputes across jurisdictional lines.

Federal MDL judges have decades of experience working cooperatively with state judges to coordinate mass tort litigation. Examples abound of state and federal judges working together on scheduling, leadership appointments, discovery, pretrial rulings, and bellwether trials. In the Vioxx litigation, Judge Fallon, who was handling the federal MDL, worked intimately with three state judges. Sometimes, the four judges would even take the bench together in joint hearings to issue consistent rulings and encourage settlement.[147] And Judge Pointer, who was handling the federal Silicone Gel Breast Implants MDL, enlisted the help of state judges in managing common discovery and resolving difficult issues of state law.[148] Other examples are numerous. Indeed, the Federal Judicial Center and the National Center for State Courts have collaborated to manage a website and put out a Pocket Guide for MDL and state judges working together on mass tort cases.[149]

Federal-state coordination is not perfect, of course, and there are instances where state and federal judges may butt heads. But incomplete aggregation has advantages, even if jurisdictional redundancy comes at the cost of some efficiency.[150] Satellite state court litigation can act as an external check on a federal MDL. If the lead lawyers in the federal MDL appear to be shirking their responsibilities or getting too cozy with the defendant in settlement discussions, and appeals to the MDL judge fall on deaf ears, other lawyers can press the issue in front of different decision-makers in state court.[151] Or if the MDL judge is moving slowly, plaintiffs can press forward to trial in state courts. Too much friction would erode the benefits of aggregation. But the opportunity to take an issue to a different judge in a different court system imposes some competitive discipline on federal MDL judges and lead lawyers and serves as a valuable pressure-relief valve if things go badly.[152]

Plus, there are affirmative advantages to incomplete aggregation. MDL creates opportunities to enlist multiple decision-makers to move mass tort litigation towards maturity. Centralizing cases in an MDL does not mean that the MDL judge has to become a bottleneck.[153] State court trials (often scheduled in coordination with the federal MDL judge) can serve as bellwethers and generate critical information about liability and case values.[154] And MDL judges can enlist the help of other federal judges in moving the litigation forward. In the 3M Combat Earplugs MDL, for example, Judge Rodgers invited senior and visiting district judges from around the Eleventh Circuit to come to Florida and try bellwether cases, allowing multiple trials to proceed simultaneously.[155] And in the National Prescription Opiate MDL, Judge Polster adopted a “hub-and-spoke” model, in which he picked a set of test cases that would yield the most useful information and remanded them to transferor judges (the spokes) ready to try them promptly as bellwethers. Meanwhile, the rest of the cases remained pending in the MDL (the hub) for future settlement negotiations.[156] This sort of strategic disaggregation expands the pool of bellwether cases beyond those that MDL judges have the authority to try themselves, speeds the development of information necessary for broader settlements, and takes advantage of the diversity of judges and juries in multiple courts to yield more accurate results.[157] In other words, MDL coordination offers the benefits of centralization and decentralization at the same time.

3.   Settlement, Not Trial, Is the Endgame

In comparing the two systems, mass tort bankruptcy proponents sometimes contrast resolution in bankruptcy with the prospect of thousands of trials in the state and federal courts after an MDL concludes.[158] This is portrayed as a nightmare that would grind the court system to a halt. The “lucky” few might win lottery-like verdicts, but most tort victims would be consigned to endless waiting in trial queues after MDL pretrial proceedings have concluded and the cases are remanded by the thousands.

In reality, remands and trials are rare. An MDL does not “necessarily” result in the remand of thousands of cases to the districts where they were originally filed for trial.[159] In fact, less than 3 percent of cases are ever remanded.[160] The nightmare scenario of transferor courts being suddenly inundated with forty thousand trial-ready cases just does not happen in the real world. When they do remand cases, MDL judges and the JPML are strategic about it. They do it in phases, with an eye towards facilitating settlement; they do not simply unleash tens of thousands of cases into endless trial queues. Indeed, even in the Asbestos MDL, where in 2008 Judge Robreno determined that no global settlement would be forthcoming and took an aggressive approach to remanding cases, only about 750 out of 185,000 cases were actually sent back to their originating districts.[161] The rest were almost all resolved in the MDL by 2013.[162]

In MDLs, as in all litigation, nearly all cases are resolved before trial. The rate of trials in mass tort MDL cases appears no different from the rate of trials in federal civil cases more generally—that is, fewer than 1 percent of cases are resolved by trial.[163] Settlement is the endgame in the overwhelming majority of cases. And in MDLs, cases tend to settle in large groups, not one by one.[164]

4.   MDLs Regularly Achieve Global Resolution

Contrary to the claims of mass tort bankruptcy proponents, global resolution of mass tort claims is not impossible in MDL. Indeed, MDLs have comprehensively resolved many mass tort disputes through a variety of approaches, including confidential inventory settlements,[165] global nonclass aggregate settlement programs,[166] and class action settlements under Rule 23.[167] What these various approaches have in common is that they create an alternative procedure for groups of claimants to resolve their disputes and receive compensation efficiently and consensually.

a.       MDL Settlement Structures

Comprehensive nonclass settlements in MDLs typically take the form of an offer (or series of offers) by the defendant to the plaintiffs’ lawyers to settle all of their claims for a lump sum. The terms of this offer are usually set forth in a “Master Settlement Agreement” or some similar document. To be clear, a master settlement agreement is not itself a settlement between the parties to the litigation. Instead, it is an agreement between the defendant and the plaintiffs’ lawyers that specifies the amount the defendant is willing to pay the lawyers’ clients and the framework through which the settlement proceeds will be allocated to clients who choose to participate.[168]

The deal can be structured as a single “global” master settlement agreement, open to all lawyers in the MDL to settle the claims of all of their clients on the same terms. Or the defendant may negotiate a separate master settlement agreement with each plaintiffs’ firm to resolve the litigation firm by firm in a series of “inventory” settlements.[169]

While each master settlement agreement is tailored to the unique dynamics of a particular mass tort dispute, they tend to share some common features. First, the master settlement agreement will define the total settlement amount to be allocated among all participating plaintiffs.[170] That amount is negotiated based on the parties’ assessments of the number and strength of eligible claims, which, of course, draws on all of the information developed through the MDL process.

Second, the master settlement agreement will specify the eligibility criteria for participation in the settlement program and the required documentation or other proof that claimants will have to provide.[171]

And third, the master settlement agreement will define the process for allocating settlement payments to participating claimants and obtaining their releases in exchange. Not all claimants will receive the same amount of compensation; their allocations will vary depending on the injuries they have suffered and the parties’ estimation of the strength of their claims. The specifics of the allocation process vary, but they often take the form of a “grid” or “points” system. Drawing on information generated through the MDL process, the parties will identify various factors relevant to compensation, such as severity of injury, duration of exposure, age, underlying medical conditions, and other proxies for claim strength. The parties will then construct a grid specifying compensation amounts for different combinations of factors.[172] Or they may assign a certain number of base points to each injury type and then adjust those points up or down depending on other relevant factors. Each claimant’s “score” is then used to determine the final compensation amount.[173]

When a plaintiffs’ law firm enters a nonclass master settlement agreement to settle its inventory or decides to participate in a global master settlement agreement, it must then present the proposed deal to its clients and obtain their consent to settle.[174] The “aggregate settlement rule” (ABA Model Rule of Professional Conduct 1.8(g) and its state counterparts) requires that, in doing so, the law firm disclose to each client all of the terms of the deal, including how much their other clients are getting and how the attorneys will be paid.[175] These disclosures are an important guarantee of transparency and horizontal equity. They make sure that all claimants know which criteria matter in determining settlement payments and how much they are getting compared to other similarly situated claimants before they consent to participate and release their claims.[176] Master settlement agreements in MDLs thus operate on an opt-in model. Each claimant must consent in order to be bound.

As private contracts, master settlement agreements do not require judicial approval.[177] MDL judges are often involved in facilitating settlement, and they will sometimes weigh in informally on the merits of the deal (and, in doing so, send a valuable signal about the fairness of the settlement).[178] But unless a class action is certified, there is no formal fairness hearing or Rule 23(e) approval required.

MDLs can also be resolved on an opt-out basis through a class action settlement. MDLs and class actions are not mutually exclusive. The JPML may consolidate competing class actions into an MDL so that a single judge can decide which (if any) to certify as a class.[179] Or, as is more common in mass torts, the defendant and lead lawyers may decide that a class action is a useful vehicle for resolving all of the individual actions that have been transferred into an MDL.[180] They thus agree that the lead lawyers will file a new settlement-only class action in the MDL district and simultaneously ask the MDL judge to certify a class and approve the settlement under Rule 23(e). Many MDLs are resolved through class action settlements negotiated within the MDL in this manner.[181] Class action settlements shift the default rule to participation and come with increased transparency and judicial supervision under Rule 23.

b.       MDL Settlements Have Tools for Closure

Mass tort bankruptcy proponents suggest that MDLs cannot achieve closure in mass tort disputes because “dissenting claimants can opt out of settlements even when super majorities favor them.”[182] Indeed, according to Casey and Macey: “Reaching global resolution is possible only when courts can compel participation.”[183] But recognizing the value of closure—and that defendants will often pay for it—plaintiffs’ lawyers in MDLs have figured out ways to bundle claims together and deliver to defendants the closure they desire without a judicial mandate.[184] The ability to bundle claims together for sale to the defendant gives plaintiffs’ lawyers leverage to negotiate for a peace premium. Indeed, private settlement agreements within MDLs frequently include a variety of contractual provisions designed to promote closure.[185] While they may not be as complete as a discharge and channeling injunction in bankruptcy, MDL settlements have a long track record of resolving substantially all of the claims in mass tort disputes.

For example, nearly all MDL settlements—whether global master settlement agreements, inventory settlements, or class action settlements—include a “walk-away provision” that allows the defendant to back out of the deal if a critical threshold of eligible claimants declines to participate.[186] These thresholds vary (from 85 percent right up to 100 percent), as do the consequences of triggering them.[187] But whatever the precise terms, a walkaway provision guarantees that the defendant will not be bound by the settlement unless it achieves its desired degree of closure. A defendant who wants complete finality for all known claimants can negotiate for a right to walk away if participation is less than 100 percent.[188]

Other ways that MDL settlements promote closure include participation bonuses and most-favored-nation clauses. Participation bonuses increase the total settlement fund by escalating amounts as claimant participation approaches 100 percent, providing extra incentive for lawyers to get the last few eligible claimants to sign on.[189] A most-favored-nation clause promises that if the defendant subsequently settles on more favorable terms with any claimant outside the master settlement framework, it will increase payments to all participating claimants to match.[190] Because topping up all participating claimants would be prohibitively expensive, a most-favored-nation clause credibly signals the defendant’s commitment to fight nonsettling claimants tooth and nail.[191] Unless they are fully prepared to litigate their claims through trial and appeal, claimants face a strong incentive to participate in the MDL settlement because they know they will not get a better deal on their own.[192]

Master settlement agreements often include terms that prevent plaintiffs’ lawyers, who tend to know more about the relative strengths of the claims in their inventories than the defendant, from funneling weak claims into the settlement while cherry-picking their best claims for continued litigation.[193] Many settlements will require lawyers to agree to recommend the deal to all of their eligible clients. And if some clients still do not want to participate, the settlement agreement will require lawyers to take all necessary steps to withdraw from representing nonsettling clients.[194] The Vioxx settlement is probably the best-known example,[195] but many other MDL settlements (e.g., AMS Mesh Products, Propulsid, Fosamax, 3M Earplugs) have followed a similar model.[196] Lawyers who fail to comply with these recommendation and withdrawal provisions risk having their entire inventory expelled from the settlement. Plaintiffs’ lawyers are thus either “all in” for their whole inventory or “all out.”[197]

While they are primarily aimed at preventing strategic action by plaintiffs’ lawyers, these sorts of lawyer recommendation and withdrawal provisions create a powerful motivation for claimants to participate in the settlement. If clients do not sign on, they will likely have to find a new lawyer to handle their cases. Lawyer recommendation and withdrawal provisions are thus powerful tools for achieving closure.

Lawyer recommendation and withdrawal provisions are controversial. They have been criticized as coercive and unethical.[198] Exactly how coercive these contractual provisions are depends on the details of their terms and the market for alternative representation.[199] If the plaintiffs who do not want to settle on the terms offered have strong claims, those should be attractive to other MDL lawyers or new market entrants; if they have weak claims, they may be unlikely to do any better on their own, given the cost of litigating a product liability case individually. And it is worth noting that plaintiffs’ lawyers cannot unilaterally drop nonsettling clients; they must seek judicial permission to withdraw from representation.[200]

But if the lawyer does successfully withdraw, it may not be easy for nonsettling plaintiffs to find a new lawyer willing and able to undertake the representation—especially if all the other experienced mass tort lawyers are entering similar settlements or if lawyer withdrawal provisions are paired with terms that alter the market for legal services.[201]

Other settlement terms promote closure going forward by altering the market for legal services. As noted above, defendants do not want the settlement payments they make today to fund new litigation against them tomorrow. They want to get the lawyers they settle with out of the business of suing them. Simply paying the lawyers not to sue in the future would violate the ethical rules that bar lawyers from agreements restricting their practice of law.[202] But some settlement terms require lawyers to cease advertising for new claims. Others require participating lawyers to forgo referral fees for nonsettling clients, thus eliminating any financial incentive to help them find another lawyer.[203] Still others, like the National Prescription Opiate Distributors Settlement, use the conflict-of-interest rules to get lawyers out of the game. By spreading payments out over an extended period and making some of those payments contingent on no new suits being filed, the Settlement made it impossible for settling lawyers to take on new clients, whose recoveries would essentially come out of their current clients’ pockets.[204] In short, there are numerous ways for defendants to contract for closure in MDL.

MDL judges also have tools to promote closure in MDL settlements. After a global settlement is consummated in an MDL, the defendant will often ask the judge to issue a so-called “Lone Pine order” for all nonsettling claims still pending.[205] A Lone Pine order requires plaintiffs to come forward with certain evidence (typically expert medical evidence of causation and injury) by a certain deadline or face summary judgment.[206] In other words, it requires the nonsettling plaintiffs to “put up or shut up.” Because they require plaintiffs’ lawyers to either make a substantial investment in nonsettling cases right away or abandon them, Lone Pine orders make participating in the settlement all the more attractive.[207]

c.       Closure Can Reach Beyond the Bounds of the MDL

The scope of a mass tort will not typically be limited to the cases actually filed in the MDL. Many mass tort claims are filed in state courts or not filed at all. Mass tort plaintiffs’ lawyers will often negotiate “tolling agreements” with the defendant to pause the statute of limitations on certain claims instead of filing them in court.[208] This “shadow docket” of unfiled claims can make up a substantial portion of mass tort lawyers’ claims inventories and, indeed, may vastly outnumber the lawsuits actually filed.[209] Any global resolution will have to take account of unfiled claims as well as filed claims.

MDL settlements routinely do so. Master settlement agreements are not limited to plaintiffs who have filed actions in the MDL. Both global settlements and inventory settlements are typically open to—and often conditioned on the participation of—all of the settling lawyer’s clients, whether their claims have been filed in the MDL, in state court, or not at all. In the Vioxx litigation, for example, almost 50,000 claimants enrolled in the settlement program, even though only about 9,600 cases were pending in the MDL at the time the settlement went into effect.[210]

MDL’s closure tools can reach beyond the jurisdictional limits of the MDL. MDL judges will often issue so-called “case census” orders requiring lawyers to identify all clients who have retained them in connection with the mass tort dispute—whether those clients have filed cases in the MDL or not.[211] This helps the parties identify the full universe of claims over which they are negotiating and sets the denominator for any settlement walkaway provision. Lawyer recommendation and withdrawal provisions apply equally to clients with state court and unfiled claims as they do to clients with pending actions in the MDL.[212] Plus, the releases obtained in MDL settlements are not limited to defendants who are named as parties in the MDL. Consensual releases in private settlement agreements can be as broad as the parties would like and can include third parties who have not been sued.

Nor is there any limit to the subject matter of claims that can be resolved in private MDL settlements. The National Prescription Opiate Distributors Settlement, for example, resolved governmental claims by state attorneys general who were never part of—and at times actively resisted—the MDL alongside the common law and federal statutory claims of the localities who had been litigating with private counsel in the MDL.[213] Thus, settlements negotiated within the MDL can and do resolve claims beyond the MDL court’s jurisdictional reach.

5.   Future Claims

Mass tort bankruptcy proponents often assert that only bankruptcy—not MDL—is capable of addressing the thorny problem of future claims.[214] Future claims and latent injuries complicate the resolution of mass torts in any forum.[215] And it is true that bankruptcy has tools for dealing with some kinds of future claims that MDL lacks.[216] But under the right circumstances, MDLs have been able to resolve mass torts involving latent injuries and future claims. None of MDL’s approaches to future claims are as complete as bankruptcy’s discharge and channeling injunction. And they do not address the problem of current claimants exhausting the defendant’s assets before future claims even arise. But they work well at providing certainty and predictability for solvent defendants who face the prospect of future liability.

MDL settlements have addressed future claims through a variety of approaches. One approach is aimed at a cash flow problem. In the asbestos litigation, many plaintiffs’ firms entered into structured settlement plans with asbestos manufacturers.[217] Under these arrangements, the defendant would agree to settle a certain number of the firm’s cases per year at amounts set forth in an agreed-upon grid, thus yielding predictability about ongoing asbestos liability that the defendant could plan and account for.[218]

Another approach to future claims involves pairing attractive upfront benefits like medical monitoring or interim payments with a back-end insurance component. The World Trade Center Disaster Site settlement, for example, paired modest, but quick interim payments for all first responders who had been exposed with an insurance policy that would pay if and when they developed cancer.[219] This strategy encourages exposure-only claimants to come forward and register with the settlement administrator for access to the benefits. And it channels future claims into the settlement process as injuries manifest.

Finally, parties in MDLs have used class action settlements to address future claims. While it is sometimes assumed that class actions involving future claims are off the table after Amchem and Ortiz, class action settlements can, in fact, successfully resolve future claims in mass tort cases if they are properly designed to avoid structural conflicts of interest between differently situated class members.[220]

In the NFL Players Concussion Litigation, for example, the MDL court certified and approved a class action settlement resolving the claims of more than 20,000 retired football players who had suffered concussions during their playing days and were at risk of developing neurological disorders.[221] Anticipating the future claims problem, the MDL judge appointed a member of the Plaintiffs’ Steering Committee as separate subclass counsel to represent the future claimants in negotiating the settlement.[222] The ultimate settlement paired upfront medical monitoring with a promise of back-end compensation from an injury trust fund if and when more serious injuries developed.[223] (Using a medical monitoring program as the upfront inducement to participate has the added benefit of mitigating harm if it leads to early detection and more effective treatment of latent diseases.) Only 202 of the approximately 20,000 class members (~1 percent) opted out of the class action settlement.[224] The Third Circuit affirmed, noting that thousands of exposure-only claimants were represented by hundreds of lawyers in the MDL, and “with so many sets of eyes reviewing the terms of the settlement . . . [we] have little problem saying that their interests were adequately represented.”[225]

C.   MDL’s Limitations in Mass Torts

The foregoing should not be read to suggest that MDLs are perfect. Mass torts present challenges in any forum. And MDL has its limitations as a procedural vehicle for delivering closure.

First, MDL is not comprehensive. The JPML’s power to transfer cases under § 1407 is limited to those filed in or removed to federal court.[226] Unlike bankruptcy, MDL has no mechanism to compel all persons who have a claim against the defendant to participate in a single proceeding. Tort victims may bring their claims in state court, or they may sit on the sidelines and watch the progress of the MDL before deciding if or where to sue. While federal MDL judges have a long and successful history of working cooperatively with state judges in handling parallel mass tort litigation,[227] there are also opportunities for friction between the state and federal proceedings. State judges may make different decisions about case management, discovery, speed of proceedings, or even substantive liability. And while MDL judges and lawyers have found ways to reach beyond the MDL’s jurisdictional bounds in crafting settlements,[228] there is no way to compel claimants to participate.

Second, MDL has no mechanism to bind dissenters. Unlike bankruptcy, participation in the ultimate resolution in an MDL is not mandatory; claimants cannot be bound without their consent. Indeed, unless a Rule 23(b)(3) class action is certified, MDL generally operates on an opt-in—not opt-out—model.[229] Claimants are only bound to a nonclass settlement negotiated in an MDL if they affirmatively opt in. This creates the potential for holdouts, who may prevent the plaintiffs from offering the defendant complete closure, even if the defendant would be willing to pay a peace premium to secure it.[230] While MDL dealmakers have crafted many contractual and case management tools for promoting closure, none are as powerful as bankruptcy’s ability to compel mandatory participation and override individual claimants’ refusal to consent.

Third, there are limits on MDL’s ability to resolve future claims. An MDL, by definition, encompasses only presently filed claims. The parties to an MDL can negotiate settlement structures that anticipate and account for future claims and create incentives for exposure-only claimants to opt in to arrangements that will compensate and bind them if and when their injuries manifest. But unless a class action is certified within the MDL, future claimants who do not opt in will not be bound. Class actions can shift the default rule and reach all future claimants who do not opt out, as in the NFL Concussion case.[231] But constructing an adequate class action settlement becomes considerably more difficult when, unlike NFL Concussion, the universe of future claimants is unknown. While Amchem and Ortiz do not rule out the use of class action settlements for unknown future claimants who may have latent injuries, certifying such classes has proven challenging.

And fourth, MDL has limited judicial supervision over settlements. Like any aggregate legal representation, MDL involves complicated principal-agent problems that are exacerbated by the number of claimants represented.[232] There is always a risk that the dealmaking lawyers will put their interests ahead of their clients’ in negotiating a settlement. Unlike bankruptcy, MDL includes no formal mechanism requiring judges to assess the fairness of settlements unless a class action is certified, though MDL judges can and do send informal signals about the fairness of settlements.[233]

Despite these limitations, MDL has been largely successful at resolving mass tort controversies. For the most part, parties in MDLs have been able to reach settlements that resolve claims en masse by channeling them into a streamlined and efficient process for doling out compensation. There is no evidence of thousands of duplicative trials or endless trial queues.[234] It is not obvious that costs and delays are greater in MDL than in bankruptcy. Indeed, many mass tort bankruptcies have stretched on for years and generated substantial costs.[235]

But even if bankruptcy is more efficient, and there are gains to be had from shifting mass tort cases out of MDL and into bankruptcy, efficiency is not enough. Recall that the justification for bankruptcy’s ability to override individual claimants’ rights—the very rights that make complete aggregation in MDL so challenging—is that doing so benefits the creditors.[236] It is the creditors’ bargain, not the debtor’s. So, the relevant question is whether tort claimants share in those gains. And the answer depends on the bargaining dynamics in the two systems.

III. Bargaining Dynamics in Bankruptcy and MDL

If tort victims are going to share in any surplus generated by bankruptcy’s efficiencies and ability to deliver closure, they need the leverage to bargain for it. But several features of bankruptcy shift bargaining leverage from plaintiffs to defendants as compared to MDL.

A.   The Extra Step in the Texas Two-Step

As noted above, the Texas Two-Step has made mass tort bankruptcy more attractive to solvent companies. Defenders of the Two-Step say that it leaves tort plaintiffs no worse off—indeed, they may be better off because they too can share in the reduced costs of resolving their claims in bankruptcy.[237] A valid Two-Step should not reduce the funds available to pay tort claimants because it includes a funding agreement under which GoodCo will pay all of BadCo’s tort liability up to the value of the legacy company.[238] If the legacy company had enough money to satisfy all its tort liability before the divisional merger, then those resources should still be available to pay tort claimants in full afterwards.[239] A Two-Step that really did reduce the funds available to tort claimants would be a fraudulent transfer.[240]

But what defenders of the Texas Two-Step miss is that it throws another obstacle in plaintiffs’ way. If BadCo is undercapitalized, the tort claimants will have to litigate and win a fraudulent transfer action in addition to litigating their claims on the merits. And the tort claimants may not be the ones deciding whether and on what terms to pursue a fraudulent transfer claim—at least in the first instance.

Fraudulent transfer claims belong to the bankruptcy estate, not the individual creditors.[241] As a result, the managers of BadCo—who will retain control over the company during the course of the reorganization in bankruptcy as a debtor in possession—will have the first crack at whether to bring a fraudulent transfer claim.[242] And the insiders managing BadCo—who are typically employees of GoodCo seconded to manage the company through bankruptcy—are unlikely to aggressively pursue fraudulent transfer claims against themselves and their employers.[243]

This reality may embolden insiders at the defendant to push the envelope on the funding agreement. As Casey and Macey recognize, the defendant already has superior information about the value of the firm’s assets and may use that asymmetry to allocate assets that are worth relatively little to tort claimants in ways that are opaque to a reviewing court.[244] Or insiders may divert value away from tort claimants openly. For example, in the J&J Texas Two-Step, the managers of LTL Management (the BadCo) allowed Johnson & Johnson Consumer, Inc., (the GoodCo) to spin off its most valuable business unit without objection. Then, after the Third Circuit dismissed the initial bankruptcy, LTL gave up its own most valuable asset—a $61 billion funding guarantee from the J&J parent corporation that the Third Circuit described as “an ATM disguised as a contract”—for essentially no consideration.[245]

Even if the creditors committee representing the tort claimants obtains derivative standing to pursue the fraudulent transfer claim and prevails on that claim, the remedy would be avoidance or rescission of the transfer.[246] That would simply leave the claimants to pursue the merits of their claims against GoodCo in the bankruptcy, not result in its dismissal. This extra step on plaintiffs’ path to recovery shifts bargaining leverage in defendants’ favor.

B.   Automatic Stay of All Litigation

Filing under Chapter 11 automatically stays all litigation against the debtor.[247] And bankruptcy courts are often willing to enter preliminary injunctions, effectively extending the reach of that stay to the debtor’s affiliated companies.[248] When the debtor is insolvent, the stay of litigation is essential to preventing a destructive race to the courthouse.[249] Even when there is enough money to go around, the pause in litigation and centralization of control in the bankruptcy court can avoid duplicative litigation and give the parties some breathing room to negotiate a consensual resolution.

But the automatic stay also takes away plaintiffs’ biggest source of leverage in mass tort litigation—the threat of jury trials and their verdicts.

Although mass tort litigation nearly always ends in settlement, bargaining over the terms of those settlements takes place in the shadow of litigation. And plaintiffs’ ability to push their cases forward toward trial is their biggest stick. The automatic stay takes that stick away and places total control over the pace of the litigation in the hands of the bankruptcy judge.[250]

MDLs have been subject to the same criticism. Plaintiffs’ lawyers often complain that transfer into an MDL derails their attempts to get their cases to trial quickly. MDL also places tremendous control over the pace of litigation in the hands of the MDL judge, who has broad discretion over case management decisions, like which discovery or motions to allow and which to put on hold.[251] This can operate as a de facto stay on the federal litigation.[252] Moreover, under Lexecon, the MDL judge lacks the power to try transferred cases.[253] Trial cannot occur until after remand, and the JPML’s ordinary (and so far unwavering) practice is not to remand cases until the MDL judge suggests it.[254] In other words, cases are stuck in the MDL until the MDL judge lets them go. Critics contend that MDL judges sometimes refuse to remand cases in an attempt to force the parties to settle.[255] This, some argue, leaves plaintiffs in the MDL disarmed in negotiations because they cannot credibly threaten trial.[256]

But bankruptcy’s limitation on plaintiffs’ ability to threaten trial is far more complete than MDL’s. While MDL judges have near-complete control over the pace of the federal litigation, they do not control the satellite litigation in state courts. MDL judges often coordinate with state court judges on scheduling and case management, but that coordination is necessarily partial. If an MDL judge moves too slowly or withholds adjudication, plaintiffs can ask a different judge in a separate court system to set an early trial date.[257] And indeed, the sometimes-faster pace of litigation in state court can provide an external push to the MDL and allow plaintiffs to maintain the threat of trial. In bankruptcy, there is no such external push as the bankruptcy judge has complete control over the pace of all claims.

And while MDL judges cannot try transferred cases without the parties’ consent, unlike bankruptcy judges, they do have the authority to try cases that have been filed directly in the MDL district.[258] MDL judges frequently use that power to conduct bellwether trials in the MDL.[259] While the primary purpose of bellwether trials is to inform settlement negotiations, they result in real judgments that impose real obligations on the defendant to pay up.[260] Bankruptcy judges, by contrast, lack the statutory authority to try personal injury cases.[261] A bankruptcy judge could conceivably seek the cooperation of an Article III district judge to hold a jury trial or selectively lift the stay on one or more cases to permit trial to go forward in state court.[262] Or the bankruptcy court could conduct a bellwether-trial-like proceeding to help estimate the value of claims for purposes of voting and plan confirmation.[263] But such efforts appear to be exceedingly rare.[264] And even if a plaintiff were to win such a bellwether trial in bankruptcy, any payment would likely be delayed until after a plan of reorganization is confirmed.[265]

In the absence of a credible threat of trial, one of tort claimants’ biggest sources of leverage in a mass tort bankruptcy is the defendant/debtor’s desire to get out of bankruptcy as soon as possible. As noted above, bankruptcy can be a painful experience for a debtor and its management. Bankruptcy is costly in terms of professional fees, and it can be devastating in terms of operations and market perceptions. Some have argued that this is part of the bankruptcy bargain; debtors only get a release after they have subjected themselves to the procedural rigors of bankruptcy.[266] I am not sadistic; I have no desire to punish debtors with the pain and expense of bankruptcy disclosures, disruption, and judicial oversight for their own sake. But the prospect that tort claimants will vote down a plan of reorganization and prolong the pain can be a powerful source of leverage. And the debtor—particularly a solvent debtor whose managers still owe a duty of loyalty to the shareholders—will have an incentive to put enough money on the table to secure the tort claimants’ votes and end the ordeal.[267]

In a Texas Two-Step bankruptcy, however, there is no pain to prolong. The debtor, BadCo, does not have any business operations to speak of. It can stay in bankruptcy indefinitely, while tort claimants vote down plan after plan, without inflicting any pain on GoodCo. Tort claimants, by contrast, need to bring the bankruptcy to a close before they see any money. And their contingency-fee lawyers can only hold out so long before they need to see some return on their investment.

In a Two-Step bankruptcy, the automatic stay, combined with the debtor-in-possession’s exclusive right to propose a plan of reorganization, shifts significant leverage from mass tort plaintiffs to defendants.[268] This combination results in an extended pause of at least four, and up to eighteen, months, during which the debtor can make a take-it-or-leave-it offer to tort victims whose rejection will inflict little pain in response.[269] Even after the exclusivity period has expired and creditors can propose their own plans, the process can drag on for years.[270] Professional fees in bankruptcy can be high, but they are easily dwarfed by the money the defendant is not spending on litigation costs and judgments in the tort system.[271]

Casey and Macey suggest that the Texas Two-Step actually benefits tort claimants because it prevents other, more sophisticated creditors from opportunistically holding up the bankruptcy to extract value at their expense.[272] Thus, by limiting the scope of the bankruptcy, the Texas Two-Step increases tort creditors’ leverage vis-à-vis other creditors. Perhaps this is a benefit (though I fail to see how the collective action problem motivating the bankruptcy can be so easily cabined to a single class of creditors[273]), but it ignores the possibility that the debtor will manipulate the process to use delay to its advantage, thus decreasing tort claimants’ leverage vis-à-vis the debtor.

C.   Voting

Voting is tort claimants’ main source of leverage in a mass tort bankruptcy.[274] To confirm a plan of reorganization, the debtor must propose a deal that is attractive enough to persuade the creditors to vote for it. Voting in bankruptcy proceeds on both a per capita and pro rata basis. The plan must win approval from a simple majority of creditors in each impaired class by number and a two-thirds majority by claim value.[275] In addition, if the plan employs a § 524(g) asbestos trust and channeling injunction, it must be approved by a 75 percent supermajority of tort claimants.[276] The threat to vote down an inadequate plan gives the tort claimants leverage to convince the defendant to put sufficient money on the table.

In theory, voting can be an excellent way to disable holdouts without undermining tort claimants’ collective leverage vis-à-vis the defendant. Indeed, I have advocated for incorporating voting into aggregate litigation in the past.[277] Under the right circumstances, voting can be a better way to preserve a group’s leverage than allowing individuals to exit.[278] Binding dissenters to the results of the vote prevents a single holdout from blowing up the deal, allowing the group to effectively negotiate for a peace premium. At the same time, the right to vote against the deal enables tort claimants to voice their opposition if they think the deal is inadequate.[279] As Casey and Macey correctly point out, the stakes of voting against a plan of reorganization are much lower than the stakes of opting out of a group settlement in an MDL.[280] Dissenting claimants can still participate in the deal, and share in any collective benefit it provides, if they are outvoted. In an MDL settlement (class or nonclass), by contrast, a claimant who wants to register dissent must be willing to reject the deal, exit the group, and litigate alone.[281] By lowering the stakes of dissenting, a voting scheme makes it easier for claimants to express their true preferences on the deal, which, in theory, should make it harder for the debtor to foist a lowball settlement on the group.[282]

While there have been experiments with similar voting mechanisms in MDLs (inspired in part by § 524(g) of the Bankruptcy Code), for the most part, they have not caught on. In its Principles of the Law of Aggregate Litigation project, the American Law Institute (ALI) advocated altering the traditional aggregate settlement rule to allow groups of clients sharing the same lawyer to precommit to be bound by a supermajority vote on settlement offers.[283] But so far only one state (West Virginia) has amended its ethics rule to adopt the proposal, and several others have rejected it.[284] Similarly, in the opioid litigation, the MDL judge certified a novel “negotiation class,” not for the purpose of litigation, but solely for the purpose of negotiating a settlement with the defendants and distributing its proceeds according to a predetermined formula.[285] Class members who did not opt out at the outset would be bound by a supermajority vote on whether to accept any deal. The Sixth Circuit, however, in a 2-1 ruling, rejected the negotiation class as beyond the scope of Rule 23.[286]

Although they have not been widely adopted, both the ALI proposal and the negotiation class recognize voting’s potential to enable tort claimants to overcome the collective action problem they face and offer the defendant peace in exchange for a premium. In short, voting schemes could be very attractive in mass tort resolution.

But—and this is the central theme of the entire field of election law—the outcome and legitimacy of any vote is highly dependent on the process of voting.[287] Unfortunately, the process of voting in mass tort bankruptcies leaves much to be desired. Here, I focus on three problems: claimants’ participation in the vote, allocation of voting rights, and control over the terms of aggregation.[288]

1.   Participation in the Vote.

For voting to legitimize binding the entire class of tort claimants to the terms of a reorganization—and to preserve their leverage in negotiating with the debtor—the results of the vote must accurately reflect their preferences. But there are reasons to doubt that an affirmative vote on a plan actually signals that the requisite majority of tort claimants consented to its terms.

As many critics note, votes on mass tort reorganization plans are generally low-turnout affairs.[289] Low turnout certainly complicates the narrative that a plan has received “overwhelming support” from 95 percent of tort victims, if only 20 percent of them actually voted.[290] I am less concerned than some about low turnout per se. After all, low turnout is a persistent feature of most political elections as well. I see no fundamental problem with binding nonvoters to the result of the vote without their affirmative consent so long as two conditions are met: (1) nonvoters had a fair opportunity to participate in the process and (2) the voters who turn out are likely to make up a reasonably representative sample of the whole. But mass tort bankruptcies often fall short on both.

To have a fair opportunity to participate in the vote, tort claimants need notice. The Bankruptcy Code requires the debtor to provide creditors with a disclosure statement containing “adequate information” to enable each creditor to make “an informed judgment.”[291] But providing adequate notice to mass tort victims can be challenging, and those challenges do not go away in bankruptcy. While it may be relatively straightforward to identify and send notices to tort claimants with pending litigation against the defendant, notifying other potential claimants who have been injured but not yet sued (let alone exposure-only claimants, but more on that below) may be far more difficult, yet they are no less entitled to vote on a plan that will determine their rights.[292]

Outside of bankruptcy, courts have refused to certify mass tort class actions where giving adequate notice to absentees appears impracticable.[293] And nothing about the bankruptcy process makes it any easier to notify mass tort victims. Due process, of course, requires balancing the benefits and burdens of procedural protections, and shortcuts on notice might be justified in bankruptcy if the alternative to reorganization is pennies on the dollar in a liquidation.[294] But when there is enough money to go around, and the alternative is not liquidation but continued litigation and potential settlement in an MDL, shortcuts on constitutionally required notice are much harder to justify.

Even among those who vote, bankruptcy has fewer guarantees than MDL that the vote tally accurately reflects their preferences. The actual voting on a plan of reorganization is often done by lawyers for the tort claimants, not the claimants directly.[295] These lawyers will typically receive “master ballots” allowing them to vote their clients’ claims en masse.[296] This practice, which is authorized by Federal Rule of Bankruptcy Procedure 3018(c), is sensible in many ways. Tort claimants will typically need legal advice in order to vote intelligently. And a lawyer who controls a bloc of votes will have leverage in negotiating with the debtor over the terms of the plan.

But when lawyers vote their clients’ claims by master ballot, there are fewer guarantees that the clients’ consent is genuine. Look no further than the dispute over ballots in J&J’s third talc bankruptcy.[297] There, one law firm notified all of its clients that it would vote their claims against the plan unless they instructed the firm otherwise. Only a handful did so, and the firm voted the remaining approximately 11,500 claims against the plan.[298] Subsequently, a second law firm, which had been co-counsel with the first before a falling out, voted the same approximately 11,500 clients in favor of the plan pursuant to a claimed power of attorney.[299] The bankruptcy court ultimately found problems with how both sets of votes were obtained and dismissed the bankruptcy in part because of “significant voting and solicitation irregularities that make it impossible to certify the vote.”[300] But the court did not disapprove of the use of master ballots.[301] And the conflict in the vote tally highlights the risk that tort claimants may not be making their own decisions on whether to accept a plan of reorganization.[302]

In MDL, by contrast, the authority to accept or reject a settlement remains firmly with the client. The aggregate settlement rule requires lawyers to obtain each client’s affirmative consent to participate in a settlement after disclosing all of its terms.[303] And courts have largely rejected attempts to delegate settlement authority to lawyers ex ante.[304] In short, MDL does more to ensure that tort claimants themselves have the opportunity to decide whether a settlement is in their interests and that claimants who sign onto a deal really favor it.

2.   Allocation of Voting Rights.

Even if claimants have a fair opportunity to participate, and the vote tally accurately reflects the preferences of those who vote, we still need to be concerned with whether the results of the vote are representative of the strength and variety of tort claims against the defendant/debtor. And that depends in large part on how voting rights are allocated.

Voting works best when the interests within a class of voters are relatively uniform. Where a class of voters is heterogeneous—where interests within the group diverge—we need to worry that the majority will vote benefits to itself and costs to the minority.[305] This is the classic tyranny-of-the-majority problem recognized in contexts from corporate law to constitutional law.[306] Bankruptcy attempts to deal with the tyranny-of-the-majority problem by placing creditors into classes of substantially similar claims and by requiring affirmative votes by both number of claims and claim value.[307] Unfortunately, neither mechanism is particularly good at dealing with heterogeneity among mass tort claimants. Because it is difficult to figure out the value of unliquidated mass tort claims, bankruptcy courts often take shortcuts on both classifying and valuing tort claims for voting purposes. And those shortcuts can attract mass filing of meritless claims and lead to undercompensation for the strongest claims.

In a mass tort, the value of victims’ claims can vary widely along many dimensions. Some may be suffering serious injuries while others have only minor impairments; some may have strong evidence of causation while others may struggle to prove the connection between the defendant’s conduct and their specific injuries; some may face affirmative defenses like assumption of the risk or statute of limitations while others may not; many different states’ substantive law may apply to different claimants; and so on. Indeed, these sorts of variables typically factor into negotiations over the grid or point system that will determine how compensation is ultimately handed out in a mass settlement or plan of distribution.

But figuring out the valuation of each tort claim can be difficult and costly, particularly in an immature mass tort. As a result, bankruptcy courts tend to lump all tort claimants into a single creditor class[308] and estimate the value of their claims uniformly at one dollar for voting purposes.[309] In some asbestos bankruptcies, courts have attempted to divide tort claimants into rough categories by type of injury (e.g., mesothelioma, lung cancer, asbestosis) for voting purposes and assign estimated values to each category based on historical settlement amounts.[310] But asbestos is a very mature tort where liability and general causation are rarely contested, and different disease categories have well-established going rates in the tort litigation system. Similar data is not readily available for less mature torts.[311] Outside of the asbestos context, I have been unable to locate a mass tort bankruptcy that used this sort of “category class” method of estimation. Instead, “one-dollar-one-vote” is by far the most popular method for estimating mass tort claims for voting purposes.[312]

The one-dollar-one-vote method of estimation essentially collapses the two-thirds majority by value requirement into a per capita vote.[313] Bankruptcy courts tend to dismiss this concern by pointing to overwhelming votes approving plans of reorganization and asserting that any failure to more accurately estimate claim values would be harmless error.[314] Doing so, however, creates a serious risk of undervaluing the strongest claims.[315] When voting rights are allocated uniformly, winning plan approval becomes a simple numbers game. And if relatively weaker claims outnumber stronger claims, the plan of distribution will need to be designed to appeal to the majority, whose claims may not have much value outside of bankruptcy.[316] This risks seriously undercompensating tort victims who have suffered the worst injuries.

Even worse, uniform valuation of claims for voting purposes invites the filing of meritless claims in bankruptcy. Defendants often complain about a “Field of Dreams” problem in MDLs—if you build it, they will come.[317] By lowering the costs of litigating, the story goes, MDL encourages plaintiffs’ lawyers to flood the zone with meritless claims in hopes that they will be swept into some global settlement without anyone ever looking closely at them.[318] There are many reasons to question this narrative, but, fundamentally, weak or meritless claims are unlikely to attract more than a nuisance-value settlement in MDL.[319] The threat of victory at trial is plaintiffs’ main source of leverage in multidistrict litigation, and meritless claims have very little chance of prevailing at trial. Most can be screened out through plaintiff fact sheets or settlement eligibility criteria. And MDL settlements are typically designed to be attractive to plaintiffs with the strongest claims—the ones who present a real threat at trial.

But in bankruptcy, where voting, not the threat of trial, is the tort claimants’ source of leverage, the Field of Dreams problem can be serious. If voting rights are handed out uniformly without regard to claim value, a meritless claim is worth just as much as a bulletproof one. Claims that would have had no value in the tort system may flood into bankruptcy.[320] Some lawyers may recruit numerous weak or meritless claims to file in the bankruptcy, giving themselves control of a large voting bloc in negotiations.[321] They may then offer these votes to the debtor in support of a lowball plan of reorganization that pays claims that would be worthless outside of bankruptcy, but systematically undercompensates claimants with serious injuries. Indeed, the defendant may actively solicit the cooperation of plaintiffs’ lawyers to drum up weak claims and enter “plan support agreements” with those lawyers to vote in favor of a plan opposed by stronger claims.[322] Unlike an MDL, where the defendant has every incentive to try to screen out as many claims as possible, once in bankruptcy, the debtor might welcome the deluge and work with a subset of plaintiffs’ counsel to design voting procedures that minimize screening of predictably friendly votes.[323] Combined with low turnout, this risk of bloc voting by weak claims makes the conclusion that all is well because of overwhelming votes in support of the plan suspect.[324]

Bankruptcy courts might try to prevent mass tort bankruptcies from becoming magnets for meritless claims by including merits-based questions on proof-of-claim forms and requiring supporting documentation, much like plaintiff fact sheets in MDL, before claimants are allowed to vote.[325] There is always a tradeoff with such procedures: the more detail required, the more it will cost to review the results. Still, well-designed proof-of-claim forms might succeed in screening out frivolous claims (e.g., those where the claimant was never exposed or has suffered no injury) before they are allowed to vote.[326] But they will be less effective at distinguishing between relatively weak but nonfrivolous claims and relatively strong claims. And that is the distinction that matters. In MDL, a plaintiff who can clear a fact-sheet screen still requires a credible threat of prevailing at trial to have much leverage. In bankruptcy, clearing the proof-of-claim screen equals the same voting power as every other tort claimant.

Because voting is the source of tort claimants’ leverage in bankruptcy, a misallocation of voting power creates a real risk of tyranny of the majority. The masses with weak claims may vote in favor of a plan that undercompensates claimants with the strongest claims. And the concern goes deeper than just the exploitation of the most severely injured victims. If the voting majority holds claims that have no value outside of bankruptcy, then they may be willing to accept a plan of reorganization that gives them something while substantially underpaying the tort claimants as a whole. In short, allocating disproportionate voting power to weak claims shifts leverage in the defendants’ favor.

3.   Control Over the Terms of Aggregation.

Finally, bankruptcy’s voting rules shift leverage away from tort claimants by placing control over the terms of aggregation firmly in the debtor’s hands. In MDL, aggregate resolution requires the consent of all claimants; if claimants aggregate, they do so on terms that benefit them. In bankruptcy, by contrast, the defendant can unilaterally impose aggregation on the claimants. This matters because control over the terms of aggregation is a source of leverage in capturing the surplus that aggregation creates.

This is perhaps easiest to see by comparing bankruptcy to the negotiation class. Closure creates value, but in order to achieve closure in an MDL, the defendant has to share enough of the resulting surplus with the claimants to get them all on board.[327] Holdouts may make achieving closure challenging, but the power to hold out is also the power to avoid being exploited.[328] Bankruptcy solves the holdout problem by allowing the defendant to impose a supermajority voting rule on the claimants. But that means the defendant only needs to share enough of the surplus to get 75 percent of the claimants on board; it can keep the rest for itself. And, as discussed above, there are ample opportunities for the defendant/debtor to manipulate the voting and allocation processes to shortchange the strongest claims.

Contrast voting in bankruptcy with voting in the negotiation class, where the aggregation occurs on the plaintiffs’ terms, not the defendant’s. The idea behind the negotiation class was to create an entity in advance of settlement negotiations that could credibly offer the defendant peace in exchange for a premium.[329] But because claimants have an opportunity to opt out of the negotiation class at the outset, the allocation formula and voting procedures have to be designed in a way that all claimants find acceptable.[330] It cannot shortchange the strongest claims without risking their defection and defeating the whole endeavor. The same is true of the ALI’s proposal to allow groups of claimants to precommit to voting on aggregate settlement offers.[331] Claimants’ consent on the front end is an important guarantor of fair process and distribution on the back end.[332] By binding themselves to group decision-making on their own terms, claimants maximize their collective leverage against the defendant. That is one reason defendants in the opioid litigation were so opposed to the negotiation class.

Without voting in MDL, aggregation happens in plaintiffs’ lawyers’ offices. And, as discussed above, in order to be able to offer the defendant peace, lawyers often incorporate closure provisions designed to get clients to sign onto the deal into MDL settlement agreements.[333] Many of these terms, especially those requiring lawyers to withdraw from representing nonsettling claimants, have rightly been criticized as coercive.[334] Indeed, Casey and Macey point to such closure provisions as a major disadvantage of MDL.[335] Exactly how coercive these closure provisions are depends on their details and on the shape of the market for legal services in the particular mass tort.[336] But some closure provisions can, indeed, be quite powerful.

Two points about MDL closure provisions are worth emphasizing in contrast to mass tort bankruptcies. First, MDL dealmakers include these sorts of terms in settlements because they need claimants’ consent to the settlement. Otherwise, claimants cannot be bound. Achieving closure in an MDL—whether through an opt-in model like a master settlement agreement or an opt-out model like a class action—ultimately requires getting all of the plaintiffs onboard with the deal. This means that MDL settlements must be designed to be attractive to the plaintiffs with the strongest claims and the lawyers who represent them. And the aggregate settlement rule ensures that claimants with strong claims have the information they need to evaluate how well they are being treated relative to other plaintiffs.[337] Closure provisions can put pressure on claimants to accept, but a determined plaintiff with a strong claim is always free to reject the deal and press ahead towards trial.[338] If too many plaintiffs with strong claims walk away, the deal will collapse. So, to achieve closure in an MDL, the parties have to build consensus among the major players on the plaintiffs’ side who represent the claimants with the strongest claims. Contrast this dynamic with bankruptcy, where voting rules give the weakest claims the same leverage as the strongest.

Second, it is plaintiffs’ lawyers who are crafting these closure provisions that pressure their clients to settle. For MDL critics, this betrayal is the worst part. Tort victims are twice victimized, first by the defendant and then again by their own lawyers.[339] But this critique misses a fundamental reality of mass litigation: It is better for plaintiffs to have their allies designing closure tools than their adversaries.

Closure in mass litigation creates value. And the party that can aggregate claims to effectively deliver closure is in the best position to capture the surplus it creates. In mass tort MDLs, plaintiffs’ lawyers are the primary aggregators. They are the ones who bundle claims together and negotiate group settlements with the defendant. If including closure provisions in their MDL settlements enables them to deliver all of their clients’ claims, they can extract a peace premium in exchange. In mass tort bankruptcy, by contrast, the defendant can unilaterally impose aggregation on the claimants by filing for Chapter 11.

Empowering plaintiffs’ lawyers in this way comes with risks, of course. Agency costs in litigation—particularly aggregate litigation—are inevitable. The plaintiffs’ lawyers may be tempted to keep a share—perhaps a large share—of the surplus for themselves.[340] Surely there is room for improvement in MDL.[341]

But plaintiffs also benefit from having powerful champions on their side.[342] And the contingency fee goes a long way toward aligning the lawyer’s incentives with those of the claimants—the more money lawyers get for claimants, the more they get for themselves. At bottom, plaintiffs are likely to do better if their own agents have the leverage that comes with the power to aggregate than if their adversaries have that power. Indeed, real-world experience bears this out. When defendants in 3M and Purdue Pharma lost the ability to unilaterally impose aggregation through bankruptcy, settlement amounts went up.[343]

D.   Future Claims

Mass tort bankruptcy’s ability to resolve future claims is held out as a major advantage over MDL. Indeed, some have gone so far as to say that, after Amchem and Ortiz, only bankruptcy can handle future claims in mass torts.[344] That claim is false. Mass torts with future claims components have been successfully resolved through settlements negotiated in MDLs.[345] But it is certainly true that bankruptcy has been used more frequently to resolve future claims, particularly where the identity of potential future claimants is unknowable, and continued litigation may exhaust the defendant’s resources before future claimants have a chance to recover.

Mass torts involving latent injuries and unknown future claimants present challenges in any procedural system. Simply shifting those claims from MDL to bankruptcy does not eliminate those challenges. Indeed, bankruptcy assumes away many of the due process problems that plagued the mass tort class action in Amchem and Ortiz. For example, as described above, bankruptcy has no better tools for dealing with structural conflicts of interest that may arise from differences among various tort victims’ claims.[346] And the challenge of providing constitutionally adequate notice to a universe of unknown potential tort claimants is no different in bankruptcy than it is in class actions.[347]

Structural protections for future claimants are weak in bankruptcy. Future claimants do not get to vote on the plan of reorganization.[348] They are not protected by the rule of absolute priority or the best interests of the creditors test.[349] Bankruptcy relies primarily on a court-appointed future claims representative to look out for the interests of future claimants in mass tort bankruptcies. Criticisms of future claims representatives are well known, and I will not rehash them all here.[350] But I do want to highlight a few features in comparison to MDL. Most importantly, the future claims representative lacks the proper incentives and leverage needed to maximize the future claimants’ recovery. The court-appointed future claims representative in bankruptcy is paid hourly, not based on how well the future claimants do.[351]

Contrast this arrangement with a properly constructed class action in an MDL that includes subclasses for present and future claimants with separate representation to provide structural assurances of adequate representation.[352] In a class action, the fees awarded to the lawyer for the subclass of future claimants will be determined based on how much future claimants will recover under the settlement. This gives the class action lawyer a powerful economic incentive to negotiate to maximize the resources devoted to future claims. And without the future subclass counsel’s assent, no deal will go through.

The future claims representative in bankruptcy has no power to block a plan that shortchanges future claimants. The future claims representative gets neither a veto nor even a vote on plan confirmation.[353] The representative can only influence the terms of the plan through persuasion.[354] Although a bankruptcy judge may be reluctant to confirm the plan over the representative’s objection, without clients, skin in the game, or a source of real leverage, the future claims representative’s power in negotiations may be limited.[355] Indeed, the future claims representative might be more focused on building a reputation as an “honest broker” and laying the groundwork for future appointments than on maximizing the future claimants’ recovery.[356] Of course, there are concerns about repeat players in MDL too.[357] But class counsel or MDL lawyers’ temptation to go along to get along with an eye toward future engagements is always tempered by their profit motive in maximizing the recovery in their contingent-fee representations.[358]

Indeed, bankruptcy’s treatment of future claims appears inferior to the class action’s treatment of them on almost every metric, except that courts tolerate it.[359] That toleration is based, in large part, on the presumption that there is not enough money to go around.[360] It is not difficult to conclude that even an imperfect process leading to a plan of reorganization that pays future claimants something is better for them than a liquidation that pays future claimants nothing.[361] Similarly, imperfect treatment of future claims in bankruptcy is better for future claimants than allowing current claimants to bleed the defendant dry in multidistrict litigation. But when there is enough money to go around—and the justification for bankruptcy is not the necessity to equitably distribute a limited fund, but rather its efficiency and potential to generate a peace premium—overriding future claimants’ preexisting rights without structural assurances that they are adequately represented becomes much harder to justify.

To state the obvious, it is not imperative that every mass tort be resolved for all time through a single comprehensive deal. There is nothing fundamentally wrong with a solvent defendant resolving current claims with current claimants and then facing litigation against future claimants in the future. As the Supreme Court said in Harrington, bankruptcy courts do not have “a roving commission” to solve all collective action problems wherever they find them “blind to the role other mechanisms (legislation, class actions, multi-district litigation, consensual settlements, among others) play in addressing them.”[362]

Even assuming that a mass tort bankruptcy creates value, without leverage or a properly incentivized representative, there is little reason to think that future tort claimants will be able to extract more out of a solvent defendant in bankruptcy than in MDL. When the defendant has enough money to pay all claims as they become due, future claimants are likely to do better by waiting to sue if and when their injuries manifest than by taking what they can get in bankruptcy.

Bankruptcy is not about protecting debtors; it is about maximizing value for creditors. If the defendant sees value in closure—and it almost certainly will—it can craft a settlement program that provides enough up-front benefits to future claimants (like medical monitoring or an insurance component) such that future claimants will want to participate in it.[363] Or it can attempt to negotiate a settlement with a properly constructed class under Rule 23 with properly incentivized class counsel for future claimants, as in NFL Concussion.[364] Either way, the defendant will probably end up putting more money on the table. But the defendant has no right to closure. If it wants peace, it must share the premium that results.

E.   Valuation

The predictable response to many of the concerns discussed above is that they all stem from a problem of valuation.[365] If we can just get the valuation right, everything else will fall into place. If we can get the valuation of the company’s assets right, then we can be sure the Texas Two-Step is not a fraudulent transfer. If we can get the valuation of each tort claim right, then we can properly allocate voting rights. If we can get the valuation of the defendant’s total liability right, then we can be sure that there will be enough money in the trust to pay current and future claims equitably—or completely when there is enough money to go around. So long as we can get the valuation right, we can design the plan of distribution to ensure that tort claimants do as well, or better, in bankruptcy than outside it.

Indeed. Valuation is the whole ballgame. But there is little reason to think that bankruptcy is more likely to get the valuation of tort claims right than the tort litigation system itself.

Mass tort bankruptcy proponents often contend that valuation of tort claims in bankruptcy will be more efficient, consistent, and fair than “lottery-like” jury verdicts based on fifty different sets of disparate state law.[366] And they point to the claims-estimation process as a valuable tool for efficient valuation.[367] So, it is worth pausing to ask: What exactly are we trying to value?

The underlying rights that mass tort claimants hold are tort claims created by state substantive law and enforceable through state and federal procedural rules in accordance with the Erie doctrine and constitutional rights to due process and a jury trial.[368] We might think that establishing a uniform nationwide grid like a workers’ compensation system would be a more predictable, equitable, and efficient system for determining tort liability and distributing compensation.[369] But bankruptcy has no warrant to change the preexisting substantive rights of creditors just because we do not like the way they are allocated.[370] The rights-holders themselves—the tort claimants—have the authority to modify their rights by agreement with the defendant.[371] And they might very often find it in their interests to do so by entering into settlements that avoid the cost and risk of valuing their claims through the jury trial system. But that is their prerogative, not the bankruptcy judge’s.

So, at bottom, when a bankruptcy court estimates the value of mass tort claims, it is supposed to be estimating their expected value in the tort litigation system, with all its variations in substantive law, procedural law, regional biases, and other warts. Determining the valuation of unliquidated tort claims is challenging to say the least. It is also costly and time consuming. But there is little reason to think that bankruptcy judges are better at this exercise than MDL judges, or that reaching accurate valuations is any quicker or less costly in a bankruptcy court than in an MDL.

In MDL, the valuation of tort claims is an iterative process that plays out over many steps as the tort matures.[372] Cases are filed and transferred. Pleadings are drafted, amended, and tested through motion practice. The parties exchange information in discovery. Fact sheets are distributed, answered, and compiled to construct databases that detail variations among claims and group them into categories. Expert witnesses are hired, deposed, and asked to participate in “science days” to educate the judge and parties. Summary judgment motions are filed, argued, and decided. Theories of liability are developed, tested, and refined. Test cases are selected for individualized discovery and potential bellwether treatment. Bellwether cases are tried and decided by juries or settled on the courthouse steps. Parallel cases proceed in state court with partial coordination and partial independence, sometimes moving faster or slower than the MDL. And all the while, the parties are engaged in negotiation, constantly incorporating new information and revising their estimates of claim value. This is, in a very real sense, the goal of an MDL (and litigation more broadly): to facilitate the exchange of information until the parties’ estimates of claim valuation converge enough that they can reach a consensual resolution.[373]

Valuation of unliquidated tort claims in bankruptcy presents serious challenges for both estimation and distribution. Those challenges are made more complicated by Code provisions that deprive the bankruptcy court of the authority to adjudicate personal injury claims; personal injury claims must be tried in an Article III (or state) court with the right to a jury trial preserved.[374]

1.   Valuation for Estimation.

Although they cannot adjudicate personal injury claims for the purposes of distribution, bankruptcy courts are empowered to “estimate” the value of tort claims for purposes of allocating voting rights and informing the negotiation and confirmation of a plan of reorganization.[375] That plan will channel all current and future tort claims to a settlement trust for payment. And the resulting estimation of aggregate liability may be used to determine how much money the debtor will contribute to the trust; that is, it will set the maximum aggregate payment to all tort claimants if the plan is confirmed.[376]

The Bankruptcy Code does not provide any specific procedures for estimating mass tort claims, so bankruptcy judges are largely left to figure it out on their own. In practice, this usually devolves into a “battle of the experts” with each side hiring economists who attempt to predict the aggregate amount of the defendant’s liability based on historical settlements and judgments in the state and federal courts and projections about the total number of current and future claims.[377] The first wave of mass tort bankruptcies mostly involved asbestos, a very mature mass tort. Asbestos liability was usually not contested, and claim values for various disease types were fairly well known based on years of litigation and settlement in the tort system.[378] But in less mature torts, the challenges of valuation are substantial.[379]

Unsurprisingly, bankruptcy judges prefer that the debtor and various creditors’ committees reach a negotiated resolution on the top-line number (which, as a practical matter, will be used to set the defendant’s maximum liability).[380] Accordingly, bankruptcy judges may direct the parties to mediation.[381] If the parties cannot agree—or need more information, as in immature torts—the bankruptcy court can conduct adversarial estimation proceedings.[382] But, as Professor Samir Parikh explains, “the idea that the bankruptcy court will, after a few days of hearings, estimate thousands of victims’ claims totaling hundreds of millions of dollars is arguably absurd.”[383]

While some have suggested conducting bellwether trials to more comprehensively inform the estimation process, as far as I have been able to discern, this has never actually happened.[384] Conducting a bellwether trial in the bankruptcy court—even for estimation—could face jurisdictional obstacles and may require the cooperation of the district court.[385] And it would, of course, entail all the costs and delays of conducting bellwether trials in the MDL.

The bankruptcy court could replicate other features of multidistrict litigation to try to estimate the value of claims. It could, for example, authorize fact discovery, conduct Daubert hearings on scientific experts, or send out plaintiff fact sheets to gather information.

Such experimentation would not be unprecedented. In the A.H. Robins bankruptcy, for example, the court appointed an expert, Francis McGovern, who crafted a 50-page questionnaire and sent it to a sample of 6,000 tort claimants. Professor McGovern then used the results to create a database of information about claims that could be compared to additional data collected about historical settlement values in the tort system.[386] But there is little reason to think that bankruptcy judges will be better or more efficient at employing MDL procedures than an MDL judge who does this sort of thing day in and day out.[387] Indeed, in A.H. Robins, the district court withdrew the bankruptcy reference, so the fact-sheet-like procedure was overseen by a district judge, not a bankruptcy judge.[388] The more bankruptcy mimics MDL procedures to value claims, the less it looks like a more “efficient and cost-effective” alternative.[389] Professor McGovern described his efforts at valuing claims as “the largest and most expensive social science survey ever conducted under the auspices of a court.”[390]

Alternatively, the bankruptcy judge could invite the participation of the MDL judge in a sort of hybrid estimation process. In the Dow Corning bankruptcy, for example, Judge Sam Pointer, who had been overseeing the Silicone Gel Breast Implants MDL, secured permission from the Chief Justice to sit by designation in the bankruptcy district and work cooperatively with the bankruptcy judge.[391] The parties contemplated having Judge Pointer selectively withdraw the reference in order to conduct either a trial on the issue of general causation or a series of bellwether-like “summary estimation jury trials,” but the plan never came to fruition.[392] Such a hybrid model might work under certain circumstances, but it is highly dependent on a cooperative relationship between the bankruptcy judge and the MDL judge. In a world where bankruptcy and MDL may develop into rivalrous systems for the resolution of mass torts, such cooperation is not guaranteed—particularly if the defendant filed for bankruptcy as a ticket out of an MDL that was not going its way.[393]

Or, as Professors Sergio Campos and Samir Parikh suggest, the bankruptcy judge could selectively lift the stay on certain claims to allow for expedited discovery and trials in state or federal court.[394] The results of these trials could serve as bellwethers and provide valuable information in estimating claims in the bankruptcy, as they did in the PG&E bankruptcy.[395] Selectively lifting the stay, however, is not a complete substitute for the information generated in an MDL. Campos and Parikh caution “that the comparative advantage of nonbankruptcy courts in determining the value of mass tort claims does not arise exclusively from trials themselves.”[396] They explain, “[n]aturally, trials provide data points for estimating the value of claims. The real value, however, may stem from the opportunity afforded to the court to accurately assess factual and legal issues, as well as from the court’s experience in doing so in other similar cases.”[397] Indeed, the bellwether process in MDL involves far more than simply allowing a handful of claims to go to trial. At least as important is the process of identifying a methodology for selecting an appropriately representative sample and working up through discovery the pool of trial-ready cases from which to choose.[398] Again, there is little reason to suspect that bankruptcy judges will be better at these sorts of case management decisions than an experienced MDL judge.

Most of these procedures are aimed at determining a top-line estimate of the debtor’s aggregate tort liability.[399] Outside of the context of asbestos bankruptcies, where claim values for different disease categories are well known from decades of litigation, bankruptcy courts have put very little effort into estimating the relative value of different claims for voting purposes. Even in Dow Corning and A.H. Robins, tort claims were estimated uniformly at one dollar for the purposes of assigning voting rights.[400] So, bankruptcy’s estimation procedures do little to bolster the bargaining position of tort victims with the strongest claims.

Whatever procedure the bankruptcy judge adopts, estimation remains an interim step. It will not result in payments to tort claimants and thus places little external pressure on the debtor. In MDL, the alternative to a negotiated resolution is trial and judgment, and if the defendant loses, it will have to pay up. In bankruptcy, the alternative to a negotiated estimation is more negotiation. “Winning” an estimation proceeding does not get the tort claimants paid; they will not see any money until a plan of reorganization is confirmed or the debtor is liquidated.

2.   Valuation for Distribution.

Valuation is also critical for distribution under the plan of reorganization. The terms of the plan are negotiated in the shadow of the estimation and voting procedures. As a result, as compared to an MDL, the parties are likely operating with less information and are in a bargaining environment where tort claimants have less leverage.

In the typical mass tort bankruptcy, the plan of reorganization creates a trust that the debtor funds with an agreed-upon amount of money or other assets, which are determined in light of the estimation of its total tort liability.[401] The bankruptcy court then enters a channeling injunction discharging the debtor from liability and channeling all tort claims to the trust, which is governed by the trust distribution procedures specified in the plan.[402] This creates a mandatory process for tort claimants. There is no way for them to opt out of the trust distribution procedures or pursue the debtor or any other released party directly.

The trust distribution procedures typically set up an administrative process to evaluate, resolve, and pay claims according to a grid system, point system, or some other form of alternative dispute resolution (ADR), much like an MDL settlement.[403] The claims administrator determines the value of each tort victim’s claim and pays the claimant accordingly. If the plan of reorganization did not allocate enough money to the trust to pay all anticipated claims in full, then the trust distribution procedures will require each tort claimant to take a haircut.[404]

As proponents of mass tort bankruptcies are quick to point out, tort claimants who are dissatisfied with the valuation that the trust’s ADR procedure places on their claims do not have to take the settlement.[405] Instead, they remain free to sue the trust in state or federal court and present their claim to a jury. Trust distribution procedures are often designed to deter the exercise of this right by making tort claimants jump through hoops and exhaust administrative remedies.[406] But at the end of the day, if claimants insist, their Seventh Amendment jury trial rights are preserved.[407] Tort claimants can have a jury liquidate the value of their tort claims.

The catch, however, is that even if they prevail at their jury trials, our tort claimants’ recoveries will still be governed by the trust distribution procedures, which will typically place caps on both compensatory and punitive damages.[408]

Such caps make sense when there is not enough money to go around. If the defendant cannot satisfy all of its tort liability, claimants who prevail at their jury trials should take a haircut, just like everyone else. And no claimant should get punitive damages if that means other claimants must go without compensatory damages.[409] But these realities highlight the dangers of estimating the aggregate value of claims incorrectly.[410] If the trust is underfunded or includes damages caps not because the defendant was insolvent, but because the parties and bankruptcy judge got the estimation wrong—or because the tort creditors lacked the leverage to negotiate for greater funding for the trust—then claimants who exercise their jury trial rights should not have to take a haircut.[411] Indeed, in Ortiz, the Supreme Court expressly rejected the idea of placing such limits on a back-end right to opt out of an artificially limited fund.[412]

The threat of trial in bankruptcy is thus vastly different from the threat of trial in MDL. In bankruptcy, tort claimants’ ability to opt out at the back end and take their claims to trial against the trust does not translate into leverage on the front end when they are negotiating the plan with the debtor.[413] The threat of back-end trials against the trust is of little concern to the debtor, whose aggregate liability is capped by the plan of reorganization.[414] This matters not only because of the risk of undercompensating the most severely injured tort victims who have the strongest claims, but even more so because it shifts leverage from tort victims collectively to the defendant. Bankruptcy has the ability to force all claimants to take from the negotiated amount, and the threat of trial does not factor into negotiating that amount; voting is the main source of leverage in bankruptcy.

In MDL, by contrast, the threat of trial is the main driver of the negotiated resolution. MDL litigants are always free to forego a global settlement offer and take their claims to trial. MDL plaintiffs rarely exercise that right, of course, but their ability to choose between the two guarantees that they do better in settlement than they expect to do at trial. And MDL dealmakers cannot design a settlement that just appeals to the majority; they need a settlement that appeals to all claimants if they want to get the strongest ones on board. A high-value claim in MDL can threaten a devastating loss at trial that will add to the defendant’s total liability unless the settlement is adjusted. A high-value claim in bankruptcy is just one vote against the plan that can safely be ignored because a back-end loss at trial will not affect the defendant’s total exposure.

F.   Forum Shopping

Even if bankruptcy skews the bargaining environment in favor of the defendant/debtor, mass tort bankruptcy proponents might take some comfort in the fact that bankruptcy is subject to close judicial oversight. After all, the bankruptcy judge must approve the plan of reorganization before it goes into effect, and bankruptcy judges can use that power to ensure that mass tort victims are not exploited.[415] MDL judges, by contrast, have no formal power to review private settlements reached in the MDL unless the settlements are part of a class action (though MDL judges can and do weigh in informally on the fairness of settlements).[416] That comfort, however, is undermined by the risk of forum shopping.

The forum shopping problem in bankruptcy is well documented.[417] In most Chapter 11 cases, the debtor picks the district in which to file. And corporate debtors have a virtually unlimited menu of options because they can reincorporate in their preferred jurisdiction shortly before filing. Indeed, in many districts, the debtor can be reasonably certain which bankruptcy judge will get the case.[418]

In MDL, by contrast, neither plaintiffs nor defendants have much opportunity to engage in forum shopping. The JPML—not the parties—picks the transferee district and handpicks the MDL judge.[419] And the parties cannot do much to structure their affairs to limit the JPML’s choices. Because MDLs are for pretrial proceedings only, the JPML is “not encumbered” by questions of personal jurisdiction or venue, and it is thus free to choose basically any federal district in the country to host an MDL.[420] Studies have shown that the JPML chooses districts favored by plaintiffs and defendants about equally.[421] And the JPML can—and sometimes does—choose a district favored by neither party.[422] The JPML’s practice is to pick the MDL judge that it thinks is best for the particular litigation.[423]

Bankruptcy gives defendants an option to escape the JPML’s chosen judge. Indeed, defendants can have it both ways. Defendant sentiment—or at least rhetoric—has turned against MDL in recent years.[424] But MDLs still offer defendants many advantages, not the least of which is the chance to win big on a dispositive pretrial motion that applies to all cases and effectively ends the federal litigation without the corresponding risk of losing big at a class-wide trial.[425] This is a valuable opportunity, and most defendants would prefer to obtain summary judgment on a common issue like preemption or general causation in the MDL over an estimation proceeding in bankruptcy.

By reducing the pain associated with bankruptcy, the Texas Two-Step lets defendants have it both ways: They can give the MDL a try and take a shot at a big win but then take the escape hatch to bankruptcy if things go poorly. Indeed, J&J appeared fairly content litigating talc claims in the MDL for years on end until it lost a major Daubert motion on general causation and failed to secure the reversal of several state court verdicts. Shortly thereafter, it spun off LTL and took it into bankruptcy. Likewise, 3M appeared content to sit tight in the Combat Arms Earplugs MDL until Judge Rodgers started moving cases off the inactive docket in waves and scheduling them for trial, prompting 3M’s subsidiary Aearo to seek friendlier terrain in bankruptcy court.[426]

By highlighting defendants’ ability to shop for a favorable bankruptcy forum, I do not mean to criticize bankruptcy judges. I have no doubt that bankruptcy judges approach their cases—however they got there—with integrity and professionalism. Indeed, bankruptcy judges have shown tremendous creativity and attention to fairness when faced with complex, thorny problems.[427] But there are structural incentives for bankruptcy judges to make their courtrooms debtor-friendly.[428] And bankruptcy judges may have a cultural predisposition to favor dealmaking over adjudication.[429] Both of these factors could create blind spots when it comes to evaluating whether tort claimants get a better outcome than they would have in litigation. Those concerns are lessened when the tort claims proceed in front of an MDL judge, who is chosen by the JPML according to the congressionally specified procedure rather than before a bankruptcy judge handpicked by the debtor.[430]

Conclusion

Bankruptcy’s mandatory nature and powerful closure tools are essential to preserving value and providing for an equitable distribution when there is not enough money to go around. But even if bankruptcy’s ability to achieve closure in mass tort controversies creates value when there is enough money to go around, there is no equitable necessity to justify overriding tort victims’ preexisting rights. To justify binding tort victims without their consent, we must be sure that they will share in the resulting surplus—that they will do better in a mass tort bankruptcy than they would in the most likely alternative, a federal MDL. As I have shown in this Article, however, MDL is not as bad as its critics contend, and moving from MDL to bankruptcy systematically shifts bargaining leverage from mass tort plaintiffs to defendants. Without leverage, tort victims cannot bargain for a share of any peace premium created by bankruptcy and, indeed, are likely to see smaller recoveries than in an MDL.

I applaud Casey and Macey’s view that “scholars should think about the optimal mechanism for resolving mass torts and develop an account of what reforms can be made to existing resolution schemes to achieve that system.”[431] And creative proposals from scholars like Bussel, Francus, Parikh, Simon, and others about how to reform bankruptcy to better handle mass torts are welcome. Perhaps reforms can preserve bankruptcy’s value-generating features without undermining tort claimants’ leverage. Or perhaps hybrid models can capture some of MDL judges’ comparative advantage in valuation while maintaining bankruptcy’s ability to yield closure. But unless and until such reforms are adopted, we need to compare the systems that we have, not the systems we wish we had. And, as they operate today, when there is enough money to go around, bankruptcy is not better than MDL for mass torts.

Although it may be an imperfect gatekeeping mechanism, the Third Circuit’s financial distress test may be the best we can do.[432] It gets at the central idea that motivates the creditors’ bargain: The debtor only gets to use bankruptcy’s tools to modify creditors’ rights over their objection when doing so necessarily works to the creditors’ collective benefit. When there is not enough money to go around, it is obvious how bankruptcy can leave creditors better off than a destructive race to the courthouse. But because bankruptcy shifts bargaining leverage from tort plaintiffs to defendants, we cannot be confident that bankruptcy leaves tort victims better off than MDL when there is enough money to go around. Bankruptcy should not be an exit strategy for defendants who think an MDL is not going their way. In such situations, bankruptcy is not an appropriate vehicle for the resolution of mass torts.

Copyright © 2026 D. Theodore Rave, Bernard J. Ward Centennial Professor of Law, University of Texas School of Law. Thanks to Ken Ayotte, Lynn Baker, Bob Bone, Andrew Bradt, Ralph Brubaker, Tony Casey, Zach Clopton, Jens Dammann, Mechelle Dickerson, Brian Fitzpatrick, Hon. Royal Furgeson, Jonah Gelbach, John Golden, Hon. Meredith Grabill, Tara Grove, Chad Husnick, Jeff Jonas, Hon. Michael Kaplan, Jonathan Lipson, Angela Littwin, Joshua Macey, Jared Mayer, Troy McKenzie, Samir Parikh, Charlie Silver, Adam Silverstein, Lindsey Simon, Jordan Steiker, Jay Westbrook, and Patrick Woolley for helpful comments and discussions. Brennan Caruthers and Madeline Love provided invaluable research assistance. Disclosure: I was an expert witness for the Official Committee of Talc Claimants in the second LTL bankruptcy. All references and claims in this article are drawn from publicly available material.

[1]. For an excellent explanation of the Texas Two-Step, see generally Michael A. Francus, Texas Two-Stepping Out of Bankruptcy, 120 Mich. L. Rev. 38 (2022).

[2]. In re LTL Mgmt., LLC, 64 F.4th 84, 106 (3d Cir. 2023); In re LTL Mgmt., LLC, 652 B.R. 433, 447 (Bankr. D.N.J. 2023).

[3].Amended Prepackaged Chapter 11 Plan of Reorganization of the Debtor, In re Red River Talc LLC, 670 B.R. 251 (Bankr. S.D. Tex. 2025) (No. 24-90505). The Fourth Circuit, for example, appears open to the Texas Two-Step, having declined multiple opportunities to dismiss a similar two-step bankruptcy in the Georgia-Pacific/Bestwall case. See Bestwall LLC v. Off. Comm. of Asbestos Claimants (In re Bestwall LLC), 71 F.4th 168, 185 (4th Cir. 2023) (affirming extension of preliminary injunction to non-debtor Georgia Pacific LLC); Off. Comm. of Asbestos Claimants of Bestwall, LLC v. Bestwall LLC, No. 19-408, 2019 WL 13512209, at *1 (4th Cir. Nov. 14, 2019) (denying interlocutory appeal of motion to dismiss for bad faith); Bestwall LLC v. Off. Comm. of Asbestos Claimants of Bestwall, LLC, 148 F.4th 233, 236 (4th Cir. 2025) (affirming denial of motion to dismiss bankruptcy for lack of subject matter jurisdiction where debtor is solvent); see also Order, Off. Comm. of Asbestos Pers. Inj. Claimants v. Aldrich Pump LLC, No. 24-129 (4th Cir. Apr. 17, 2024) (denying interlocutory appeal of motion to dismiss similar Texas-Two Step). The Fifth Circuit has not yet weighed in on the maneuver.

[4]. In re Red River Talc LLC, 670 B.R. at 307.

[5]. In re 3M Combat Arms Earplug Prods. Liab. Litig., No. 19-md-2885, 2022 WL 17853203 (N.D. Fla. Dec. 22, 2022); see J. Maria Glover, Due Process Discontents in Mass-Tort Bankruptcy, 72 DePaul L. Rev. 535, 567–68 (2023).

[6]. See Informational Brief at 7–8, 11–13, In re Aearo Techs. LLC, No. 22-02890-JJG-11, (Bankr. S.D. Ind. July 26, 2022), Dkt. No. 12.

[7]. See, e.g., Findings of Fact, Conclusions of Law, and Order (I) Confirming the Fourth Amended Joint Chapter 11 Plan of Reorganization of Endo International plc and its Affiliated Debtors and (II) Approving the Disclosure Statement with Respect Thereto, In re Endo Int’l plc, No. 22-22549 (Bankr. S.D.N.Y. Mar. 22, 2024); Findings of Fact, Conclusions of Law, and Order (I) Confirming Fourth Amended Joint Plan of Reorganization (with Technical Modifications) of Mallinckrodt PLC and its Debtor Affiliates Under Chapter 11 of the Bankruptcy Code, In re Mallinckrodt lc, No. 20-12522 (Bankr. D. Del. Mar. 2, 2022).

[8]. In re Boy Scouts of Am. & Del. BSA, LLC, 642 B.R. 504 (Bankr. D. Del. 2022).

       [9]. 144 S. Ct. 2071 (2024).

[10]. See, e.g., Kane v. Johns-Manville Corp., 843 F.2d 636, 638 (2d Cir. 1988) (asbestos liability claims); In re Dow Corning Corp., 244 B.R. 721, 731–32 (Bankr. E.D. Mich. 1999) (breast implant liability claims). See generally S. Elizabeth Gibson, Fed. Jud. Ctr., Case Studies of Mass Tort Limited Fund Class Action Settlements & Bankruptcy Reorganizations (2000) (comparing methods for mass tort resolution across three class actions and four bankruptcies).

     [11]. Scholarship on mass tort bankruptcy is extensive. For some classics, see generally S. Elizabeth Gibson, Judicial Management of Mass Tort Bankruptcy Cases (2005); Alan N. Resnick, Bankruptcy as a Vehicle for Resolving Enterprise-Threatening Mass Tort Liability, 148 U. Pa. L. Rev. 2045 (2000); Mark J. Roe, Bankruptcy and Mass Tort, 84 Colum. L. Rev. 846 (1984).

[12]. See Troy A. McKenzie, Toward a Bankruptcy Model for Nonclass Aggregate Litigation, 87 N.Y.U. L. Rev. 960, 979–86, 999–1002 (2012).

[13]. Seeid. at 1002–08.

[14]. See Andrew D. Bradt, Zachary D. Clopton & D. Theodore Rave, MDL Strikes Back, 110 Corn. L. Rev. 1855, 1856 (2026).

[15]. See, e.g., Francus, supra note 1, at 43–47; Adam Levitin, The Texas Two-Step: The New Fad in Fraudulent Transfers, Credit Slips (July 19, 2021), https://creditslips.org/2021/07/19/the-texas-two-step/ [https://perma.cc/FS5K-2AM6]; Lindsey D. Simon, Bankruptcy Grifters, 131 Yale L.J. 1154, 1157–59 (2022); Melissa B. Jacoby, Shocking Business Bankruptcy Law, 131 Yale L.J.F. 409, 411–12 (2021); Jonathan C. Lipson, “Special”: Remedial Schemes in Mass Tort Bankruptcies, 101 Tex. L. Rev. 1773, 1777–79 (2023).

[16]. See, e.g., Edward J. Janger, Aggregation and Abuse: Mass Torts in Bankruptcy, 91 Fordham L. Rev. 361, 370 (2022).

[17]. See, e.g., Ralph Brubaker, Mandatory Aggregation of Mass Tort Litigation in Bankruptcy, 131 Yale L.J.F. 960, 983–86 (2022); Pamela Foohey & Christopher K. Odinet, Silencing Litigation Through Bankruptcy, 109 Va. L. Rev. 1261 (2023).

     [18]. Anthony J. Casey & Joshua C. Macey, In Defense of Chapter 11 for Mass Torts, 90 U. Chi. L. Rev. 973 (2023).

[19]. Id. at 989–91.

[20]. Id. at 973, 1012–17.

[21].Harrington v. Purdue Pharma L.P., 144 S. Ct. 2071, 2092 (2024) (Kavanaugh, J., dissenting).

[22]. Contra Abbe R. Gluck, Elizabeth Chamblee Burch & Adam S. Zimmerman, Against Bankruptcy: Public Litigation Values Versus the Endless Quest for Global Peace in Mass Litigation, 133 Yale L.J.F. 525 (2024). For the purposes of this Article, I want to set aside debates about noneconomic values like voice and participation, not because they are unimportant, but because I want to consider bankruptcy on its own terms. For a taste of those debates, compare id. and Foohey & Odinet, supra note 17 with Anthony J. Casey & Joshua C. Macey, Bankruptcy by Another Name, 133 Yale L.J.F. 1016 (2024).

[23]. See D. Theodore Rave, Governing the Anticommons in Aggregate Litigation, 66 Vand. L. Rev. 1183, 1192–98, 1223–26 (2013).

[24]. SeeIn re LTL Mgmt., LLC, 64 F.4th 84, 111 (3d Cir. 2023) (“[R]esort to Chapter 11 is appropriate only for entities facing financial distress. This safeguard ensures that claimants’ pre-bankruptcy remedies—here, the chance to prove to a jury of their peers injuries claimed to be caused by a consumer product—are disrupted only when necessary.”); Richard A. Nagareda, The Preexistence Principle and the Structure of the Class Action, 103 Colum. L. Rev. 149, 164–67 (2003) (arguing that class action settlements cannot be used to unilaterally alter plaintiffs’ preexisting rights without their consent).

     [25]. Rave, supra note 23, at 1192–96.

[26]. See Thomas H. Jackson, Bankruptcy, Non-Bankruptcy Entitlements, and the Creditors’ Bargain, 91 Yale L.J. 857, 860 (1982); Douglas G. Baird, A World Without Bankruptcy, 50 L. & Contemp. Probs. 173, 183–84 (1987). See generally Thomas H. Jackson, The Logic and Limits of Bankruptcy Law (1986) [hereinafter Jackson, Logic and Limits] (laying out creditors’ bargain theory).

     [27]. Casey & Macey, supra note 18, at 973 (emphasis added); see also id. at 1012 (arguing for “a baseline rule against any bankruptcy plan that leaves tort claimants worse off than they would have been outside of bankruptcy”).

[28]. See infra Part II.A.

[29]. See Alexandra D. Lahav, The Continuum of Aggregation, 53 Ga. L. Rev. 1393, 1394–95 (2019).

[30]. See, e.g., Daniel J. Bussel, The Mass Tort Claimants’ Bargain, 97 Am. Bankr. L.J. 684, 698–701 (2023); Ralph Brubaker, Assessing the Legitimacy of the “Texas Two-Step” Mass-Tort Bankruptcy, 42 Bankr. L. Letter No. 8 1, 8–12 (2022) [hereinafter Brubaker, Two-Step Part I].

[31]. See Rave, supra note 23.

[32]. See, e.g., Casey & Macey, supra note 18, at 989–90, n.61; Bussel, supra note 30, at 704–05, 709–10 (criticizing the Third Circuit’s decision to characterize J&J’s Chapter 11 reorganization as filed in “bad faith” because J&J was not in financial distress).

[33]. See, e.g., Charles Silver & Lynn A. Baker, Mass Lawsuits and the Aggregate Settlement Rule, 32 Wake Forest L. Rev. 733, 760–63 (1997).

[34]. See id. at 763–67; Samuel Issacharoff & D. Theodore Rave, The BP Oil Spill Settlement and the Paradox of Public Litigation, 74 La. L. Rev. 397, 413–15 (2014); Sullivan v. DB Invs., Inc., 667 F.3d 273, 339 n.9 (3d Cir. 2011) (Scirica, J., concurring).

     [35]. Rave, supra note 23, at 1195–98.

[36]. Id. at 1198–1201.

[37]. See generally Richard A. Nagareda, Mass Torts in a World of Settlement (2007) (describing the evolution of methods used by mass tort counsel to achieve settlement).

[38]. See Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 807 (1985) (“[A] chose in action is a constitutionally recognized property interest possessed by each of the plaintiffs.”); see also D. Theodore Rave, Tort Claims as Property Rights, 69 DePaul L. Rev. 587 (2020).

     [39]. Ortiz v. Fibreboard Corp., 527 U.S. 815, 821 (1999).

     [40]. 521 U.S. 591, 623, 627 (1997).

     [41]. 527 U.S. 815 (1999).

[42]. See, e.g., Deborah R. Hensler, Asbestos Litigation in the United States: Triumph and Failure of the Civil Justice System, 12 Conn. Ins. L.J. 255, 271–73, 279 (2006).

[43]. See, e.g., McKenzie, supra note 12, at 1002–10; William Organek, Mass Tort Bankruptcy Goes Public, 77 Vand. L. Rev. 723, 737–39 (2024).

[44]. See McKenzie, supra note 12, at 1002–09.

[45]. See, e.g., Ralph Brubaker, Mass Torts, the Bankruptcy Power, and Constitutional Limits on Mandatory No-Opt-Outs Settlements, 23 Fla. St. U. Bus. Rev. 111, 124 (2024) [hereinafter Brubaker, No-Opt-Outs]; Ralph Brubaker, Assessing the Legitimacy of the “Texas Two-Step” Mass-Tort Bankruptcy (Part III): The Constitutional Limits of the Bankruptcy Power, 44 Bankr. L. Letter No. 10 1, 9–15 (2024) [hereinafter Brubaker, Two-Step Part III]. See generally Thomas E. Plank, The Constitutional Limits of Bankruptcy, 63 Tenn. L. Rev. 487, 491–92 (1996) (“Under the Bankruptcy Clause, Congress may only enact legislation that regulates the relationship between an insolvent debtor and her creditors.”). There is a robust debate about where the constitutional limits of bankruptcy lie, but the equitable necessity justification captures its intuitive core. For a helpful overview of the debate, see generally Jonathan C. Lipson, Debt and Democracy: Towards a Constitutional Theory of Bankruptcy, 83 Notre Dame L. Rev. 605 (2008).

[46]. See Zechariah Chafee Jr., Bills of Peace with Multiple Parties, 45 Harv. L. Rev. 1297, 1311 (1932) (“The fund or limited liability is like a mince pie, which can not be satisfactorily divided until the carver counts the number of persons at the table.”).

     [47]. 527 U.S. 815, 838–39 (1999) (“The concept driving this type of suit was insufficiency, which alone justified the limit on an early feast to avoid a later famine. . . . The equity of the limitation is its necessity.”); see also Brubaker, No-Opt-Outs, supra note 45, at 124 (“And the bankruptcy exception to claimants’ due-process rights has traditionally been understood as necessitated by the same limited-fund justification that the Court concluded was not present in Ortiz.”).

[48]. In re Aldrich Pump LLC, No. 20-30608, 2023 WL 9016506, at *20 (Bankr. W.D.N.C. Dec. 28, 2023).

[49]. See, e.g., Samuel Issacharoff, Governance and Legitimacy in the Law of Class Actions, 1999 Sup. Ct. Rev. 337, 340–41 (1999) (describing due process challenges in aggregate litigation).

[50]. See, e.g., Jonathan C. Lipson, The Rule of the Deal: Bankruptcy Bargains and Other Misnomers, 97 Am. Bankr. L.J. 41, 43 (2023) (describing “bankruptcy’s strong appetite” for dealmaking, which can “distort or deviate from positive law”).

[51]. See, e.g., Brubaker, No-Opt-Outs, supra note 45, at 114; McKenzie, supra note 12, at 1018–19.

     [52]. Kane v. Johns-Manville Corp., 843 F.2d 636 (2d Cir. 1988).

[53]. In re Dow Corning Corp., 255 B.R. 445 (E.D. Mich. 2000).

[54]. In re PG&E Corp., 46 F.4th 1047 (9th Cir. 2022).

[55]. See Bussel, supra note 30, at 707; Brubaker, Two-Step Part III, supra note 45, at 8.

[56]. See, e.g., 11 U.S.C. § 1125; McKenzie, supra note 12, at 1007–12.

[57]. See Jared A. Ellias, Employee Bankruptcy Trauma (John M. Olin Ctr. for L., Econ., & Bus. at Harv. L. Sch., Discussion Paper No. 1106, 2023), https://laweconcenter.law.harvard.edu/wp-content/uploads/2024/11/Ellias_1106.pdf [https://perma.cc/2MKT-92J2].

[58]. See, e.g., Gibson, supra note 10, at 25–26.

[59]. See, e.g., Michael Francus, Designing Designer Bankruptcy, 102 Tex. L. Rev. 1205, 1224–27 (2024). Prepackaged bankruptcies or “prepacks” are deals where creditors agree on the terms of reorganization and line up votes in support of the plan before filing for bankruptcy. In the mass tort context, a prepack is essentially the bankruptcy equivalent of a settlement class action. See generally Ronald Barliant, Dimitri G. Karcazes & Anne M. Sherry, From Free-Fall to Free-For-All: The Rise of Pre-Packaged Asbestos Bankruptcies, 12 Am. Bankr. Inst. L. Rev. 441 (2004) (tracing the history and mechanisms that account for the rise of pre-packaged asbestos bankruptcies).

[60]. Seegenerally Francus, supra note 1; see also Samir D. Parikh, Mass Exploitation, 170 U. Pa. L. Rev. Online 53, 58–59 (2022).

     [61]. Tex. Bus. Orgs. Code Ann. § 1.002(55)(A) (effective Jan. 1, 2006).

     [62]. Parikh, supra note 60, at 59.

     [63]. Casey & Macey, supra note 18, at 1009.

[64]. Id. at 1009–10; Parikh, supra note 60, at 59, 68–70; see also Francus, Two-Stepping, supra note 1, at 42–44 (arguing that divisional mergers are fraudulent transfers).

[65]. See Francus, supra note 59, at 1229–30.

[66]. See Casey & Macey, supra note 18, at 994–96.

[67]. See, e.g., Nagareda, supra note 24, at 157 (“Mandatory treatment of class members’ claims—denial of the right to opt out—is warranted only when conditions antecedent to the class itself make class members’ rights interdependent, and, in doing so, necessitate a choice between competing preexisting rights.”).

[68]. See generally, e.g., Jackson, Logic and Limits, supra note 26 (explaining creditors’ bargain theory). See also Bussel, supra note 30, at 697.

     [69]. 11 U.S.C. § 1129(b)(2).

     [70]. Id. § 1129(a)(7)(ii).

[71]. See Ralph Brubaker, Taking Exception to the New Corporate Discharge Exceptions, 13 Am. Bankr. Inst. L. Rev. 757, 759–61 (2005).

     [72]. Casey & Macey, supra note 18, at 973. Other mass tort bankruptcy proponents share the same premise. See Francus, supra note 59, at 1234 n.171 (“[A] properly designed Two-Step adds value to a debtor business but ensures that any value goes to the tort victims until they are paid in full. Thus, shareholders do not receive any extra value created by using a Two-Step (instead of a class action or ordinary bankruptcy), tort victims do.”); Bussel, supra note 30, at 697 (“The goal of the Mass Tort Claimants’ Bargain is to preserve the basic nonbankruptcy economic rights of the claimants while substituting an efficient collective administrative process to realize those substantive rights.”).

[73]. E.g., In re LTL Mgmt. LLC, 64 F.4th 84 (3d Cir. 2023); In re Aldrich Pump, LLC, No. 20-30608, Adv. No. 20-03041, 2021 WL 3729335 (Bankr. W.D.N.C. Aug. 23, 2021); In re DBMP LLC, No. 20-30080, 2021 WL 3552350 (Bankr. W.D.N.C. Aug. 11, 2021); In re Bestwall LLC, 605 B.R. 43 (Bankr. W.D.N.C. 2019); see alsoIn re Aearo Techs. LLC, 642 B.R. 891 (Bankr. S.D. Ind. 2022) (not technically a Two-Step, but similarly structured).

[74]. See, e.g., Casey & Macey, supra note 18, at 998–99.

[75]. See, e.g., In re Aldrich Pump LLC, No. 20-30608, 2023 WL 9016506, at *20–21 (Bankr. W.D.N.C. Dec. 28, 2023).

[76]. See, e.g., Janger, supra note 16; Brubaker, Two-Step Part III, supra note 45, at 15–16.

[77]. Cf. Gluck, Burch & Zimmerman, supra note 22, at 551–61(discussing some of the prosocial values inherent to mass tort litigation that are lacking in bankruptcy proceedings); Foohey & Odinet, supra note 17, at 1265–69, 1313–24 (noting how the use of bankruptcy to resolve mass torts damages public trust in judicial institutions and harms tort victims by limiting forms of justice to monetary compensation); Jonathan C. Lipson & Pamela Foohey, The End(s) of Bankruptcy Exceptionalism: Purdue Pharma and the Problem of Social Debt, 46 Cardozo L. Rev. 861, 867–72 (2025) (discussing the ways the Purdue Pharma bankruptcy upset constitutional norms and failed to account for “social debt”).

     [78]. Nagareda, supra note 37, at 158–59 (“Absent the ability to alter unilaterally class members’ preexisting rights to sue in tort . . . settlement designers must purchase those rights by way of the benefits promised to class members for remaining in the settlement. Class members’ preexisting rights to sue truly must be purchased rather than simply appropriated, in keeping with their status as a form of property.”).

[79]. See infra Parts III.C–E.

[80]. See, e.g., Foohey & Odinet, supra note 17, at 1277–78; Janger, supra note 16, at 370.

[81]. In re LTL Mgmt., LLC (LTL 1.0), 637 B.R. 396, 410 (Bankr. D.N.J. 2022) (“[T]he Court simply cannot accept the premise that continued litigation in state and federal courts serves best the interest of their constituency. Many of these cases . . . have been pending for a half dozen or more years and remain years away from trial dates, not to mention the substantial delays they face in the inevitable appeals process.”).

     [82]. Casey & Macey, supra note 18, at 981 (“Procedural limitations can make resolution slow and costly. Holdouts can disrupt the entire settlement. The prospect of future claimants who will not be bound by the MDL results makes it impossible to reach full resolution. Judges may be unfamiliar with the law they are applying. Procedural rules are applied inconsistently.”); id. at 1000 (“Without a mechanism like bankruptcy to force claimants into a single tribunal, separate courts will engage in redundant discovery, different plaintiffs will hire multiple attorneys to handle similar claims in various geographic regions, and courts and attorneys will have to educate themselves about different tort regimes.”).

     [83]. In the J&J talc bankruptcy, for example, Judge Kaplan explained that:

The multidistrict litigation . . . will at best produce a handful of bellwether trials later in 2022, offering some insight into the strength of the cases, but will also necessarily return nearly 40,000 cases to federal courts across the country to await pre-trial proceedings and eventual trials and appeals. Notwithstanding the pre-trial work undertaken through the MDL, the fact remains that plaintiffs and defendants will be forced to relitigate causation, and damages, and apportion liability among defendants in every case, which will be both costs prohibitive and “burden the tort system with unnecessarily drawn-out litigation.”

LTL 1.0, 637 B.R. at 412 (citing Debtor’s expert report by Charles Mullin, at 9 [hereinafter Mullin Report]). And he estimated that just defending those pending claims through trial “would cost up to $190 billion.” Id. at 419. Casey and Macey similarly assume that tens of thousands of individual cases will have to be tried and worry that outlier verdicts in even a small handful of those cases could destroy the defendant. Casey & Macey, supra note 18, at 987, 997–98.

     [84]. Samir Parikh, The New Mass Torts Bargain, 91 Fordham L. Rev. 447, 478 (2022) (“Deterrence is unrealized because there are significant lottery effects.”); Casey & Macey, supra note 18, at 999 (“[O]utside of bankruptcy, luck plays a large role in determining who gets paid. It is inequitable to advantage one tort claimant over another simply because that tort claimant’s injury became evident at an earlier point in time. Similarly, some tort claimants may receive unusually sympathetic juries who award large damages.”); LTL 1.0, 637 B.R. at 414 (“In the eyes of this Court, the tort system produces an uneven, slow-paced race to the courthouse, with winners and losers.”).

[85]. See LTL 1.0, 637 B.R. at 414 (“Of equal concern to this Court is the capacity for the state and federal courts to protect future claimants, whose claims may surface in the next half century given the acknowledged latency period for the types of cancer at issue. The needs of these victims are wholly ignored by the current rush to secure judgments against Debtor in the federal and state courts.”); In re LTL Mgmt., LLC (LTL 2.0), 652 B.R. 433, 451 (Bankr. D.N.J. 2023) (“In short, the Court remains unconvinced that the procedural mechanisms and notice programs offered in the tort system can protect future claimants’ rights in the same manner as the available tools in the bankruptcy system.”).

     [86]. Casey & Macey, supra note 18, at 1005 (“Reaching global resolution is possible only when courts can compel participation.”).

     [87]. Harrington v. Purdue Pharma L.P., 144 S. Ct. 2071, 2092 (2024) (Kavanaugh, J., dissenting) (quoting Casey & Macey, supra note 18). Justice Kavanaugh’s take on mass tort bankruptcy was much rosier: “Bankruptcy, on the other hand, reduces administrative costs and allows all of the affected parties to come together, pause litigation elsewhere, invoke procedural safeguards including discovery, and reach a collective resolution that considers both current and future victims.” Id.

     [88]. Parikh, supra note 84, at 476.

[89]. Id. at 477–78.

[90]. Id. at 478.

[91]. Id.

[92]. Id.; see also id. at 478–79 (“Because of extreme lottery effects, erosion of individual victims’ rights, and a lack of predictability, some scholars have concluded that the MDL structure must be redesigned.”).

     [93]. Casey & Macey and Parikh rely on the work of several scholars who have been highly critical of MDL, including Elizabeth Chamblee Burch & Margaret S. Williams, Perceptions of Justice in Multidistrict Litigation: Voices from the Crowd, 107 Corn. L. Rev. 1835 (2022); Elizabeth Chamblee Burch, Judging Multidistrict Litigation, 90 N.Y.U. L. Rev. 71 (2015); Abbe R. Gluck & Elizabeth Chamblee Burch, MDL Revolution, 96 N.Y.U. L. Rev. 1 (2021); L. Elizabeth Chamblee, Unsettling Efficiency: When Non-Class Aggregation of Mass Torts Creates Second-Class Settlements, 65 La. L. Rev. 157 (2004); Howard M. Erichson, MDL and the Allure of Sidestepping Litigation, 53 Ga. L. Rev. 1287 (2019); Linda S. Mullenix, Aggregate Litigation and the Death of Democratic Dispute Resolution, 107 Nw. U. L. Rev. 511 (2013); Martin H. Redish & Julie M. Karaba, One Size Doesn’t Fit All: Multidistrict Litigation, Due Process, and the Dangers of Procedural Collectivism, 95 B.U. L. Rev. 109 (2015); Charles Silver & Geoffrey P. Miller, The Quasi-Class Action Method of Managing Multi-District Litigations: Problems and a Proposal, 63 Vand. L. Rev. 105 (2010). In LTL 1.0, Judge Kaplan relied primarily on the Debtor’s expert economist Charles Mullin for his portrayal of MDL. See, e.g., LTL 1.0, 637 B.R 396, 412 (Bankr. D.N.J. 2022) (citing Mullin Report). However, while I disagree with many of his conclusions in LTL 2.0, Judge Kaplan listened thoughtfully to a more balanced description of how MDL works with expert witnesses testifying on both sides. 652 B.R. 433, 442 (Bankr. D.N.J. 2023). Full disclosure: I was one of those witnesses, along with former JPML member retired Judge Royal Furgeson and prominent mass tort defense lawyer Sheila Birnbaum. Id.

[94]. See generally Lynn A. Baker & Andrew D. Bradt, MDL Myths, 101 Tex. L. Rev. 1521 (2023) (responding to critics’ misconceptions about MDL).

[95]. See, e.g., Parikh, supra note 84, at 494 (“Current structural anomalies are allowing debtors to fashion a distortive new bargain. But bankruptcy is still the optimal venue for resolving many mass tort cases.”).

     [96]. Sergio Campos & Samir Parikh, Due Process Alignment in Mass Restructurings, 91 Fordham L. Rev. 325, 355 (2022). Parikh also suggests appointing a committee of three future claims representatives, empowering them to vote on the plan of reorganization, requiring judicial approval of any law firm they hire, and extending § 524(g)’s special provisions for asbestos bankruptcies to cover all mass torts with future claims components. Parikh, supra note 84, at 496–502.

     [97]. Bussel, supra note 30, at 701–02. Bussel advocates “assuming a clean slate” and “imagin[ing] how we might construct a broadly accessible bankruptcy-based solution for . . . mass tort problems” with major changes to current practices. Id. at 688, 754–55.

     [98]. Casey & Macey, supra note 18, at 1012–17; Simon, supra note 15, at 1207–10.

     [99]. Casey & Macey, supra note 18, at 1018–19.

[100]. Id. at 1011. Casey & Macey also suggest making fraudulent transfer actions easier to win, heightening the standard for preliminary injunctions staying litigation against third parties, and replacing incumbent managers with a trustee. Id. at 1014–20. Francus, similarly, suggests replacing BadCo’s management with a trustee in Two-Step bankruptcies. Francus, supra note 59, at 1259.

   [101]. Francus, supra note 59, at 1257–58 (advocating a new mass tort section of the Bankruptcy Code); Parikh, supra note 84, at 499–502 (advocating amending and extending § 524(g)); Bussel, supra note 30, at 695–97 (advocating new statutory framework to regulate mass tort resolution).

[102]. See, e.g., Andrew D. Bradt & D. Theodore Rave, The Information-Forcing Role of the Judge in Multidistrict Litigation, 105 Calif. L. Rev. 1259 (2017) (addressing MDL governance failures and suggesting that judges inform claimants about the fairness of proposed MDL settlements).

   [103]. Much of this Section is drawn from my expert report in LTL 2.0.

[104]. See Glover, supra note 5, at 553.

   [105]. 28 U.S.C. § 1407.

[106]. See Lexecon Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S. 26 (1998).

[107]. See Thomas H.L. Forster, Note, Out of the “Black Hole”: Toward a New Approach to MDL Procedure, 100 Tex. L. Rev. 1227, 1228 (2022) (noting that in 2019 the JPML remanded only 2.34 percent of MDL cases); see also Elizabeth Chamblee Burch, Remanding Multidistrict Litigation, 75 La. L. Rev. 399, 400–01 (2014).

[108]. See Andrew D. Bradt, “A Radical Proposal”: The Multidistrict Litigation Act of 1968, 165 U. Pa. L. Rev. 831, 839 (2017).

[109]. Seeid.

[110]. See Zachary D. Clopton, MDL as Category, 105 Corn. L. Rev. 1297 (2020). For up-to-date statistics on pending actions in MDLs, see Pending MDLs, JPML, https://www.jpml.uscourts.gov/pending-mdls-0 [https://perma.cc/72GP-E7VX].

   [111]. Bradt, supra note 108, at 843; see also D. Theodore Rave, Management and Judging in Multidistrict Litigation, 42 Rev. Litig. 291, 292 (2023).

[112]. See John G. Heyburn II, A View from the Panel: Part of the Solution, 82 Tul. L. Rev. 2225, 2240–43 (2008).

[113]. See generally J. Alex Grimsley, Fed. Jud. Ctr., Annotated Manual for Complex Litigation, Fourth (2025) (detailing case management tools).

[114]. See, e.g., Amchem Prods., Inc. v. Windsor, 521 U.S. 591 (1997); Issacharoff, supra note 49.

[115]. See Andrew D. Bradt & D. Theodore Rave, It’s Good to Have the “Haves” on Your Side: A Defense of Repeat Players in Multidistrict Litigation, 108 Geo. L.J. 73, 78–79 (2019) [hereinafter Bradt & Rave, Defense of Repeat Players]; Andrew D. Bradt & D. Theodore Rave, Aggregation on Defendants’ Terms: Bristol-Myers Squibb and the Federalization of Mass-Tort Litigation, 59 B.C. L. Rev. 1251, 1296 (2018) [hereinafter Bradt & Rave, Defendants’ Terms]; Bradt & Rave, supra note 102, at 1269–73.

   [116]. Bradt & Rave, Defendants’ Terms, supra note 115, at 1258; Andrew D. Bradt, The Long Arm of Multidistrict Litigation, 59 Wm. & Mary L. Rev. 1165, 1213 (2018).

[117]. See Uber Techs., Inc. v. Jud. Panel on Multidistrict Litig., 131 F.4th 661, 669 (9th Cir. 2025).

   [118]. Bradt & Rave, Defendants’ Terms, supra note 115, at 1296.

[119]. See Andrew D. Bradt, Something Less and Something More: MDL’s Roots as a Class Action Alternative, 165 U. Pa. L. Rev. 1711, 1712 (2017).

[120]. See Marc Galanter, Why the “Haves” Come Out Ahead: Speculations on the Limits of Legal Change, 9 L. & Soc’y Rev. 95, 99–100 (1974) (discussing advantages that repeat-player litigants have over one-shot litigants like tort claimants).

[121]. See id.

[122]. See, e.g., Grimsley, supra note 113, § 10.22.

[123]. See David L. Noll, What Do MDL Leaders Do? Evidence from Leadership Appointment Orders, 24 Lewis & Clark L. Rev. 433, 456 (2020).

[124]. See, e.g., Bolch Jud. Inst., Guidelines and Best Practices for Large and Mass-Tort MDLs 64 (2d ed. 2018).

[125]. See, e.g., Eldon E. Fallon, Common Benefit Fees in Multidistrict Litigation, 74 La. L. Rev. 371, 374–75 (2014).

[126]. See Bradt & Rave, Defense of Repeat Players, supra note 115, at 96.

[127]. See, e.g., Bradt, supra note 108, at 836, 865.

[128]. See D. Theodore Rave & Francis E. McGovern, A Hub-and-Spoke Model of Multidistrict Litigation, 84 L. & Contemp. Probs. 21, 37 (2021).

[129]. See, e.g., Robert G. Bone, Settlement and the “Good” Judge, 38 Rev. Litig. 211 (2019).

[130]. Cf. Ortiz v. Fibreboard, Corp., 527 U.S. 815, 857 (1999) (rejecting class action settlement in part because it provided the same compensation to class members whose claims varied substantially).

   [131]. Francis E. McGovern, An Analysis of Mass Torts for Judges, 73 Tex. L. Rev. 1821, 1843–45 (1995).

   [132]. Margaret S. Williams, The Effect of Multidistrict Litigation on the Federal Judiciary Over the Past 50 Years, 53 Ga. L. Rev. 1245, 1271 (2019) (finding that cases in MDLs last about twice as long as the average civil case “because they are more complex”).

   [133]. Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993).

   [134]. In the Zoloft litigation, for example, Judge Rufe excluded the plaintiffs’ general causation expert under Daubert and then granted summary judgment for the defendant on all of the cases in the MDL. SeeIn re Zoloft (Sertraline Hydrochloride) Prods. Liab. Litig., 858 F.3d 787 (3d Cir. 2017) (affirming summary judgment). Many other MDLs have also been resolved in one fell swoop by dispositive motions on common issues. See Douglas G. Smith, Resolution of Common Questions in MDL Proceedings, 66 U. Kan. L. Rev. 219, 232–43 (2017) (collecting examples of MDL courts granting summary judgment).

[135]. See, e.g., Margaret S. Williams, Jason A. Cantone & Emery G. Lee III, Fed. Jud. Ctr., Plaintiff Fact Sheets in Multidistrict Litigation Proceedings: A Guide for Transferee Judges 7 (2019).

[136]. See Exhibit A, In re 3M Combat Arms Earplug Prods. Liab. Litig., No. 19-md-2885, 2022 WL 2292862 (N.D. Fla. 2022), Dkt. No. 3076-1.

[137]. See, e.g., Alvin K. Hellerstein, James A. Henderson, Jr. & Aaron D. Twerski, Managerial Judging: The 9/11 Responders’ Tort Litigation, 98 Corn. L. Rev. 127, 146–48 (2012) (describing the database of claims that laid the groundwork for settlement).

   [138]. 523 U.S. 26 (1998).

   [139]. By signing a Lexecon waiver, the parties to a transferred action agree to waive objections to personal jurisdiction and venue in the transferee court, allowing the MDL judge to try the case. See Bolch, supra note 124, at 22. Some direct filing stipulations do not allow for trial in the MDL district and instead specify a district where the case will be transferred for trial at the close of pretrial proceedings. Id.

[140]. See Andrew D. Bradt, The Shortest Distance: Direct Filing and Choice of Law in Multidistrict Litigation, 88 Notre Dame L. Rev. 759, 763 (2012).

[141]. See Alex Berenson, A Mistrial Is Declared in 3rd Suit over Vioxx, N.Y. Times (Dec. 13, 2005), https://www.nytimes.com/2005/12/13/business/a-mistrial-is-declared-in-3rd-suit-over-vioxx.html [https://perma.cc/3LBQ-UY6Y].

[142]. See, e.g., Eldon E. Fallon, Jeremy T. Grabill & Robert Pitard Wynne, Bellwether Trials in Multidistrict Litigation, 82 Tul. L. Rev. 2323 (2008); Eldon E. Fallon, Bellwether Trials, 89 UMKC L. Rev. 951, 951–53 (2021); Alexandra D. Lahav, Bellwether Trials, 76 Geo. Wash. L. Rev. 576 (2008); Robert G. Bone, Statistical Adjudication: Rights, Justice, and Utility in a World of Process Scarcity, 46 Vand. L. Rev. 561, 569–76 (1993).

[143]. Cf. Adam S. Zimmerman, The Bellwether Settlement, 85 Fordham L. Rev. 2275, 2277 (2017) (discussing how settlement outcomes can provide parties with “building blocks” of data to settle outstanding cases).

[144]. See, e.g., Rave, supra note 111; David L. Shapiro, Federal Rule 16: A Look at the Theory and Practice of Rulemaking,137 U. Pa. L. Rev. 1969, 1975 (1989).

[145]. See 28 U.S.C. § 1407(a).

   [146]. About half the states have these sorts of “State MDL” procedures. See Zachary D. Clopton & D. Theodore Rave, MDL in the States, 115 Nw. U. L. Rev. 1649, 1656–63 (2021). Indeed, in some states, like New York, New Jersey, and Texas, the opportunity to coordinate with a federal MDL is a factor in favor of creating a state consolidation. See id. at 1669; D. Theodore Rave & Zachary D. Clopton, Texas MDL, 24 Lewis & Clark L. Rev. 367, 384 (2020).

[147]. See Jeremy T. Grabill, Judicial Review of Private Mass Tort Settlements, 42 Seton Hall L. Rev. 123, 144–45 (2012).

[148]. See Sam C. Pointer, Jr., Reflections by a Federal Judge: A Comment on Judicial Federalism: A Proposal to Amend the Multidistrict Litigation Statute,73 Tex. L. Rev. 1569, 1571 (1995).

   [149]. Fed. Jud. Ctr., Coordinating Multijurisdictional Litigation: A Pocket Guide for Judges (2013), https://www.fjc.gov/content/coordinating-multijurisdiction-litigation-pocket-guide-judges [https://perma.cc/2C8X-XFF4].

[150]. See, e.g., Robert M. Cover, The Uses of Jurisdictional Redundancy: Interest, Ideology, and Innovation, 22 Wm. & Mary L. Rev. 639 (1981); Alexandra D. Lahav, Recovering the Social Value of Jurisdictional Redundancy, 82 Tul. L. Rev. 2369 (2008).

[151]. See Zachary D. Clopton & D. Theodore Rave, Opioid Cases and State MDLs, 70 DePaul L. Rev. 245, 261–62 (2021).

[152]. See id.

[153]. See Rave & McGovern, supra note 128, at 22.

[154]. See, e.g., J. Maria Glover, Mass Litigation Governance in the Post-Class Action Era: The Problems and Promise of Non-Removable State Actions in Multi-District Litigation, 5 J. Tort L. 3, 30–31 (2012); Clopton & Rave, supra note 151, at 246–47.

[155]. See Suggestions of Remand at 6–7, In re Nat’l Prescription Opiate Litig., MDL No. 2804 (J.P.M.L. Nov. 19, 2019); Case Management Order No. 21 at 1–2, In re 3M Combat Arms Earplug Prods. Liab. Litig., No. 19-md-2885 (N.D. Fla. Aug. 24, 2021).

   [156]. Id.; see also Rave & McGovern, supra note 128, at 39–42 (describing hub-and-spoke model).

   [157]. Rave & McGovern, supra note 128, at 35; see also Burch, supra note 107, at 407–09 (explaining advantages of decentralized decision-making for state-law claims); Lahav, supra note 150 (same).

[158]. See Ralph Brubaker, Assessing the Legitimacy of the “Texas Two-Step” Mass-Tort Bankruptcy (Part II), 43 Bankr. L. Letter No. 4 1, 7 (2023) [hereinafter Brubaker, Two-Step Part II] (describing this “canard”).

[159]. ContraLTL 1.0, 637 B.R. 396, 412 (Bankr. D.N.J. 2022) (“The multi-district litigation . . . will also necessarily return nearly 40,000 cases to federal courts across the country to await pre-trial proceedings and eventual trials and appeals.”).

   [160]. Forster, supra note 107, at 1228.

[161]. See Eduardo C. Robreno, The Federal Asbestos Product Liability Multidistrict Litigation (MDL-875): Black Hole or New Paradigm?, 23 Widener L.J. 97, 156, 180 (2013).

[162]. Id.

   [163]. Admin. Off. U.S. Cts., Table C-4—U.S. District Courts—Civil Cases Terminated, by Nature of Suit and Action Taken—During the 12-Month Period Ending June 30, 2022, U.S. Courts: Statistical Tables for the Federal Judiciary (June 30, 2022), https://www.uscourts.gov/statistics/table/c-4/statistical-tables-federal-judiciary/2022/06/30 [https://perma.cc/4KLS-VFKR] (showing that 0.7% of all federal cases reached trial (either jury or bench); among diversity of citizenship tort actions, 0.1% of “Health Care/Pharma” cases and 0.3% of “Other Personal Injury” cases reached trial in a twelve-month period); see also Nora Freeman Engstrom, The Diminished Trial, 86 Fordham L. Rev. 2131, 2135–40 (2018) (noting that the real trial rate is even lower because federal data overcounts “trials” to include any proceeding where evidence is presented (i.e., Rule 23 hearings or Daubert hearings)).

[164]. See Lynn A. Baker & Charles Silver, In Defense of Private Claims Resolution Facilities, 84 L. & Contemp. Probs. 45 (2021).

[165]. See id. at 55–59 (describing inventory settlements in the GM ignition switch litigation).

[166]. E.g., Settlement Agreement, In re Vioxx Prod. Liab. Litig., No. 05-md-1657 (E.D. La. Nov. 9, 2007) [hereinafter Vioxx Settlement].

[167]. E.g., In re Nat’l Football League Players Concussion Inj. Litig., 821 F.3d 410 (3d Cir. 2016).

[168]. See Baker & Silver, supra note 164, at 56.

[169]. See, e.g., Lynn A. Baker, Mass Torts and the Pursuit of Ethical Finality, 85 Fordham L. Rev. 1943, 1945–46 (2017).

[170]. See Baker & Silver, supra note 164, at 56.

[171]. See, e.g., Vioxx Settlement, supra note 166, §§ 2.1–2.9 (defining injury and proximity “gates” that claimants must clear to be eligible for payment).

[172]. See, e.g., U.S. Durom Cup Settlement Program Agreement Between Zimmer, Inc. and Claimants’ Liaison Counsel 4–6 (Feb. 11, 2016), https://www.elizabethchambleeburch.com/_files/ugd/884328_99c7f1d7cafd4d08bdd074d9d3d010c3.pdf?index=true [https://perma.cc/UZZ4-RTHE] (defining compensation grid).

[173]. See, e.g., Vioxx Settlement, supra note 166, §§ 3.1–3.5 (defining points system for claim valuation).

   [174]. Model Rules of Pro. Conduct r. 1.2(a) (A.B.A. 2025).

[175]. See, e.g., Howard M. Erichson, A Typology of Aggregate Settlements, 80 Notre Dame L. Rev. 1769 (2005).

[176]. See Baker & Silver, supra note 164, at 58.

[177]. See id. at 65.

[178]. See Bradt & Rave, supra note 102, at 1284–92.

[179]. See Catherine R. Borden, Fed. Jud. Ctr., Managing Related Proposed Class Actions in Multidistrict Litigation (2018).

[180]. See, e.g., In re Oil Spill by Oil Rig “Deepwater Horizon”, 910 F. Supp. 2d 891, 900–02 (E.D. La. 2012).

[181]. E.g., Brown v. Am. Home Prods. Corp. (In re Diet Drugs (Phentermine, Fenfluramine, Dexfenfluramine) Prods. Liab. Litig.), Nos. 1203, 99-20593, 2000 WL 1222042 (E.D. Pa. Aug. 28, 2000); In re Nat’l Football League Players Concussion Inj. Litig., 821 F.3d 410 (3d Cir. 2016).

[182].Harrington v. Purdue Pharma L.P., 144 S. Ct. 2071, 2092 (2024) (Kavanaugh, J., dissenting) (quoting Casey & Macey, supra note 18).

   [183]. Casey & Macey, supra note 18, at 1005.

[184]. See D. Theodore Rave, Closure Provisions in MDL Settlements, 85 Fordham L. Rev. 2175 (2017); Elizabeth Chamblee Burch & Margaret S. Williams, Repeat Players in Multidistrict Litigation: The Social Network, 102 Corn. L. Rev. 1445, 1504–09 (2017).

[185]. See Rave, supra note 184 (describing publicly available settlements).

[186]. Seeid. at 2179–81.

[187]. Id.

   [188]. Exercising the walkaway option is the defendant’s prerogative, so if 9,999 out of 10,000 plaintiffs sign on, the defendant is not forced back into litigation with all 10,000 because of one nonsettling plaintiff. In practice, defendants almost never walk away when they have achieved substantial closure. See Baker, supra note 169, at 1951.

[189]. See Rave, supra note 184, at 2183–84 (describing Propulsid and World Trade Center settlements).

   [190]. Nagareda, supra note 24, at 218.

[191]. See Rave, supra note 184, at 2185–86.

[192]. See id.

[193]. See id. at 2189–96.

[194]. See Baker, supra note 169, at 1962–63.

   [195]. Vioxx Settlement, supra note 166, §§ 1.2.6.2, 1.2.8.1, 1.2.8.2.

[196]. See Rave, supra note 184, at 2191–96.

[197]. Id.

[198]. See, e.g., Howard M. Erichson & Benjamin C. Zipursky, Consent Versus Closure, 96 Corn. L. Rev. 265, 284–90 (2011).

   [199]. Rave, supra note 184, at 2196.

   [200]. MDL judges have approved many withdrawal requests without comment or fanfare, presumably finding implicitly that the withdrawal would not prejudice the client or that continued representation would impose an undue hardship on the lawyer funding the case on a contingency-fee basis. But MDL judges are not rubber stamps, and many judges have rejected plaintiffs’ lawyers’ requests to withdraw. See, e.g., McDaniel v. Daiichi Sankyo, Inc., 343 F. Supp. 3d 427, 431–33 (D.N.J. 2018); In re FEMA Trailer Formaldehyde Prods. Liab. Litig., No. 07-1873, 2011 U.S. Dist. LEXIS 105581 (E.D. La. Sep. 16, 2011); cf. Order, In re DePuy Orthopaedics, Inc. Pinnacle Hip Implant Prods. Liab. Litig., No. 11-md-2244-K (N.D. Tex. Dec. 17, 2019) (outlining procedure for lawyers seeking to withdraw without substituting new counsel).

   [201]. Rave, supra note 184, at 2196–2200.

   [202]. Some settlements require participating lawyers to affirm that they have “no present intention” of pursuing similar litigation against the defendant in the future. But such terms are essentially toothless, as the lawyer’s intentions could change moments after the settlement. Id. at 2197.

   [203]. Vioxx Settlement, supra note 166, § 1.2.8.2.

[204]. See Memorandum from Lynn A. Baker & Samuel Issacharoff to Plaintiffs’ Exec. Comm., Nat’l Prescription Opiate Litig., U.S. Dist. Ct. N.D. Ohio, MDL No. 17-MD-2804 1–2 (Aug. 6, 2021), https://nationalopioidsettlement.com/wp-content/uploads/2021/08/Baker-Issacharoff-Opinion-Letter-for-Opioids-MDL-PEC-Janssen-and-Distributors-8_6_21F-PDF.pdf [https://perma.cc/PSF2-SVMH].

[205]. See Grabill, supra note 147, at 155.

[206]. See Nora Freeman Engstrom, The Lessons of Lone Pine, 129 Yale L.J. 2, 43–46 (2019); Elizabeth Chamblee Burch, Nudges and Norms in Multidistrict Litigation: A Response to Engstrom, 129 Yale L.J.F. 64, 77 (2019).

[207]. See Rave, supra note 184, at 2185.

   [208]. D. Theodore Rave, Multidistrict Litigation and the Field of Dreams, 101 Tex. L. Rev. 1595, 1601 (2023).

[209]. Id. In the Zantac litigation, for example, the case census process revealed more than 150,000 unfiled claims, compared with fewer than 2,500 filed in the MDL. In re Zantac (Ranitidine) Prods. Liab. Litig., 644 F. Supp. 3d. 1075, 1096 (S.D. Fla. 2022).

[210]. SeeIn re Vioxx Prods. Liab. Litig., 802 F. Supp. 2d 740, 762 (E.D. La. 2011) (49,893 claimants enrolled); JPML, Statistical Analysis of Multidistrict Litigation Fiscal Year 2008 22, https://www.jpml.uscourts.gov/sites/jpml/files/JPML_Statistical%20Analysis%20of%20Multidistrict%20Litigation_2008.pdf [https://perma.cc/5ZKU-WHAJ] (9,580 cases pending as of September 30, 2008).

[211]. See Rave, supra note 184, at 2181–82 (citing Master Settlement Agreement § 1.01, In re Actos Prods. Liab. Litig., No. 11-md-2299 (W.D. La. Apr. 28, 2015) (“The purposes of the registration requirements . . . are to allow the Parties and the Courts to identify filed and unfiled cases and claims connected to ACTOS products.”)).

[212]. See, e.g., Vioxx Settlement, supra note 166, §§ 1.2.7, 1.2.8, 17.1.35 (requiring lawyers to recommend the settlement to all clients, whether or not their claims were filed in the MDL, and to withdraw from representing any who rejected the settlement).

   [213]. Distributor Settlement Agreement Between Janssen Pharmaceuticals and Settling States, Settling Distributors & Participating Subdivisions (Mar. 25, 2022), https://nationalopioidsettlement.com/distributor-janssen-settlement-documents/ [https://perma.cc/SQ4F-HGEH].

[214]. See, e.g., Casey & Macey, supra note 18, at 981; Resnick, supra note 11, at 2045–46.

[215]. See, e.g., Nagareda, supra note 37.

[216]. See 11 U.S.C. § 524(g) (providing mechanisms for channeling future asbestos claims into trust).

[217]. See Stephen J. Carroll, Deborah Hensler, Allan Abrahamse, Jennifer Gross, Michelle White, Scott Ashwood & Elizabeth Sloss, RAND, Asbestos Litigation Costs and Compensation: An Interim Report 25 (2002).

[218]. See, e.g., id.; Paul D. Carrington, Asbestos Lessons: The Unattended Consequences of Asbestos Litigation, 26 Rev. Litig., 583, 601 (2007).

   [219]. World Trade Center Litigation Settlement Process Agreement, As Amended at 32–36, 87, In re World Trade Ctr. Disaster Site Litig., 834 F. Supp. 2d 184 (S.D.N.Y. 2011) (No. 21-mc-00100).

   [220]. As the American Law Institute (ALI) has explained, avoiding structural conflicts of interest—those that “would present a significant risk that the lawyers for claimants might skew systematically the conduct of the litigation so as to favor some claimants over others on grounds aside from reasoned evaluation of their respective claims or to disfavor claimants generally vis-à-vis the lawyers themselves”—is key to legitimating resolution of cases on a class action basis. Am L. Inst., Principles of the Law of Aggregate Litigation § 2.07(a)(1)(B) (2010).

[221]. In re Nat’l Football League Players Concussion Inj. Litig., 821 F.3d 410, 420, 429–30 (3d Cir. 2016).

[222]. Id. at 429.

[223]. Id. at 423–45.

[224]. Id. at 425.

[225]. Id. at 433.

   [226]. 28 U.S.C. § 1407(a).

[227]. See supra Part II.B.2.

[228]. See supra Part II.B.3.

[229]. See, e.g., Bradt & Rave, supra note 102, at 1271.

[230]. See, e.g., Rave, supra note 23, at 1198–1201.

   [231]. In re Nat’l Football League Players Concussion Inj. Litig., 821 F.3d 410, 410–11 (3d Cir. 2016).

[232]. See, e.g., Burch, Judging Multidistrict Litigation, supra note 93, at 135; Silver & Miller, supra note 93, at 117–18.

[233]. See Bradt & Rave, supra note 102, at 1295–1301.

   [234]. The rate of trials in mass tort MDLs appears no different from the rate of trials in federal civil cases more generally—that is, less than 1 percent of cases are resolved by trial. Seesupra note 163 and accompanying text.

[235]. See Gibson, supra note 10; Randi Love, Advisers Reap $500 Million from Longest Asbestos Bankruptcy, Bloomberg L. (July 28, 2025), https://news.bloomberglaw.com/bankruptcy-law/advisers-reap-500-million-from-longest-asbestos-bankruptcy [https://perma.cc/RDH8-KBDB] (noting that the Georgia-Pacific/Bestwall Texas Two-Step bankruptcy has been pending for almost eight years and generated almost half a billion dollars in professional fees and costs).

[236]. See supra Part I.B.

   [237]. Casey & Macey, supra note 18, at 1007–10.

[238]. See Parikh, supra note 60, at 68–70 (explaining how funding agreement attempts to address fraudulent transfer risk).

[239]. Id.; see also Francus, supra note 59, at 1234 & n.171 (explaining how funding agreement protects tort claimants); Bussel, supra note 30, at 694 (“The two-step should not be constructed or viewed as a device to relieve the primary tortfeasor of any liability to claimants at all.”).

[240]. See, e.g., Francus, supra note 1, at 43; Levitin, supra note 15.

[241]. See, e.g., In re DBMP LLC, No. 20-30080, 2012 WL 3552350, at *18–20, 28 (Bankr. W.D.N.C. Aug. 11, 2021).

[242]. See id.

   [243]. Francus, supra note 59, at 1231–32, 1247 (“The result is that management might decline to pursue a fraudulent transfer action . . . [or settle one] on the cheap. . . . And because the debtor’s management control such fraudulent-transfer claims, tort victims will not be able to bring those cases on their own if management fall short.”); see also Jared A. Ellias, Ehud Kamar & Kobi Kastiel, The Rise of Bankruptcy Directors, 95 S. Cal. L. Rev. 1083, 1110 (2022) (noting that bankruptcy directors may settle claims for a fraction of their value).

   [244]. Casey & Macey, supra note 18, at 978, 1014–16; see also Brubaker, Two-Step Part I, supra note 30, at 5–6 (describing how Two-Step structurally subordinates tort claimants to all other creditors of the business).

[245]. LTL 2.0, 652 B.R. 433, 448 (Bankr. D.N.J. 2023). After the Third Circuit held that the generosity of the funding agreement meant that LTL was not in financial distress, LTL agreed with J&J to terminate the funding agreement and replace it with a much less generous funding agreement (up to $30 billion) with the GoodCo that was not guaranteed by the J&J parent corporation. Id. The ostensible justification for the termination and substitution of the funding agreement was J&J’s contention that the dismissal of the first bankruptcy drew the continuing enforceability of the funding agreement into question. Id. at 438–40.

[246]. See 11 U.S.C. §§ 544(b), 548.

   [247]. Id. § 362(a).

[248]. See, e.g., Lipson, supra note 50, at 71. As Lipson notes, the standard for obtaining such a preliminary injunction requires a showing of likelihood of success in confirming a plan of reorganization, not a likelihood of success on the underlying merits. Id. at 75. Harrington’s rejection of nonconsensual third-party releases has not eliminated bankruptcy courts’ ability to extend preliminary injunctions against third-party claims where litigation of “those claims would interfere with the debtor’s reorganization efforts.” In re Parlement Techs., Inc., 661 B.R. 722, 728 (Bankr. D. Del. 2024).

[249]. See McKenzie, supra note 12, at 1004.

[250]. See Bussel, supra note 30, at 713–14.

   [251]. Like any district court judge, MDL judges have tremendous discretion over case management. See Fed. R. Civ. P. 16; Rave, supra note 111, at 298–300, 304–08. That discretion is emphasized further in the newly adopted Rule 16.1, which provides guidance on case management in MDLs. Fed. R. Civ. P. 16.1.

[252]. See, e.g., Rave & McGovern, supra note 128, at 37.

   [253]. Lexecon, Inc. v. Milberg Weiss Bershad Hynes & Lerach, 523 U.S. 26 (1998).

[254]. See JPML R. P. 10.1(b), 10.3(a); Burch, supra note 107, at 400–01.

[255]. See, e.g., Elizabeth Chamblee Burch, Mass Tort Deals: Backroom Bargaining in Multidistrict Litigation 99–133 (2019) (arguing MDL judges “nudge” settlements); Parikh, supra note 84, at 476–77.

[256]. See, e.g., Silver & Miller, supra note 93, at 123–24.

[257]. See Clopton & Rave, supra note 151, at 261–62, 270–71.

[258]. See Bradt, supra note 140, at 796 n.206.

[259]. See, e.g., Bolch, supra note 124, at 21–22.

[260]. See Fallon et al., supra note 142.

   [261]. 28 U.S.C. § 157(b)(5).

[262]. See 11 U.S.C. § 362(d)(1).

[263]. See 28 U.S.C. § 157(b)(2)(B).

[264]. See, e.g., Yun Park, Insurers Object to Litigation ‘Test Cases’ in Diocese’s Ch. 11, Law360 (Aug. 5, 2025) (quoting insurers’ counsel’s assertion that “[i]n 200-plus mass tort asbestos, pharmaceutical and product liability bankruptcies, our research has not revealed any where a bankruptcy court adopted ‘test cases’”).

   [265]. Bankruptcy courts generally lack statutory authority to make final distributions to creditors in Chapter 11 proceedings except pursuant to the terms of a confirmed plan of reorganization. See Czyzewski v. Jevic Holding Corp., 580 U.S. 451, 464–68 (2017); see also Francus, supra note 59, at 1232 (“[B]ecause the funding agreement conditions payment on the victims voting to confirm a plan, victims have no way to obtain payment during the pendency of the bankruptcy.”).

[266]. See, e.g., Janger, supra note 16, at 381 (“Bankruptcy courts have an extraordinary power to discharge debts, while granting juridic entities continued life. Compliance with process is the price of that power.”); Edward J. Janger & Adam J. Levitin, The Proceduralist Inversion—A Response to Skeel, 130 Yale L.J.F. 335, 336 (2020) (arguing that procedural protections are the link between prebankruptcy entitlements and bankruptcy distributions); Jacoby, supra note 15, at 411 (“Bankruptcy à la carte extracts the tools of Chapter 11 meant to be available only as part of a package deal and redistributes the benefits.”).

[267]. See Bussel, supra note 30, at 698–99.

[268]. See id. at 714–15.

   [269]. By default, the Bankruptcy Code gives debtors the exclusive right to propose a plan of reorganization for four months after filing a petition, and it authorizes bankruptcy courts to extend the period of exclusivity up to eighteen months. 11 U.S.C. § 1121(d)(2)(A). In practice, the hiatus from litigation might last even longer; Georgia-Pacific took its BadCo subsidiary Bestwall into bankruptcy in 2017, and the case remains pending in the bankruptcy court as of this writing eight years later. See Bestwall LLC v. Off. Comm. of Asbestos Claimants of Bestwall, LLC, 148 F.4th 233, 248 n.2 (4th Cir. 2025) (King, J., dissenting).

[270]. See, e.g., Love, supra note 235 (noting that the Georgia-Pacific/Bestwall Two-Step bankruptcy has been pending for nearly eight years).

   [271]. The total professional fees in J&J’s first two talc bankruptcies have been estimated at $178 million ($116 million for LTL 1.0 and $62 million for LTL 2.0) during the twenty-one months it spent in bankruptcy (including appeals). See Evan Ochsner, J&J Unit’s Failed ‘Two-Step’ Talc Bankruptcies Cost $178 Million, Bloomberg L. (Oct. 4, 2023), https://news.bloomberglaw.com/bankruptcy-law/j-j-units-failed-two-step-talc-bankruptcies-cost-178-million [https://perma.cc/W8MG-BCF4]. Before filing for bankruptcy the first time, J&J’s talc litigation defense costs—not including judgments and settlements—were running $10–20 million per month. In re LTL Mgmt. LLC, 64 F.4th 84, 94 (3d Cir. 2023). Thus, J&J avoided something on the order of $210–420 million in defense costs alone during LTL’s bankruptcies, again not including any judgments or settlements it might have paid. Plaintiffs, of course, recovered nothing in that time period. See also In re Aldrich Pump, No. 20-30608, 2023 WL 9016506, at *10 (Bankr. W.D.N.C. Dec. 28, 2023) (noting that “for the past three years, both the Debtors and the Affiliates have enjoyed a respite from the tort system and a ‘payment holiday’ from the $100 million-a-year costs they were previously incurring”).

   [272]. Casey & Macey, supra note 18, at 1007–09.

   [273]. Casey and Macey say that

[b]ankruptcy’s procedural rights are designed to protect creditors from opportunistic maneuvers by debtors and other creditors (especially those with more sophistication). They thus provide a benefit when a creditor risks taking a loss, but are counterproductive when they protect creditors who face no such risk. Because mass tort bankruptcies are designed to resolve only tort liability (not other financial claims), there is no reason to open the process to the participation of other financial creditors.

Id. at 1008. Notably, other mass tort bankruptcy proponents start from the opposite assumption that bankruptcy’s strength is that it recognizes that the problem cannot be confined to tort claimants alone and brings to the table other stakeholders who are necessarily impacted. See Resnick, supra note 11, at 2061; S. Elizabeth Gibson, A Response to Professor Resnick: Will This Vehicle Pass Inspection?, 148 U. Pa. L. Rev. 2095, 2099 (2000); cf. Francus, supra note 59, at 1236 (noting that the asset partitioning created by a Two-Step bankruptcy is incomplete because the pools of assets created by the funding agreement overlap).

[274]. LTL 1.0, 637 B.R. 396, 416–17 (Bankr. D.N.J. 2022) (“It is appropriate to note that the true leverage remains where Congress allocated such leverage, with the tort claimants who must approve of any plan employing a § 524(g) trust by a 75% super majority.”).

   [275]. 11 U.S.C. § 1126(c).

   [276]. 11 U.S.C. § 524(g)(2)(B)(ii)(IV)(bb); see alsoIn re Quigley Co., 346 B.R. 647, 653 (Bankr. S.D.N.Y. 2006) (explaining voting rules in asbestos bankruptcies). In non-asbestos mass tort bankruptcies that are not governed by § 524(g), courts have generally required “overwhelming consent” of the claimant class in excess of the 75% required by § 524(g). See Bussel, supra note 30, at 695–96, 696 n.34 (citing In re Purdue Pharma L.P., 69 F.4th 45, 78-79 (2d Cir. 2023), rev’d sub nom., Harrington v. Purdue Pharma L.P., 144 S. Ct. 2071 (2024)).

   [277]. Rave, supra note 23, at 1247–50.

[278]. Id. at 1233.

[279]. Id.See generally Albert O. Hirschman, Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States (1970) (describing “exit” and “voice” as alternative levers of governance).

   [280]. Casey & Macey, supra note 22, at 1040–41.

[281]. Id.

[282]. See Rave, supra note 23, at 1233, 1247–49. As scholars have shown in the analogous context of sovereign debt, collective action clauses in sovereign bonds that bind bondholders to a supermajority vote on any restructuring plan preserve bondholders’ collective leverage better than “exit consents,” which aim to deter holdouts by making individual litigation less attractive. See, e.g., Ran Bi, Marcos Chamon & Jeromin Zettelmeyer, The Problem that Wasn’t: Coordination Failures in Sovereign Debt Restructurings 16–17 (IMF Working Paper No. 11/265, 2011); Sean Hagan, Designing a Legal Framework to Restructure Sovereign Debt, 36 Geo. J. Int’l L. 299, 343–44 (2005).

   [283]. A.L.I., Principles of the Law of Aggregate Litigation § 3.17 (2010).

[284]. See W. Va. R. Pro. Conduct § 1.8 cmt. 17 (2015) (adopting a version of the ALI proposal for mass tort cases); N.Y.C. Bar Ass’n Comm. on Pro. and Jud. Ethics, Aggregate Settlements – Formal Opinion 2009-06 (2010) (rejecting ALI proposal).

[285]. SeeIn re Nat’l Prescription Opiate Litig., 332 F.R.D. 532 (N.D. Ohio 2019); Francis E. McGovern & William B. Rubenstein, The Negotiation Class: A Cooperative Approach to Class Actions Involving Large Stakeholders, 99 Tex. L. Rev. 73 (2020).

[286]. In re Nat’l Prescription Opiate Litig., 976 F.3d 664, 676–77 (6th Cir. 2020).

[287]. See Samuel Issacharoff, Pamela S. Karlan, Richard H. Pildes, Nathaniel Persily & Franita Tolson, The Law of Democracy: Legal Structures of the Political Process 1 (6th ed. 2022) (“[E]lections do not take place in a legal or institutional vacuum. . . . At the heart of a democratic political order lies a process of collective decisionmaking that must operate through pre-existing laws, rules, and institutions. The kind of democratic politics we have is always and inevitably itself a product of institutional forms and legal structures.”).

   [288]. Not coincidentally, these problems track major challenges in political elections, namely access to the franchise, apportionment of representation, and gerrymandering of legislative districts.

[289]. See, e.g., Melissa B. Jacoby, Sorting Bugs and Features of Mass Tort Bankruptcy, 101 Tex. L. Rev. 1745, 1756–57 (2023).

[290]. See, e.g., Lipson, supra note 50, at 91, 94 (noting that less than 20 percent of eligible claimants voted on the Purdue Pharma plan); Lipson & Foohey, supra note 77, at 896–97.

   [291]. 11 U.S.C. §§ 1125(a)(1), (b).

[292]. See 11 U.S.C. § 1126(a); Grady v. A.H. Robins Co., 839 F.2d 198, 203 (4th Cir. 1988) (holding that “Congress has created a contingent right to payment” for future tort claimants).

[293]. See, e.g., Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 628–29 (1997) (rejecting class certification).

[294]. Cf. Mullane v. Cent. Hanover Bank & Tr. Co., 339 U.S. 306, 317–19 (1950) (explaining that due process requires notice that is reasonable under the circumstances, not notice that is unreasonably expensive given the stakes); Mathews v. Eldridge, 424 U.S. 319, 335 (1976) (holding that due process requires balancing the private interest, the risk of error, and the costs of additional procedures).

[295]. See, e.g., Gibson, supra note 11, at 130 (“It is a common practice in mass tort bankruptcy cases to send master ballots directly to all lawyers known to be representing persons with tort claims against the debtor and to allow them to vote to accept or reject the plan on behalf of each client who has authorized them to do so.”); Lipson, supra note 50, at 91–92.

[296]. See Gibson, supra note 11, at 130.

[297]. In re Red River Talc LLC, 670 B.R. 251, 286–92 (Bankr. S.D. Tex. 2025).

[298]. Id. at 286–97.

[299]. Id. at 286–92, 290 n.308.

[300]. Id. at 278.

[301]. Id. at 286 (“Master ballots are incredibly useful in bankruptcy cases. The Master Ballot used in this case adheres to the Bankruptcy Code and it could have worked here.”).

   [302]. The practice of treating clients’ failure to respond to notice of how the firm intends to vote as “negative consent” was not limited to these two firms and appeared to be common in the J&J bankruptcy. See id. at 295. Other law firms claimed a “power of attorney” to vote their clients’ claims based on language in their retainer agreements giving them authority to “manage and handle . . . claims as [the lawyer] deem[s] proper and to investigate and prosecute them, with or without filing a lawsuit, in any manner [the lawyer] deem[s] advisable.” Id. at 281, 286. The bankruptcy court found such a general delegation of authority insufficient for a lawyer to vote his client’s claims but approved of voting by master ballot under more specific powers of attorney. Id. at 284–86 (“[A] specific power of attorney granting authority to vote is required.”).

   [303]. Model Rules of Pro. Conduct r. 1.8(g) (A.B.A. 2025).

[304]. See, e.g., Tax Auth., Inc. v. Jackson Hewitt, Inc., 898 A.2d 512 (N.J. 2006).

[305]. See Rave, supra note 23, at 1250–51. See generally McGovern & Rubenstein, supra note 285.

[306]. See D. Theodore Rave, Two Problems of Fiduciary Governance, in Fiduciary Government 49, 50–51 (Evan J. Criddle, Evan Fox-Decent, Andrew S. Gold, Sung Hui Kim & Paul B. Miller eds. 2018) [hereinafter Rave, Two Problems of Fiduciary Governance]; D. Theodore Rave, Fiduciary Voters?, 66 Duke L.J. 331, 343–47 (2016).

[307]. See 11 U.S.C. §§ 1122(a), 1126(c).

   [308]. In theory, the bankruptcy rules allow judges to separate tort claimants into different voting classes based on the relative value of their claims. Fed. R. Bankr. P. 3013. But, as Bussel has explained, this classification approach “is virtually never used.” Bussel, supra note 30, at 726. One likely reason why not is that constructing appropriate subclasses in advance of the vote requires knowing a great deal about the relative valuation of the different claims. More on that below in Part III.E. Bussel argues that bankruptcy courts should be required to assess the propriety of classification before soliciting votes and suggests that tort claimants might be sorted into voting classes based on how they wish to be treated in the plan of distribution. Id. at 727–28. If they want quick payments in exchange for a minimal evidentiary showing, they opt into one voting class; if they want access to more substantial payments under a grid or point system, they opt into a separate voting class; and so on. See id. at 727. The feasibility of such a self-sorting approach may be limited by the fact that the different categories of relief available under the plan of distribution are going to be negotiated in the shadow of the vote. But subclassing would certainly be an improvement over the status quo in most mass tort bankruptcies.

[309]. See, e.g., Menard-Sanford v. Mabey (In re A.H. Robins Co.), 880 F.2d 694, 696, 698 (4th Cir. 1989) (affirming the district court’s assignment of one dollar value to each unliquidated tort claim for voting purposes); see also Jacoby, supra note 289, at 1758 (noting ubiquity of this practice).

[310]. See, e.g., In re Paddock Enters., LLC, No. 20-10028, 2022 WL 1746652 (Bankr. D. Del. May 31, 2022); In re Quigley Co., 346 B.R. 647, 654–56 (Bankr. S.D.N.Y. 2006).

[311]. See, e.g., In re A.H. Robins, 880 F.2d at 697–98 (reasoning that “any attempt to evaluate each of the 195,000 individual claims for voting purposes would cause intolerable delay”).

   [312]. My research assistant compiled a list of mass tort bankruptcies going back to 2005. Of the sixteen cases where we were able to discern an estimation process for voting purposes, thirteen uniformly valued claims at one dollar, and three (all asbestos bankruptcies) used the category class method. Results on file with author. This pattern is consistent with Gibson’s findings about estimation for voting in mass tort bankruptcies prior to 2005. Gibson, supra note 11, at 132–33.

[313]. See Bussel, supra note 30, at 724–25; Brubaker, Two-Step Part I, supra note 30, at 14. Casey and Macey admit that the one-dollar-one-vote method is “problematic” and suggest that “it is likely insufficient to meet the Code’s requirement that courts estimate claims.” Casey & Macey, supra note 22, at 1049–50. They suggest that bankruptcy courts should “develop a uniform and robust claims-estimation process.” Id. at 1050. Unfortunately, they do not identify any reforms that would deal with the valuation challenges discussed below. See infra Part III.E.

[314]. See, e.g., Kane v. Johns-Manville Corp., 843 F.2d 636, 645, 647 (2d Cir. 1988); In re A.H. Robins, 880 F.2d at698; see also Bussel, supra note 30, at 721–24 (noting the trend and cataloging cases where over 90 percent of mass tort claimants voted in favor of reorganization).

[315]. See, e.g., In re Quigley Co., 346 B.R. 647, 656, 659 (Bankr. S.D.N.Y. 2006) (rejecting plan because per capita vote was close enough to call into question whether the two-thirds requirement would have been satisfied if claims had been accurately valued).

[316]. See Bussel, supra note 30, at 724–25; Gibson, supra note 273, at 2111–13; S. Todd Brown, How Long Is Forever This Time? The Broken Promise of Bankruptcy Trusts, 61 Buff. L. Rev. 537, 557 (2013) (“These attorneys who can deliver sufficient votes to satisfy the supermajority vote requirement exercise considerable control over the design of the trust, appointments to leadership roles within the trust, and the distribution procedures that define the process for reviewing and paying claims.”).

[317]. See, e.g., Multidistrict Litigation Subcommittee, Multidistrict Litigation Subcommittee Report, in April 1, 2020 Meeting of the Advisory Committee on Civil Rules 145, 145–46 (2020), https://www.uscourts.gov/sites/default/files/04-2020_civil_rules_agenda_book.pdf [https://perma.cc/KD7K-7XY2] (describing proposals to address “what might be called the ‘Field of Dreams’ problem”).

[318]. See, e.g., Samir D. Parikh, The Alchemist’s Inversion, 110 Corn. L. Rev. 1693, 1697 (2026).

   [319]. Rave, supra note 208, at 1606–12.

[320]. See Bussel, supra note 30, at 717 & n.95.

   [321]. Some have said this is what happened in LTL 2.0. In that case, a handful of lawyers recruited large numbers of new talc claimants who had never filed claims in the MDL but were willing to sign plan support agreements. 652 B.R. 433 (Bankr. D.N.J. 2023). Together, they negotiated a plan of reorganization prior to LTL’s second bankruptcy filing that the MDL leadership found grossly inadequate. Id. The problem may have been even worse in J&J’s third bankruptcy (Red River Talc), which was structured as a prepackaged deal. In re Red River Talc LLC, 670 B.R. 251 (Bankr. S.D. Tex. 2025) (No. 24-90505); see, e.g., Adam Levitin, If You’re Gonna File in Texas, You Gotta Have Your Votes in Hand, Credit Slips (Sep. 24, 2024), https://creditslips.org/2024/09/24/if-youre-gonna-file-in-texas-you-gotta-have-your-votes-in-hand/ [https://perma.cc/3NMV-HRHW]; Adam Levitin, Stuffing the Chapter 11 Ballot Box With “Junk” Claims, Credit Slips (Apr. 26, 2024), https://creditslips.org/2024/04/26/stuffing-the-chapter-11-ballot-box-with-junk-claims/ [https://perma.cc/LES2-4T53]; see alsoIn re Glob. Indus. Techs., Inc., 645 F.3d 201, 204–07, 207 n.14 (3d Cir. 2011) (describing an increase from 169 pre-bankruptcy silica claims to 5,125 voting silica claims, a significant number of which appeared to be meritless).

[322]. See, e.g., Bussel, supra note 30, at 717–19; see also Lipson, supra note 50, at 50; Douglas G. Baird, Bankruptcy’s Quiet Revolution, 91 Am. Bankr. L.J. 593, 606 (2017).

   [323]. In J&J’s third talc bankruptcy, for example, the master ballot sent to plaintiffs’ lawyers required them to list all of their clients, identify their disease, and certify that they had used talc in the perineal area. It expressly directed lawyers not to submit supporting documentation, and it promised that if the certifications were completed, the debtor would not object to the client’s claim “for voting purposes only.” The debtor did, however, reserve the right to object to claims when it came time for payment. Exhibit A, Master Ballot for Counsel to Holders of Channeled Talc Personal Injury Claims Voting on Prepackaged Chapter 11 Plan of Reorganization of the Debtor at 5–6, In re Red River Talc LLC, 670 B.R. 251 (Bankr. S.D. Tex. 2025) (No. 24-90505), Dkt No. 430-1.

[324]. Contra In re Purdue Pharma L.P., 633 B.R. 53, 61 (Bankr. S.D.N.Y. 2021) (rejecting objections to a plan’s confirmation based on inadequate turnout due to overwhelming support by those votes that were cast).

[325]. See Gibson, supra note 11, at 75–77. Bussel suggests imposing filing fees on mass tort claimants, placing small claims in a separate voting class, or sanctioning lawyers who bring meritless claims in bankruptcy. Bussel, supra note 30, at 719. But none of those approaches get at the problem of how to figure out which claims are the wheat and which are the chaff.

[326]. See, e.g., In re Imerys Talc Am., Inc., No. 19-10289, 2021 WL 4786093, at *10–12 (Bankr. D. Del. Oct. 13, 2021) (disqualifying master ballot voting of 15,719 claims).

[327]. See Nagareda, supra note 37, at 159 (“Class members’ preexisting rights to sue truly must be purchased rather than simply appropriated . . . .”).

   [328]. Rave, Two Problems of Fiduciary Governance, supra note 306, at 50.

   [329]. McGovern & Rubenstein, supra note 285, at 79.

[330]. Id. at 79–80.

   [331]. A.L.I., Principles of the Law of Aggregate Litigation § 3.17 (2010).

   [332]. Rave, supra note 23, at 1229, 1246–56

[333]. See supra Part II.B.4.b.

[334]. See, e.g., Burch, supra note 255, at 138; Burch & Williams, supra note 184; Erichson & Zipursky, supra note 198.

   [335]. Casey & Macey, supra note 22, at 1040–41, 1051 (“In contrast [to bankruptcy], MDL is widely known and criticized for using coercive terms to force dissenting claimants to join settlements.”).

   [336]. If other lawyers stand ready and able to represent nonsettling claimants, their choice is largely preserved; if the settlement disrupts the market for legal services to interfere with referrals and new entrants, their choice is much more limited. See Rave, supra note 184, at 2196–201.

[337]. See supra note 175 and accompanying text.

[338]. See Bradt & Rave, Defense of Repeat Players, supra note 115, at 114.

   [339]. Burch, supra note 255, at 202–03; Burch & Williams, supra note 93, at 1880–81.

   [340]. Bradt & Rave, Defense of Repeat Players, supra note 115, at 101 n.178; Rave, supra note 23, at 1216 n.122.

   [341]. I have, for example, argued for increased judicial oversight of MDL settlements, Bradt & Rave, supra note 102, at 1264–65, and in favor of the ALI’s voting proposal, Rave, supra note 23, at 1252–54.

   [342]. Bradt & Rave, Defense of Repeat Players, supra note 115.

   [343]. 3M had set aside approximately $1 billion to fund its earplug liabilities in bankruptcy; it settled for $6.01 billion in the MDL after the bankruptcy was dismissed. See Brendan Pierson, 3M Agrees to Pay $6 Billion in US Military Earplug Lawsuit Settlement, Reuters (Aug. 29, 2023), https://www.reuters.com/legal/3m-co-agrees-pay-6-billion-earplug-lawsuit-settlement-2023-08-29/ [https://perma.cc/47PQ-X5BK]. In Purdue, the Sacklers increased their contribution to the deal from $6 billion to $6.5 billion after the Supreme Court eliminated their ability to obtain nonconsensual third-party releases. See Geoff Mulvihill, Purdue Pharma and Owners to Pay $7.4 Billion in Settlement of Lawsuits Over the Toll of OxyContin, Associated Press (Jan. 23, 2025), https://apnews.com/article/purdue-pharma-sackler-settlement-opioid-lawsuits-ea6c89aa9cafc8fdd18fabfad503eeea [https://perma.cc/P58D-48V5].

[344]. See, e.g., Brubaker, No-Opt-Outs, supra note 45, at 116 (settlement of future claims “just cannot be done outside of bankruptcy”); see also id. at 117 (“[B]ankruptcy is the only process that can effectively resolve future mass-tort liability.”).

[345]. See, e.g., In re Nat’l Football League Players Concussion Inj. Litig., 821 F.3d 410, 429–30 (3d Cir. 2016).

   [346]. Bankruptcy requires only “substantial similarity” to place creditors in a class, not the “predominance and superiority” required by Rule 23(b)(3). See Fed. R. Civ. P. 23; Gibson, supra note 273, at 2107. And mass tort bankruptcies tend to lump all tort creditors into a single class, ignoring differences that would require subclasses with separate representation in a class action. See sources cited supra note 308 and accompanying text. This is not simply a difference in the governing procedural rules, as bankruptcy’s lack of subclasses ignores the due process foundations of Rule 23’s requirements. See Gibson, supra note 273, at 2107–08; Brubaker, Two-Step Part I, supra note 30, at 15 (calling this the “original sin of mass-tort bankruptcies”).

[347]. See, e.g., Gibson, supra note 11, at 79–80; Ralph Brubaker, Back to the Future Claim: Due Process in and Beyond the Mass Tort Reorganization (Part I), 34 Bankr. L. Letter No. 11 1 (2014).

[348]. See 11 U.S.C. § 524(g)(2)(B)(ii)(IV)(bb); Parikh, supra note 84, at 484 n.290 (“Neither future claimants—beneficiaries of the trust—nor their appointed representative are included in the voting class.”).

[349]. See Resnick, supra note 11, at 2078 (arguing that Code should be amended to make future demands count as statutory “claims” and thus trigger these protections); Brubaker, Two-Step Part I, supra note 30, at 16 (arguing that future claimants who cannot vote should not be deemed to have accepted the plan and thus equity should receive nothing unless future claims are paid in full, though acknowledging that this “would require a dramatic change in the prevailing practice”).

[350]. See, e.g., Frederick Tung, The Future Claims Representative in Mass Tort Bankruptcy: A Preliminary Inquiry, 3 Chap. L. Rev. 43, 59 (2000) (“The fundamental and unavoidable problem is that the purported principals play no part in choosing or monitoring their agent or even in deciding whether to retain one at all.”); Nagareda, supra note 37, at 176–79; Campos & Parikh, supra note 96.

   [351]. Yair Listokin & Kenneth Ayotte, Protecting Future Claimants in Mass Tort Bankruptcies, 98 Nw. U. L. Rev. 1435, 1451–52, 1457–60 (2004).

[352]. See, e.g., In re Nat’l Football League Players Concussion Inj. Litig., 821 F.3d 410 (3d Cir. 2016).

[353]. See Gibson, supra note 11, at 68.

[354]. See id.

[355]. See Nagareda, supra note 37, at 176–79.

[356]. Seeid.; Tung, supra note 350, at 72, 75–76; Parikh, supra note 84, at 490–91.

[357]. See generally Burch & Williams, supra note 184; Silver & Miller, supra note 93.

[358]. See Nagareda, supra note 37, at 177 (“For a mass tort plaintiffs’ lawyer, however, the incentive to become a repeat player operates in consonance with the incentive provided by contingency fee arrangements to maximize the recovery for a given client or group thereof.”). Nagareda suggests that merely shifting to a contingent fee for the future claims representative would not cure the problem. “Contingent compensation would not supply the leverage that the futures representative lacks by virtue of the absence of actual clients.” Id. at 176–77.

[359]. See, e.g., Campos & Parikh, supra note 96, at 354 (“Rule 23 may not represent a flawless approach, but it is a viable one. Unfortunately, few of the safeguards and practices delineated above are apparent in the FCR selection process.”).

[360]. See, e.g., Vancouver Women’s Health Collective Soc’y v. A.H. Robins Co., 820 F.2d 1359, 1364 (4th Cir. 1987) (“A bankrupt estate’s resources are always limited and the bankruptcy court must use discretion in balancing these interests when deciding how much to spend on notification.”).

[361]. See, e.g., Tung, supra note 350, at 54, 61.

   [362]. Harrington v. Purdue Pharma L.P., 144 S. Ct. 2071, 2073 (2024).

[363].World Trade Center Litigation Settlement Process Agreement, As Amended at 32–36, 87, In re World Trade Ctr. Disaster Site Litig., 834 F. Supp. 2d 184 (S.D.N.Y. 2011) (No. 21-mc-00100) (pairing upfront interim payments to all qualified claimants with an insurance policy that would pay later if claimants developed cancer).

   [364]. In re Nat’l Football League Players Concussion Inj. Litig., 821 F.3d 410 (3d Cir. 2016).

[365]. See, e.g., Casey & Macey, supra note 18, at 991 & n.66, 1012–17; Resnick, supra note 11, at 2081–82.

[366]. E.g., Harrington v. Purdue Pharma L.P., 144 S. Ct. 2071, 2090 (2024) (Kavanaugh, J., dissenting); Parikh, supra note 84, at 477–79. Others are less sanguine. Bussel cautions against rushing to bankruptcy until there has been “a significant incubation period in the tort system in which the litigation matures” into a track record of settlements and judgments that can be used as baselines for valuation. Bussel, supra note 30, at 705–06.

[367]. See, e.g., LTL 1.0, 637 B.R. 396, 409 (Bankr. D.N.J. 2002).

[368]. Cf. Phillips Petroleum Co. v. Shutts, 472 U.S. 797, 807 (1985) (explaining that legal claims are constitutionally protected property).

[369]. See, e.g., Francus, supra note 59, at 1245 (“The benefits of coordinating creditor collection and imposing a workers’ compensation scheme come with little or no downside.”).

[370]. See Butner v. United States, 440 U.S. 48, 55–57 (1979) (holding that bankruptcy courts must follow state property law); Brubaker, supra note 17, at 970–71.

[371]. See Nagareda, supra note 24, at 156–57, 195–96.

[372]. See McGovern, supra note 131, at 1843.

[373]. See Bradt & Rave, supra note 102, at 1290–91; Bone, supra note 129, at 236–38.

[374]. See 28 U.S.C. § 1411(a); see also 28 U.S.C. § 157(b)(2)(B) (excluding “liquidation or estimation of contingent or unliquidated personal injury or wrongful death claims” from core bankruptcy proceedings).

[375]. See Gibson, supra note 11, at 89–90; Resnick, supra note 11, at 2082.

[376]. See Gibson, supra note 11, at 89; Campos & Parikh, supra note 96, at 337 & n.79.

   [377]. Gibson, supra note 11, at 68, 91–92, 100–04; Parikh, supra note 84, at 492.

   [378]. Gibson, Management, supra note 11, at 82.

[379]. SeeIn re Dow Corning Corp., 211 B.R. 545, 562 n.16 (Bankr. E.D. Mich. 1997) (noting that estimation of mass tort liability “rests on the shaky foundation that judges can accurately estimate the results of a series of extremely speculative problems”); Bussel, supra note 30, at 705–06 (suggesting that access to bankruptcy be conditioned on a “significant incubation period in the tort system in which the litigation matures before forced collective resolution is invoked”). See generally Nat’l Bankr. Rev. Comm., Chapter 2: Business Bankruptcy, in Bankruptcy: The Next Twenty Years 303 (1997) (discussing complexity of valuing claims in mass torts).

[380]. See Brubaker, Two-Step Part I, supra note 30, at 12–13.

   [381]. Mediation is hardly unique to bankruptcy. If determining a top-line estimate of aggregate liability would be useful in reaching settlement, the MDL judge could direct the parties to mediation. Indeed, MDL judges frequently appoint special masters to help with settlement negotiations. Or the parties could engage a mediator of their own accord.

[382]. See 11 U.S.C. § 502(c).

   [383]. Parikh, supra note 84, at 492.

   [384]. My research assistant compiled a list of all mass tort bankruptcies since 2005, searched them for the overinclusive terms “bellwether,” “trial,” and “jury,” and did not locate any example of a bankruptcy court conducting a bellwether trial for estimation.

[385]. See Stern v. Marshall, 564 U.S. 462, 502–03 (2011) (explaining how the bankruptcy court’s decision on the state law counterclaim violated Article III of the Constitution); In re Dow Corning Corp., 211 B.R. 545, 579 (Bankr. E.D. Mich. 1997) (addressing due process concerns).

   [386]. Gibson, supra note 10, at 195.

[387]. Contra Resnick, supra note 11, at 2084 (asserting that “there is no reason to believe that a bankruptcy judge would be any better, nor any worse, than other judges” in valuing tort claims).

   [388]. Gibson, supra note 10, at 190.

[389]. See Casey & Macey, supra note 22, at 1018.

   [390]. Francis E. McGovern, Resolving Mature Mass Tort Litigation, 69 B.U. L. Rev. 659, 686 (1989).

   [391]. Gibson, supra note 11, at 14–15.

   [392]. The bankruptcy court rejected this plan, and the parties ultimately reached an agreement on the debtor’s aggregate liability for purposes of confirming a plan of reorganization. SeeIn re Dow Corning Corp., 211 B.R. 545, 558–60 (Bankr. E.D. Mich. 1997); Gibson, supra note 10, at 221–23.

[393]. See, e.g., In re 3M Combat Arms Earplug Litig., No. 19-md-2885, 2022 WL 17853203 (N.D. Fla. Dec. 22, 2022); see also Bradt, Clopton & Rave, supra note 14, at 1858.

   [394]. Campos & Parikh, supra note 96, at 330–31, 357–58.

[395]. SeeIn re PG&E Corp., No. 19-30088, 2019 WL 3889247, at *1–2 (Bankr. N.D. Cal. Aug. 16, 2019) (lifting stay as to eight plaintiffs for expedited trials to “provide an important data point that most likely will facilitate resolution of the wildfire tort claims in this case”). While bellwether trials (whether in the MDL court, in transferor courts, or in parallel state proceedings) are a standard feature of MDL practice, they do not appear to be common in mass tort bankruptcies. Casey and Macey assert that it is “easy to find” examples of bankruptcy courts referring tort cases to other state or federal courts for trial. Casey & Macey, supra note 22, at 1033. But, unlike the PG&E bankruptcy, none of the sixteen cases they cite involve anything like a bellwether process. Most of their examples did not even involve mass torts, and the three that did involved lifting the stay to allow insurance disputes or claims against nondebtors to go forward. See Desimone Hosp. Servs., LLC v. W. Va.-Am. Water Co., No. 14-cv-14845, 2014 WL 1577051 (S.D.W. Va. Apr. 16, 2014) (claims against nondebtor); In re Roman Cath. Church for Archdiocese of New Orleans, No. 21-1238, 2021 WL 3772062 (E.D. La. Aug. 25, 2021) (counterclaim by nondebtor against creditor); Roman Cath. Diocese of Rockville Ctr. v. Certain Underwriters at Lloyds, London & Certain London Mkt. Cos., 634 B.R. 226 (S.D.N.Y. 2021) (insurance coverage dispute).

   [396]. Campos & Parikh, supra note 96, at 358.

[397]. Id.

[398]. See, e.g., Fallon, Grabill & Wynne, supra note 142, at 2343.

[399]. See Gibson, supra note 11, at 89–90.

[400]. See Gibson, supra note 10, at 198, 225.

[401]. See, e.g., Brubaker, Two-Step Part I, supra note 30, at 12–13.

[402]. See, e.g., Simon, supra note 15, at 1167–68.

[403]. See Lloyd Dixon, Geoffrey McGovern & Amy Coombe, RAND, Asbestos Bankruptcy Trusts: An Overview of Trust Structure and Activity with Detailed Reports on the Largest Trusts (2010); cf. D. Theodore Rave, Settlement,ADR, and Class Action Superiority, 5 J. Tort L. 91, 98 (2012) (explaining how class action settlements create forms of ADR). Unless the tort is very mature, however, this process will be based on less information than would be available in an MDL.

   [404]. There may also be limits on the timing of payment. For example, the trust may make interim payments and give the claimants the right to come back for a second-round distribution if there is enough money left over after other claimants are paid.

   [405]. Casey & Macey, supra note 22, at 1035 (“[A]fter a mass-tort bankruptcy settlement, individual victims still have the right to opt for litigation in state tort systems. That statement may surprise most readers because critics of bankruptcy routinely get this wrong. . . . If the individual claimants disagree with the liability and damages determination made by the fund administrators, they retain the right to litigate the matter in the tort system against the settlement trust long after the bankruptcy is over.”).

[406]. See, e.g., Bussel, supra note 30, at 740.

[407]. See, e.g., LTL 1.0, 637 B.R. 396, 413 (Bankr. D.N.J. 2022).

[408]. See id. (“The trust distribution procedures (‘TDP’) and plans, however, will usually place timing restrictions and caps on compensatory and punitive damage recoveries. These limitations are critical to the process since one of Congress’s primary intentions in creating § 524(g) was to ensure uniform treatment of all claimants.”); Brubaker, No-Opt-Outs, supra note 45, at 119–20.

[409]. See 11 U.S.C. § 726(a)(4) (subordinating claims for punitive damages).

[410]. See Brubaker, No-Opt-Outs, supra note 45, at 120 (“If it turns out that the amount set aside to pay all of the mass-tort claims is insufficient to fully pay all of the claims in full, then they will not all be fully paid.”).

   [411]. In theory, if there is enough money to go around, the trust might not place limits on compensatory damages (punitive damages may be another matter), and claimants who successfully litigate against the trust can achieve a full recovery. But there is little reason to think that a trust distribution plan would be designed without caps. An uncapped plan could easily unravel, as the strongest claimants would refuse settlement offers and litigate against the trust—precisely the type of adverse selection the defendant is trying to avoid. And the claimants with the strongest claims will have little leverage to bargain for such a process. They can find themselves outvoted by claimants with weaker claims who want to see quicker guaranteed payments, and who would be quite content to agree to damages caps that their claims would never reach.

   [412]. Ortiz v. Fibreboard Corp., 527 U.S. 815, 847 n.23 (1999) (“It is no answer in this case that the settlement agreement provided for a limited, back-end ‘opt out’ in the form of a right on the part of class members eventually to take their case to court if dissatisfied with the amount provided by the trust. The ‘opt out’ in this case requires claimants to exhaust a variety of alternative dispute mechanisms, to bring suit against the trust, and not against Fibreboard, and it limits damages to $500,000, to be paid out in installments over 5 to 10 years . . . despite multimillion-dollar jury verdicts sometimes reached in asbestos suits.”); see also Brubaker, No-Opt-Outs, supra note 45, at 117–22. But see Bussel, supra note 30, at 741 (arguing that “[o]verwhelming consent” of claimants should override individual right to opt out for jury trial).

   [413]. Casey and Macey assert that MDL settlements “utilize similar administrative procedures as bankruptcy, but without the option to go to trial,” citing the Vioxx settlement. Casey & Macey, supra note 22, at 1036. That is not quite right in the narrow sense that some MDL settlements do, indeed, include back-end opt-out rights. E.g., Brown v. Am. Home Prods. Corp. (In re Diet Drugs (Phentermine, Fenfluramine, Dexfenfluramine) Prods. Liab.Litig.), Nos. 1203, 99-20593, 2000 WL 1222042 (E.D. Pa. Aug. 28, 2000). More broadly, though, it misses the fundamental point. The choice about whether or not to participate in an MDL settlement comes at the front end. Claimants who are unhappy with an MDL settlement (perhaps because it would underpay their high-value claims) need not opt in to that settlement to begin with. In bankruptcy, by contrast, claimants who thinks the settlement would underpay their high-value claims may be outvoted by claimants with lower value claims and thus bound to go through the administrative procedures set up by the bankruptcy trust before they have any opportunity to opt out and take their (likely capped) claims to court.

[414]. See Brubaker, No-Opt-Outs, supra note 45, at 119; Brubaker, Two-Step Part I, supra note 30, at 11–13; Brubaker, Two-Step Part II, supra note 158, at 7; Brubaker, Two-Step Part III, supra note 45, at 9–10.

[415]. See 11 U.S.C. §§ 1128, 1129.

   [416]. Bradt & Rave, supra note 102 at 1264–65.

[417]. See, e.g., Lynn M. LoPucki & William C. Whitford, Venue Choice and Forum Shopping in the Bankruptcy Reorganization of Large, Publicly Held Companies, 1991 Wis. L. Rev. 11; Adam J. Levitin, Purdue’s Poison Pill: The Breakdown of Chapter 11’s Checks and Balances, 100 Tex. L. Rev. 1079, 1128–50 (2022); Anthony J. Casey & Joshua C. Macey, Bankruptcy Shopping: Domestic Venue Races and Global Forum Wars, 37 Emory Bankr. Devs. J. 463, 478–79 (2021) (noting that “liberal venue rules make [forum shopping] especially easy for bankruptcy cases”).

[418]. See, e.g., Adam J. Levitin, Judge Shopping in Chapter 11 Bankruptcy, 2023 U. Ill. L. Rev. 351; cf. Steve Vladeck, 18. The Growing Abuse of Single-Judge Divisions, One First (Mar. 13, 2023), https://www.stevevladeck.com/p/18-shopping-for-judges [https://perma.cc/57KK-268G] (discussing judge shopping in public law cases).

   [419]. 28 U.S.C. § 1407.

   [420]. Bradt, The Long Arm of Multidistrict Litigation, supra note 116, at 1212 (quoting In re FMC Corp. Pat. Litig., 422 F. Supp. 1163, 1165 (J.P.M.L. 1976)).

   [421]. Zachary D. Clopton & Andrew D. Bradt, Party Preferences in Multidistrict Litigation, 107 Calif. L. Rev. 1713 (2019).

[422]. E.g., In re Silicone Gel Breast Implants Prod. Liab. Litig., 793 F. Supp. 1098 (J.P.M.L. 1992).

[423]. See Heyburn, supra note 112, at 2239–41.

[424]. See, e.g., Andrew Bradt & Calen Bennett, Adult Supervision? Appellate Review, Mandamus, and the Federal Rules in Multidistrict Litigation, 50 Fla. St. U. L. Rev. 187, 190–91 (2022).

[425]. See Rave, supra note 208, at 1606–08.

[426]. In re 3M Combat Arms Earplug Litig., No. 19-md-2885, 2022 WL 17853203 (N.D. Fla. Dec. 22, 2022); cf. Samuel Issacharoff & Troy A. McKenzie, Managerialism and its Discontents, 43 Rev. Litig. 1, 21–22 (2023) (discussing 3M’s strategy).

[427]. See, e.g., LTL 1.0, 637 B.R. 396, 429 (Bankr. D.N.J. 2022) (“In permitting this case to proceed going forward, this Court stands prepared to employ its limited equitable authority under § 105(a) to facilitate and assist Debtor and all tort claimants to achieve a fair and just result, consistent with the social policies and objectives intended by Congress in enacting the Bankruptcy Code.”).

[428]. See Lynn M. Lopucki, Courting Failure: How Competition for Big Cases Is Corrupting the Bankruptcy Courts 138–39 (2005).

[429]. See Issacharoff & McKenzie, supra note 426; Andrew D. Bradt, Zachary D. Clopton & D. Theodore Rave, Dissonance and Distress in Bankruptcy and Mass Torts, 91 Fordham L. Rev. 309, 314, 317–18 (2022).

[430]. Cf. Bussel, supra note 30, at 708 (suggesting that a national body akin to the JPML be created to decide whether mass tort defendants are eligible for bankruptcy and, if so, to assign the judge).

   [431]. Casey & Macey, supra note 22, at 1022.

[432]. See In re LTL Mgmt., LLC, 64 F.4th 84, 101–02 (3d Cir. 2023).