!@41@! The Rule of the Deal: Bankruptcy Bargains

and Other Misnomers

Jonathan C. Lipson*

Practice under chapter 11 depends on bargaining and negotiation to produce a plan of reorganization for the corporate debtor. Because this bargaining occurs behind closed doors, its dynamics are opaque. In most cases, this is not a problem because we infer that quasi-private ordering will produce greater economic recoveries for more claimants, even as it may occasionally distort or deviate from positive law (the “rule of the deal”).

This article assesses the rule of the deal in the normatively difficult bankruptcy of opioid-maker Purdue Pharma. The most notorious deal in that case was the nonconsensual “release” of hundreds of individuals and entities for direct liability for their alleged role in Purdue Pharma’s twice-confessed federal drug- marketing crimes, including the Sackler family, who own the company and controlled it until shortly before bankruptcy.

Less attention has been paid to deals that made those releases virtually inevitable from the outset. This article fills that gap, explaining how “ex ante” and “ex post” bankruptcy bargains—a governance change, a stipulation, and a settlement with the Department of Justice—made it practically impossible for creditors to resist those releases. This had implications for the rule of law, and for the economic outcomes of the case.

 !@42@!

Table of Contents

 !@43@!

The conduct of bankruptcy proceedings not only should be right but must seem right.[1]

Freedom of contract means, among other things, never having to say you are sorry.[2]

Is chapter 11 governed by the rule of law—or the rule of the deal?

By “rule of law” I adopt Margaret Jane Radin’s straightforward definition: “first, there must be rules; second, those rules must be capable of being followed.”[3] “Rule of the deal,” by contrast, describes bankruptcy’s strong appetite for bilateral bargains—deals—among small groups of strategically placed stakeholders that can distort or deviate from positive law. These bankruptcy bargains may be “ex ante,” as in the “restructuring support agreement” (RSA) early in a case, or “ex post,” as in settlement or asset-sale agreements midway through a case. They are log-rolled through the case to gain momentum toward a “grand bargain” that makes a plan of reorganization implementing those deals practically inevitable.

These bankruptcy bargains are not usually troublesome. They have long been the lifeblood of chapter 11 practice. They are also the subject of a large body of literature, which focuses mostly on their economic implications, and which typically asks whether they are coercive, or efficient, or—trickier—both?[4]

But they can be problematic in what I will call “social debt” bankruptcies. Social debt is financial liability for serious (e.g., criminal) misconduct, often involving violations of health and safety laws, made unsustainable due to persistent governance failures of transparency and accountability.

The chapter 11 reorganization of opioid maker Purdue Pharma, the subject of this article, is perhaps the most notorious example of a “social !@44@!  debt” bankruptcy. In addition to other opioid makers (e.g., Endo[5]), other examples include organizational liability for large-scale sexual assault (e.g., over thirty Catholic organizations,[6] The Boy Scouts of America[7]), and alleged contributions to the crisis of gun violence (e.g., InfoWars[8]). “Social debt” is, in short, Ronald Coase’s famous problem of social cost—externalities—writ very large.[9]

The most important questions in social debt bankruptcies will involve transparency and accountability: how did the misconduct occur, and do we have some confidence that those responsible have, in fact, been held accountable in a legally and socially acceptable way? Often, these questions will be answered by collateral litigation, such as individual prosecutions or tort lawsuits. Those other processes involve adversity and confrontation over basic questions of liability where the rule of law is often at a premium. By their nature, they will produce information about the underlying allegations, and determine individual liability or innocence.

The rule of the deal, by contrast, focuses on maximizing creditor payouts. The rule of law and the rule of the deal are both necessary features of the legal system—almost everything settles—but social debt bankruptcies require more of the former than the latter. !@45@!

Unfortunately, Purdue Pharma was the opposite. The case has received extraordinary attention, including congressional scrutiny,[10] front page coverage in The New York Times,[11]and even three episodes of Last Week Tonight with John Oliver[12] for “nondebtor releases” (NDRs) that would give Purdue Pharma’s owners, the wealthy and secretive Sackler family, the “global peace” they have long sought for their alleged role in the opioid crisis (the “Sackler Releases”).[13]

A nondebtor release “operate[s] as a bankruptcy discharge arranged without a filing and without the safeguards of the Bankruptcy Code.”[14] Thus, as others have observed, Purdue Pharma would give the Sacklers the benefits of bankruptcy without its burdens, including obligations of transparency (financial disclosure) and accountability (giving most assets to creditors).[15] !@46@!

The Sackler Releases are as hot as bankruptcy gets. Critics call them an “outrage,”[16] “lawless,”[17] “shocking,”[18] a “grift,”[19] and flatly unconstitutional.[20] Among other things, they are not authorized by any provision of the Bankruptcy Code, may exceed bankruptcy court authority, and may violate due process.[21] Proponents counter that they are “necessary” and that creditors voted for them by large majorities when they approved Purdue Pharma’s plan of reorganization.[22] In any case, they may say, governments—who were the most assertive of Purdue Pharma’s 600,000 creditors—were actively involved, ameliorating concerns about the “social” nature of the debt at issue in the case.

As of this writing, it is not clear whether the Sacklers will “get away with it.”[23] Although the Bankruptcy Court in Purdue Pharma approved the Sackler Releases, the District Court reversed on grounds that the Bankruptcy Code does not permit them.[24] The case is pending before the Second Circuit Court of Appeals, which may use it to clarify its position on NDRs, having previously worried that they might be “abused,” but neither forbidding nor permitting them.[25]

Although Purdue Pharma has generated great heat, there has been less light. Wanting is an explanation of the dynamics that produced the Sackler !@47@!  Releases, to guide courts in future appeals in Purdue Pharma and in other similar cases.[26]

This article aims to fill the gap. It draws on primary case documents to show how bankruptcy bargains in Purdue Pharma made the Sackler Releases practically inevitable from the outset: (i) the selection of corporate agents, professionals, and a judge likely to implement a settlement framework that the Sacklers had agreed to with a small group of Purdue Pharma’s 600,000 creditors before bankruptcy; (ii) a preliminary injunction and stipulation in which creditor representatives ceded important leverage against the Sacklers; and (iii) a settlement in 2020 with the Department of Justice (DOJ) that, as a practical matter, foreclosed all options other than the Sackler Releases.

Existing literature has no account for bankruptcy bargains in social debt cases. Nor could it, I argue here, because the logic of bankruptcy theory, doctrine, and practice seek to solve a problem of remedy—priority in right of distribution—whereas social debt bankruptcies may also involve the predicate question of liability. In Purdue Pharma, for example, the key question for many was about the truth of allegations that the Sacklers had knowingly sought to “turbocharge” the market for opioids, even after their company had previously pled guilty to federal drug-marketing charges.[27]

Perversely, perhaps tragically, Purdue Pharma’s bankruptcy bargains not only undercut transparency and accountability; they were also a bad deal. It is not clear, for example, that it was the Sackler Releases that added value, but instead the resistance by creditors to them, as would be the case in any ordinary bargaining. Here, court-ordered mediations did result in the Sacklers increasing the face amount of their proposed contribution, but the net present value did not increase much because the payment period was !@48@!  extended, while the projected value of the company—which creditors would “own” through a trust structure—declined significantly, according to monthly operating reports in the case.[28]

Nor should this be surprising. Chapter 11 already gives debtors in possession significant advantages, in particular control of decisions about a plan of reorganization and whether (and how) to litigate estate causes of action. Early bargains in Purdue Pharma further tilted the playing field in favor of the deal the Sacklers wanted, so resistance to it was hampered. Creditors were divided about that deal, and the subsidiary deals that would lead to it. The Sacklers would understandably want to take advantage of chapter 11’s deal culture and a fragmented creditor body. The rule of the deal could force them together. But this assumed that a deal on terms framed by the Sacklers was the only option.

It did not have to be this way. Chapter 11 has mechanisms that can address problems of transparency and accountability. In Purdue Pharma, for example, these could have included a bellwether trial against the Sacklers; more fulsome disclosure of the results of investigations by estate fiduciaries (or the appointment of an examiner to do so); or simply a line-item on the ballot permitting creditors to opt out of the Sackler Release—thus treating it as the contract term it purports to be. None of these things happened, fueling public concern about the case.

This article makes three contributions. Part 1 describes the theory, doctrine, and practice of ex ante and ex post bankruptcy bargains. It shows that the former can be coercive and the latter are subject to standards focused on asset values and distributions. The deeper problem is that these mechanisms solve problems of remedy where liability is assumed or not controversial. They cannot, however, answer questions of liability in social debt bankruptcies. Part 2 describes key bankruptcy bargains in Purdue Pharma leading to the Sackler Releases. Part 3 exposes contradictions of the rule of the deal as implemented in Purdue Pharma, the most important of which was that creditors “agreed” to the Sackler Releases. It also offers brief thoughts on what could have happened differently.

1.  The Law and Theory of the Deal !@49@!

Chapter 11 reorganization favors consensus over conflict.[29] But, it can be a coercive process.[30] It forces all stakeholders in the debtor into a single forum in which supermajority voting can radically alter individual rights through a confirmed plan of reorganization. That plan, and deals leading to it, presumptively seek to maximize the value of the debtor’s assets, through doctrines that focus chiefly on distributive rights, such as the relative priorities of creditors’ claims to the debtor’s value. These doctrines work reasonably well when underlying liability is not in serious dispute. But, they tell us little about how to resolve disputed liability for social debt.

1.1  The Law and Practice of Bankruptcy Bargains

Bankruptcy bargains can be struck at roughly two different moments: the beginning or middle of the case. We can call them “ex ante” and “ex post” because they will precede or embody a case-resolving deal, which is then implemented in a plan of reorganization.[31] Ex ante bargains often seek to set the rules of the road, such as through “restructuring support agreements” or case stipulations.[32] Ex post bankruptcy bargains may settle major disputes or sell major assets. The first will guide the trajectory of the case; the second will for most practical purposes determine the distributive outcome.

1.1.1 Ex Ante Bargains—Restructuring Support Agreements

Debtors may enter bankruptcy with prenegotiated deals in hand or strike them at the beginning of the case. These may take the form of !@50@!  prebankruptcy bond workouts,[33] case financings,[34] “back-stop” agreements,[35] or, more generally, “restructuring support agreements” (RSA).[36]

RSAs mark a significant development in practice, a “quiet revolution,” according to Professor Douglas Baird.[37] Under an RSA, key stakeholders may pre-commit to support the debtor’s proposed plan of reorganization, provided that the case and the plan satisfy certain criteria.[38] Bankruptcy courts will approve an RSA if it reflects a reasonable exercise of business judgment by management.[39]

In the past, negotiations with creditors would be more open-ended, would involve a broader group of creditors, and would be embodied in the plan itself, instead of earlier deals approved by the court. Now, “[i]nstead of bargaining that all are free to join, there is a sequence of two-party bargains, beginning with the key players (typically key creditors and the debtor). Each bargain fixes the share of the participating creditor.”[40] These bargains are the building blocks of the rule of the deal.

Critics worry that these ex ante bargains may be “coercive.”[41] They are often developed by a small group of key stakeholders, and presented on an “emergency” basis to a judge who has little familiarity with the background facts, but good reason to worry that if she fails to approve the deal, the case may collapse before it even starts.[42] They are then log-rolled through other constituencies whose support the original participants believe they need.

This presents several concerns, the most obvious of which is that those in any given deal may have little interest in protecting those outside of it. !@51@!  Professor Melissa Jacoby has argued that chapter 11 is a “package deal,” and thus that “à la carte” deals for particular constituencies violate the premises and spirit of reorganization.[43] It is, however, not clear how much we can or should expect creditors to be their siblings’ keepers. Judge Goldblatt recently observed that “nothing in the law [] requires holders of syndicated debt to behave as Musketeers.”[44] He was responding to the observation (by a state court judge) that syndicated lenders may sometimes evince a “spirit of … all for one, one for all.”[45]

Perhaps. But who, realistically, expects lenders negotiating the deal to act as musketeers for tort creditors? What incentive would they have to bring the tort creditors into an RSA?

Being left out of an ex ante bargain is not the only problem. Perhaps a greater problem (because potentially more coercive) is that ex ante bargains can mute, rearrange, or displace formal protections for all creditors built into the Bankruptcy Code.

Take the power to seek an examiner. Section 1104(c) of the Bankruptcy Code provides that a bankruptcy court “shall” order the appointment of an examiner “to conduct such an investigation of the debtor as is appropriate, including an investigation of any allegations of fraud, dishonesty, incompetence, misconduct, mismanagement, or irregularity in the management of the affairs of the debtor of or by current or former management of the debtor … .”[46]

Congress created the role of the examiner to provide “special protection for the large cases having great public interest … to determine fraud or wrongdoing….”[47] Examiners’ reports in cases such as Enron, Worldcom, and Lehman Brothers performed “an important public service in explaining the spectacular and unanticipated collapses of these firms.”[48]

Yet, prior research has found that examiners are “vanishingly rare in most cases, large or small”[49] because “neither managers nor investors !@52@!  (creditors) want this sort of oversight of the process.”[50] Courts and system participants “worry that examiners could simply be expensive fishing expeditions.”[51]

Still, creditor protections like examiners “could ameliorate the problems of agency cost and systemic integrity that concerned Congress when it created chapter 11, and that persist today in new but perhaps equally problematic forms.”[52] Even the credible threat of their use can be a bargaining chip whose value can be compromised by an RSA. To empower insiders to contract them away at the beginning may thus be a problem.

Yet, they are often contracted away. In the chapter 11 reorganization of opioid maker Endo, for example, the debtor agreed in its RSA to object to any motion seeking the appointment of an examiner, or converting or dismissing the case.[53] If such relief were sought and approved, it would be a “termination event” under the RSA.[54] A later effort to appoint an examiner would not necessarily end the case—but it would threaten a foundational deal in it.[55] As discussed in Part 2, below, Purdue Pharma had a similar dynamic.

Problematic or not, appellate courts have little enthusiasm for second-guessing a bankruptcy court’s approval of an ex ante bargain. In the recent LATAM Airways case, for example, certain creditors aligned with management entered into a “backstop” agreement under which they agreed to support exit financing for the debtors in exchange for millions of dollars in fees.[56] 

On appeal, appellants argued that these agreements had the effect of resolving certain key questions about the debtor’s reorganization, which should be addressed under a plan of reorganization. The appellees (the debtor) argued that the orders approving the backstop agreement were not appealable final orders. The District Court, sitting in an appellate capacity, considered the question “a close one,” but agreed with the debtor—these were questions for a later day.[57] The later day was the confirmation of a plan !@53@!  that largely implemented the backstop agreement, and which was virtually impossible to challenge.

Ex ante bankruptcy bargains are not always problematic. Their coercive costs must be measured against their benefits, including that they can reduce time and expense, and encourage the creativity of the parties to the deals. That creativity can advance chapter 11’s larger consensus goal of maximizing wealth for all. To the extent the interests of stakeholders are adequately represented—and that will often be a key issue—they presumably agree because they perceive these deals to provide certainty in the reorganization process and, in any case, there may be no realistic alternative.

But in cases involving social debt, ex ante bargains can be problematic because they involve a level of public interest that may be in tension with the private-ordering goals of the RSA.

1.1.2 Ex Post Bargains

Ex post bargains are major transactions during a case short of or leading to a plan of reorganization, such as the compromise of a disputed significant claim or the sale of material assets of the debtor. While there are concerns about coercion here, as well, doctrine focuses on distribution: Will the deal alter or impair priority in right of payment? This is obviously salient in the ordinary commercial case. But it may be tangential when there are analytically prior and important questions of liability, as may be the case in social debt bankruptcies.

The Supreme Court’s opinion in TMT Trailer Ferry expressed the standard for settlements, and in so doing recruited the language of priority, by requiring that they be evaluated to determine whether they are “fair and equitable.”[58] This phrase, derived from over one hundred years of caselaw, is synonymous with absolute priority, and is codified in Bankruptcy Code section 1129(b).[59]

In TMT, a bankrupt shipping company had granted liens that a bankruptcy trustee investigated and initially concluded were avoidable as fraudulent transfers.[60] Later, the trustee decided to compromise the claim by paying it in full, over time, under a plan of reorganization. Over objections from unsecured creditors, the trial court (acting as a bankruptcy court) approved these settlements.[61] !@54@!

Before the Supreme Court, the central question was whether this compromise should have been approved. If the fraudulent transfer claim had merit, the settling creditor in TMT would have lost its lien (and thus priority), increasing distributions to (junior) general unsecured creditors. In reversing the lower courts, the Court stated that “a proposed compromise forming part of a reorganization plan [must be] fair and equitable.”[62] To make this determination, the TMT Court held that the lower court should “compare the terms of the compromise with the likely rewards of litigation.”[63]

This advice, to compare the terms of the settlement to the prospects of litigation, leaves much to the imagination. While courts should assess the likelihood of successful litigation on the merits, the “most important consideration” is priority.[64] Courts have thus concluded under TMT that they should not “conduct a mini-trial on the merits of the settlement or otherwise resolve disputed issues of law or fact underlying the settlement.”[65] Instead, they should “canvass the issues and see whether the settlement ‘fall[s] below the lowest point in the range of reasonableness.’”[66] 

Bankruptcy courts do this well as to claims against the debtor and estate causes of action. They routinely assess disputes that involve a fight over distribution in an environment where underlying liability—e.g., unpaid loans—is not seriously in dispute. In TMT, the litigation was an avoidance !@55@!  action. In the leading Court of Appeals decision, Iridium, it was a priority-skipping settlement.[67] It is less clear, however, whether or how a bankruptcy court can assess potentially novel or normatively difficult direct causes of action against nondebtors, such as those at issue in Purdue Pharma.

Concerns about coercion persist as to ex post bargains, too. Case-transforming sales of assets have sometimes been considered problematic on grounds that they were a plan “sub rosa,” a plan of reorganization in substance, but without the procedural protections of disclosure, voting and so on which we ordinarily see at the apex of a case.[68]

Although ex post bargains have become routine, the Supreme Court drew the line in Jevic, striking a “structured dismissal” which would have reordered the priority of final distributions outside of a plan.[69] The Court’s main concern in Jevic was not distributive, but instead procedural: “the distributions at issue here more closely resemble proposed transactions that lower courts have refused to allow on the ground that they circumvent the Code’s procedural safeguards.”[70] The Court did not specify what those safeguards are, but I have argued elsewhere they had to include mechanisms that preserve participation, predictability, and procedural integrity, such as plan disclosure, classification, and voting.[71]

Justice Breyer recognized that “consent” might have overcome much of what was wrong in Jevic.[72] But, outside a plan, “consent” in the chapter 11 context can be hard to determine.[73] In any case, the whole point of !@56@!  nonconsensual nondebtor releases—the core issue in Purdue Pharma—is that they are (obviously) not consensual.

     1.2 The Special Problem of Personal Injury/Wrongful Death Tort Claims

Bankruptcy is chiefly a remedial scheme—a “special remedial scheme,” the Supreme Court has said.[74] Among other things, this implies that it is not designed to determine disputed primary liability. Indeed, for creditors who timely file a proof of claim (a simple document asserting the debt), the system presumes the debtor’s liability.[75] To litigate whether the debtor is liable for each and every debt that it obviously owes would be wasteful.

This presumption of liability works well with debts arising from broken contracts, such as unpaid loans or for tort liability determined by other courts. It does not work as well with social debt that may be disputed, because such claims will often involve alleged torts. For purposes of payouts, however, the Bankruptcy Code makes no distinction between contract and tort: both usually produce unsecured claims of equal rank against the debtor’s assets.[76] Because those assets are probably worth far less than aggregate claims, unsecured creditors can generally hope at best for a pro rata share of their claims, in cash, in new debt, or in stock of the reorganized debtor.

In theory, Congress tried to provide a different process for the determination of contingent and unliquidated personal injury tort or wrongful death claims, which may be at issue in social debt cases, by assuring that such claims would be determined by a jury trial in a U.S. district court, rather than a bankruptcy court.[77] In fact, however, such trials !@57@!  are rare because even a victory is unlikely to significantly improve the payout: the winning plaintiff simply takes her judgment and gets in line to share, pro rata, with all other unsecured creditors.[78]

Although Congress apparently “never contemplated that the bankruptcy process might be used for the resolution of mass tort claims,”[79] the exceedingly broad scope of relief provided by the bankruptcy discharge (and the ability to address future claims) made it appealing to mass tortfeasors (in particular, those with asbestos or medical device liability) in the early 1980s.[80] By 1994, when Congress enacted the “asbestos amendments” to authorize nondebtor releases in cases involving such liability—the only express statutory authorization for NDRs—it understood that chapter 11 could (and perhaps should) aggregate and resolve liability for broad-scale negligence or products liability.[81]

Today, bankruptcy courts openly claim that bankruptcy is better than the tort system. Before being reversed, Judge Kaplan denied a motion to dismiss the chapter 11 case of talc debtor LTL because, among other reasons, “[t]he tort system has struggled to meet the needs of present claimants in a timely and fair manner. The system is ill-equipped to provide for future claimants. The Court has no reason to believe this will differ for the talc plaintiffs here.”[82] !@58@!

1.3. The Theory of the Deal

Although the Third Circuit disagreed, there may be truth to Judge Kaplan’s assertion. Professor Jay Westbrook has observed that wealth maximization is a “consensus” goal that few scholars would challenge.[83] Plaintiffs who insist on “pristine due process rights” may do so at the expense of “substantial monetary compensation.”[84]

1.3.1 Contract Debates in Bankruptcy

Creative deal-making in bankruptcy finds strong advocates among “contractualists,” loose adherents of “law and economics.” This position is probably the dominant body of bankruptcy scholarship, and insists that bankruptcy is and should be governed by a norm of economic efficiency above all else.

The most influential early expression of contractualism is the so-called “creditors’ bargain” articulated by Professors Thomas Jackson and Douglas Baird and their followers.[85] The creditors’ bargain is hypothetical, “a bargain that rational creditors would have reached ex ante to [] divide the debtor’s assets.”[86] Bankruptcy law, on this view, should seek to maximize wealth by permitting the parties to renegotiate in the light of their prebankruptcy entitlements, which bankruptcy should take as given.

While no bankruptcy scholar would wholly disclaim free contract or wealth maximization, many have been skeptical of the creditors’ bargain model. Critics have objected that it lacks empirical support,[87] normative !@59@!  force,[88] or basis in law.[89] Before she became a United States Senator, Elizabeth Warren argued that bankruptcy law in fact does and should reflect “competing—and sometimes conflicting—values.”[90] Wealth maximization is one such policy, but so too are process values of participation, outcome predictability and procedural integrity,[91] all of which can prevent abuse, and none of which is costless. More recent scholars continue to contest contractualism,[92] asserting that it is a “fallacy.”[93]

Nevertheless, the economic position has strong appeal given the distributive orientation of bankruptcy doctrine. The diffusion of countervailing positions has made it difficult for opponents of neoclassical economics to marshal a consolidated response. It is easy enough to challenge the predicates of economic analysis—few truly believe in “homo economicus” unbound[94]—but “freedom of contract” is a catchy cry, and realist alternatives struggle to gain traction.[95]

1.3.2 The “New Bargaining Theory” and the Problem of Holdouts

The latest salvo in these debates is reflected in the view that chapter 11 implements a “New Bargaining Theory,” as Professor Anthony Casey would have it. This holds that chapter 11 exists to solve problems of “incomplete contracting,” and should do so by creating a “renegotiation framework designed to minimize the parties’ ability and incentives to hold !@60@!  each other up.”[96] The challenge, on this view, is that there will be strong incentives for some creditors to hold out, which “will lead to a collapse in bargaining where no deal is ever struck.”[97]

Combating holdouts, on this view, “relies heavily on judicial discretion and procedural measures.”[98] But this discretion can permit coercion and, in any case, will lead judges to focus on the distributive consequences of the deals presented to them, rather than whether particular individuals should (or should not) be subjected to judicial process to determine liability.

Professors David Skeel and George Triantis, otherwise supporters of free contract, have called bankruptcy policy toward contracting “incoherent” because the governing standards are vague.[99] Professor Stephen Lubben has argued that the “panic” over holdouts—and thus the need to coerce them with bankruptcy bargains—is exaggerated.[100]

Nevertheless, bankruptcy bargains drove the reorganization of Purdue Pharma. The Sackler Releases are arguably the most notorious example of that, but in order to understand them, it is necessary to understand the deals that foreclosed all other possibilities. Those other possibilities included litigation against the Sacklers to determine direct liability, or investigations and reports that estate fiduciaries could and should have produced. Those forms of scrutiny might have produced greater transparency, increased recoveries for creditors, and instilled confidence in the process. The rule of the deal, however, undercut those efforts.

2.  The Sacklers’ Social Debt and Bankruptcy Bargains

Maximizing wealth may be the overarching goal of bankruptcy, but for many survivors of serious misconduct, social debt is hard to cash out. It is not difficult to see that Purdue Pharma created social debt, unmanageable social costs with profound public spillovers. The overdose crisis is a !@61@!  “national public health care emergency,”[101] an “epidemic” in the words of the U.S. Department of Health and Human Services.[102]

The allegations against Purdue Pharma and the Sacklers have been told often and vividly, and warrant only a summary here to frame the deals that would resolve them in bankruptcy. I do not claim that the allegations are true—rather, that Purdue Pharma’s bankruptcy bargains would ensure that we could not know in a legally actionable way whether they were true before the Bankruptcy Court would give the Sacklers the releases they sought.

2.1 The Sacklers, Purdue Pharma, and the Opioid Crisis

Purdue Pharma is a closely-held company, at all relevant times owned and controlled by the Sackler family, among the wealthiest in the nation.[103] Purdue Pharma introduced OxyContin in 1995.[104] Following an aggressive and successful marketing campaign, annual sales of OxyContin reached $1 billion in 2001; it became “‘the most prescribed brand-name narcotic medication’ in the U.S.,”[105] contributing significantly to an overdose crisis that to date has claimed over 500,000 lives.[106]

In 2007, Purdue Pharma and certain executives—but not the Sacklers—pleaded guilty to federal charges of misbranding OxyContin with the intent !@62@!  to defraud or mislead.[107] Purdue Pharma had, according to the DOJ, falsely “marketed and promoted OxyContin as less addictive, less subject to abuse and diversion, and less likely to cause tolerance and withdrawal than other pain medications.”[108]

Despite a “corporate compliance agreement,” Purdue Pharma continued to engage in this misconduct, and was subject to a “veritable tsunami” of tort litigation.[109] In 2017, sixty-four cases brought by governmental plaintiffs were consolidated in a multidistrict litigation in the Northern District of Ohio (MDL).[110] Discovery in the MDL revealed evidence of the involvement of members of the Sackler family in Purdue Pharma’s misconduct. Those family members were added as defendants in hundreds of the 3000+ cases pending against Purdue Pharma.[111] By the middle of 2019, forty-nine states’ attorneys general had filed new or amended lawsuits naming specific members of the Sackler family and/or Sackler-related entities.[112]

These “direct” claims asserted liability “under various unfair trade practices and consumer protection laws that make officers, directors and managers who are responsible for corporate misconduct personally liable for their actions.”[113] Because Purdue Pharma was a closely-held company, the Sacklers—who dominated its board of directors—may also have had independent liability on “alter ego” or similar theories.[114]

Absent bankruptcy, the only way the Sacklers could achieve the “global peace” they sought—that is, to eliminate their direct and personal liability !@63@! —was to litigate every allegation to a preclusive final judgment or to settle. While either would be expensive, the Sacklers had the money.[115] But allegations leaked in the bankruptcy (and the Sacklers’ prebankruptcy efforts to suppress bad facts) suggest that another concern was transparency: The Sacklers wanted to shield as much information as possible.[116] Merits litigation would risk economic losses and divulge even more incriminating information.

Thus, they understandably wanted to settle. But regular courts could not give them the peace they wanted. In the late 1990s, the Supreme Court made it very difficult to certify mandatory “limited fund” class action settlements, which would bind all current and potential plaintiffs without individualized assent or adjudication.[117] Nor, presumably, did they wish to commence bankruptcy in their own names in order to discharge debt.[118]

But nondebtor releases in a bankruptcy of Purdue Pharma might work. In March of 2018, certain members of the Sackler family negotiated a framework for a “global settlement” with certain parties to the MDL (the !@64@!  “Sackler Settlement Framework”).[119] The Sackler Settlement Framework had three basic elements, reflected in an unsigned term sheet filed at the beginning of the bankruptcy: (i) the Sacklers would “give” Purdue Pharma to creditors; (ii) they would pay $3 billion (later increased to $5.5 billion) into creditor trusts over an extended period; and (iii) in exchange, they would receive “comprehensive releases.”[120] The Sacklers thus “made it clear well before [Purdue Pharma] filed for chapter 11 bankruptcy that they would contribute toward Purdue’s bankruptcy estate only if they received blanket releases that would put ‘all of the litigation behind them.’”[121] Purdue Pharma commenced its bankruptcy in September 2019 in order to obtain these releases.

2.2 Bankruptcy Bargains Implementing the Sackler Settlement Framework

The Sackler Releases were the apex “deal” in the Purdue Pharma case. Although they have received the most attention, they would not have been possible without three predicate deals, bargained for before and during bankruptcy: (i) prebankruptcy governance changes that assured the Sacklers of corporate agents, professionals, and a bankruptcy judge likely to support their settlement goals; (ii) a preliminary injunction that shielded the Sacklers from tort lawsuits asserting direct liability during the case; and (iii) a controversial settlement with the DOJ which made the releases virtually inevitable.

2.2.1 Governance Changes—Shopping for Corporate Agents, Professionals and a Judge

To implement the Sackler Settlement Framework, the Sacklers needed two things: (i) credible corporate agents and professionals to manage the chapter 11 process; and (ii) a bankruptcy judge likely to support their desire to settle rather than to litigate. Governance changes made by the Sacklers !@65@! —that is, “deals” between Purdue Pharma and its owners—apparently led to both.

The Sacklers caused Purdue Pharma to hire the law firm of Davis Polk & Wardwell as “restructuring counsel” in March of 2018, around the time they struck the initial settlement deal with certain plaintiffs.[122] Turnaround expert Steve Miller joined the company’s board in July 2018 as members of the Sackler family left it.[123] Both Davis Polk and Miller are longtime repeat players in the chapter 11 system.[124] It appears that both supported the Sackler Settlement Framework from the outset. Before bankruptcy, they coordinated closely with the Sacklers about messaging in support of it.[125] As the Sacklers resigned from Purdue Pharma’s board in 2018–19, Miller brought on two other “bankruptcy directors” to form a “special committee” to negotiate the Sackler Settlement Framework.[126]

The Sacklers (presumably with the help of Davis Polk) found the second critical piece in United States Bankruptcy Judge Robert Drain. Judge Drain may have been attractive because he was known to be a sophisticated judge with a strong preference for negotiated resolutions. He had self-published a !@66@!  novel in which opiates played a role,[127] and had previously approved the sort of NDRs the Sacklers would seek.[128]

Purdue Pharma and the Sacklers could only draw Judge Drain if “venue” were proper. Fortunately for the Sacklers, bankruptcy has liberal venue rules. Section 1408 of the Judicial Code permits a corporate debtor to “bootstrap” into the bankruptcy courts of any district where at least one affiliate has “domicile.”[129] Domicile, in turn, exists in any state in which an entity debtor is formed.[130]

Purdue Pharma was a Delaware limited partnership, with a New York general partner.[131] Thus, venue would have been proper in the Southern District of New York (SDNY). But there was a catch: There are nine bankruptcy judges in the SDNY, with varying views about nondebtor releases, the key to the global peace they sought.[132] The SDNY is, however, !@67@!  divided into two divisions, Manhattan and White Plains,[133] and Judge Drain happened to be the only judge in the White Plains Division at the time. If the Sacklers commenced their case in White Plains, rather than Manhattan, Judge Drain was certain to be their judge.

To achieve this involved a simple governance change. On March 1, 2019, about a year after first striking the settlement framework, and six months before declaring bankruptcy, it appears that the Sacklers, acting as shareholders of Purdue Pharma’s New York general partner, authorized the company to change its registered corporate agent to White Plains, New York (and naming restructuring counsel Davis Polk as such).[134] This gave Purdue Pharma a connection to White Plains, enabling it to commence its case in the White Plain’s bankruptcy court. As Professor Adam Levitin has observed, Purdue Pharma sought Judge Drain not “because he is a great judge”—although he may be—but instead because “they think he will be a great judge for them.”[135]

Forum shopping has long been a concern in chapter 11 cases. Indeed, before nondebtor releases, it may have been the hottest topic in bankruptcy. Professors Lynn LoPucki and Bill Whitford first spotted the issue in their groundbreaking 1991 study, Venue Choice and Forum Shopping in the Bankruptcy Reorganization of Large, Publicly Held Companies, where they “unexpectedly discovered extensive forum shopping.”[136] The power to choose the forum for a large bankruptcy reorganization “determines where hundreds or thousands of parties will go to court,” they said, “and may be !@68@!  determinative of the outcomes of cases.”[137] LoPucki has gone on to decry the persistence of what he now calls “court competition,” which he says has resulted in a “corrupt”[138] and “lawless” ^[139] chapter 11 system.

As with other debates in bankruptcy, concerns about forum shopping have met a contractualist response. Professors David Skeel and Ken Ayotte, for example, counter that because creditors rarely seek to change venue, they accept the dominant venue choices.[140] The choice of forum is, as scholars such as Professor Robert Rasmussen or Professor Alan Schwartz might argue, simply an option on the menu of reorganization choices available to parties trying to resolve the corporate debtor’s distress.[141] Here, of course, the Sacklers were arguably trying to choose not merely a forum, but a particular judge. And, they sought to resolve not merely Purdue Pharma’s financial distress, but theirs, too.

The choice of corporate agents, professionals, and judge had two critical implications for all that would follow. First, as in all chapter 11 cases, Purdue Pharma—and not creditors—would control all practical decisions about whether to sue or settle claims that the estate may have had.[142] But this meant that many causes of action were almost certain to be settled without litigation. Because management of the debtors were appointed by the Sacklers, and Miller had publicly supported the deal, it was difficult to imagine Purdue Pharma would do anything other than settle according to the Sackler Settlement Framework.

This is not entirely surprising. Professor Jared Ellias and his colleagues have shown that bankruptcy directors like Miller and his colleagues appear to favor the interests of shareholders, even though the law of priority would !@69@!  seem to require them to consider the interests of creditors.[143] Indeed, although Purdue Pharma’s solvency was never tested in litigation, it is reasonably well-established that directors of companies in the “zone of insolvency” owe some sort of fiduciary duties to or for the benefit of creditors.[144] While a small group of plaintiff representatives had initially agreed to the Sackler Settlement Framework, the vast majority had not. The sequence of bankruptcy bargains that followed would give others little choice.

Purdue Pharma’s counsel also muddied the picture. Once Purdue Pharma commenced its bankruptcy (in September 2019), lead counsel, Davis Polk’s Marshal Huebner, offered shifting visions of his role. At certain times, he spoke as if he were a neutral, a “fiduciary for all,”[145] which would necessarily include the Sacklers. At other times, however, he spoke as if he were a crusader avenging the Sacklers’ misdeeds.[146] Sometimes he claimed that Purdue Pharma was the “plaintiff” against the Sacklers as “defendants,” although the only salient complaint filed (as explained below) was the one !@70@!  against victims to enjoin direct litigation against the Sacklers.[147] After Judge McMahon struck the Sackler Releases, Huebner stated that he was a “vessel” for the victims.[148] It is not clear how one law firm, much less one lawyer, can be a “fiduciary for all,” who sues the victims for whom he is nevertheless a “vessel.”

This may reflect the complex position of counsel to any debtor in possession. In any mass tort case, the debtor may be managed by at least some individuals implicated in the underlying harm, but is by law required to be a fiduciary for creditors.[149] A social debt bankruptcy, however, heightens concerns about who is acting for whom.

Second, the Sackler Settlement Framework would function in certain respects like a restructuring support agreement, shaping all that would follow. It was the first step in the sequence of bargains that led to the Sackler Releases.[150] To be sure, the initial Sackler Settlement Framework was with a small group of plaintiffs, not a single creditor. But they largely acted in unison in the case, and were quick to support the Sackler Settlement Framework whenever it was challenged.[151]

Purdue Pharma proceeded as a series of negotiations or mediations that had the effect of inducing or forcing all other creditors to capitulate to Purdue Pharma’s plan implementing the Sackler Settlement Framework. They could negotiate improvements to it, but they could not realistically change the fact that that was the only option. Creditors who wanted a “day in court” against the Sacklers, or some public account of their alleged misconduct before they voted on the plan, would get neither.

2.2.2 The Preliminary Injunction and the Case Stipulation

Still, the Sacklers had at least one other problem: because they were not debtors in bankruptcy, the “automatic stay” of civil litigation would not shield them (it would only protect Purdue Pharma).[152] Thus, they would need a preliminary injunction to halt pending litigations against them during !@71@!  the bankruptcy, in anticipation of making it permanent with the Sackler Releases.

Although no provision of the Bankruptcy Code permits such injunctions, courts authorize them under section 105 of the Bankruptcy Code[153] in order to shield management of a debtor, who might otherwise be distracted from efforts to restructure if they have to defend litigation personally.[154] At the outset of the bankruptcy, Purdue Pharma sought such a preliminary injunction to shield the Sacklers and scores of entities that they owned or controlled from the hundreds of suits against them.[155]

The Sackler Settlement Framework—the first bargain in the sequence—was the rationale. Counsel to Purdue Pharma argued that the preliminary injunction was needed to negotiate a deal with the Sacklers, and that only a deal with the Sacklers would avoid the “tragedy of the commons”[156] of direct litigation, replacing it with “billions of dollars of cash and critical resources … to address the opioid crisis.”[157]

Creditors, the most active of which were governments, split on whether to support the preliminary injunction. The so-called “consenting” parties had agreed to the Sackler Settlement Framework at the outset, and so supported the injunction.[158] They had already provisionally made peace with the Sacklers. The so-called “nonconsenting” States, which included aggressive !@72@!  plaintiffs such as New York and Massachusetts, had not. They objected to the preliminary injunction on grounds that it interfered with their police powers to prevent Purdue Pharma from continuing its illicit activities and to continue their direct suits against the Sacklers.[159] This dynamic—creditor division over whether to settle or sue—would be a subsidiary theme in the case, which Purdue Pharma and the Sacklers would be able to exploit.

Judge Drain supported the injunction for two reasons, both rooted in the Sackler Settlement Framework. First, litigating against the Sacklers would be costly, “potentially murderous” litigation.[160] The “prospect” of a plan that implemented the Sackler Settlement Framework, by contrast, warranted the injunction in order to give the parties a “clear shot” to negotiate, free from the distraction of litigation on the merits.[161]  It would be an opportunity for “due diligence,” the judge said, as in any other deal.[162] Whether the Sacklers would contribute, and if so how much, would be !@73@!  “based on all the facts, including all the claims that your clients and every other plaintiff has asserted against them,” the judge stated.[163]

Second, the injunction would be woven into other bankruptcy bargains that would lead to the releases. One of the most important was an ex ante bargain in the form of a case stipulation under which the Sacklers would produce information that would enable creditors to negotiate the settlement.[164] Case stipulations like Purdue Pharma’s are not inherently problematic. Here, however, the stipulation included a protective order which would prevent parties from publicly reporting what they learned through due diligence.[165]

Although confidentiality orders appear to have become common in chapter 11 cases, “there is a long-standing presumption of public access to judicial records.”[166] Bankruptcy Code § 107 enhances this presumption by providing that, subject to certain exceptions, “a paper filed in a case under this title and the dockets of a bankruptcy court are public records and open to examination by an entity at reasonable times without charge.”[167] Transparency has historically been an important value in corporate reorganization.[168]

In Purdue Pharma, media organizations later challenged the sealing and redaction of information produced in discovery pursuant to the protective order. Although Judge Drain ultimately agreed with the movants about the need to release some information (the names of investment advisors to the Sacklers), he waited until shortly before confirming the plan of reorganization to do so.[169] Any information revealed by then would have come after creditors had to vote on Purdue Pharma’s plan of reorganization. !@74@!

Moreover, creditors who joined the stipulation ceded important powers that would have given them leverage in negotiations with the Sacklers. Among other things, the Official Committee of Unsecured Creditors (UCC)—the only statutory fiduciary for creditors in the case—agreed not to terminate Purdue Pharma’s exclusive right to file a plan for the first eight months of the case.[170] This meant that Purdue Pharma, run by bankruptcy directors appointed by the Sacklers, controlled the decision whether to propose a plan that would release the Sacklers largely free of concern that creditors might propose a plan that did anything else.

The UCC also ceded the powers to seek the appointment of an examiner or trustee in the case, or to convert or dismiss the bankruptcy.[171] As noted above, these powers can be credible threats to lever information, or to fundamentally alter the course of the case. The UCC would later complain that the Sacklers failed to produce adequate information despite the stipulation.[172]

The details of the Sackler Settlement Framework would be negotiated through four court-ordered mediations, each producing interlocking ex post bankruptcy bargains.[173] These mediations modestly improved payouts; the Sacklers raised their offer from $3 billion to about $5.5 billion, which sounds significant, but they also doubled the payout period.[174] They even promised certain noneconomic remedies, such as a “document repository” that would reveal some information about the Sacklers’ role at Purdue Pharma.[175] But these were predicated on first giving the Sacklers the releases they sought. Because creditors would have little information about the claims being released—and could not litigate against the Sacklers anywhere—they would have to grant the releases without knowing what they were giving up.

Certain government creditors appealed the preliminary injunction to the United States District Court. Twenty-four of the objecting States had sued the Sacklers before bankruptcy, and wanted to continue to do so. They opposed the Sackler Settlement Framework because, they said, they wanted !@75@!  transparency. “It’s critical,” Massachusetts Attorney General Healey said, “that all the facts come out about what this company and its executives and directors did, that they apologize for the harm they caused, and that no one profits from breaking the law.”[176]

On appeal, District Judge McMahon rejected these concerns. Like Bankruptcy Judge Drain, she believed that the Sackler Settlement Framework required interim peace to produce the global peace the Sacklers sought (ironically, she would be the same judge who later struck the releases).[177] Evincing faith in the bargaining process, she noted that “[i]t cannot [] be presumed that [the objecting States] will object to a final form of settlement.”[178] Judge McMahon viewed the mere possibility of a Purdue Pharma plan based on the Sackler Settlement Framework as grounds for sustaining the injunction.

While preliminary injunctions in bankruptcy require a showing of a likelihood of success on the merits, here the success would be confirmation of plan—not the merits of the allegations against the Sacklers. But serious allegations against the Sacklers had only recently come to light in the multidistrict litigation. These, in turn, led to the direct lawsuits against the Sacklers stayed by the preliminary injunction. There had been no judicial determination on the merits of those claims, but plaintiffs were gaining on the Sacklers before bankruptcy.[179] Shutting down all efforts to hold the Sacklers accountable in some public and transparent way fueled concerns that the Sacklers were using Purdue Pharma’s bankruptcy to “get away with it.”[180]

There were at least two ways to address these concerns early in the case without dismissing it entirely. The first and perhaps most obvious would have been to permit a bellwether litigation to determine whether the Sacklers, themselves, should bear responsibility for Purdue Pharma’s misconduct. In a bellwether trial, the claims of some members of a large group of claimants are tried in order to “provide a basis for … settlement or !@76@!  for resolving common issues.”[181] Test cases are one way that courts can resolve mass claims “within a reasonable time frame, with a reasonable degree of consistency, without bringing the system to a screeching halt, or removing the hope of justice and compensation within their lifetimes from the expectations of our citizens.”[182]

Several states argued that this could have happened within the initial 180-day injunction period.[183] If they won, the trial would only determine liability—not give the plaintiff better payment rights. But it would provide insight into the strength of claims that the Sacklers played a central role at Purdue Pharma and in the company’s contribution to the opioid crisis. This, in turn, would enhance parties’ ability to price the releases that the Sacklers sought.[184]

Judge Drain rejected this because, he said, even a test case would produce a litigation explosion. “Why would I just do this one?” he asked rhetorically at the first preliminary injunction hearing.[185] Drawing an analogy to Dr. Strangelove, he said he feared that “people will want to advance so they can say I’m going to be next… . They want the next doomsday machine.”[186] !@77@!

Second, Judge Drain could have appointed an independent examiner to investigate and report on those allegations, so that creditors and the public in general would have an independent report on the mounting allegations against the Sacklers. An examiner was proposed early in Purdue Pharma.[187] Many personal injury victims apparently wanted one.[188] But Judge Drain forcefully rejected these suggestions.[189] It was “idiotic” to call for an independent examination and public report, he said, because of “the commitment as part of the injunction to have a full account, and the examinations that are going on.”[190]

But there was no “account” of the direct claims against the Sacklers and the results of investigations that were conducted remained concealed by the !@78@!  case stipulation and protective order (and claims of attorney-client privilege). The bankruptcy bargains supporting, and supported by, the preliminary injunction foreclosed opportunities for transparency and accountability, leading instead to the releases that would permanently shield the Sacklers.[191]

2.2.3 The DOJ Settlement

The third deal, which “sealed” all others, was a controversial settlement between Purdue Pharma and the DOJ which had a so-called “poison pill” that practically assured the Sacklers would get a plan with the “global peace” they sought. This was a coercive ex post bankruptcy bargain with significant distributive consequences.

Purdue Pharma under the Sacklers was not chastened by the 2007 criminal plea, and continued to violate drug marketing laws. Before bankruptcy, the DOJ opened another investigation into allegations that Purdue Pharma had continued to engage in substantially the same conduct,[192] notwithstanding a corporate compliance program intended to prevent it.[193]

In November 2020, during the bankruptcy, Purdue Pharma agreed to plead guilty to a criminal information filed by the DOJ in the United States District Court for the District of New Jersey. In its plea agreement, the company (though, again, not the Sacklers) “admitted to substantial deliberate wrongful conduct” including “aggressive efforts to boost opioid sales.”[194] !@79@!

The Purdue Pharma-DOJ settlements established that the federal government had $18 billion dollars in claims, while a separate civil settlement with the Sacklers required them to pay only $225 million.[195] Purdue Pharma and the DOJ also agreed that the DOJ would have a $2 billion criminal forfeiture judgment with “the status of an allowed superpriority administrative expense claim … with priority over any and all claims and administrative expenses of any kind.”[196] This made the United States the most powerful creditor in the case because the criminal forfeiture judgment would be paid ahead of Purdue Pharma’s other creditors, in particular, personal injury and governmental creditors.[197] It was, for all practical purposes, the only true priority creditor (although, as discussed below, legal fees exceeding half-a-billion dollars would also have priority, but junior to the “super priority” claims of the United States[198]).

In the settlement, the DOJ agreed to give over $1 billion of those payment rights to state and local governments.[199] This may have induced many states, who were often active creditors in the case, to support the DOJ settlement.[200] !@80@!

If that “gift” was a carrot, the DOJ settlement also had a “stick”: a so-called “poison pill.” The term “poison pill” derives from corporate governance practice, and is shorthand for various measures that a corporate board could take to deter a hostile acquisition of a corporation.[201] Poison pills can be coercive because they can constrain value-maximizing choices.[202]

Here, the DOJ settlement had several conditions that would lead to a plan that implemented the Sackler Settlement Framework: any deviation from those conditions would cause the DOJ’s claim to balloon back up to its full amount, essentially eliminating recoveries for any creditors other than the United States. In particular, it required Purdue Pharma to reorganize as a “public benefit company,” or similar entity,[203] which meant that Purdue Pharma had to confirm a plan of a particular sort—its case could not be dismissed, and it could not liquidate under chapter 7.

The same “nonconsenting” States that opposed the preliminary injunction also objected to the DOJ settlement on grounds that it was a “plan sub rosa”—a plan of reorganization in fact, but without the procedural protections of disclosure and an informed, supermajority vote, as required by the Bankruptcy Code.[204]

Judge Drain rejected the sub rosa plan concerns because, he said, “[i]t’s just simply not the case that a major resolution in the case which limits people’s options going forward … constitutes a sub rosa plan.”[205] This may be true in theory, but glossed over the fact that the DOJ settlement would make any alternative to a plan proposed by Purdue Pharma economically irrational. Those alternatives would have included dismissal or conversion of the case to a chapter 7 liquidation or the appointment of a chapter 11 trustee. In any of these cases, the Sacklers’ bankruptcy directors would no longer control Purdue Pharma, potentially threatening the Sackler Releases—and the Sacklers.

The DOJ settlement had three key effects. First, Purdue Pharma—and indirectly its creditors—would bear almost all liability for the alleged !@81@!  misconduct. While there is little doubt that Purdue Pharma should have borne this liability, it also meant that, other than the $225 million the Sacklers were paying as part of their civil settlement, they would likely have no further liability at all, except to the extent they so agreed under Purdue Pharma’s plan.

Second, the poison pill, and the priority of the United States’ claim rendered it something like a “dominant secured creditor,” but with a twist. In a 2004 article, Professor Jay Westbrook identified the procedural power that a dominant secured creditor—one with a lien on all of a debtor’s assets—can wield over the bankruptcy process.[206] The concern, Westbrook correctly noted, was that while bankruptcy is intended to respect the property entitlements of secured creditors, the process was to be governed by estate fiduciaries.[207]

The poison pill, and the United States’ priority rights in Purdue Pharma’s assets (through its forfeiture power), gave the DOJ powers as great as those of a dominant secured creditor, but with the complication that the standards by which it could disrupt the process and take control were contingent on an outcome—an alternative to a Purdue Pharma plan—that this very settlement made highly unlikely. The DOJ thereafter would wield great but uncertain power over the case, accelerating momentum to confirm a plan that would release the Sacklers.

Third, it strongly suggested that the Sacklers would not be prosecuted individually. Of course, Purdue Pharma’s bankruptcy could not achieve that result formally,[208] although the DOJ press release all but said so.[209] But, more fundamentally, the deal towards which the parties were negotiating contemplated Sackler payments over many years. To charge the Sacklers after the plan was confirmed would threaten to disrupt those payments: why would the Sacklers continue to pay if they were being charged anyway? To prosecute, in other words, would be to invite significant political !@82@!  blowback from the States, who would not want to lose the billion-dollar carrot they were getting under the DOJ settlement.

This invites a question: why did the DOJ want to settle? The subsequent administration took a different approach, and through the United States Trustee challenged the Sackler Releases.[210] Did the right hand not know what the left hand was doing? It is not entirely clear, but it appears that Purdue Pharma and the Sacklers worked together to persuade the DOJ in 2020 to undertake no individual prosecutions.[211] The subsequent administration may simply have had a different view about the propriety of nondebtor releases.

The DOJ settlement did not explicitly require a Purdue Pharma plan to release the Sacklers. Yet, only the most naïve would fail to see that it was a critical step toward that goal because it made any alternative to a plan proposed by the debtors—which the Sacklers still owned—economically irrational. Indeed, Judge Drain ultimately approved the DOJ settlements because he said that they paved the way for the parties to negotiate the Sackler Releases.[212] Moreover, the effectiveness of Purdue Pharma’s plan and Purdue Pharma’s sentencing were made contingent upon one another.[213] The DOJ settlement was, as Gerald Posner and I noted in The New York Times, likely to be the Sacklers’ “last poison pill.”[214] !@83@!

2.3 Why Did Creditor Representatives Agree?

While it is not hard to understand why the Sacklers and Purdue Pharma would want the three deals described above, they still had to have some creditor support. If the Sackler Releases, and the bankruptcy bargains leading to them, were so problematic, why did creditors agree to them?

Obviously many creditors, in particular the so-called “consenting” parties, believed that they had come to some sort of resolution with the Sacklers. The Sackler Settlement Framework was good enough for them.

But what of the others? In part, the ordinary science of chapter 11 practice worked against them. As explained above, before bankruptcy, the Sacklers chose corporate directors, professionals, and a judge who were highly likely to support the deal they wanted. They rendered resistance to the preliminary injunction and the DOJ settlement futile. But, there were also other constraints on creditor representatives that reduced their incentives or capacity to challenge the deals that led to the releases. Purdue Pharma and the Sacklers did not necessarily create these constraints, but they exploited them to convert the Sackler Settlement Framework into Purdue’s plan of reorganization.

2.3.1 The Official Creditors’ Committee

To act as a check on the debtor in possession, large chapter 11 cases such as Purdue Pharma will have an official committee of unsecured creditors appointed by the United States Trustee, and which “ordinarily consist[s] of the persons, willing to serve, that hold the seven largest claims against the debtor of the kinds represented on such committee.”[215] In Purdue Pharma, there was only one official committee of unsecured creditors (the UCC), and its ability to aggressively represent creditors was limited for three reasons.

First, its membership was internally divided between four personal injury creditors, four commercial creditors (including LTS Lohman, an opioid patch manufacturer, and Blue Cross-Blue Shield), and the Pension Benefit Guaranty Corporation.[216] These did not necessarily represent the same “kind” of creditors because, among other reasons, some commercial creditors were (or may have become) codefendants in other opioid litigations; the personal injury creditors were always plaintiffs.[217] Positions taken, or information revealed, in the Purdue Pharma chapter 11 case might !@84@!  hurt commercial creditors in other opioid litigations (such as the multidistrict litigation), which were progressing against many defendants.

Ryan Hampton, a personal injury creditor and opioid activist who chaired the UCC until he resigned in anger after Purdue Pharma’s plan was confirmed, wrote that the outcome of the case—in particular, the release of the Sacklers—was “the opposite of what I and many other victims sought: We repeatedly called for transparency into the process, accountability for the Sacklers who had owned the company and reparations for the millions of people affected by the OxyContin-fueled drug epidemic.”[218] While Hampton may have called for these things, the UCC appears to have openly pursued few of them.

Second, not only were there potential frictions within the UCC, but it omitted arguably the most important kind of creditor: government entities. The United States Trustee, which appoints the committee, had interpreted the Bankruptcy Code to exclude governmental units from serving on official committees of unsecured creditors.[219] This meant that the United States and hundreds of government creditors—in particular the states—would have no collective, official representation in the case.

This was not surprising, because government actors have generally not been permitted to serve on official creditors’ committees.[220] But, as explained below, this left different public actors to operate through various ad hoc committees. As indicated by the objections of the nonconsenting States, those ad hoc committees did not always agree with one another or the UCC. This left creditor groups highly fragmented.

Third, the UCC was the only statutory committee. Bankruptcy courts have the power to appoint other official committees and representatives, and this has been important in mass tort cases, especially those involving long-tail claims, such as those arising from asbestos exposure. Future claimants are presumed to have different views about reorganization than current creditors. Recognizing the need to enable future claimants to participate in reorganization formally, Congress in 1994 explicitly conditioned the availability of nondebtor releases in asbestos cases on the !@85@!  protections of a future claims representative.[221] But, the rules about NDRs involving asbestos liability do not apply to other types of debt. There was no other formal protection for individual creditor interests in Purdue Pharma, a gap the Sacklers and the debtor were able to exploit.

2.3.2 Ad Hoc Committees

Rather than statutory committees or official representatives of creditors, Purdue Pharma relied on about a dozen “ad hoc” committees to represent many different creditor groups.

Ad hoc committees are “loose affiliation[s]” of creditors, formed to represent self-identified interests.[222] Importantly, ad hoc committees “do not purport to represent the interests of any other parties in interest.”[223] Unlike official committees, they are self-selecting, and are not appointed by the United States Trustee. Neither they nor their professionals are statutory fiduciaries; nor are they or their professionals subject to the Bankruptcy Code’s special provisions regarding conflicts of interest.[224]

Three ad hoc committees were especially important in Purdue Pharma: (1) the committee representing the so-called “consenting” parties; (2) the committee representing so-called “nonconsenting” states; and (3) the committee representing individual personal injury claimants.

The consenting parties group was important because it was initially comprised of eighteen plaintiffs (mostly governments) which had apparently agreed to the Sackler Settlement Framework before bankruptcy.[225] The nonconsenting states, by contrast, were probably the most aggressive and !@86@!  successful opponents of the Sackler Settlement Framework early in the case. As noted above, they challenged the preliminary injunction and the DOJ settlements, albeit unsuccessfully. In the end, all of these States capitulated to the settlement framework. Even the “nine” holdouts who won the appeal before the District Court capitulated to the Sackler Releases thereafter, during the further appeal to the Second Circuit, in exchange for a modest increase in real consideration.[226]

The ad hoc committee of personal injury claimants was, in retrospect, the most problematic of the ad hoc committees. This committee claimed that it was “comprised of 60,761 personal injury claimants.”[227] While its initial disclosures indicated that it had had between 8 and 13 individual members, they were later amended to remove all member names in order to protect allegedly confidential information. It was not clear who constituted this committee or how they were appointed. It was thus not possible to determine whether they or their counsel could adequately represent scores of thousands of personal injury creditors in the case.

The personal injury claimants’ ad hoc committee was also problematic because its counsel apparently took the lead in drafting “trust distribution procedures” that would make it difficult for individuals to recover for their claims against Purdue Pharma, and agreed without reservation to release the Sacklers, even though they would retain at least $6 billion which individuals could have tried to pursue outside of bankruptcy absent the preliminary !@87@!  injunction discussed above.[228] Among other things, counsel to this group advocated for procedures that would have eliminated their putative clients’ right to a jury trial against the debtors (later modified), and did eliminate it as to the Sacklers.[229] Those procedures also imposed significant burdens of proof on individuals, including that they have evidence of original prescriptions for OxyContin.[230] Because many victims died without having had a prescription, or cannot find the one they had, this would eliminate the claims of many creditors who might otherwise qualify for the meagre amounts available under the plan (topping out at around $48,000).[231]

The ultimate problem with the ad hoc committees is that they, too, were constrained by their own deals with Purdue Pharma: in many cases, the debtors agreed to pay their lawyers’ fees in connection with the decision to support the Sackler Settlement Framework and the plan that embodied it.[232] While estate professionals are paid by the estate, their compensation is different from that of the professionals of ad hoc committees. The fees of the former are subject to various standards under the Bankruptcy Code and the !@88@!  fee application process, which requires estate professionals to submit fairly detailed statements about their work and how it benefitted the estate.[233]

The payment of the fees of professionals retained by ad hoc committees in Purdue Pharma, by contrast, was agreed to largely through various deals struck in the case, including the plan of reorganization.[234] It is not clear what legal or evidentiary standard governs such payments. Notably, in some cases, the fees were incurred prior to bankruptcy. Absent those special deals, such fees would presumptively be unsecured claims (if recoverable from the debtors) or payable from creditors pursuant to whatever arrangement particular creditors had with their lawyers (e.g., contingency fee arrangements). Given the modest payments to personal injury creditors and the long duration of payment terms under the plan, creditors’ lawyers would otherwise have faced significant collection challenges. Thus, just as the Sacklers were able to buy off the initial consenting parties, whose bankruptcy bargain jumpstarted the case, Purdue Pharma (largely with Sackler funding) induced professionals representing ad hoc committees to capitulate to the Sackler Settlement Framework.

In addition to these soft conflicts, there have long been concerns about the distortions that arise from repeat play in chapter 11. Large cases tend to be dominated by a small number of judges, law firms and distress professionals who appear together frequently in a few select districts that some argue seek big cases.[235] This was part of the concern about Judge Drain: until he announced his retirement, six years before his term ended,[236] he was (and may have continued to be) a repeat player in large chapter 11 cases because he was a respected judge who would try to facilitate deals sought by major parties in a case.[237] !@89@!

Ordinarily, encouraging deals is not problematic. Indeed, it is what chapter 11 was designed to do. But, social debt calls for more than mediated deals struck in secret. While most disputes settle—including many involving high-stakes social problems—they do so within a framework of adversity which provides greater confidence that responsible individuals are being held accountable according to the rule of law. That is a very different framework from the one created under chapter 11, especially as it is practiced according to the rule of the deal.

3.  Fallacies of the Rule of the Deal

Compromise implies contradiction. Litigants decide, for whatever reasons, that peace matters more than principle. This is rarely problematic. Cases involving social debt, however, demand more. This Part argues that the rule of the deal as applied in Purdue Pharma rested on, and reflects, a series of fallacies and contradictions which undermined rule-of-law values and creditor recoveries.

3.1 The Contradictions of the Rule of the Deal

The contractualist position discussed in Part 1 has long been criticized on grounds of intellectual inconsistency: if private ordering is so great, why recruit a public mechanism—a bankruptcy court—to implement deals that rational actors should accept?[238] Professor Diane Dick has argued that chapter 11 theory rests on an “Efficiency Fallacy,” the “assumption that negotiations naturally lead to efficient restructuring outcomes.”[239]

The typical response is that the power to impose NDRs reduces the incidence of holding out and information failures. But holding out is always a problem in collective action, and private ordering has developed capacities to manage it.[240] Moreover, the equal and opposite problem—taking the rights !@90@!  of others without consent, as the Sackler Releases would do—may be at least as problematic.

Some fallacies in Purdue Pharma were obvious. Judge Drain insisted that the Sackler Releases were not an adjudication on the merits.[241] But that was true only in the formal sense that none of the important judicial mechanisms of transparency and accountability embedded in our system—such as motion practice and a trial on the merits—led to it. As noted above, the preliminary injunction that halted direct litigation against the Sacklers was justified not by reference to the merits of those claims, but instead to the probability that a plan of reorganization would be confirmed.

It was false as a substantive matter because the outcome would be the same as a judgment for the Sacklers: the releases in Purdue Pharma would permanently enjoin all civil efforts to hold them accountable for their role in the opioid crisis. It has the force of a judgment on the merits, without any of the process required by law to reach a judgment on such an important question.[242]

But there were other, equally problematic contradictions. Here, it appears that creditors holding out did not destroy value, but instead drove the Sacklers to pay more. At the same time, without the ability to obtain releases, the Sacklers had little incentive to agree to participate at all. Determining how to set the right incentives to maximize recoveries has long been a central challenge for all economically-oriented legal spheres, including chapter 11. Whether assent—or resistance—creates value in chapter 11 is a complex and contested question.

Social debt complicates the analysis, because at least some of the problems in these cases are difficult to cash out. Still, if chapter 11 is about bargaining and deals, then actual assent has to matter as to claims against nondebtors. That, of course, is what was lacking in Purdue Pharma.

3.1.1 Creditor Support for the Releases

Consider first the problematic claim that creditors supported the Sackler Settlement Framework. !@91@!

Judge Drain stated at the confirmation hearing that “an unprecedented number” of votes were cast on the plan.[243] But this was not quite true,[244] and in any case, what he really meant was that Purdue Pharma’s plan was supported by the vast majority of creditors who voted.[245] But fewer than 20% of the 618,194 claimants entitled to vote—and fewer than 50% of the subset of claimants with personal injury claims—cast any ballot on Purdue Pharma’s plan.[246] The United States, in its capacity as the largest (and first priority) creditor, joined most creditors and submitted no ballot.[247]

Moreover, there are questions about who actually voted, and how their votes were counted. It appears that the Purdue Pharma plan permitted voting by “master ballot” by attorneys representing four or more clients.[248] The plan valued personal injury claims for voting purposes at $1.00/claim, which could not have been accurate if (as allowed) they could be worth up to $48,000 per claim.[249] The Bankruptcy Code presumes that claims are allowed in the amounts stated in the proof of claim.[250] Where claims are contingent or unliquidated, their value may be estimated for plan voting purposes.[251] In Purdue Pharma, however, there was no effort to estimate the value of these claims before plan voting. Instead, the Purdue Pharma Bankruptcy Court reversed this presumption, and placed the burden of showing an entitlement to vote in an amount greater than $1.00/claim on the individual claimants. !@92@!

Other courts have had to grapple with these issues, and sometimes come out the other way. In In re Quigley, for example, the Bankruptcy Court refused to confirm a plan where there were concerns that a vote, cast mostly on a master ballot, was “manipulated” and “tainted” by side deals cut by the corporate parent of a debtor, which sought the benefit of nondebtor releases.[252] These deals “led the court to conclude … that the vote was “tainted,” the plan was not proposed in “good faith” under section 1129(a)(3), and that the favorable votes induced by the parent’s side deals should be designated and not counted under section 1126(e).[253]

Purdue Pharma’s creditors also lacked meaningful information about the direct claims against the Sacklers that would be released under the NDRs. To solicit votes on a plan, the bankruptcy court must first approve a “disclosure statement” which contains “adequate information” to enable creditors to make an informed decision whether to vote for or against it.[254] In Purdue Pharma, the disclosure statement devoted about forty pages to discussions of the estate’s claims against the Sacklers (e.g., for fraudulent transfers), investigations of those claims undertaken by both debtor’s counsel and UCC counsel, and why the Sacklers were difficult collection targets (having offshored about half of the funds they’d stripped out of the company before bankruptcy). But it gave creditors’ only two pages on third-party direct claims, and focused almost entirely on why such claims would be futile. Notably, it appears that the disclosure statement does not contemplate the sort of consumer fraud types of claims that were asserted in the direct actions, but instead more conventional tort analysis.[255] !@93@!

Confusingly, Purdue Pharma claimed in the disclosure statement that the value of the direct claims against the Sacklers (and other nondebtors) was “unknowable” and then booked them for plan purposes at $0.[256] Judge Drain overruled objections to the disclosure statement on these grounds, and approved it notwithstanding this omission.[257]

But the direct claims asserted in hundreds of prebankruptcy lawsuits against the Sacklers were “knowable.” As previously discussed, the parties could have taken certain steps to attribute some value to these claims, whether through bellwether lawsuits or otherwise, but the procedural path of Purdue Pharma’s chapter 11 case foreclosed those opportunities. The disclosure statement told creditors none of this.

Moreover, the creditor representatives noted above (many of whom had capitulated to the Sackler Settlement Framework early on) exhorted creditors to vote for the plan. The UCC, for example, sent a “plan support letter” urging this.[258] Plan support letters, like many of the mechanisms used in Purdue Pharma, are not in themselves unusual or necessarily problematic. Sometimes, courts will seek to present a balanced picture to creditors and permit the parties to include letters for and against the plan (if there is opposition).[259]

Here, the problem was that the only statutory fiduciary for creditors appears to have acceded to the Sackler Settlement Framework out of resignation, not principle. Its plan support letter argued, in essence, that litigation would be difficult and time-consuming; the Sacklers may be judgment proof; therefore, the available deal was better than any alternatives.[260] But this simply restated the debtors’ own assertions. There !@94@!  was almost no analysis of the merits of the direct claims against the Sacklers, or of the evidence on which the UCC was basing its recommendation.

As in electoral democracy, a failure to vote has no effect on the votes cast, so could not prevent confirmation of the plan. The problem in Purdue Pharma, however, is that elections do not determine individual rights as to third parties: contracts and courts do.[261] While nearly 20% of creditors may have voted for Purdue Pharma’s plan, the great majority of creditors of Purdue Pharma—that is, holders of about 500,000 claims—expressed no view one way or the other. Buried in boilerplate was the proviso that a failure to vote would be deemed a vote for the plan.[262]

It is no answer to say that chapter 11 plans routinely include nondebtor releases, as Judge Wiles has observed.[263] Even if that is true—and it is an empirical question not yet answered—this seems especially problematic in cases involving social debt. The severity of the allegations against the Sacklers required more than chapter 11 deal-making, because the social and political stakes were so high. The outcry surrounding the case shows that the general public had significant misgivings about what Purdue Pharma’s bankruptcy sought to accomplish, even if within the confines of the rule of the deal.

Nor does it help to say that it was a settlement. If the plan really was a settlement, it should only have released creditors who assented to it. A voluntary settlement “cannot possibly ‘settle,’ voluntarily or otherwise, the !@95@!  conflicting claims of [those] who do not join in the agreement,” the Supreme Court has said.[264] “Of course,” the Court has admonished, “parties who choose to resolve litigation through settlement may not dispose of the claims of a third party … without that party’s agreement.”[265] That, however, is what Purdue Pharma’s plan ultimately does.

Moreover, there are questions about the representation of individual claimants in the mediations that produced the settlements. Judge Drain accepted the results of the mediation because, he reasoned, “[t]he people representing the personal injury claimants in the mediation were some of the most effective personal injury lawyers in the world, which means that they are aggressive, creative, knowledgeable and responsible in the pursuit of their clients’ claims.”[266] He supported his findings by reference to statements submitted in support of the plan by noted personal injury lawyers, such as Jayne Conroy, who asserted that the payments under the trust distribution procedures reflected a “settlement premium.”[267]

That may or may not be true, but a review of the mediator’s report cited by Judge Drain shows that the only counsel representing personal injury claimants in the mediation were the lawyers retained by the personal injury ad hoc committee, led by the Wall Street firm White and Case.[268] Ms. Conroy’s name is nowhere to be found in the mediator’s report on which the Bankruptcy Court relied.[269]

Other judges have rejected Purdue Pharma’s approach, stating that consent “by silence” will be ineffective.[270] In SunEdison, for example, Judge Bernstein observed that “[a]bsent a duty to speak silence does not constitute consent.”[271] Frequently, chapter 11 plans with NDRs require some evidence of assent, consistent with the contractual—not electoral—nature of the instrument, by giving creditors the option to grant releases or not.[272] !@96@!  Creditors who opt out (or do not opt in) can still “have their day in court” against the nondebtors.

Indeed, the term “release” is a misnomer. Except in the world of chapter 11, a release is a contract term, and is enforceable (or not) subject to the usual rules that govern contracts.[273] The confirmation of the Purdue Pharma plan had the effect of a verdict in the Sacklers’ favor, but the process leading to it neither determined the merits of those claims nor gave creditors insight into what the plan would take from them. It was a deal without any pretense of individualized assent; so the assertion that creditors agreed to release the Sacklers is simply wrong. Many did. But it is hard to infer assent from the far greater number who did not vote, when they had neither meaningful information, nor realistic choice, about the direct claims they would be “releasing.”

3.1.2 Sackler Separation from Purdue Pharma

Contradictory views about the relationship between the Sacklers and Purdue Pharma were a second class of fallacy that supported the deals in Purdue Pharma.

On one hand, Purdue Pharma and the Sacklers had separate legal identities—that’s why Purdue Pharma was in bankruptcy and the Sacklers were not. Thus, the Bankruptcy Court found that the Sacklers’ assets were simply out of creditors’ reach, a fact Judge Drain said he found “incredibly frustrating.”[274] Moreover, they were distinct for purposes of the attorney-client privilege.[275] !@97@!

Yet, while Purdue Pharma and the Sacklers were separate for these purposes, they were indistinguishable for many others. They were, for example, “inextricably intertwined” for purposes of the preliminary injunction;[276] the Sacklers were Purdue Pharma’s “de facto CEO” for purposes of the DOJ Settlement;[277] Purdue Pharma was, in the eyes of personal injury lawyer Michael Moore, the “Sackler Company.”[278] This was because the evidence indicated that they were “heavily involved in decisions on how to market and sell opioids … and oversaw sales and marketing budgets and corresponding upward trends in OxyContin prescribing.”[279]

Moreover, Judge Drain seemed to believe there was an identity between claims against the debtors and claims against beneficiaries of the Sackler Releases—that is, all were essentially “derivative” and not “direct.” The Sackler Releases “cover claims that are truly derivative of the Debtors’ claims,”[280] he found, because he apparently viewed them as akin to veil-piercing or breach of duty claims.[281]

But Judge Drain could not have known that. As discussed above, the direct claims against the Sacklers were only coming to light as a result of the opioid multidistrict litigation. The preliminary injunction he granted at the outset of the case foreclosed all efforts by courts of competent jurisdiction to determine the merits of those claims. In any case, the consumer protection statutes under which the Sacklers were being sued would apparently impose independent liability, even if the defendant was the director of a corporation with joint liability.[282] Thus, under well-established precedent, direct and !@98@!  particularized creditor claims are not property of the estate—they are property of the creditor and cannot be asserted or settled by the estate.[283] 

Similarly, the Bankruptcy Court seemed to believe that, even if the Sacklers had independent liability for the debtors’ misconduct, creditors could not enhance payouts because the Sacklers would then have reimbursement or indemnity claims against the debtors under limited partnership or similar agreements between the Sacklers and the debtors. Any effort to recover from the Sacklers on direct claims, the court said, would have been little more than a “second fork in the pie.”[284] Thus, as the court appeared to see it, creditors who retained the right to recover directly from the Sacklers would get an “extra” helping.

This assumed that there was one “pie.” But that would be true only if they did not have separate legal identities. If they were separate from Purdue Pharma, then there were as many potential pies as there were defendants to sue. That’s what joint and several liability is. Thus, the right to sue the Sacklers would not have been a second fork in the same pie. It would have been the same (or a similar) fork in other pies. Some creditors may therefore have recovered more than others. But that is always true outside bankruptcy, and a credible threat of liability may have led the Sacklers to pay more to get creditors’ actual assent to the releases.

Judge Drain is a sophisticated jurist, who surely understood that shareholders and directors (and their assets) are separate and distinct from the corporations they own or manage. How then could he have believed there was only one pie—not many?

Much of his analysis of the direct claims came via a discussion of jurisdiction, and whether litigation against the Sacklers “‘might have any conceivable effect on the bankruptcy estate.’”[285] A “conceivable effect” on the estate would, in Judge Drain’s view, result from claims that the Sacklers could assert against Purdue Pharma for indemnification or contribution under limited partnership or similar agreements governing the debtors and their roles as shareholders and directors.[286]

But the Bankruptcy Court made no findings on any of these issues. Nor could it, because no terms giving the Sacklers such rights were considered by the court. If they were, surely the Sacklers’ claims should have been disallowed under sections 502(d) and (e) of the Bankruptcy Code, which !@99@!  bar recovery by creditors against whom the estate has causes of action (e.g., for fraudulent transfer, as here) and disallow contingent claims for reimbursement or contribution.[287] Any Sackler claims that survived may also have been subject to equitable subordination.[288] Moreover, it appears that some of the underlying state laws in issue would forbid insurance coverage for, or indemnification and contribution claims by, defendants in the position of the Sacklers.[289]

The fact that a legal matter may have a “conceivable effect” on a bankruptcy estate for jurisdictional purposes says nothing about the underlying merits of the claim. To have jurisdiction over a matter does not provide a rule of decision by which the matter may be resolved.[290] Here, the Bankruptcy Court applied no rule other than that the Sackler Releases were part of a “settlement” embodied in a plan of reorganization. But the direct claims were not the estate’s to settle and, in the absence of individualized assent, any claim that they were being “settled” in the bankruptcy rested on a series of problematic contradictions.

The rule of law depends, in part, on consistent application of basic principles, especially within a case.[291] As Radin asserts, the rule of law requires rules which are applied without contradiction.[292] To implement the nonconsensual “release” in Purdue Pharma—an oxymoron, itself—the Bankruptcy Court had to accept a number of contradictions or fallacies. While that may be the price of compromise and settlement in ordinary commercial reorganizations, the social character of liability for contributing to the overdose crisis made this difficult to accept. As explained in the next Part, those contradictions may also have undermined creditors’ recoveries. !@100@!

3.2 The Economic Failings of Purdue Pharma

During a six-hour reading of his plan confirmation order, Judge Drain declared himself “bitter” about the result in Purdue Pharma, which he then spelled out for greater emphasis: “B-I-T-T-E-R.” [293] He was bitter, he said, because the Sacklers did not contribute as much as he had hoped.

It is a fair point. It is true that, during the case, the Sacklers increased their offer from $3 billion payable over nine years to $5.5 billion over eighteen years. But the dynamics of the case, reflecting the rule of deal, created little incentive for them to pay more.[294]

The problem was one of basic litigation strategy. As J.B. Heaton has observed, the economics of settlement will be the product of background legal or social pressure.[295] Here, those who rejected the deal created the pressure. If it was not enough, it was because the debtors, creditor groups allied with the Sacklers, and the judge created or permitted conditions that made it difficult to exact more from the Sacklers. Thus, Purdue Pharma was also problematic as measured against the distributive standards that ordinarily govern under the rule of the deal, for three reasons.

First, although the Sacklers increased the face amount of their contribution during the case, its present value actually rose little, if at all, because their payment period was doubled from 9 to 18 years.[296] Media reports estimated that the Sacklers would be able to make the promised payments solely from income on the funds they withdrew from Purdue Pharma.[297] If so, they would keep as principal about $10 billion in proceeds from conduct that many, including Purdue Pharma, consider criminal. !@101@!

Second, while the Sacklers’ real contribution increased modestly, the value of the debtors (as stated in their monthly operating reports) appears to have declined significantly. Recall that the Sackler Settlement Framework requires not only a cash contribution from the Sacklers, but also that creditors are paid from the proceeds of Purdue Pharma’s future sales.

At the outset of the case, in October 2019, the debtors (other than the holding company parent entity) showed consolidated equity of almost $2 billion.[298] By December of 2022, however, that had dropped to about $1.4 billion.[299] Thus, while the optics of the Sackler contribution may have improved, economic reality for creditors has deteriorated.

Third, counsel to the debtors frequently made a point of saying that funds would go only to opioid abatement.[300] While some funds may ultimately serve that important purpose, it should be noted that, as of this writing, no dollars have gone to opioid abatement. But, over $740 million—a little over half the reported value of the debtors—has been spent on professional fees, in particular attorneys, through December of 2022, with the meters still running.[301] And this does not include hundreds of millions of additional dollars that will be paid to personal injury lawyers from trusts created by the plan.[302] Early estimates pegged total professional fees at around $1.2 billion—or roughly half the present value of what creditors can expect if the Sackler Releases are reinstated, and the Sacklers keep their promises.[303]

Moreover, problematic insiders have received significant payments during the case. Board member Peter Boer, for example, appears to have received over $2 million in directors’ fees.[304] But, he does not appear to have been a member of the board’s “special committee” of bankruptcy directors !@102@!  discussed above. Instead, he was appointed in 2008, and served on the board while Purdue Pharma was apparently committing its second set of confessed crimes.[305] Equally bad: there is evidence that he advised the Sacklers on how to strip assets out of Purdue Pharma in July 2007, after the company’s first criminal plea, in May 2007.[306]

Contrast that with payouts for personal injury creditors. Individual claimants can recover no more than about $48,000.[307] All personal injury claims (that is, as a class) are capped at $750 million.[308] As noted above, the trust distribution procedures would also require adult claimants to show that they had a prescription for OxyContin.[309] This is problematic, because it is likely that many claimants either did not have a prescription or cannot find the one they had.

Similarly, case insiders frequently asserted that they were advancing the “public interest” through Purdue Pharma’s bankruptcy. At the first hearing on the preliminary injunction, counsel to the debtors argued that special protection was needed so that Purdue Pharma’s assets could go “towards the public good.”[310] At the hearing on plan confirmation, another attorney for Purdue Pharma said that the plan was in “the best interest of all creditors and the American public more generally, to help abate the opioid crisis.”[311]

It is, however, important to note that these and many other attorneys are billing upwards of $1700/hour.[312] These may be the standard rates for BigLaw in big cases. But, the actual public servants in the case—the scores of lawyers from the offices of State attorneys general and the DOJ, who are actually tasked with acting in the public interest—were making nothing close to that. !@103@!

To be fair, the professionals in Purdue Pharma faced significant challenges.  Their rates were presumably “market.” Professional fees have long been a sore point in chapter 11 practice and scholarship.[313] Without wading into that debate, the important point here is simply that cases involving social debt should probably be especially sensitive to concerns that the costs of the bankruptcy process are justified by the benefits creditors can expect to see. Here, many of the conditions that gave rise to worries about the character of the process may also have undermined the purported goal of the rule of the deal, which is to maximize creditor recoveries.

As to the larger question: Why didn’t Judge Drain put greater pressure on the Sacklers? He said he believed he had no choice. “It is incredibly frustrating,” he stated in his ruling on plan confirmation. He further explained,

that the law recognizes, albeit with some exceptions, although fairly narrow ones, the enforceability of spendthrift trusts. It is incredibly frustrating that people can send their money offshore in a way that might frustrate U.S. law. It is frustrating, although a long-established principle of U.S. law, that it is so difficult to hold board members and controlling shareholders liable for their corporation’s conduct.[314]

He believed he had no power to change that; he could not “impose what I would like on the parties.”[315]

This is true in a formal sense. But, from my perspective, the economic outcomes might have been better if he had allowed greater pressure on the Sacklers from the outset. That pressure could have come from decisions as simple as permitting a bellwether litigation to proceed against the Sacklers at the beginning of the case. In addition, or instead, he could have signaled neutrality about the appointment of an examiner, or deferred approval of the DOJ settlement until plan confirmation. The threat of greater forced disclosure, or the risk of case conversion or dismissal (practically eliminated by the DOJ settlement), might have led the Sacklers to pay more to induce greater participation. A disclosure statement that provided greater !@104@!  independent insight into the merits of claims against the Sacklers—or a plan that permitted creditors to opt out of the Sackler Release—might have required the Sacklers to contribute more to win necessary support.

Hindsight is easy, and I do not want to suggest that Judge Drain did anything other than what he thought was best under conditions as he saw them. Chapter 11’s deal culture was an attractive environment in which the Sacklers could seek to obscure and distract from the allegations against them while contributing to a plan of reorganization that might help ameliorate a crisis in which their company (and allegedly they) played a significant role. Judge Drain is a respected judge because he understands and operates effectively in that environment.

This may have made his court attractive to large companies with complex restructuring problems. In ordinary commercial cases, that culture can facilitate bargaining that produces creative, and perhaps value-maximizing, resolutions superior to those offered by the ordinary legal system, or bargaining outside of it.

But, social debt bankruptcies are different because the normative stakes are much higher. In Purdue Pharma, those stakes also required the involvement of political actors at every level of government, who were divided horizontally and vertically. They, in turn, had uneasy relationships with personal injury and commercial creditors, who had their own conflicting incentives. It would have been very difficult for the Sacklers (or their corporate agents and professionals) to ignore those divisions and to fail to exploit them.

To bring a deal together under such conditions was undoubtedly difficult. It took great effort, which Judge Drain (and, later, Judge McMahon) lauded. And, if a deal is your goal, then perhaps the outcome is commendable. But perhaps that is also evidence that a deal—at least, one framed by the Sacklers—was not an appropriate starting and ending point. !@105@!

Conclusion

Neither social debt nor the rule of the deal is new. Both have long been features of the legal system. I have used Purdue Pharma as a case study not because I believe it is representative of other cases, but because I hope it is not. The tragedy of Purdue Pharma is that even modest efforts to provide transparency and accountability, such as through a bellwether litigation, a litigation risk analysis, more fulsome disclosure, or line-item consent, might have produced a better outcome, allaying the concerns of victims and survivors whose demands for dignitary justice went largely unheeded and paying more, sooner. To be sure, there are other social debt bankruptcies, and whether (or to what extent) they present similar problems of transparency and accountability will require future study.

While I obviously think the Sackler Releases are impermissible, it is possible (indeed, I believe probable) that the Second Circuit or the Supreme Court will think otherwise. Purdue Pharma and the Sacklers made strong arguments to the Second Circuit that the Sackler Releases were “necessary” to provide funding to abate the opioid crisis. No one would doubt the need; whether the rule of the deal was the right path is another matter.

In the meantime, this article has shown that the Sackler Releases were virtually inevitable from the outset, reflecting strategic choices by Purdue Pharma and the Sacklers before and throughout the bankruptcy, choices which Judge Drain largely support, and which creditors could not effectively challenge. Under these conditions, the rule of the deal was a threat to the rule of law. It was also a poor deal in the bargain.

…