This series is being managed by the Bankruptcy Roundtable and Xiao Ma, SJD at Harvard Law School, xma [at] sjd [dot] law [dot] harvard [dot] edu.
Check the HLS Bankruptcy Roundtable periodically for additional contributing posts by academics and practitioners from institutions across the country.
Bankruptcy has a public and a private side. The reorganization of a private company in chapter 11 has implications for the public, and, in some reorganizations, the public interest is quite substantial. The recent bankruptcy of the third largest crypto exchange in the world, FTX, represents just the kind of corporate restructuring where the public interest is front and center. Yet the public priority embedded in these proceedings has the potential to be overlooked. In this work, we aim to change that by shining light on the stakes, the costs, and the allocative decisions to be made in what will no doubt be described as one of the most consequential legal proceedings to happen in the world of crypto. Specifically, the outcome of these proceedings will help clear up what it means to hold crypto as a form of property, as well as the custodial v. proprietary nature of the relationship between crypto exchange companies and their customers as to rights in crypto assets. The answers to these questions will not only help resolve this bankruptcy but they will also guide lawmakers and regulators as they seek a way to regulate and police the crypto market in the future. As such, we question whether the private value capturing model that is chapter 11 is the right framework—particularly when it comes to the allocation of who bears the costs—for these largely public-oriented matters.
By Ingrid Bagby, Michele Maman, Anthony Greene, and Marc Veilleux (Cadwalader Wickersham & Taft LLP)
Note: This post is the second post in a series of posts on bankruptcies of cryptocurrency companies and the emerging issues they pose. The first post can be read here (by Megan McDermott).
This series is being managed by the Bankruptcy Roundtable and Xiao Ma, SJD at Harvard Law School, xma [at] sjd [dot] law [dot] harvard [dot] edu.
Check the HLS Bankruptcy Roundtable periodically for additional contributing posts by academics and practitioners from institutions across the country.
Crypto-watchers and bankruptcy lawyers alike have speculated how customer claims based on digital assets such as cryptocurrencies should be valued and measured under bankruptcy law. However, a crypto-centric approach to valuing claims raises a number of issues. For example, measuring customer claims in cryptocurrency and making “in-kind” distributions of these assets could lead to creditors within the same class receiving recoveries of disparate USD value due to fluctuation in cryptocurrency prices. Moreover, the administrative burden associated with maintaining, accounting for, and distributing a wide variety of cryptocurrencies as part of a recovery scheme or plan may prove costly and complex. Equity holders also might challenge the confirmability of a plan where valuations and recoveries are based on cryptocurrency rather than USD, as a dramatic rise in cryptocurrency values may allow for a return of value to equity.
A recent dispute in the Celsius bankruptcy proceedings as to whether a debtor is required to schedule claims in USD, or whether cryptocurrency claims can be scheduled “in-kind,” may serve as a preview of things to come on these issues. In Celsius, each Debtor’s schedule of unsecured creditors’ claims (Schedule E/F) listed customer claims by the number of various forms of cryptocurrency coins and account types, rather than in USD. Subsequently, a group of Celsius preferred shareholders filed a motion directing the Debtors to amend their Schedules to reflect customer claims valued in USD, in addition to cryptocurrency coin counts.
Ultimately, the Debtors and the Series B Preferred Holders were able to consensually resolve the motion by the Debtors agreeing to amend their schedules by filing a conversion table reflecting the Debtors’ view of the rate of conversion of all cryptocurrencies listed in the Debtors’ schedules to USD as of the petition date. However, it remains to be seen whether scheduling of claims in cryptocurrency and providing conversion tables will become the norm in similar cases involving primarily crypto-assets. Practitioners and creditors should expect further issues to arise in the claims resolution process in crypto-related cases as claimants and liquidation trustees (or plan administrators) wrestle with how to value claims based on such a volatile asset, subject to ever-increasing regulatory scrutiny. For now, the bankruptcy process continues to run on USD.
Reorganization proceedings, in contrast to liquidation sales, constitute a rather recent development in insolvency law. Embodied by Chapter 11 in the US, this trend has been further brought to light by the European Directive 2019/1023 of 20 June, 2019 on restructuring and insolvency, that requires EU Member States to set up a preventive reorganization framework. Reorganization plans typically involve complex interplays between competing stakeholders’ interests, and the classification of claims for the purpose of voting on the reorganization plan is both an illustration of these tensions and a mechanism designed to address them. Choices of European policymakers, especially in France where the classification of claims represents a major change, can offer new perspectives on best practices for modern reorganization plans. The classification of claims is an important feature of reorganization proceedings. First, it appears as a countermeasure to the debtor-in-possession and other debtor-friendly rules. Second, it promotes the adoption of a plan against hold-out problems from hostile minority or out-of-the-money creditors. Third, it provides guarantees of fairness and viability of the plan, through the consent of a representative majority of creditors.
Given the importance of claims’ classification, the design of classes has drawn a lot of attention during the implementation of the EU Directive in France. Debtor-in-possession proceedings leave the debtor with significant power over classification, with the potential for abuse through “gerrymandering”, i.e., the strategic classification of claims to create an artificially accepting impaired class, ensuring the adoption of a potentially unfair plan. Bankruptcy statutes fail to provide clear and binding criteria to restrict such strategies. Similarly, Chapter 11 case law – although precedents, notably Matter of Greystone III Joint Venture out of the 5th Circuit, have suggested a ban on gerrymandering – has been reluctant to challenge debtors’ classifications. Instead, policymakers have set protective rules for creditors, mainly the best-interest-of-creditors test and the absolute priority rule. However, these protections may be circumvented and difficult to enforce.
In this context, procedural design is suggested as a means of reconciling debtor-friendly rules with effective protections for creditors, thus ensuring a balance of interests in reorganization proceedings. First, effective judicial review over the classification of claims appears desirable and is addressed under French law with (i) the appointment of a trustee, with limited powers, assisting the debtor in possession, and (ii) an early, dedicated and fast-tracked appeal against the classification of claims, allowing the judicial resolution of disputes over classification before the adoption of the plan. A similar result may be achieved through a reinforcement of classification hearings. Second, hostile classification strategies could be avoided through prepackaged plans, in which the debtor negotiates with its creditors prior to filing for Chapter 11. French law provides for a dedicated two-stage framework through conciliation proceedings – confidential negotiations under the supervision of a court-appointed professional – followed by fast-tracked reorganization proceedings, where the plan can be adopted through a vote in classes that have been designed within the conciliation negotiations. Such proceedings are consistent with the modern negotiated, deal-approach to reorganizations while making use of insolvency mechanisms against hold-out problems, thus favoring a preventive and pragmatic solution to distressed situations.
By Brigid K. Ndege (Lewis Brisbois Bisgaard & Smith LLP) and Christian Conway (Clark Hill PLC)
Parties have long questioned whether the existence of two programs—the Bankruptcy Administrator program and the U.S. Trustee program—to administer bankruptcy cases fails to meet the U.S. Constitutional requirement for uniformity in bankruptcy law. In 2017, an increase in quarterly fees by Congress brought this dormant constitutional issue to the forefront because it illustrated the lack of uniformity between these two programs. After the fee hike, debtors in regions administered by the U.S. Trustee program paid significantly more in quarterly fees than debtors in regions administered by the Bankruptcy Administrator program. The drastic difference in fees for debtors in the two programs resulted in legal challenges to the constitutional uniformity of the fee hike. This eventually led to a circuit split, with the Fifth and Fourth circuits holding that the fee increase was constitutional and the Second and Third circuits holding that the fee increase was not constitutional. Although the U.S. Supreme Court resolved this ensuing circuit split in Siegel v. Fitzgerald, by unanimously holding that the fee hike was unconstitutional, the Court declined to address whether the dual bankruptcy system was constitutional and the appropriate remedy for debtors who paid more fees under the fee increase.
By Mark Roe and William Organek (Harvard Law School)
Note: This is the seventh in a series of posts on the Texas Two-Step, the bankruptcy of LTL Management, and the future of mass tort bankruptcies. Check the HLS Bankruptcy Roundtable throughout the summer for additional contributing posts by academics from institutions across the country.
Earlier posts in this series can be found here (by Jin Lee and Amelia Ricketts), here (by Jonathan C. Lipson), here (by Jared A. Ellias), here (by Anthony Casey and Joshua Macey), here (by David Skeel), and here (by Ralph Brubaker).
Considerable attention is now being paid to the Texas Two-Step in bankruptcy. The Two-Step anticipates the movement of assets and liabilities from one corporate entity to another, via a divisive merger that splits the assets and liabilities of the original entity. After the movement of the assets and liabilities, the liabilities sit in one entity (often a subsidiary of a larger enterprise). Meanwhile, the liabilities are separated from many of the assets (in the most controversial form of the Two-Steps), which sit in another entity. The entity with the bulk of the liabilities then files for bankruptcy. The Two-Step is central to the Johnson & Johnson (“J&J”) bankruptcy of a subsidiary, aiming to separate the talc liabilities from J&J’s extensive assets.
When assets are transferred from a firm that is thereby rendered insolvent, or when the assets are transferred with actual intent to hinder, delay or defraud creditors, the transferred assets can, of course, be recovered by the transferring firm as a fraudulent transfer. Bankr. Code § 548. Similar transactions are regularly accomplished under corporate structures as spinoffs: the firm moves assets into a subsidiary, for example, and then “spins” off the subsidiary’s stock to the firm’s stockholders. After the spinoff the old stockholders own two companies, one with the assets (and possibly some of the liabilities of the just-created subsidiary) and the other with the liabilities (and any remaining assets) of the original company.
The Texas divisive merger statute creates a fraudulent transfer conundrum, because it says movements of assets pursuant to a divisive merger are not transfers. If there’s no transfer, there’s no fraudulent transfer liability, as there must first be a transfer for there to be liability.
If the bankruptcy process were ousted of power to control fraudulent transfers, then the debtor firm would have more freedom to move assets and liabilities in ways that would allow the firm to escape liability. No judge would get to the meaty issues (e.g., was the transferring firm insolvent? were the transfers done with intent to hinder, delay, or defraud creditors? is the bankrupt firm with the bulk of the liabilities an alter ego or successor of the original firm?) because there’d never have been a triggering transfer.
The Texas statute itself is clear on its face that there’s no transfer under Texas law:
When a merger takes effect . . . all rights, title and interests to all . . . property owned by each . . . party to the merger is allocated . . . as provided in the plan of merger without . . . any transfer or assignment having occurred . . .
Commentatorshavenoted that a Two-Stepping debtor might make such an argument. While it has not yet been explicitly raised in the LTL bankruptcy, the commentators anticipated correctly that such an argument was coming. In another pending Texas Two-Step bankruptcy, In re DBMP, the debtor made this argument at length. In an oral ruling1 delivered a little more than a week ago, on July 7, Judge Craig Whitley agreed with the key plain meaning premises of the Two-Step argument but ultimately rejected it as facilitating “wholesale fraud.”
The court began by accepting the debtor’s interpretation that, under a plain meaning reading of the Texas statute, no transfer occurred; and under a plain meaning reading of section 548 of the Bankruptcy Code, a transfer is a necessary predicate for a fraudulent transfer to have occurred. Hence, a plain meaning construction of section 548 and the Texas statute means no fraudulent transfer exposure.
Judge Whitley saw where such a plain meaning reading led, but refused to go there, rejecting the debtor’s conclusion. Going down the plain meaning route would, he said, lead to absurd results, leaving plaintiffs with “no recourse whatsoever.” And such a reading would contradict another provision of the Texas statute, which states that a divisive merger is not meant to “abridge any . . . rights of any creditor under existing law,” Tex. Bus. Orgs. Code § 10.901. Finally, Judge Whitley went deep: such a plain language reading of the Texas statute would run contrary to longstanding general principles of Anglo-American fraudulent transfer law.
Judge Whitley’s conclusion was in our view correct. But the conclusion can be reached more directly—by a plain meaning reading of the Bankruptcy Code.
The bankruptcy courts have already been instructed by Congress not to pay attention to the Texas statute, even when the state statute declares that an asset disposition in a divisive merger is not a transfer. True, considerable policy issues (e.g., supremacy of federal law, the extent to which state property and contract law is incorporated into the bankruptcy process) could well be brought to bear if the Bankruptcy Code were unclear here. But the most straightforward way to answer the question raised in DBMP comes from a parsimonious textual analysis of the Code.
Here’s the Bankruptcy Code’s two-step ouster of the Texas Two-Step:
The text of the fraudulent transfer statute, section 548, begins: “The trustee may avoid any transfer . . . of an interest of the debtor in property . . . ” (emphasis added). If we stop there, a basis might be had for examining state law for whether a transfer cognizable under the Bankruptcy Code has taken place. Property is, after all, transferred under state law.
But the Code does not say that state law governs whether a movement of property is a bankruptcy transfer. The Code itself defines the term “transfer” and does so independently of state laws’ appellations, leading to Step 2 of the ouster.
Section 101(54) defines what a “transfer” is for bankruptcy purposes, such as section 548 (governing fraudulent transfers). It states: “The term ‘transfer’ means . . . each mode, direct or indirect, absolute or conditional, voluntary or involuntary, of disposing of or parting with (i) property; or (ii) an interest in property.”
The Code thereby instructs bankruptcy courts to conclude that a transfer has occurred for each “mode . . . of disposing of . . . property . . . .” A divisive merger under Texas law is surely a “mode . . . of disposing of . . . property. . . .” As a result, for bankruptcy purposes a Texas two-step is a transfer, whatever the Texas authorities decide to call it. And, therefore, the Texas Two-Step should have no import in bankruptcy for determining whether there’s been a transfer for bankruptcy purposes. Property has been disposed of. Thus, for Code purposes there is a “transfer.” The first statutory predicate to considering whether there has been a fraudulent transfer has been satisfied and the court could then go on to the other, meaty fraudulent transfer issues. The Texas Two-Step is a transfer because the Bankruptcy Code says it is.
While we reach an identical conclusion to that of Judge Whitley, and we do not fault his reasoning from the bench, our analytic path is better in the long run for bankruptcy decisionmaking. The court’s reliance upon the best way to interpret the potential contradictions of the Texas Business Organizations Code is a precarious foundation for the ruling. One could imagine another bankruptcy court, faced with the Texas statute’s contradictions (“it’s not a transfer” vs. “it’s not in derogation of any other right”), interpreting and concluding differently.2Another judge might not consider such a result as absurd as Judge Whitley and we do. And yet another bankruptcy court could feel compelled to certify questions to the Texas Supreme Court on how to interpret the Texas divisive merger statute and its impact.
The new mass tort bankruptcies present major issues of policy and statutory construction for which answers will not be assured and apparent. But the question of whether there is a transfer for bankruptcy purposes is clear. A divisive merger is a disposition of property and, hence, the Code says it’s a transfer, thereby triggering the opening prerequisite to there being a fraudulent transfer.
The Code says so. Plainly.
1: A recording of the hearing is embedded in the linked PDF, which PDF may need to be downloaded in order to access the recording; the relevant portion of the hearing begins around 20:15 in the recording.
2: See Curtis W. Huff, The New Texas Business Corporation Act Merger Provisions, 21 St. Mary’s L.J. 109, 122-25 (1989).
By Kelly DiBlasi and Jason George (Weil, Gotshal & Manges LLP)
Increased competition, rising fuel costs, and the lingering effects of the COVID-19 pandemic have caused financial distress for airlines and other aviation companies and led to an increase in their use of Chapter 11 to address these issues. Chapter 11’s many tools, including the automatic stay, DIP financing, and the ability to reject burdensome contracts or bind dissenting creditors, are available to both domestic and foreign companies. However, one underappreciated aspect of Chapter 11 is its flexibility in different stages of a restructuring with varying levels of stakeholder consent. For example, a company at imminent risk of running out of cash, defaulting on obligations, and/or confronting creditors collecting significant overdue payables can use Chapter 11 to prevent creditors from exercising remedies against the company and its property. Other companies may already have agreement with key stakeholders on the major components of a reorganization plan, only needing an in-court process to implement certain aspects of the plan that cannot be accomplished out of court, such as binding dissenting creditors. Chapter 11 can accommodate all such scenarios. This flexibility, combined with the tools and benefits described above, makes Chapter 11 an attractive option for airlines and other aviation companies in distress. As the aviation industry continues to experience financial distress, it is anticipated that more companies in the industry will look at Chapter 11 as a viable option to pursue a restructuring.
Financially distressed companies often seek refuge in federal bankruptcy court to auction valuable assets and pay creditor claims. Mass tort defendants – including Purdue Pharma, Johnson & Johnson, Boy Scouts of America, and USA Gymnastics – introduce new complexities to customary chapter 11 dynamics. Many mass tort defendants engage in criminality that inflicts widescale harm. These debtors fuel public scorn and earn a scarlet letter that can ultimately destroy the value of an otherwise profitable business. Scarlet-lettered companies could file for bankruptcy and quickly sell their assets to fund victims’ settlement trusts. This Article argues, however, that this traditional resolution option would eviscerate victim recoveries. Harsh public scrutiny has diminished the value of the resources necessary to satisfy claims, creating a discount that must be borne by victims.
My public benefit proposal charts a new course. Instead of accepting fire sale prices and an underfunded settlement trust, the scarlet-lettered company emerges from bankruptcy as a corporation for the public benefit. This modified reorganization offers victims the greatest recovery. The continued operation preserves value during a transition period, after which the going concern can be sold efficiently. Further, assets that have been tainted by corporate criminality are cleansed behind a philanthropy shield and sold to capture the value rebound. The victims’ collective is the owner of the new company and can participate in a shareholder windfall if the reorganized company experiences strong post-bankruptcy performance.
At the forefront of a new trend in aggregate litigation, this Article proposes a public benefit alternative to traditional resolution mechanisms. This approach delivers utility that will support application in a variety of contexts, assuming certain governance safeguards are maintained. In our new age of greater personal and corporate accountability, more scarlet-lettered companies will emerge and ultimately land in bankruptcy. The need to address the disposition of tainted assets will be paramount in compensating mass tort victims trying to reassemble fractured pieces. This Article explains a new phenomenon and reconceptualizes resolution dynamics in a way that will have policy implications that transcend aggregate litigation.
The full article will be available at 117 Nw. U. L. Rev. ___ (forthcoming 2022) and can be accessed here.
On July 9, 2021, Southern District of New York Judge J. Paul Oetken held that §503(c) of the Bankruptcy Code prohibits board-appointed officers from receiving payments under a key employee retention plan (KERP) “absent a particularly strong showing that they do not perform a significant role in management,” as such officers are “insiders” within the definition of §101(31)(B)(ii). In re LSC Communications, Inc., 631 B.R. 818 (S.D.N.Y. 2021). This was the case even though the Bankruptcy Court found that the employees in question were officers “in title only.” Id. at 820. Judge Oetken’s ruling affects six employees who received KERP payments under the reorganization plan of LSC Communications, Inc. (“LSC”), which filed for Chapter 11 relief in April 2020.
In its initial order, the Bankruptcy Court approved payments under the KERP for all 190 covered employees. U.S. Trustee William K. Harrington appealed the decision to the District Court, claiming that, as LSC was incorporated in Delaware, the court should apply Delaware state law, which provides that any person appointed by a corporation’s board of directors is an officer. In response, LSC countered that the employees could not be considered insiders because they lacked any significant decision-making authority.
Judge Oetken wrote that case law regarding who constitutes an “officer” is “less than clear,” noting that the Bankruptcy Code does not define the term or provide any insight regarding its meaning. Id. at 824. Although Judge Oetken rejected the Trustee’s argument that state law exclusively applies when determining whether an employee is an officer under the Bankruptcy Code, he concluded that the Bankruptcy Court erred by looking beyond the fact that the six employees were appointed by LSC’s board. Even if a court chooses to undertake a more expansive analysis, he wrote, the fact that the six employees were appointed by the board and would be deemed officers under Delaware corporate law should “weigh heavily in concluding that the employees are officers for Bankruptcy Code purposes.” Id. at 826. In supporting his position, Judge Oetken claimed that giving more weight to objective criterion like appointment by the board provides greater clarity to the parties than the kind of “functional, non-exhaustive test” used by the Bankruptcy Court, although he noted that this kind of test may still be appropriate in some cases. Id. at 825.
This case, and its somewhat unusual holding, seems to present a scenario that has become exceedingly rare. In September, the Government Accountability Office (GAO) published a report finding that debtors regularly avoid restrictions in the Bankruptcy Code by paying bonuses before or after filing for bankruptcy or by designating bonuses during bankruptcy as “incentive” payments. The report stated that fewer than 1% of debtors requested court approval to pay employee bonuses during 2020 and, when debtors did seek court approval, the courts approved “nearly all” of those requests. Maria Chutchian wrote in a recent Reuters article that creditors rarely challenge such bonus payments in court, noting debtors will often arrange private deals with creditors before filing for bankruptcy. In the same article, she discusses the Mallinckrodt restructuring, in which the pharmaceutical company’s creditors chose to challenge bonus payments to executives facing accusations of misconduct related to the opioid epidemic. The judge allowed the payments, stating that “mere allegations” of misconduct did not suffice to render them inappropriate.
Finally, Jared Elias concluded in an article published in 2019 that, although fewer companies have used court-approved bonus plans in the years since the Bankruptcy Code’s bonus provisions were reformed in 2005, the overall level of executive compensation has remained the same. This lends support to the GAO’s analysis and the anecdotes collected by Reuters that KERPs are rarely used because parties have ample opportunity to achieve the same end while avoiding the scrutiny of §503(c). He argues that bankruptcy and district court judges are poorly equipped to police payments made before or after a debtor files for bankruptcy, and that the U.S. Trustee’s office and creditors struggle to do so due to information asymmetries and limited resources.
Given the above, it appears that Judge Oetken’s decision this past summer may ultimately prove to be an anomaly. In most cases, these bonus payments simply will not become a part of the bankruptcy plan. Debtors can avoid scrutiny by making payments before or after filing for bankruptcy. If they decide they must make the payments during bankruptcy, they can reduce the scrutiny they will incur by recharacterizing them as incentive payments, thereby avoiding the heightened restrictions that apply to KERPs. And, even if they decide to include the payments in their restructuring plan, it seems unlikely that creditors will challenge the payments—particularly if counsel has chosen to arrange a deal behind the scenes to avoid a dispute in court.
By Diana Bonfim (Banco de Portugal; Catholic University of Portugal – Catolica Lisbon School of Business and Economics) and Gil Nogueira (Bank of Portugal – Research Department)
How does corporate reorganization affect labor outcomes in bankruptcy? The existing literature argues that corporate reorganization affects the reallocation of labor because it retains workers in bankrupt firms. In some cases, bankrupt firms remain alive for too long and retain workers inefficiently. In other cases, reorganization reduces the probability of inefficient liquidation.
In this paper we show that resource retention is not the only determinant of labor outcomes in bankruptcy. The decision process in bankruptcy creates a principal-agent problem between firms’ claimholders and other stakeholders (e.g., workers, suppliers). Claimholders decide bankruptcy outcomes but other stakeholders with limited say in the bankruptcy process are also affected by these outcomes.
Workers are among these stakeholders. They use job contracts with firms as a form of insurance in times of adversity. In the absence of corporate reorganization, workers lose these job contracts and experience persistent costs of job loss. Reorganization improves labor outcomes because it reduces the probability that workers lose the insurance provided by job contracts when the costs of job loss are high.
We test this hypothesis empirically using data from Portuguese reorganization cases. The institutional setting has several features that help design an adequate empirical strategy. First, reorganization cases are randomly allocated across judges. We use this random assignment as a source of variation in the probability of reorganization that is not affected by other factors that also influence workers’ careers. Second, Portuguese firms report financial statements annually, which we use to check whether reorganization affects labor reallocation to more productive or profitable firms. Finally, we link this data to a rich administrative employer-employee matched dataset, which allows us to track workers who eventually change jobs. This dataset is unique because it contains rich job descriptors. We use this data to establish a relationship between corporate reorganization and the scarring effect of bankruptcy on workers’ job functions.
We uncover three main findings. First, we measure the effect of corporate reorganization on the sorting of workers to productive and profitable firms. In five years, only about 20% of the workforce remains in reorganized firms. Many workers from reorganized firms find jobs with new employers. We find no evidence that reorganization affects the reallocation of labor to efficient or profitable firms.
Second, reorganization is an important source of labor insurance against negative productions shocks. In the short term, reorganization increases the probability that workers are employed. In the long term, reorganization increases wages and reduces the scarring effect of job downgrading that is often observed in recessions. Reorganization reduces the probability that workers move to less skill-intensive occupations and increases occupation wage premia.
Third, we show that reorganization improves job transitions to new employers. Reorganization increases the average time it takes to leave a firm that files for bankruptcy by one year. Reorganization reduces the probability that workers move to low-paying jobs and increases the probability that workers find high-paying jobs with new employers.
Overall, our results show that corporate reorganization is an important source of labor insurance in bankruptcy, thereby mitigating the scarring effect of job loss. The full article is available here.