[*PG1]OPTIMAL BANKRUPTCY IN A NON-OPTIMAL WORLD
Abstract: Consumer bankruptcy insures individuals against misfortune. Like other forms of insurance, bankruptcy reduces an individual’s incentive to guard against misfortune and provides her with an incentive to overstate her need for relief. The “first-best,” or optimal, bankruptcy system, like the first-best tax or public assistance system, solves these moral hazards without any loss of efficiency. In bankruptcy, this first-best approach would deny relief to debtors responsible for their own distress and reduce the deserving debtors’ obligations to an amount commensurate with their ability to pay. While the Bankruptcy Code tries (in part) to follow this first-best approach, such a utopian system requires omniscient judges who can perfectly determine which debtors deserve relief and how much a deserving debtor can pay. Real bankruptcy judges have interpreted the Bankruptcy Code to implement a second-best, or feasible, bankruptcy system that accounts for the limited information that they possess.
The truly destitute have little to fear from their creditors. Their poverty prevents their creditors from seizing anything of value, and the days when default meant imprisonment, enslavement, or even death have long since passed.1 Bankruptcy protects those with something left to lose—a home, a car, future income, etc.2 Without con[*PG2]sumer bankruptcy, creditors could reach these assets,3 and debtors would be worse off. Bankruptcy therefore provides debtors a benefit—debt relief—which has economic value to the debtor only to the extent that the debtor otherwise could have paid the debt. Because an ideal bankruptcy system would provide this benefit only after the debtor has suffered some misfortune, bankruptcy can be viewed as similar to a public insurance program.4
If private insurance markets functioned perfectly, society would not need a consumer bankruptcy system to provide this form of insurance. Debtors could instead rely on private contracts to insure against risks such as illness and unemployment that trigger financial distress.5 Yet the world is not so perfect. Although consumers currently can purchase various forms of health, property, and credit insurance6 that, like bankruptcy, relieve them of their debt after they have suffered some reversal of fortune, private insurance may be unable to achieve perfect outcomes because of certain well-recognized market failures. For example, contracting costs may be too high to allow consumers to negotiate effective insurance against all risks.7 Consumers may also suffer from a host of cognitive or volitional failures that prevent them from purchasing appropriate insurance or re[*PG3]sisting the siren song of easy credit.8 Finally, individuals may have greater information about their vulnerability to financial distress than do lenders or insurers, and lenders and insurers may refuse to contract for fear of attracting high-risk individuals.9
The bankruptcy laws may be viewed as a response to these market failures. Under this view, the “optimal” bankruptcy system could be defined as the set of legal rules that best approximates the insurance contract that the consumer would purchase if the market failures did not exist. Unfortunately, a claim that judges should attempt to implement the optimal bankruptcy system provides little guidance because the literature lacks a good description of such a system. The literature on consumer bankruptcy has focused on issues such as whether debtors should be entitled to a complete or limited discharge of their debts10 and whether, and to what extent, debtors should be entitled to keep certain types of property after bankruptcy.11 More recently, scholars have focused on other policy issues such as the percentage of a bankrupt debtor’s income that should be exempt from [*PG4]seizure,12 the extent to which the debtor should have a choice between complete and limited discharges of debts,13 and the limitations on a debtor’s freedom to reaffirm pre-bankruptcy debts.14 Missing from the literature is a broader inquiry into the bankruptcy system that debtors and creditors would themselves choose in the absence of market failure.
To fill this gap in the literature, this Article looks to, and adapts, current theories regarding the optimal structure of public assistance programs and taxation.15 Bankruptcy has obvious similarities to these programs. Like bankruptcy, public assistance programs provide benefits to individuals based on their need.16 Although the type of benefit is different in bankruptcy than in public assistance programs—the forgiveness of debts owed to creditors rather than food stamps or welfare payments financed by the public fisc—the systems are otherwise quite similar. Indeed, public assistance programs are frequently justified as a form of public insurance,17 and both the optimal bankruptcy system and the optimal public assistance program would confer their benefits on the truly needy without blunting the [*PG5]incentives for beneficiaries to engage in desirable behavior—for example, to work, to save, etc.18
But bankruptcy is not just about giving relief to debtors; it is about taking from them too. Like taxation, bankruptcy often requires individuals to pay certain sums of money, although the payments in bankruptcy go to creditors rather than to the public fisc. The goal of optimal taxation mirrors that of an optimal public assistance program: it seeks to extract payments from those capable of paying without diminishing the incentives to engage in productive behavior.
In constructing theories of an optimal public assistance or tax program, economists usually begin by describing the “first-best” system, or the system that society should adopt if the government were omniscient.19 In both cases, the optimal system would impose a fixed obligation to pay (for taxation) or a fixed benefit (for public assistance) based not on the individual’s actual income or assets, but on the individual’s ability to earn.20 While this first-best solution is easy to describe, it is also trivial in the sense that it has limited application in the real world where the government is far from omniscient. The solution is, however, still valuable because it establishes a benchmark for evaluating feasible second-best solutions. Deviations from that benchmark are then justified by the imperfections in the governmental processes that run the programs.
The application of optimal tax and public assistance theory to bankruptcy leads to an interesting insight. In many ways, the Bankruptcy Code appears designed to achieve the first-best solution. This goal is particularly evident in Chapter 13, which gives debtors not a complete discharge of their obligations, but merely an adjustment of debts,21 replacing one set of debt obligations with another that is, ideally, based on the debtor’s ability to pay.22 If that goal were attainable, the bankruptcy relief conferred would be optimal. It would provide the debtor with insurance against financial distress with none of the efficiency losses associated with diminishing the debtor’s incentives to produce and save.
[*PG6] Yet, as in taxation and public assistance, the first-best solution in bankruptcy is also an unrealistically utopian solution because judges lack the information necessary to accurately measure a debtor’s ability to pay and to determine whether the debtor deserves relief. Because of this, judges cannot prevent all debtor misbehavior. Just as the opossum plays dead to ward off predators, bankruptcy may cause some debtors to exaggerate their plight in order to avoid repayment or gain a greater amount of relief; call this the “opossum problem.” It occurs, for example, when a debtor conceals assets or hides his ability to earn by working less than he is capable of working. In addition, just as the grasshopper fails to plan for the coming winter,23 the possibility of bankruptcy relief may cause some debtors to live beyond their means or engage in other negligent or even willful misbehavior that makes financial distress more likely; call this the “grasshopper problem.” Unlike the opossum, the grasshopper’s need is real, though it could have been avoided. In a world with such problems, the optimal bankruptcy system is a second-best solution—meaning one that accounts for the limited ability of courts to identify the grasshoppers and opossums among us and for the inevitable distortion of debtor incentives.
It is here, however, that the Bankruptcy Code is notably deficient. In many instances, the Code gives little guidance to courts that must adjust the utopian goal to the realities of the actual world. In response, courts have improvised. They have seized on the discretion granted them explicitly or implicitly by the Bankruptcy Code and have attempted to develop a second-best bankruptcy system that accounts for their own limitations.24 In many respects, these improvisations parallel solutions developed in implementing tax and public assistance programs.25 This Article concludes that most of this judicial improvisation is both necessary as a matter of policy and consistent with the language of the existing Bankruptcy Code.26
[*PG7] Bankruptcy judges do not deny relief to the spendthrift or the negligent27 and, despite frequent discussion of the grasshopper problem,28 few, if any, academics argue that bankruptcy judges should. Perhaps this should not surprise us. Neither public insurance programs nor private insurance contracts deny relief to the negligent, and the standard explanations for this approach apply to bankruptcy as well.29 Both public insurance programs and private insurance contracts, however, deny relief for willful misconduct.30 Here bankruptcy, or more specifically Chapter 13, provides an exception, in the form of a superdischarge that would even relieve the debtor of a judgment for a willful and malicious tort such as sexual assault.31 Most bankruptcy judges limit the impact of this anomaly by using the good faith standard of Chapter 13 to restrict access to the superdischarge to those debtors who have received other punishment and who try to repay as much as possible.32
Recently, scholars have paid more attention to the opossum problem in the form of proposals for means testing that would test whether a debtor can repay any of his debts.33 Rather than engage in this debate, this Article focuses on the only major consumer bankruptcy chapter that requires debtors to repay out of future income: Chapter 13.34 Though the text of Chapter 13 appears to invoke the utopian approach of adjusting debts based on a debtor’s ability to pay, judges resist this approach because they are skeptical of their own ability to estimate a debtor’s future income.35 This causes some judges to adopt a strained reading of the Code that effectively creates a tax on the debtor’s income in bankruptcy.36 This approach is unfortunate because judges could address their concerns in ways that do less violence to the plain meaning of the Bankruptcy Code and better use [*PG8]what abilities they do have.37 Specifically, judges could rely on subsequent modifications to a bankruptcy plan when they can provide a workable estimate of the debtor’s future income, and should dismiss a debtor’s bankruptcy petition when they cannot.38
Part I provides a brief outline of the optimal taxation and public assistance literatures and the theory of the second-best. Part II applies the lessons of this literature to the bankruptcy context and defines the utopian and second-best bankruptcy systems. Part III compares these bankruptcy systems to the existing Bankruptcy Code and argues that, although several aspects of Chapter 13 of the Bankruptcy Code appear to try to implement a utopian bankruptcy system, judges should instead consider the limits of their abilities and work to implement the second-best system within the constraints of the current language of the Bankruptcy Code.39
This Article seeks to define the optimal bankruptcy procedure in the abstract and to determine the extent to which this procedure should guide judicial analysis of existing legislation. To accomplish this goal, this Article analogizes the search for an optimal bankruptcy procedure to a well-known problem in economics: the search for an optimal program of progressive taxation and public assistance.40
For at least the past one hundred years, economists have tried to define optimal progressive taxation and public assistance programs to deal with the problems of poverty and inequality.41 Modern analyses of these redistributive programs view them either as a form of altruism or insurance.42 The insurance argument takes one of two forms. [*PG9]First, many forms of public assistance, such as unemployment insurance, may provide real insurance to existing individuals as the high-income individuals of today may be poor tomorrow and thus may themselves benefit from the relief provided.43 Second, other redistributive policies, such as progressive taxation, may insure the individual against the lottery of birth that grants some individuals the ability to earn very little and others the ability to earn a great deal.44 Regardless, the concept is roughly the same. The government should choose the redistributive program that an individual would choose before he knew the outcome of the risk against which the system is designed to insure, the system that he would choose behind a “veil of ignorance.”45
Economic articles on taxation or public assistance usually proceed in two steps. First the author defines the first-best system, or the system that the government should implement if it were omniscient and could directly solve the opossum and the grasshopper problems described above.46 But economists generally do not believe that the government is omniscient, and they focus more heavily on the description of the second-best system that is optimal in a world of opossums, grasshoppers, and a government that can only imperfectly identify them.47
An omniscient government can adopt an elegantly simple first-best tax or transfer system that allows it to achieve whatever social goals it desires without any loss of efficiency: the first-best system imposes on individuals a lump-sum tax or provides to individuals a lump-sum transfer based on their ability to pay.48 To implement this system, the government projects how much each individual would earn if he worked the efficient number of hours and then requires the individual to pay an amount equal to any projected earnings in excess of the amount that the government decides the individual should consume. If the individual’s projected earnings are less than the expenses that the government thinks proper, the government gives the individual a transfer equal to the shortfall. For example, assume that the government determines that all individuals should consume $50,000 worth of goods and services. If the government determines that a high-ability individual should work hard enough to earn $90,000, it will require that individual to pay a tax of $40,000 regardless of the amount that he earns. If the government determines that a low-ability individual should only have to work hard enough to earn $10,000, it will give that individual a transfer of $40,000 regardless of the amount that he actually earns.
Each individual may work more or less than the government estimated, but the amount that the individual works, and therefore the amount that he actually earns, does not affect his tax or transfer; that is the meaning of a lump-sum tax. Because individuals’ actual earnings do not affect their tax or transfer, their return from working an additional hour, their take-home pay, matches the social return, the amount that they produce, and they will work the efficient number of hours that the omniscient government predicted. For example, the high-ability individual could work harder and earn $100,000. If he did, he would be entitled to keep the extra $10,000 that he earned. He would, however, not value the additional $10,000 as much as the leisure that he would have to sacrifice in order to earn it. If he did value the additional $10,000 more than the leisure he had to sacrifice to get it, then it would not have been efficient for him to work only enough to earn $90,000, and the omniscient government would have determined that he should work enough to earn $100,000. Therefore, [*PG11]the high-ability individual ends up earning the $90,000 that the omniscient government predicts.
This does not mean that the first-best tax will not affect the number of hours that the individual will work. In fact, the tax will likely cause the individual to work more hours than he would have because he will be less wealthy and will therefore be willing to sacrifice more leisure in order to gain more income. For example, if the government requires our high-ability individual to pay a tax of $90,000 and he earns only $90,000, he would be left with nothing to eat. To avoid starvation, he would be willing to work a great deal in order to earn just a little more than the government will take. This wealth effect, however, is irrelevant to the question of whether the tax is efficient.49 A tax is efficient if, given the current amount of tax the individual must pay, the individual values an hour of leisure more than the amount of money he could produce by working an additional hour. A lump-sum tax is efficient because an individual is free to work another hour and keep the entire amount that he produces, but he chooses not to do so.
The government can use a lump-sum tax to achieve as much progressivity as it desires without any loss of efficiency.50 Each individual would work an efficient number of hours if the high-ability individual paid a tax of $40,000 and the low-ability individual received a transfer of $40,000 or if the high-ability individual paid no tax and the low-ability individual received no transfer. To determine the amount of each tax or transfer, one must consider the purpose of the program. Assume that the program is motivated by insurance so that the government chooses the program that an individual would choose if he did not know how he would fare in a lottery of birth that gives some individuals the ability to earn $90,000 and some the ability to earn just [*PG12]$10,000.51 Assume further that the happiness that an individual derives from an additional dollar of consumption depends only on how much he is consuming,52 and that as the individual consumes more each additional dollar yields less happiness. If these assumptions hold, the government should set the first-best tax or transfer such that the individual will always consume the same amount regardless of his ability to earn;53 the government should structure the program so that the low-ability individual and the high-ability individual will each consume $50,000. Because each is consuming the same amount, each would place the same value on an additional dollar of consumption and there would be no gain from redistributing wealth from one to the other. The government would demand that the high-ability individual pay a tax equal to $40,000, or the amount by which his projected earnings of $90,000 exceed $50,000 and offer a transfer to the low-ability individual of $40,000 or the amount by which his projected earnings of $10,000 fall below $50,000.
The first-best system described above effectively assumes away the moral hazards (the grasshopper and opossum problems) that redistributive programs create by assuming that the government is omniscient. If the government cannot identify the high-ability individuals, these individuals will have an incentive to “play ‘possum” and claim a low earnings ability in order to pay a lower tax or receive a larger transfer. For example, the political debate over whether welfare re[*PG13]cipients can in fact support themselves54 is an argument over the importance of the opossum problem. It is a debate over whether the recipients are truly needy or whether they can support themselves. In addition, if the government cannot distinguish between those whose low-ability is a result of bad luck and those who failed to invest sufficiently in the future, individuals will have an incentive to behave like the grasshopper. The debate over whether welfare laws lead to more teen pregnancies55 is a debate over the grasshopper problem. All would agree that the teenage mother is needy, but some would argue that she would not have become needy but for the prospect of relief.56 To the extent that the government cannot identify either the opossums or the grasshoppers, the optimal tax is a second-best tax.
The optimal taxation literature sometimes assumes that each individual receives his earnings ability randomly through a lottery of birth;57 it effectively ignores the grasshopper problem. One’s ability to earn, however, often depends on prior choices such as the decision to seek further education or to work hard and earn a promotion. Although a lump-sum tax based on an individual’s earnings ability would not distort his decision to work after the tax is implemented, the individual will have much less incentive to develop the ability to earn if he knows that the tax will be implemented. If the taxation system will leave everyone with the same amount of money to consume, there is no incentive to become skilled.58 For example, assume that in order to become a high-ability individual capable of earning $90,000 a year, an individual must endure several years of rigorous study. If one knew that the tax structure was such that one would consume $50,000 [*PG14]regardless of how much one could earn, there would be no incentive for one to study.
An omniscient government can solve the grasshopper problem. If some individuals have low earnings ability solely because of hard luck and others have low earnings ability solely because of their poor choices, the government could provide relief for the unfortunate but not the lazy. More realistically, both hard work and fate play a role in determining an individual’s earnings ability. An omniscient government, however, can still implement a first-best tax or transfer system by assessing an arbitrarily high tax, effectively a punishment, whenever the individual does not invest sufficiently in his future. Because individuals will fear this penalty, they will not shirk.
If the government cannot separate the grasshoppers from the unfortunate, then the only way to make the individual bear the full costs of shirking is to make the individual bear the full costs of adverse events beyond his control as well. The only way to provide the individual with the proper incentives to study and generate the ability to earn $90,000 is never to provide relief when he may earn only $10,000, even if this low earnings ability could have resulted from misfortune rather than misbehavior. Because this would destroy the insurance that redistributive programs are designed to provide, the government again faces a trade-off. It must balance its citizens’ need for insurance against the fact that this insurance will cause its citizens to shirk. What results is a system of only partial insurance. Low-ability individuals are asked to pay less than high-ability individuals, but not enough less to make them equally well-off. The resulting system is still progressive, just not as progressive as it would have been if the government could solve the grasshopper problem directly. Likewise, public insurance programs offer the individual only partial insurance against his loss. For example, unemployment insurance typically pays about one-half of the unemployed’s former wages.59
Even if the government could not estimate an individual’s earnings ability, a lump-sum tax or transfer would still have the desirable effect on each individual’s work incentives because the individual [*PG15]would retain all of his last dollar earned regardless of the size of the tax or transfer.60 If the government cannot identify those who can earn more, however, all individuals must pay the exact same amount of tax or else those who are asked to pay more would “play ‘possum” and claim that they have a low earnings ability. The government could not meet its progressivity goals with such a system.
The optimal tax literature therefore focuses on the second-best tax or transfer, or the tax that utilizes only the information that the government can actually observe.61 In the basic optimal taxation model the government can observe the individual’s actual income but nothing else.62 A precise description of the second-best tax is quite complicated and requires strong mathematical assumptions.63 This Article, however, need only describe the second-best tax as a tax that generally64 increases with the individual’s actual income; the second-best tax is an income tax. Because an income tax deprives individuals of some of the benefit of an additional hour of work, it will discourage them from working and thus lead to some inefficiency.65 An income tax allows for some progressivity, however, as the high-ability individual would have to earn, and therefore consume, very little in order to “play ‘possum” and emulate the low-ability individual. This second-best tax reflects a compromise between the goals of efficiency and [*PG16]progressivity.66 Because society must compromise, the second-best tax is neither as progressive nor as efficient as the first-best tax.67
This Part applies the results of the optimal taxation and public assistance literature to bankruptcy and demonstrates that the first-best, or utopian, bankruptcy system adjusts the debts of a deserving debtor to an amount that is commensurate with the debtor’s potential to repay.68 Like the first-best tax, however, this utopian bankruptcy system relies on information that is not realistically available.69 Therefore, one needs to define a feasible second-best bankruptcy system.
In a world of limited information, we may not want judges to search for all of the grasshoppers among us for fear that they may mistakenly deny relief to the deserving. In addition, the ability of judges to identify the opossums depends on the information that they have at their disposal, and on this matter reasonable minds may disagree. If one follows the fairly pessimistic assumptions of the optimal taxation literature and assumes that judges can only observe a debtor’s actual income, the optimal bankruptcy system resembles an income tax.70 This Article argues, however, that judges can sometimes use a debtor’s earnings history to provide a workable estimate of an individual’s potential earnings.71 When they can, judges should instead adopt an adjustment of debts approach to bankruptcy tempered by subsequent modifications of the amount of debt when the estimation proves grossly inaccurate.
Like progressive taxation or public assistance, bankruptcy policy may be justified either as a means of providing insurance for debtors or as fulfilling certain altruistic goals of society.72 While this Article [*PG17]focuses on the insurance that bankruptcy provides, the policy prescriptions would not radically change if altruism were stressed.73 Notwithstanding the commercial success of Las Vegas and Atlantic City, consumers generally do not like risk and purchase insurance even though, on average, they pay more to the insurers than the insurers pay out in claims.74 The insurer writes these contracts because the consumer pays a premium that exceeds the expected payments on the insurance contract. The insurer can charge premiums that the consumer is willing to pay because the insurer writes contracts with many different consumers and is therefore fairly indifferent to the risk that any one consumer may suffer an unfortunate event that requires compensation.75 In other words, the insurer can effectively “diversify” the risk away.76 Consumers choose contracts that balance the insurance that they receive against the premium that they must pay.
Bankruptcy effectively makes creditors the insurer of their debtors by transferring wealth from creditors to debtors, through a reduction in debts, after debtors have suffered some misfortune.77 Many [*PG18]creditors willingly supply this insurance by extending credit while knowing that their debtors may file for bankruptcy. Like the insurance company, the large creditor contracts with many consumers and can diversify away much of the risk of any one consumer’s suffering the adverse circumstances that lead to default.78 Just as the insurer demands compensation for the risk through premiums, most creditors demand compensation for the risk of default through higher interest rates.79
Because lending markets are highly competitive80 and money can be readily invested outside the consumer lending market,81 debtors are likely to bear most if not all of the cost of bankruptcy protection in the form of higher interest rates or reduced access to credit.82 [*PG19]Therefore, the optimal bankruptcy procedure is the bankruptcy procedure a rational debtor would include in a hypothetical contract made at the time of borrowing if the debtor knew that the amount of insurance and the structure of the contract would affect the other terms of credit offered by the creditor. The optimal bankruptcy system, like the optimal tax, is the system that an individual consumer would choose at the time of borrowing behind a veil of ignorance as to his future condition.83
In assessing the debtor’s ability to pay, the optimal bankruptcy procedure would likely consider several factors including the value of the debtor’s assets and any unusual expenses that the debtor faces. This Article follows the general approach of the optimal taxation and public assistance literature, however, and focuses solely on the debtor’s earnings.84 One can extend the analysis to consider the debtor’s assets and expenses, but the gain in realism would not be worth the resulting analytical complications at this time.85 The vast majority of bankrupt debtors do not have,86 and likely never had,87 significant assets. In addition, while unexpected expenses, such as a hospital bill, are often cited as a cause of bankruptcy,88 these expenses are often one-time events that affect the total obligations of debtors but not their expenses going forward. Therefore, they do not affect the total amount that debtors should pay in bankruptcy. In addition, [*PG20]many bankruptcy scholars have abandoned faith in the ability of bankruptcy judges effectively to estimate a debtor’s reasonably necessary expenses,89 and both the National Bankruptcy Review Commission’s proposed reforms90 and proposed legislation91 would limit consideration of the debtor’s particular expenses in the great majority of filings.
The analogy between optimal bankruptcy and optimal taxation or redistribution is imperfect. Many of the limitations of the analogy are discussed below, but two points deserve special note. First, when implementing a tax or public assistance program, the government may transfer as much wealth to an individual as it wishes, provided that it raises enough money to balance its budget. By contrast, the bankruptcy judge can do no more than forgive all of the debtor’s obligations; even in a hypothetical bankruptcy regime it would be odd for a judge to order a creditor to transfer additional funds to a debtor in default.92
Second, while the optimal tax defines how all members of society should be treated, only an extremely small fraction of the population files for bankruptcy.93 The number of bankruptcy filings is even small when compared to the number of individuals who experience the un[*PG21]fortunate events that are often listed as causing the financial distress that leads to bankruptcy: unemployment, divorce, etc.94 If consumers are risk-averse, they should want insurance against all shocks, not just those that are severe enough to land them in bankruptcy. Therefore, the fact that bankruptcy provides no insurance to debtors except in those very rare circumstances when they fail to pay their obligations in full presents somewhat of a puzzle.
This puzzle is not unique to bankruptcy. Perhaps part of the answer95 can be found in an argument economists use to explain why many insurance policies only cover losses above some amount.96 This literature suggests that the cost of verifying the insured’s loss prevents the insurance contract from covering small losses.97 Economists extend the same logic to bankruptcy and suggest that because bankruptcy is a costly process, a court should only inquire as to the debtor’s circumstances when the debtor’s income is particularly low.98 [*PG22]When the court does inquire as to debtors’ circumstances, however, the court should fully insure debtors so that they always consume the same amount after this investigation.99 Therefore, subject to competing goals, bankruptcy should try to ensure that a debtor consumes the same amount regardless of the severity of loss that led to bankruptcy.100
If judges were omniscient and could identify the grasshoppers and opossums, they could implement a utopian bankruptcy system that, like the first-best tax, provides debtors with full insurance against adverse events and does not lead to any inefficiency.101 To avoid the grasshopper problem, this utopian bankruptcy system would deny relief when debtors’ financial distress resulted from their own misbehavior. For those debtors deserving of relief, the utopian bankruptcy system would reduce their debt to a new amount based on their ability to repay, thus eliminating the opossum problem.
A utopian bankruptcy system would insure debtors against misfortune without encouraging them to borrow an excessive amount or to misbehave in other ways that make financial distress too likely. An omniscient judge could achieve this goal by always granting relief for financial distress resulting from misfortune and never granting relief for financial distress resulting from misbehavior.102
Unfortunately, this guideline may prove of little practical use because often both misbehavior and misfortune will jointly cause [*PG23]financial distress.103 For example, a temporary period of unemployment might not lead to bankruptcy unless the debtor has failed to save a sufficient amount to withstand a temporary loss of income.104 Moreover, a firm may fire a worker in part because of an economic downturn and in part because the worker was not quite as diligent as his co-workers who were retained.
Still, the utopian bankruptcy system can accommodate these mixed-cause shocks as well by focusing on the reasonableness of the debtor’s choices. Just as the first-best tax assesses an arbitrarily large amount against those taxpayers who do not invest sufficiently in their future, the utopian bankruptcy system denies relief for these mixed-cause shocks to those debtors who borrow an excessive amount or who otherwise take actions that unreasonably increase the probability of financial distress. Effectively, utopian bankruptcy denies relief to the negligent.105 Because debtors realize that they will not receive relief if they are negligent, they will only borrow an appropriate amount and take the proper care to avoid financial distress.
The utopian bankruptcy system, like the first-best tax or transfer system, provides individuals with no more relief than they need. In addition, just as the first-best public assistance program gives recipients a lump-sum transfer based on their need and the first-best tax system assesses a lump-sum amount based on taxpayers’ ability to pay,106 the utopian bankruptcy system leaves debtors with a lump-sum obligation based on their ability to pay. A debt obligation is a lump-sum when the amount debtors must pay does not vary with their actual income.107 Therefore, the utopian bankruptcy system merely adjusts debtors’ obligations to a new (lower) level consistent with their ability to pay. That is, in the utopian bankruptcy system the judge projects how much the debtor would earn if the debtor worked the [*PG24]efficient amount and decides how much that debtor should retain for consumption.108
For example, assume that an omniscient judge determines that if the debtor could have specified the amount in advance, he would have chosen to consume $3,000 per month in bankruptcy. If the omniscient judge determines that the debtor still has a relatively high earnings ability and thus could earn $7,000 per month if he worked as many hours as he should, then the judge would require the debtor to pay $4,000 per month. If the judge determines that the debtor has suffered a setback that leaves the debtor only able to earn $4,000 per month if he works as many hours as he should, then the judge would require the debtor to pay only $1,000 per month. The debtor’s new obligations equal the amount by which his projected income exceeds the amount that the judge determines the debtor should consume.109 Because the debtor’s new obligations are set with regard to his projected or potential income, the debtor retains the last dollar of his actual income and therefore he will work the efficient amount that the omniscient judge predicted.110
The optimal tax literature generally focuses on the number of hours that a debtor works, but the analysis can readily be applied to other questions as well. Perhaps because the issue strikes so close to home, law professors ask how much a debtor should be required to repay if he chooses an occupation that offers more attractive non-pecuniary benefits over one that offers a higher salary.111 For example, assume that an individual can choose to work in private practice at a salary of approximately $200,000 a year or in academia at a salary of approximately $100,000 a year. According to the logic of optimal taxation, even though the debtor would earn double the monetary income by working in private practice, he should still pay the same amount that he would have paid had he worked in academia.112 If the amount that debtors must repay depends on their choice of occupa[*PG25]tion, bankruptcy may distort their choices and debtors may choose to work as professors even if they would have preferred the added compensation of private practice to the non-pecuniary benefits of academia.
To say that a fixed repayment does not distort a debtor’s choice does not mean that it will not affect the debtor’s choice. A law professor, who under normal circumstances would find the compensation more than adequate, may be forced to work in the private sector if required to pay too much to her creditors. Because of this, Professor Gross suggests that the repayment amount should be set with regard to the debtor’s occupation at the time of filing so as to preserve the debtor’s freedom of choice.113 A large required repayment, however, only forces the professor into private practice by affecting her wealth, or lack thereof. Clearly one of the benefits of wealth is that it affords an individual greater choices, and a lower repayment increases the debtor’s wealth. This is also true of any other transfer of wealth, however, and one still needs a method for determining whether the debtor or her creditors are entitled to this wealth.
In a utopian bankruptcy system, the judge would allow the debtor to keep as much wealth as the debtor would have included in a hypothetical contract chosen at the time of borrowing from behind the veil of ignorance as to the debtor’s actual future condition.114 Note that this does not necessarily mean that debtors must always earn as much as they possibly can. It is easier to understand this point if one first focuses on the question of how many hours debtors should work. Debtors would not agree to a contract that could force them to work twenty-four hours a day, seven days a week; all debtors would gladly pay a slightly higher interest rate in exchange for some leisure. Likewise, at least some debtors would contract in advance for the right to work in a position that does not yield the highest pecuniary earnings available because they would find the more remunerative alternative occupation intolerable. Whereas in practice it may be extremely difficult to determine what position the debtor would have chosen, in a utopian world an omniscient judge could accomplish this task. Therefore, the omniscient judge would base the debtor’s required repayment on the occupation the debtor would have agreed to choose in the event of default. The judge, however, will require the [*PG26]debtor to repay the same amount regardless of the occupation that the debtor actually chooses.
If a judge were truly omniscient and could predict the future, the debtor would never experience an unexpected shock. One might, however, instead adopt a weaker definition of an all-knowing judge and assume that although he can observe the debtor’s innate earnings ability, he cannot necessarily predict whether the debtor will experience the same unfortunate shocks that justify bankruptcy: unemployment, illness, divorce, etc.115 If the judge cannot predict future shocks, a bankruptcy system that merely provides the debtor with a new, albeit reduced, debt obligation and no prospect of further relief cannot be optimal. Although such a system may provide the debtor with the appropriate incentives to work hard after filing for bankruptcy, it provides no insurance against future misfortune. Therefore, the utopian bankruptcy system must allow for repetitive bankruptcy filings or modifications of the debts created by the original bankruptcy filing.
The prospect of further distress replicates the justification for bankruptcy but also replicates the grasshopper problem; this further distress may arise either from misfortune or from misbehavior. An omniscient judge can apply the same solutions that are used to solve the grasshopper problem created by the existence of bankruptcy relief.116 Just as before, if the debtor’s future losses are always caused entirely by misbehavior or entirely by misfortune, an omniscient judge would simply deny any further relief in the former case and always grant full relief in the latter.117 In the more realistic scenario in which losses are often caused by both misfortune and misbehavior, the omniscient judge would again focus on whether the debtor took [*PG27]sufficient care to avoid future losses and deny relief to those that did not.
In the utopian bankruptcy system, an omniscient judge solves the grasshopper and opossum problems by identifying and punishing those who misbehave. In the real world, judges are not omniscient. Because of this, and because some individuals may be unable to control their behavior, society may not want to punish all of the grasshoppers. Society, however, does need some method of catching the opossums. How the judge should search for the opossums among us depends on the information that she has at her disposal. A pessimist would assume that judges cannot observe any relevant facts about the debtor except his actual income and that he is indebted.118 Under this view the second-best bankruptcy system looks very much like the second-best tax; it is effectively an income tax. The guarded optimist would assume that although judges are not omniscient, they can sometimes provide a workable estimate of a debtor’s potential income, for example, by using the debtor’s past income, level of education, or some other observable predictor. Under this view, the second-best bankruptcy system invokes some of the features of the utopian bankruptcy system, but with some adjustments for the limitations of judicial ability.
The above description of the second-best public assistance program assumed that the government could not determine if an indi[*PG28]vidual took the appropriate steps to avoid her current state of need.119 This assumption probably understates the government’s ability. For example, one might believe that the government can identify those individuals who left school early without a good justification or those workers that were fired due to shoddy attendance at work. Nevertheless, public assistance programs generally will not deny relief on these grounds or similar findings that an individual’s distress resulted from her own negligent behavior.120 Perhaps this refusal to deny relief to the negligent is an unfortunate result of a “Samaritan’s dilemma;”121 altruists cannot credibly commit to denying relief to those that misbehave because the altruists would be unable to live with the resulting suffering.
Private insurance contracts, however, generally take the same approach as public insurance programs.122 Private insurance contracts, like public assistance or bankruptcy, create a grasshopper problem because the insured is insulated from the full consequences of her actions that increase the probability or extent of loss.123 A property owner protected by fire insurance may be less inclined to buy a sprinkler system. A law student whose laptop is insured through a homeowner’s policy may take fewer precautions against theft. A motorist fully insured against liability may drive more carelessly.
If insurers can observe undesirable behavior before any harm results, they can adjust their premiums accordingly. If the insurer can do so perfectly, the insured will once again bear the full costs of their behavior and they will choose the efficient level of care;124 there will be no grasshopper problem. For example, motorists have a strong incentive to drive at a safe speed because their insurance rates will increase if they receive a speeding ticket. As a practical matter, however, the insurer cannot continuously monitor all of the insured’s behavior, and thus this approach cannot completely solve the grasshopper problem.
[*PG29] Alternatively, insurers could just exclude coverage for a loss if they discover that the insured failed to take sufficient care and was therefore negligent. As long as the insurer sets the standard of care correctly, the insured would then always maintain the proper level of care for fear that if he took less care he would receive no insurance at all. This is the same principle as the tort law’s use of a negligence theory of liability.125 Thus, there would be no reason for the debtor to buy insurance against liability based solely on negligence as it would always be cheaper to take the efficient level of care than to buy insurance against the loss.126
At least from the beginning of the twentieth century, however, individuals have been able to purchase insurance against the consequences of their own negligence,127 and today such insurance is common. Property insurance protects the insured against accidental loss.128 Fire insurance will pay a claim even if the insured negligently started the fire.129 In addition, individuals buy insurance against liability based on negligence principles. Individuals buy homeowner’s insurance policies that include protections against liability that generally requires a finding of negligence.130 Today, many states effectively require motorists to buy insurance against the liability that results from their own negligence,131 and many motorists voluntarily buy insurance that far exceeds the minimum amount required by statute.
Scholars advance two primary arguments for why the market offers insurance against negligent behavior; both are based on human imperfections. First, an individual might buy insurance based on the belief that the risk of liability cannot be avoided by taking the proper level of care. A system that denies relief due to negligent behavior re[*PG30]lies on judges and juries to make difficult decisions about whether the individual took the proper level of care, and the trier of fact may sometimes incorrectly deny relief.132 Second, and more controversially, individuals might buy insurance based on the belief that they can’t always maintain the proper level of care.133 To understand the distinction between these two explanations, consider why drivers buy more than the minimum required liability insurance. The demand for insurance stems at least in part from the fear that a jury would wrongly find that the driver negligently caused an accident. The demand, however, may also stem from the fear that the driver may become momentarily distracted by his child and fail to see a red light.
The insurer who issues a policy that protects the insured against his own negligence must charge higher premiums to account for the increased chance that the insured will suffer a loss and file a claim. In order to mitigate this grasshopper problem, and thus pay lower premiums, the insured agrees to face some of the consequences of his negligence; the insured only buys partial insurance.134 For example, private insurance contracts often require the insured to pay a deductible to cover part of any loss.135 In addition, the insurance policy may not cover all of the harms that are likely to result from the insured’s negligence. Some forms of automobile liability insurance decline to cover the damage to the negligent motorist’s own vehicle. Health insurance will not pay the insured for the pain of the insured’s illness. Finally, the fact that the insured will need insurance in the future also gives the insured an incentive to take care. For example, a motorist found to have negligently caused an accident will pay higher insurance premiums in the future. If the insured has too many accidents, the insured may be unable to buy insurance altogether.
If one views bankruptcy as just another form of insurance, then one might reject the utopian goal of denying relief to negligent debtors. Although bankruptcy judges may be wise, they are clearly not omniscient and therefore will make errors. For example, to the extent that many bankrupt debtors borrowed too much to allow them to withstand a financial shock such as unemployment or divorce,136 a bankruptcy judge would be forced to rule on whether a debtor’s borrowing choices were reasonable in light of the debtor’s financial pros[*PG31]pects. A judge viewing the debtor’s decision with the benefit of hindsight may incorrectly rule against the debtor.
In some cases, however, debtors will have clearly spent their way into bankruptcy, leaving little room for judicial error. Nevertheless, the second-best bankruptcy system may grant relief to the spendthrift if society believes that some debtors are just unable to resist the temptation of easy credit. Certainly there is support for this theory in the literature. Professor Jackson argues that the need for bankruptcy is largely based on cognitive and volitional failures among debtors that prevent them from making the right credit decisions.137 Others argue that society should encourage creditors to restrict access to credit, presumably because debtors are unable to determine when borrowing is in their self-interest.138
This does not mean that the second-best bankruptcy system ignores the grasshopper problem. But because this system cannot distinguish between the grasshoppers and the truly unfortunate, society must make bankruptcy less attractive so that individuals have some incentive to avoid financial distress.139 The repeat nature of credit markets, like the repeat nature of insurance markets, will provide some help. Just as the insured takes care to avoid the increased premiums that follow an accident, the debtor will take care to avoid the increased interest rates and loss of credit that follow bankruptcy.
The threat of reduced access to credit, however, is only a partial solution, and just as insurance contracts require the payment of deductibles and insure only part of the debtor’s loss, the second-best bankruptcy must also offer only partial insurance. That is, the second-best bankruptcy system must be less generous than the utopian bankruptcy system so that debtors will face some of the consequences of actions that make financial distress more likely. Exactly how much relief the second-best bankruptcy system should provide is a matter of compromise. In the private insurance context, society does not need to worry about the correct level of partial insurance so long as the insurance contract has no negative effects on third parties;140 the insured and the insurer will agree to the most mutually beneficial terms. [*PG32]Unfortunately, the optimal bankruptcy system, like the optimal tax and public assistance programs, is only part of a hypothetical contract, and therefore the correct level of partial insurance is debatable.141
So far, this Article has treated the grasshopper problem as a function of negligence. Financial distress, however, results from far more culpable conduct as well. At the other extreme, the debtor may face a judgment for a willful tort such as sexual assault. Here the analogy to private insurance again provides some guidance.
While private insurance contracts protect debtors from the consequences of their own negligence, these contracts do not protect debtors from the consequences of their own willful misconduct. Private insurance contracts almost always contain some clause excluding liability for intentional acts.142 Even in the absence of such a clause, courts will often find an implied exception for intentional conduct,143 and a contract that purported to insure an individual against willful misconduct would likely be held void as against public policy.144 Therefore, if a homeowner intentionally sets fire to his house, fire insurance will not reimburse him.145 If a motorist intentionally hits a pedestrian, automobile insurance would not pay for the damages that would result146 and would certainly not prevent the driver from serving time in prison.
Public insurance programs often take this same approach. For example, unemployment insurance will deny benefits if the unemployed has engaged in willful misconduct such as a work-related fel[*PG33]ony or misdemeanor.147 Therefore, one might conclude that bankruptcy should deny relief for willful misconduct as well. This assumption, however, is discussed more fully below.148
The utopian bankruptcy system assumes that the judge can perfectly estimate a debtor’s earnings ability or potential earnings. To the extent that one rejects this assumption, one rejects the feasibility of a bankruptcy system based on debt-adjustment just as economists generally reject the feasibility of a progressive lump-sum tax.149 If a judge cannot distinguish the high-ability debtors from the low-ability debtors, a debt-adjustment system must leave all debtors with the same amount of debt or else those with a high-ability will just “play ‘possum” and claim that they cannot repay anything.150 Such a scheme would fail to meet the insurance goals assumed to underlie bankruptcy. Therefore, one must search for a feasible second-best approach. What one considers feasible depends on one’s assumptions about the actual ability of judges to identify the opossums among us.
Even though pessimists would reject the assumption that judges can observe a debtor’s potential earnings, they would likely concede that the bankruptcy judge can observe the debtor’s actual earnings after the bankruptcy filing. This is essentially the same assumption that underlies the analysis of the second-best tax,151 and a similar result follows. According to the pessimist, the second-best bankruptcy [*PG34]system would increase the debtor’s required repayment as the debtor’s actual income increased; the second-best bankruptcy system would effectively include an income tax.152 This system allows some limited relief because the high-ability debtor would need to earn, and therefore consume, very little in order to “play ‘possum” by emulating the low-ability individual. The precise characteristics of this effective income tax, like the precise characteristics of the second-best income tax, balance the individual’s need for insurance against the unfortunate consequences that income taxes have on an individual’s desire to work.153
A guarded optimist would believe that the pessimist understates the ability of judges to estimate a debtor’s potential income. Some debtors will work in occupations that yield a fairly regular or stable stream of income over a number of years, and the income history of these debtors, along with other observable characteristics such as level of education, may provide a fairly good basis for estimating their potential earnings. The judge, however, can never perfectly estimate future or potential earnings; the size of a debtor’s bonus or the number of overtime hours available to him may change from year to year. More seriously, the debtor may lose his job or may earn a promotion or a substantial raise.
Because the guarded optimist assumes a level of information greater than that underlying the pessimist’s view of the second-best bankruptcy system and yet less than that needed to implement the utopian bankruptcy system, the guarded optimist would choose a bankruptcy system that blends the two approaches. Recall that the utopian bankruptcy system bases debtors’ required repayment solely on their potential income and that the pessimist’s second-best bankruptcy bases debtors’ required repayment solely on their actual in[*PG35]come.154 An intermediate approach would base the debtor’s required repayment on the judge’s estimate of the debtor’s potential income, but then allow for further modifications if the debtor’s actual income deviated sharply from this estimate. If judges never modify the initial obligation, this is just the utopian bankruptcy system as the required repayment is based solely on the debtor’s projected earnings.155 If judges always modify the obligation when the debtor’s actual income deviated from the estimate, this is just the pessimist’s bankruptcy system as the required repayment is based solely on the debtor’s actual earnings.156
Recall that if judges are not truly omniscient in that they cannot perfectly predict the future, the utopian bankruptcy system includes modifications or subsequent bankruptcies to account for unexpected shocks.157 The utopian system would always adjust debtors’ required repayment when the change in their circumstances was entirely due to fate and never adjust debtors’ required repayment when the debtors themselves caused the change or when the debtors failed to work hard enough to avoid the ensuing financial difficulties.158 That is, the utopian bankruptcy system adopts the same negligence standard to solve the grasshopper problem in the context of modifications that it adopts to solve the grasshopper problem in general.159
The guarded optimist, however, does not advocate the strict use of a negligence standard in bankruptcy and would not advocate for its use in the modification context either. Just as judges may be unable to determine whether debtors’ negligent misbehavior resulted in their initial financial distress, they may be unable to determine if the debtors’ misbehavior led to the need for further relief.
Recognizing the limited ability of a judge to determine if the change in the debtor’s circumstances resulted from fate or the debtor’s own efforts, or lack thereof, the guarded optimist might advocate a system that modifies a debtor’s required repayment only after a substantial change in the debtor’s circumstances. Many small changes in a debtor’s income, such as the income resulting from a few extra hours of overtime or a slightly larger bonus, are probably the direct result of the debtor’s own efforts. Bankruptcy should not mod[*PG36]ify a debtor’s required repayment in response to these changes so that the debtor retains the incentive to work hard.160 Of course, some small changes in financial condition, such as costs associated with a broken small appliance, are the result of chance and therefore, ideally, should be insured. Nevertheless, by definition these changes are small and therefore the debtor has little need for the insurance.
By contrast, larger changes in a debtor’s financial condition may involve a greater element of chance. Workers may to a large extent earn promotions, raises, or large bonuses. Their success, however, also depends on the efforts of their co-workers and the overall success of their firm. Likewise, workers who shirk are more likely to be fired. Nevertheless, macroeconomic forces clearly play a role as is evidenced by the cyclical nature of unemployment.161
The claim that larger changes in the debtor’s financial condition are more likely to result from chance than are small changes is an empirical assertion and no proof is offered. But even if this assertion is incorrect, one might still prefer a system that only modifies the debtor’s obligations following a large change in the debtor’s circumstances. These large changes will occur only infrequently and involve large sums of money. Therefore, it will be cost effective for a judge to at least conduct an inquiry as to whether the change was caused by fate or the debtor’s own efforts and perhaps judges are correct just often enough to make such an inquiry worthwhile.162
Part II described the optimal bankruptcy procedure in the abstract. This Article seeks, however, to offer guidance to real judges who must interpret an existing Bankruptcy Code. Using the analysis of Part II as a guide, this Part argues that courts should interpret the existing Bankruptcy Code in a way that accounts for the moral hazards created by the bankruptcy system. Most courts have already shown a great willingness to improvise when interpreting an imperfect code, and a few have even been willing to ignore the plain meaning of the text of the code.163 In many cases, however, the language of the code is either ambiguous or expressly invites bankruptcy judges to use their discretion to account for debtor misbehavior. In such circumstances, courts could address many of the opossum and grasshopper problems identified in this Article while still remaining fairly faithful to the text of the code.
Since this Article discusses how courts should interpret specific provisions of the Bankruptcy Code, Part III-A provides a brief introduction to collections law that will place these provisions in the proper context. Part III-B discusses how courts should address the grasshopper problem, and Part III-C discusses how they should address the opossum problem.
A complex web of state and federal laws regulate collections in modern America. Most consumers who fail to repay their debts probably do not file for bankruptcy.164 Instead, they rely on state and federal non-bankruptcy laws to protect them from their creditors.165 For example, federal law limits the manner in which a collections [*PG38]agency may contact a debtor166 and limits the amount of a debtor’s wages that creditors can garnish to no more than 25% of the debtor’s after-tax income.167 Some state laws restrict garnishment even further or prohibit it altogether.168 Both state and federal laws also limit the amount and type of physical property that the sheriff can seize to satisfy a creditor’s claim.169 Though property exemption laws vary considerably from state to state,170 they usually offer at least some protection to debtors for certain types of assets such as their homes,171 their cars,172 and their tools.173 Any theory of bankruptcy must consider the existence of these non-bankruptcy laws.
While non-bankruptcy collections law remains extremely important, scholars now focus more heavily on bankruptcy.174 As a practical matter, there are two forms of consumer bankruptcy: a Chapter 7 liquidation and a Chapter 13 adjustment of debts.175 A substantial majority of bankrupt Americans choose a liquidation under Chapter 7.176 Chapter 7 provides debtors with a full discharge of most unsecured debts,177 thereby absolving debtors of the need to repay out of their [*PG39]future income. In theory, the Chapter 7 debtor receives this discharge in exchange for a damaged credit reputation and the loss of any assets not protected either by state property exemptions or certain bankruptcy specific exemptions available in a few states.178 In reality, manipulable179 and generous180 property exemptions ensure that only a tiny fraction of Chapter 7 debtors forfeit any assets to their general unsecured creditors.181
Anyone can file for relief under Chapter 7; one need not even be insolvent. Nevertheless, Chapter 7 does not offer effective relief to some debtors either because they would lose significant assets182 or [*PG40]because they have significant debts that they could not discharge in Chapter 7.183 Moreover, there is some evidence that bankruptcy judges and lawyers in some jurisdictions steer debtors away from Chapter 7.184 Finally, the code allows judges to dismiss Chapter 7 filings that are a substantial abuse of the Bankruptcy Code.185 Congress added this provision in 1984 following extensive lobbying by creditors186 who were concerned with what they perceived to be a large increase in the number of bankruptcy filings187 and who wanted to force more debtors to either choose Chapter 13 or avoid bankruptcy altogether. This provision, however, does not appear to have had its desired effect as the number of bankruptcy filings has sharply increased,188 the ratio of Chapter 7 to Chapter 13 filings has remained fairly stable,189 and the rate at which bankrupt debtors repay unsecured creditors remains close to zero.190 As a consequence, the con[*PG41]sumer credit industry is again pushing for reforms, commonly called “means testing,” that would force more debtors into Chapter 13.191 Whether these reforms would have any effect remains an open question.192
Unlike Chapter 7, Chapter 13 is not open to all debtors. Chapter 13’s predecessor, Chapter XIII of the Bankruptcy Act of 1898,193 was restricted to “wage earners.”194 In 1978, Congress broadened eligibility for Chapter 13 to all individuals with regular income and debts below certain ceilings.195 As discussed below, the meaning of the term “regular income” has important implications for the feasibility of a reorganization-based approach to bankruptcy.196
Debtors filing under Chapter 13 propose plans that require them to make a series of payments for a period of up to five years.197 The judge will approve the plan if it is proposed in good faith, is feasible, and meets certain repayment requirements.198 If debtors make their payments, the judge grants them a discharge of their remaining debts, with relatively few exceptions.199 Chapter 13 offers debtors a superdischarge that is broader than the discharge available in Chapter 7.200
[*PG42] Chapter 13’s predecessor, Chapter XIII, required debtors to obtain the consent of each of their secured creditors and the majority of their unsecured creditors before judges would approve their plan.201 As a consequence, virtually all Chapter XIII plans offered full repayment to all creditors, though often without interest.202 Today, Chapter 13 plans that propose full repayment, even without interest, appear to be more the exception than the norm,203 and a series of code provisions determine the amount that a debtor must repay.204
Court fees205 and priority claims, including certain tax and family law claims,206 must be paid in full. The debtor will also lose any collateral held by a secured creditor unless the plan provides for the full repayment of the secured claim, though perhaps without interest, or the secured creditor agrees to accept less.207 Because, however, the [*PG43]amount that the debtor must pay the secured creditor depends on the value of the debtor’s physical assets rather than the value of her earnings, this Article focuses on how much the debtor must pay her unsecured creditors.
There are two primary tests for how much debtors must pay their general unsecured creditors in Chapter 13: the “best interests of the creditors” test of � 1325(a)(4) and the “disposable income” test of � 1325(b). An unsecured creditor may object under � 1325(a)(4) if it does not receive as much as it would have received in a Chapter 7 liquidation.208 Though there is some evidence that debtors with significant non-exempt assets are more likely to choose Chapter 13,209 the near complete lack of repayment for unsecured creditors in Chapter 7210 implies that this “best interests of the creditors” test may mean little in practice.
Between 1978 and 1984, the only explicit test for determining the debtor’s payment to unsecured creditors was the “best interests of the creditors” test of � 1325(a)(4).211 Because so few debtors have any non-exempt assets,212 and because creditor consent was no longer required, judges accustomed to Chapter XIII plans that proposed full repayment213 began to see Chapter 13 plans that proposed little or no repayment for unsecured creditors. It seems that this change was too much for some judges, and many refused to confirm plans that paid too little to creditors on the grounds that they were not proposed in good faith as required by the Bankruptcy Code.214 While this use of the good faith requirement was controversial at the time, it became even more so after the Bankruptcy Amendments and Federal Judgeship Act of 1984.215
[*PG44] As part of the Bankruptcy Amendments and Federal Judgeship Act of 1984,216 Congress added the “disposable income” test of � 1325(b). Section 1325(b) states that a court may not confirm a plan over the bankruptcy trustee’s or a creditor’s objection unless the debtor proposes full repayment or the debtor proposes to pay into the plan all of his projected disposable income for a period of three years.217 This “disposable income” test has generated a substantial amount of controversy,218 and despite its central role in determining the debtor’s required repayment and frequent litigation about its meaning, substantial questions remain.219
The confirmation of a plan does not end the Chapter 13 process. A debtor who subsequently finds his plan too onerous has several options. Within bankruptcy, the debtor can apply for a hardship discharge220 or convert the case to Chapter 7.221 Provided the case had not been previously converted from another chapter, the debtor may dismiss the plan222 and either re-file or rely on federal and state non-[*PG45]bankruptcy laws for protection.223 Finally, � 1329 allows the debtor, the Chapter 13 trustee, and the unsecured creditors to seek a modification of the plan prior to the completion of payments.224 While parties may seek modifications for many reasons,225 this Article focuses on requests for a modification of the required repayment in response to a change in the debtor’s financial condition.
Part II implies that in managing the grasshopper problem, courts should treat negligent and willful misbehavior separately. To a large extent, the existing bankruptcy system does just that. Bankruptcy grants relief to debtors who spend too much or engage in other negligent behavior, but Chapter 7 of the Bankruptcy Code denies relief for willful misbehavior. By its terms Chapter 13, however, will grant a discharge for many forms of willful misbehavior. Courts have struggled to reconcile Chapter 13’s expanded discharge with sound policy goals.
Few, if any, commentators argue that courts should deny bankruptcy relief to the spendthrifts or the negligent, and proposals to partially eliminate Chapter 13’s expanded discharge have generated little public comment.226 This silence is explained in part by bankruptcy’s similarities to, and differences from, standard insurance.
The utopian bankruptcy system denies relief to the debtor who negligently borrows too much or fails to take sufficient precautions to avoid financial distress; essentially, utopian bankruptcy denies relief to the negligent.227 By contrast, the existing bankruptcy system does not [*PG46]distinguish between the negligent and the unfortunate. Some debtors are in bankruptcy because they could not control their spending habits, others because they suffered some unavoidable expense, and still others because of a combination of misfortune and misbehavior.228 They all receive the same discharge, however, and bankruptcy law makes little or no effort to distinguish between them.
Chapter 7 does allow the bankruptcy judge to dismiss a filing if granting relief would be a substantial abuse of the Bankruptcy Code,229 and in looking for abuse courts do inquire into the totality of the circumstances.230 Courts focus, however, on whether the debtor can in fact pay his debts rather than on whether the debtor should have acted more responsibly to avoid incurring such debt in the first place.231 That is, courts focus on whether the debtor needs bankruptcy relief rather than on why the debtor needs bankruptcy relief. Therefore, the substantial abuse provision appears designed primarily to deal with the opossum problem rather than the grasshopper problem.232
[*PG47] Although numerous scholars have written about the grasshopper problem of bankruptcy,233 there are no serious proposals for denying a bankruptcy discharge to those who overspend or otherwise negligently cause their own financial distress. Even the credit industry has not tried to limit the grasshopper’s access to bankruptcy,234 preferring instead to lobby for means testing that would identify those debtors who can pay more. As a matter of theory, means testing requires only that the debtor prove that he needs relief; means testing does not ask the debtor to explain why he needs relief. Therefore, like the substantial abuse provision, means testing appears aimed primarily at the opossum problem rather than the grasshopper problem.
In light of the analysis of Part II, this approach is unsurprising. The first lasting bankruptcy law in the United States,235 the Bankruptcy Act of 1898236 was passed at about the same time that courts began to allow individuals to insure against liability for negligent misconduct.237 Because bankruptcy is another form of insurance, the justifications for insurance that protects the negligent would seem to apply to bankruptcy as well. Many, if not most, bankrupt debtors are unable to pay their debts both because they suffered some unfortunate shock such as divorce or unemployment and because they borrowed too much to allow them to withstand such a shock.238 Society may not trust the ability of bankruptcy judges to determine if these debtors behaved prudently and therefore deserve relief. Moreover, even when it is clear that debtors spent their way into bankruptcy, society may not want to deny relief because many believe that some debtors just cannot resist the temptation of easy credit.239
[*PG48] But the second-best bankruptcy system does not ignore the grasshopper problem altogether. Precisely because courts cannot or will not identify the negligent, the second-best bankruptcy system offers only partial relief to the unfortunate as well.240 Debtors receive only partial insurance to the extent that they are worse off after bankruptcy than they would have been had they behaved in the proper manner. In practice, the bankrupt debtor will emerge with severely damaged credit, and, possibly, a damaged social reputation. On the other hand, the debtor will have enjoyed, at least temporarily, a higher standard of living than if he had behaved prudently. In addition, the debtor may still retain some of the assets purchased on credit. Therefore, bankruptcy should ask the debtor to repay something to his creditors.
Though debtors in Chapter 7 pay virtually nothing to their unsecured creditors,241 they often reaffirm debts to secured creditors so that they may retain their home, their car, or other assets pledged as collateral.242 Whether this system causes the debtor to suffer enough is open to debate, and if means testing forced debtors to pay more in bankruptcy, it could reduce the level of insurance they receive. Whether this means testing is warranted remains controversial as commentators strongly disagree as to the importance of the grasshopper problem in bankruptcy.243
Because neither private contracts nor public insurance programs relieve individuals of the consequences of their willful misbehavior,244 Part II suggests that bankruptcy should not do so either. Traditionally, bankruptcy adopted this approach. The Bankruptcy Act of 1898 [*PG49]barred the discharge of debts arising from most forms of willful misconduct whether under a Chapter VII liquidation245 or a Chapter XIII wage earner’s plan.246 Today, Chapter 7 largely retains and expands the exceptions to discharge of the Bankruptcy Reform Act of 1898.247
By contrast, Chapter 13 provides a superdischarge that is broader than the discharge available in Chapter 7 and can forgive, among other things: i) loans procured by fraud or false pretenses, ii) debts for fraud, embezzlement, or larceny, and iii) debts for willful and malicious injury.248 Because of the superdischarge, and because tort claims often do not count toward the dollar limitations of Chapter 13,249 the reported Chapter 13 case law is populated in part by debtors facing judgments or claims for fraud,250 embezzlement,251 misappro[*PG50]priation,252 aggravated assault,253 intentional shooting,254 sexual assault,255 sexual abuse of a minor,256 and other willful misconduct.
Congress left little record to indicate why it adopted the superdischarge. The commission appointed to study bankruptcy recommended that Congress retain the existing exceptions to discharge in both Chapter 7 and Chapter 13, and the legislative history reveals no justification for why this recommendation was rejected.257 The superdischarge has changed little since 1978,258 however, and a recent study by the National Bankruptcy Review Commission did not suggest a substantive change to this policy.259 This continued support for the superdischarge must be reconciled with the law’s general hostility toward insurance that protects an individual from the consequences of his willful misconduct.260
The traditional explanation for the superdischarge is that it induces debtors to file under Chapter 13 and thereby increases the repayment for all creditors.261 Given that in 1978, Chapter 13 did not require the debtor to repay any more than he would have in Chapter 7,262 this explanation for the original superdischarge is questionable. It is at least possible, however, that Congress thought that debtors would pay more,263 and some may support the continued existence of [*PG51]the superdischarge if the disposable income requirement now results in a greater repayment under Chapter 13.264
Even if the superdischarge maximizes the repayment to the creditor, however, the same could be said of an insurance policy that covers willful torts; the victim could demand payment from a solvent insurance company rather than an insolvent tortfeasor. While bankruptcy effectively mandates that the creditor insure the debtor against the debtor’s willful misconduct, non-bankruptcy insurance law generally prohibits insurance policies that cover willful misconduct.265 Courts refuse to allow such insurance policies on public policy grounds, reasoning that these insurance contracts will lessen the deterrence that judgments for willful torts are designed to deliver.266 Because the superdischarge also lessens the deterrent effect of the tort law, one needs a justification for why bankruptcy requires this insurance.
Recall that there are two justifications for insurance policies that cover negligence. First, judges and juries will sometimes err when deciding whether a defendant was negligent and therefore society should allow the individual to insure against this error.267 For example, a court could wrongfully find that a motorist negligently drove at an excessive speed and caused an accident. This same argument applies to intentional misconduct, though perhaps with lesser force. For example, a court may wrongfully find that a motorist intentionally struck a pedestrian when in fact he did so by accident. There is no reason to believe, however, that Congress changed its view of the competency of the judiciary in the late 1970s.
By contrast, the period surrounding the passage of the Bankruptcy Reform Act of 1978 did see a change in the law’s basic assump[*PG52]tions about the ability of individuals to control their own actions, the second explanation for insurance that covers negligent behavior.268 To the extent that individuals are simply unable to control their own actions, the deterrence value of judgments becomes much less important relative to the other goals that these judgments serve, such as compensating the victim.
This shift in view of individual responsibility can be most clearly seen in the criminal law. For example, the insanity defense traditionally focused on whether the defendant understood the nature and quality of his actions or the wrongfulness of his actions.269 In 1962, however, the American Law Institute wrote the Model Penal Code, which expanded the insanity defense to include defendants who could appreciate the wrongfulness of their actions but who were unable to control their actions and to conform their conduct to the requirements of law.270 The D.C. Circuit adopted this approach in 1972,271 and during the 1960s and 1970s a significant minority of state legislatures and courts, and almost all of the federal circuit courts,272 followed suit. Similarly, this period witnessed an expansion of the doctrine of diminished capacity that reduces a homicide from murder to manslaughter.273 This culminated in the now infamous “Twinkie defense” in which an individual shot and killed both the mayor of San Francisco and a supervisor but was found guilty of only manslaughter after arguing that he was suffering from depression as evidenced by an excessive consumption of Twinkies.274 The “Twinkie defense” and John Hinckley’s acquittal of the attempted murder of President Reagan helped spark a backlash against this trend.275 This backlash came too late, [*PG53]however: the Bankruptcy Reform Act of 1978 was passed in the same year as the homicide in the Twinkie case.276
Regardless of whether these changes are related to the creation of the superdischarge, other forms of insurance continue to refuse to shield individuals from the consequences of their willful misbehavior,277 and therefore the superdischarge remains an anomaly. Courts have reacted to this anomaly by turning to another section of the code, the requirement that a plan be proposed in good faith,278 to limit the availability of the superdischarge.
A minority of courts have refused to use this good faith test to question the debtor’s pre-petition conduct,279 either because they believe that Congress intended a narrow definition of good faith,280 or that the debtor is merely making use of a statutory right.281 A significant majority of courts, however, read good faith more broadly and will consider the nature of the debt sought to be discharged.282 [*PG54]Generally, the courts that use the good faith standard to police pre-petition willful misconduct do not flatly deny access to the superdischarge. Rather, consistent with the public policy goal of retaining adequate deterrence, these courts sometimes will grant a discharge of debts arising from willful misconduct, but only after the debtor has shown that he has received significant punishment for his actions or is sufficiently remorseful.283 Perhaps this inquiry into the debtor’s particular circumstances provides an explanation for why the superdischarge protects bankrupt debtors from the consequences of their willful misconduct while other insurance policies may not. While tort liability plays an important role in deterring willful misconduct, society also relies on other tools for deterrence, such as criminal liability. Because bankruptcy is a judicial process, a judge can review the debtor’s circumstances and determine if the debtor has received enough punishment from other sources to satisfy the need for deterrence.
In conducting the good faith analysis, however, courts do not just focus on the punishment the debtor has received; they consider the nature of the victim and the financial impact that the plan would have on the victim.284 Consequently, debtors guilty of severe pre-petition misconduct may be forced to make a greater effort at repayment than would other debtors.285
[*PG55] Because this Article focuses on the standard consumer lending transaction, it largely ignores the interests of the creditor. In the standard consumer lending transaction, this approach is justified because, in the long run, the debtor largely bears the cost of bankruptcy in the form of higher interest rates and reduced access to credit.286 This analysis, however, obviously does not apply to the tort victim. To the extent that bankruptcy law discharges judgments for torts, it truly shifts the loss from tortfeasors to tort victims. In addition, tort victims are poorly situated to bear this risk as they cannot easily diversify it away like the consumer lender can. In short, the tort victim is a very poor insurer of the tortfeasor and a rational bankruptcy policy must balance the interests of both creditor and debtor.287
Even if society decides that an individual deserves some relief, it still must determine how much relief is appropriate. Ideally, bankruptcy would only grant the debtor as much relief as he needs and require the debtor to repay an amount commensurate with his ability to pay. To the extent that one’s ability to pay is related to his earning ability, Chapter 7 fails to accomplish this task. While many debtors repay some amount in Chapter 7, this amount is not based on income. Rather, a tiny minority of debtors repay unsecured creditors out of their non-exempt assets, and a larger number of debtors reaffirm secured loans.288 Chapter 7 may still play a role in an optimal bankruptcy system if restricted to debtors whose incomes are so low that a court would have them repay nothing in bankruptcy.289 Perhaps [*PG56]this is why much of the current policy debate over means testing focuses on whether a debtor can repay his debts.290 This section focuses, however, on those debtors who should repay something in bankruptcy and asks how bankruptcy should collect this amount.
Recall from Part II that the utopian bankruptcy system leaves the debtor with some debts that he must repay, but reduces these debts to an amount commensurate with his potential earnings.291 Recall too that the pessimist rejects the underlying assumption of utopian bankruptcy—that judges can estimate a debtor’s potential earnings—and therefore insists that society should adopt a second-best bankruptcy system that bases the debtor’s required repayment on his actual earnings.292
This tension between the utopian and pessimistic views of bankruptcy is reflected in the cases arising under Chapter 13. The language of the Bankruptcy Code invokes the utopian bankruptcy system in that it bases the debtor’s required repayment on his estimated or projected earnings.293 Many courts, however, are skeptical of their own ability to estimate a debtor’s future income and therefore seek to create a system that bases the debtor’s repayment at least in part on his actual earnings. In order to do so, some courts are willing to significantly stray from the plain meaning of � 1325(b). This approach is unfortunate because these courts could better account for their own limitations through other provisions of the code.294 For instance, if judges are concerned that their estimates may, on occasion, significantly understate the debtor’s actual earnings, then they can order the debtor to report her income to the bankruptcy trustee and rely on the interested parties to seek a modification of the plan.295 Likewise, if judges believe that the debtor’s income is simply too uncertain to estimate, they should dismiss the debtor’s bankruptcy petition on the grounds that the debtor lacks a regular income and therefore is not eligible for relief under Chapter 13.296
Academics sometimes criticize Chapter 13 for discouraging debtors from working by depriving them of any earnings above some fixed amount.297 If Chapter 13 seized all of the debtors’ disposable income, or all of their actual income above some allowance for what the court determines are their reasonably necessary expenses, then this would be true. Such a system would effectively operate as a prohibitive tax on any income earned above the fixed amount, and the debtor would have no incentive to work any more than required to earn that fixed amount. To the extent that a Chapter 13 bankruptcy emulates the utopian bankruptcy system by leaving debtors with a fixed or lump-sum obligation based on their projected earnings, however, it does not discourage them from working at all.298
By its terms, the “disposable income” test of � 1325(b) does not base a debtor’s required repayments on her actual disposable income.299 Like the first-best tax or the utopian bankruptcy system, � 1325(b) bases the debtor’s required repayment on her estimated or projected disposable income; � 1325(b) requires the debtor to propose a plan in which she uses all of her projected disposable income to repay her debts.300 A plain reading of this section requires the judge to project the debtor’s income at the time of confirmation and set the debtor’s repayment obligations equal to the amount by which this projected income exceeds the debtor’s reasonably necessary expenses.301 A plain reading of this statute requires the judge to invoke the utopian bankruptcy system.
Some courts follow the plain meaning of this section. In 1994, in In re Anderson, the U.S. Court of Appeals for the Ninth Circuit overruled the district and bankruptcy courts’ finding that the debtor failed to meet the requirements of � 1325(b)(1)(b) because the debt-or refused to sign a best efforts certification that would in effect bind him to pay all of his actual disposable income to his creditors.302 Relying in part on the language of the statute,303 the court held that [*PG58]� 1325(b)(1)(b) requires only that debtors pledge payment of all of their projected, rather than actual, disposable income, and that therefore trustees may not require debtors to sign best efforts certifications.304
Although some jurisdictions have followed Anderson,305 a few courts have required debtors to promise a repayment contingent on their actual income.306 Unfortunately, the reasoning employed by these courts is not always clear. At least one court that required a debtor to sign a best efforts certification simply ignored the word “projected,”307 as did another court that required debtors to increase plan payments whenever they were able to secure additional overtime work or when they actually received benefits from a profit sharing plan.308 Another court required the debtor to pay an amount equal to his projected disposable income plus half of any amount by which his actual disposable income exceeded his projected disposable income.309 Although this court explained that � 1325(b)(1)(b) did not require a debtor to pay all of his actual disposable income,310 it did not explain why the debtor’s actual disposable income was relevant at all.311
[*PG59] Much more perplexing are the decisions within the Ninth Circuit that have, subsequent to Anderson, required debtors to include a disposable income clause in their plan.312 A disposable income clause appears to be functionally equivalent to a best efforts certification in that debtors promise to repay with all of their actual disposable income.313 In the 2001 decision In re James, the Bankruptcy Court for the District of Idaho explicitly acknowledged the binding precedent in Anderson, yet still based its denial of confirmation in part on the debtor’s failure to include a disposable income clause in his plan.314 Given the abundant evidence that the debtor’s plan was not proposed in good faith,315 one would like to assume that the court’s insistence on a disposable income clause was merely an isolated occurrence of harmless error. Unfortunately, this does not appear to be the case. The James court references other bankruptcy court decisions within the Ninth Circuit that ignore the binding precedent in the jurisdiction, including a decision by the Bankruptcy Appellate Panel for the Ninth Circuit that states, in dicta, that a disposable income clause is required whenever a trustee or the holder of an unsecured claim objects to the plan.316 Moreover, one case cited suggests that, at least in [*PG60]Idaho, debtors are routinely forced to include such clauses in their plans.317 One suspects that even if these plans are not required for confirmation, many debtors may simply agree to their inclusion to prevent the trustee from invoking a host of objections.318
Some of the courts that require disposable income clauses try to distinguish their facts from those of Anderson by stating that such a clause is only required when a trustee or a creditor objects to a plan.319 This distinction, however, clearly fails because the trustee objected in Anderson as well;320 the requirement of a trustee or creditor objection determines whether � 1325(b) applies at all, not whether projected or actual disposable income must be promised.321 Jackson, one of the decisions cited by James, recognizes the futility of trying to distinguish Anderson and claims that the law on this point is unclear.322 While this may be true generally, this is not true within the Ninth Circuit, where Anderson should serve as binding precedent for lower courts.323
What is consistent in each of the decisions that base debtors’ required repayment on their actual income is a pessimism about the judge’s ability to accurately estimate debtors’ earnings. This pessimism has caused judges to ignore the plain meaning of the statute and, in the case of the bankruptcy courts within the Ninth Circuit, the binding precedent in their jurisdiction as well. The judges in these cases almost invariably express frustration at their inability to accurately forecast debtors’ earnings. For example, the court may believe that the debtor may earn future promotions or raises or may be able [*PG61]to work additional overtime.324 The clearest expression of this pessimism can be found, however, in the context of the recently extended Chapter 12 of the Bankruptcy Code, “Adjustment of Debts of a Family Farmer With Regular Annual Income.”325 Chapter 12 contained a projected disposable income test that, for the purposes of this Article, is worded identically to � 1325(b).326 In 1994, in Rowley v. Yarnall, the U.S. Court of Appeals for the Eighth Circuit noted that the plain meaning of the statute required it to base the debtor’s repayment only on the debtor’s projected income, yet chose to ignore the word “projected” in the statute.327 It did so because it feared that the debtor would merely predict that disposable income will be zero, and thus render the entire disposable income test meaningless.328 Effectively, the court feared that it would be unable to project debtors’ income and that therefore debtors would have an incentive to “play ‘possum” and grossly understate their future income.329
To the extent one shares the skepticism of these courts in their ability to estimate a debtor’s future income, one would reject a debt- adjustment bankruptcy system as doomed to failure. This, however, does not justify an interpretation that ignores the plain meaning of � 1325(b). Courts can address their concerns by adopting readings of other sections that do less violence to the plain meaning of the Bankruptcy Code.330 If a court believes that it can provide a fairly good estimate of the debtor’s earnings, but that there is some risk of a big [*PG62]change in the debtor’s circumstances, it can rely on future modifications to account for this problem.331 If the court truly believes that it does not have a sufficient basis to estimate the debtor’s future income, then it should dismiss the Chapter 13 filing due to a lack of regular income.332
Even the guarded optimist of Part II does not believe that judges are omniscient and therefore rejects the utopian approach of basing the debtor’s required repayment solely on projected income. Just because judges are not omniscient, however, does not mean that they have no ability to project the debtor’s income. Rather, one might reasonably believe that judges can often use the debtor’s past income to provide a fairly good estimate of the debtor’s future income, but that on occasion this estimate will prove fairly inaccurate. As discussed in Part II, this greater confidence in the ability of judges leads one to advocate an intermediate approach between the utopian bankruptcy system that adjusts the amount of an individual’s debts based on projected income and the pessimist’s bankruptcy system that is effectively a tax on actual income. The optimist’s bankruptcy system adjusts debts based on the debtor’s projected income but then adjusts these debts again when the debtor’s actual income deviates drastically from the projected amount.333 This is at least a partial description of the approach that Chapter 13 actually takes. To understand how bankruptcy reacts to a drastic change in debtors’ financial condition, it is best to separate an improvement in debtors’ condition from a deterioration in their condition.
The extremely low completion rate of Chapter 13 bankruptcies334 suggests that judges often overestimate debtors’ ability to pay.335 A fall [*PG63]in debtors’ ability to pay, however, does not necessarily result in a failed Chapter 13 bankruptcy. If debtors’ actual income falls sharply below the projected amount, debtors may seek a modification to reduce their required repayment336 or, in an extreme case, may seek a hardship discharge.337
The prospect of post-petition relief creates a further grasshopper problem: to the extent that bankruptcy insulates debtors against an additional fall in their earning ability, they have less incentive to work hard. Because of this, some have argued that courts should not grant a modification when the decline in the debtor’s financial condition is due to the debtor’s own actions.338 This, however, is just the utopian solution to the grasshopper problem that is generally rejected both inside and outside of bankruptcy due to a concern that judges will be unable to distinguish those who deserve relief from those who do not.339
Though few, if any, reported cases deny relief to debtors on the grounds that they caused their own financial distress, the high failure rate of Chapter 13 plans suggests that this may occur in practice or that debtors sometimes find it too costly to seek a modification. This does not mean, however, that debtors receive no relief for the subsequent shock that they endure. That is, Chapter 13 does not require debtors to make the payments under their plan. Debtors may be able to reduce their required repayment without obtaining judicial approval by converting their filing to Chapter 7340 or by simply dismissing their case341 and relying on non-bankruptcy protections.
Perhaps more interesting is the fact that � 1329 explicitly allows the Chapter 13 trustee or an unsecured creditor to seek a modification to increase the debtor’s required repayments.342 Although the mechanics of the Bankruptcy Code do not work well in this context, most courts are willing to play an interstitial role and allow for modifications that increase the required payments even when the standards to be applied are unclear at best.343 They are willing to do so because even those courts that follow the plain meaning of projected disposable income are at best guarded optimists.
Congress amended � 1329 to allow creditors to share in the improvement of the debtor’s circumstances,344 but provided no standard for determining the amount by which the repayments should increase. Though � 1329 has occasionally been used to force debtors to pay following an appreciation in their assets,345 this Article focuses on the debtor’s income. When assessing how much of the increased income the debtor must pay to the creditor, the natural measure is the projected disposable income test of � 1325(b). It is not at all clear, however, whether the projected disposable income test applies to a proposed modification. For example, though � 1329 specifically references four code sections that apply to modifications, including � 1325(a), it does not directly mention the disposable income test of � 1325(b).346 Some courts have therefore held that the disposable in[*PG65]come test does not apply to modifications.347 Even if one does apply � 1325(b) to a proposed modification, there are further complications.348 The most significant among these is that � 1325(b) is restricted to projected disposable income and therefore arguably should not capture unexpected increases in the debtor’s actual income.349
Despite these problems, most courts are generally willing to use the projected disposable income test to allow creditors to capture some of the benefits of an unanticipated increase in the debtor’s financial circumstances.350 Moreover, even some of the courts that reject the use of the disposable income test in the modification context would still allow a bankruptcy court to consider an increase in the debtor’s income when ruling on a creditor’s requested modification to increase payments under a plan.351
This willingness to modify plans erodes the importance of the distinction between projected disposable income and actual disposable income. As one bankruptcy judge and commentator put it, “[i]f [*PG66]� 1325(b) is applied at modification after confirmation, the result refused in Anderson comes in through the back door . . . .”352 Because judges are not omniscient, however, this ability to modify a plan when a projection of income proves drastically wrong is central to the justification for initially basing debtors’ required repayment on their projected income. In fact, Anderson, the leading case holding that the projected disposable income test of � 1325(b) does not require debtors to promise to repay with all of their actual disposable income, based its holding in large part on the ability of creditors to seek a subsequent modification should the debtors’ circumstances improve.353
Before one decries Chapter 13 for capturing debtors’ actual income and thereby removing their incentive to work, one should note the general lack of reported cases in which courts increase debtors’ required payments in response to an increase in debtors’ earned income.354 In addition, most, if not all, of the reported cases in this category involve a drastic increase in income caused at least in part by circumstances beyond the debtor’s control.355 In Arnold,356 the court modified the plan payments after the debtor, a paper product salesman, had an increase in income of over 150%.357 While salespeople are compensated through commissions precisely because their own efforts are important in generating sales, larger economic forces and corporate decisions play a strong role as well. In Louquet,358 the court modified the plan payments after the debtor, a self-employed insur[*PG67]ance adjuster, had an increase in income of approximately 45%.359 In Powers,360 the debtor, a card dealer, also had an increase in income of about 45%.361 In addition, this income was reported only after her employer adopted a tip-allocation method that made it much harder to hide cash tips.362
Still, because the income of these individuals also depended on their own efforts, the possibility of a modification may discourage similarly situated debtors from working to increase their income. Critics of Chapter 13 often point to the debate over the proper structure of public assistance programs for the idea that the debtor must retain some of the marginal dollar earned.363 Economists have found, however, that critics of public assistance programs overstated the effective tax rates these programs create by ignoring how they are actually implemented. In calculating earnings to be counted against benefits, welfare caseworkers generally deduct numerous expenses, and errors in favor of the beneficiary often creep into the calculations.364 This results in recipients being allowed to keep some of the additional dollars earned. This effect is perhaps more pronounced in bankruptcy as each of the above debtors retained at least half of the increase in his earnings. In Louquet, the debtors’ income grew by almost $1,000 per month while the trustee only sought an increase in payments of $350 per month.365 In Arnold, the debtor’s monthly income increased by over $10,000 and yet his payments were only increased by $700 per month.366 The debtor’s gross income in Powers increased by approximately $1,000 per month but her payments increased by only $500 [*PG68]per month because she was able to claim an increase in her expenses.367
Moreover, even if the increase in debtors’ payments matches the increase in debtors’ income, debtors may still have an incentive to expend the effort necessary to increase their income. The reason for this is outlined in Arnold; as long as the underlying cause of the increase in income—say a promotion, a customer contact, or a raise—has a continuing impact after the termination of the debtor’s bankruptcy plan, the threat of a modification would not capture all of the debtor’s benefit from this change.368
A bankruptcy system that adjusts one’s debts based on one’s projected income would be administratively burdensome if modifications were often necessary. Therefore, the appropriateness of this approach depends on the ability of judges to forecast a debtor’s earnings. To estimate the debtor’s income over the three years required by the projected disposable income test,369 courts typically multiply the debtor’s current monthly income by thirty-six.370
This estimation method would fare poorly if the debtor’s income varied significantly from month to month. Section 109 of the Bankruptcy Code, however, restricts access to Chapter 13 to an individual with regular income.371 If the regular income test limited Chapter 13 to those debtors with a fairly stable aggregate income, judges could accurately estimate debtors’ future income by using their past income. One might therefore believe that an adjustment of debts approach to bankruptcy would work quite well.
Unfortunately, one can interpret the phrase “an individual with regular income” to mean an individual with some or sufficient regular income rather than an individual with regular total income. To understand the distinction, consider a salesperson who earns a salary of [*PG69]$20,000 per year and who earned no commissions in four of the five previous years and commissions of $100,000 in the other year. While this individual has some regular income, the aggregate amount of his income is not regular.
One can read � 109(e) to allow debtors with irregular total income access to Chapter 13.372 The Bankruptcy Code defines “[an] individual with regular income” to mean an “individual whose income is sufficiently stable and regular to enable such individual to make payments under a plan under Chapter 13 of this title . . . .”373 Therefore, as long as debtors have sufficient regular income to satisfy the various repayment provisions of Chapter 13, they should be allowed to file in Chapter 13.374
Prior to the addition of the projected disposable income test, judges had to determine that debtors would have sufficient income to make their payments,375 but judges did not need to consider debtors’ income when determining how large those payments should be unless they considered the debtor’s income in their good faith analysis.376 The best interests of the creditors test asks only if the unsecured creditor would have received more in a Chapter 7 liquidation and thus focuses only on debtors’ non-exempt assets.377 A few courts did consider debtors’ income when deciding whether their plans are filed in good faith, but others did not.378 Therefore, in many jurisdictions the minimum required repayment was effectively determined by the value of the debtor’s non-exempt assets, if any,379 and the value of any assets pledged by the debtor to secured creditors as collateral.380 This allowed judges to adopt a broad interpretation of the term “regular income.” One court even ruled that a debtor had sufficiently regular and stable income even though future income was not readily ascertainable with any degree of certainty.381
After the addition of the projected disposable income test, at least one court has held that the debtor’s income must be substan[*PG70]tially certain in amount and reasonably predictable.382 Most courts, however, continue to use an extremely broad definition of “regular income,” and their focus remains on whether the debtor is likely to have the income necessary to make the payments called for under the debtor’s proposed plan.383 Some courts do deny debtors access to Chapter 13 for lack of regular income, but these decisions usually involve questions of whether debtors had any income at all because the debtors were either unemployed384 or dependent on the support of friends or family385 to complete their plan. It should be noted that even these debtors are sometimes found to have regular income.386
The addition of the projected disposable income test may make such a broad interpretation of “regular income” inappropriate. As a matter of policy, a debt-adjustment bankruptcy system requires that judges be able to estimate the debtor’s income and therefore requires the debtor to have fairly stable income. One does not need to resort to public policy arguments, however, to justify a change in the interpretation of “regular income.” One can argue that the language of the Bankruptcy Code requires that judges dismiss plans filed by debtors with uncertain aggregate income because these debtors will be unable to “make payments under a plan under Chapter 13 . . . .”387
Consider how the projected disposable income test would apply to our hypothetical salesperson. Assume that the judge believes that the salesperson’s earnings history indicates that there is an 80% chance that the salesperson will earn no commissions in a given year and a 20% chance that the debtor will earn $100,000 in commissions. Although courts typically just multiply the debtor’s current monthly income by thirty-six, this is only a presumptive guide and does not preclude other estimation methods.388 While judges should not in[*PG71]clude the entire $100,000 commission,389 they should not ignore the prospect for earning this amount either. Financial analysts often value corporations based on their projected income or projected cash flow;390 in doing so they may discount uncertain earnings more heavily, but they would not ignore them. A firm’s, or an individual’s, projected or expected earnings is simply the average amount that it can earn. In the case of our salesperson, the projected income consists of the $20,000 salary and an average commission of $20,000 (20% multiplied by $100,000). If our salesperson is asked to repay an amount based on a $40,000 projected income and the salesperson has at least some reasonably necessary expenses, the salesperson will be unable to make payments the 80% of the time when he earns only $20,000. This would be true even if the court built in a small cushion for unexpected developments.391 Because the salesperson lacks sufficient stable income to make the payments that are required based on his projected income, his filing should be dismissed.392
A debtor denied access to Chapter 13 effectively has two choices: file under Chapter 7 or rely on non-bankruptcy law for protection. [*PG72]First, consider how the debtor would be treated under non-bankruptcy law. Creditors would be able to garnish the debtor’s actual income subject to state and federal limitations;393 non-bankruptcy law effectively taxes the debtor’s actual income. Note that this is the second-best bankruptcy system, or the optimal bankruptcy system when a judge cannot estimate the debtor’s potential or projected income.394
Like all income taxes, garnishment may discourage the debtor from working. The marginal rates created by garnishment law, however, are not unreasonably large by the standards of income taxation. For example, a single debtor living in Virginia and earning $25,000 a year has a marginal tax rate of approximately 25.7% once all state and federal taxes are considered.395 If the debtor’s wages are garnished, the total effective marginal rate will rise to approximately 44%.396 While this rate is quite high, it is only slightly higher than the 43% marginal rate that the debtor would pay without garnishment if the debtor earned $300,000 per year, 397 and well below some historical federal tax rates which have reached as high as 94%.398 This comparison is not meant to imply that debtors in financial distress should pay the same effective tax rate as the very wealthy; current garnishment rates may be too high to allow low-income debtors to maintain a sufficient standard of living. Rather, it is merely meant to show that society is sometimes willing to accept the work disincentives associated with high marginal rates in order to achieve other goals.
Of course, garnishment will have some effects that are not marginal in nature. The garnishment process may prove so administratively burdensome to the debtor’s employer that the employer may decide to simply fire the debtor. If garnishment leaves the debtor with too little take-home pay, the debtor may prefer to forego work altogether.399 Non-bankruptcy law, however, is designed to address these [*PG73]non-marginal effects as well. For example, federal law now prohibits terminations in response to a garnishment order, though some doubt the effectiveness of this prohibition.400 In addition, note that debtors would only refuse to work if they have some other means of support such as the assistance of family and friends or public assistance; debtors would not choose to forego work if this meant starvation. Therefore, garnishment must always leave the debtor with a sufficient amount so that the debtor would not prefer to rely on this outside source of support. This problem is likely to be particularly severe for the low-income debtor who may earn little more than is available from public assistance or the non-market economy. Perhaps for this reason federal garnishment law also prohibits garnishment that would leave the debtor with take-home pay less than thirty times the federal minimum wage.401
Unfortunately, simply dismissing Chapter 13 filings will not necessarily lead to the optimal result. These dismissals may just result in more Chapter 7 filings, and Chapter 7 ignores the debtor’s income entirely. Moreover, as a matter of theory, one cannot even conclude that the current federal limitations of garnishment, which allow the debtor to keep most of each additional dollar earned, offer debtors a greater incentive to work than a system that seizes all of their actual disposable income. To see why this is the case, it is necessary to return to the literature on public assistance.
For years economists complained that public assistance programs that reduced a recipient’s benefit with each additional dollar earned effectively created a 100% tax rate and thereby discouraged the recipient from working.402 The same could be said of a disposable income clause as debtors are required to repay with any amount above their reasonably necessary expenses. When economists actually studied how public assistance programs were implemented, however, they [*PG74]found that recipients did retain some of the additional income they earned because welfare caseworkers were liberal in allowing the recipients to deduct certain expenses.403 The same is likely to be true in bankruptcy; there is at least anecdotal evidence that debtors retain a large portion of the benefit of an improvement in their circumstances when their plans are modified to increase the required repayments.404
Still, the marginal tax rate created by a disposable income clause is likely to be extremely high. But, as a matter of theory, one cannot conclude that a collection system with a high marginal tax rate discourages work more than one with a low marginal rate. Consider a public assistance program again. If recipients are allowed to retain more of their income, the current recipients may very well work more. This change, however, would also make public assistance attractive to more individuals, and these new recipients would work less than they did before. Because benefits are still partially reduced as earnings increase, these individuals retain less of their marginal dollar earned and thus have less incentive to work. In addition, the public assistance gives them additional wealth, further reducing their need for income and their incentive to sacrifice leisure. To determine whether the lower rate resulted in more hours worked one must compare the additional hours worked by the original recipients and the reduced hours worked by the new recipients.405
This same logic would apply in bankruptcy. A system that allows bankrupt debtors to retain more of their disposable income may encourage bankrupt debtors to work more. This change may, however, also make bankruptcy attractive to more debtors and thereby encourage others to file. Because bankruptcy would seize some of the disposable income of these new bankrupt debtors, they would keep less of their marginal earnings than they did before. In addition, bankruptcy increases their wealth and thus reduces their incentive to work.
Therefore, the correct method of taxing a bankrupt debtor’s income depends critically on empirical assumptions. This is not a decision, however, that courts are asked to make in the context of imple[*PG75]menting Chapter 13. The language of the Bankruptcy Code makes it clear that Chapter 13 should only be open to those debtors with sufficient regular income to make the payments required by a plan filed under that chapter. Today, the language of the Code requires that these payments be set with regard to the debtor’s projected or expected income. Therefore, those whose income is highly uncertain are unlikely to earn enough to be able to repay an amount based on their projected income. Dismissing these cases may not lead to the optimal result, but it will not lead to a result that would justify a departure from the plain meaning of the statute.
Bankruptcy resembles a public insurance program in that it provides assistance—in the form of debt relief—to individuals who have suffered some misfortune.406 Bankruptcy resembles a tax in that it frequently compels individuals to make payments out of their future income. Like public insurance or progressive taxation, bankruptcy creates moral hazards. Because bankruptcy makes financial distress less painful, it creates a grasshopper problem by encouraging debtors to engage in behavior, such as excessive spending, that makes distress more likely.407 Bankruptcy also creates an opossum problem by encouraging debtors to claim more relief than their actual circumstances require.408 With enough information, a judge can solve both the opossum and grasshopper problems and implement a utopian bankruptcy system that, like the first-best public insurance and tax systems, results in zero loss of efficiency. To do this, the judge would reduce the debts of deserving debtors to an amount commensurate with each debtor’s ability to repay.
While such a utopian bankruptcy system provides a useful benchmark, it is a poor model for public policy because it requires omniscient judges who can precisely determine which debtors are deserving of relief and how much they can earn to repay their debts.409 Of more relevance is the second-best, or feasible, bankruptcy system, which accounts for the limited information real judges possess and the inevitable moral hazards that result from these limitations.410 In [*PG76]the absence of explicit code provisions designed to solve these moral hazard problems, judges have seized on the discretion granted them explicitly or implicitly by the Bankruptcy Code to implement important elements of this second-best bankruptcy system.411 The judicial approaches to the moral hazard problems resemble the approaches taken in actual public insurance and taxation systems.
While the utopian bankruptcy system would instruct judges to solve the grasshopper problem by denying relief to the negligent, judges have not tried to do so.412 This approach mirrors that taken by both private insurance contracts and public insurance programs that operate in an imperfect, or second-best, world.413 Insurance generally protects individuals against the consequences of their own negligence because judges cannot always distinguish between the negligent and the prudent—and individuals may not be able to conform to the judges’ expectations even if the judges could make this distinction.414
Neither public insurance programs nor private insurance contracts, however, provide relief for willful misconduct, whereas bankruptcy does. Chapter 13 of the Bankruptcy Code contains a superdischarge that will even relieve the debtor of a judgment for a willful and malicious tort.415 Judicial resistence to the superdischarge is predictable because, in affording relief to the willful and malicious, bankruptcy departed from the approach taken in other public insurance programs.416 Invoking the good faith standard of � 1325(a)(3), judges have limited the superdischarge to those debtors who have received sufficient punishment (satisfying deterrence concerns) and have made a significant effort to repay their victims.417 In this way, judges have pushed bankruptcy law toward the approach that public insurance programs take toward the grasshopper problem.418
To solve the opossum problem, the utopian bankruptcy system would instruct the judge to reduce a debtor’s obligations to an amount commensurate with the debtor’s ability to repay. Chapter 13 appears to embrace this approach by reducing a debtor’s obligations to an amount commensurate with the debtor’s projected income.419 [*PG77]Judges resist this approach, however, because they are skeptical of their own ability to project income and thus fear that a debtor will “play ‘possum” and understate future income. This skepticism causes some judges to adopt a strained reading of the Bankruptcy Code to adopt a second-best approach, a system that creates an effective tax on the debtor’s actual income in bankruptcy. This approach mirrors the structure of actual taxation and public assistance programs. In general, citizens are taxed not on what they can earn but what they actually earn. Likewise, public assistance programs frequently reduce the benefits paid as the recipient’s actual income increases, but make no attempt to measure a recipient’s actual abilities.
In the bankruptcy context, however, judges may be able to account for their limited information in ways that do less violence to the plain meaning of the Bankruptcy Code. If judges can use the debtor’s income history to provide at least a workable projection of the debtor’s future income, they can base the debtor’s required payments on this projection, not on actual income. This approach would be much closer to the ideal solution to the opossum problem and it should be favored, provided that the debtor’s income history allows the judge to make a fairly accurate projection.
In the unique context of Chapter 13, which should apply only to debtors with regular income,420 the debtor’s past income may provide the judge with sufficient information to project future income. If the judge does not have sufficient information—i.e. if the debtor does not have predictable income—the statutory structure suggests that the judge respond by holding Chapter 13 relief to be unavailable rather than by imposing a tax on actual income within bankruptcy.421
If debtors denied relief in Chapter 13 rely on state law for protection, then they will in fact receive a tax on their actual income in the form of a state garnishment proceeding; this is the appropriate collection mechanism for an individual with irregular income. A debtor also has a right, however, to file under Chapter 7 of the Bankruptcy Code. This chapter ignores debtors’ future income—probably their most valuable asset—when setting their required repayment. This Article suggests that, despite this anomaly, Chapter 7 may still play a role in an optimal bankruptcy system if restricted to those debtors who would repay nothing in Chapter 13.
[*PG78] Whether current bankruptcy law contains adequate mechanisms for limiting Chapter 7 to these debtors is questionable,422 though this is a topic for another day. Nevertheless, the analogy between bankruptcy and public insurance is likely to yield valuable insights.423 The search for the proper tools to deal with the problems of poverty and inequality has represented one of the most vibrant areas of economic research for over one hundred years. Although economists have long agreed on a fairly simple first-best approach, they continue to vigorously debate the precise form of the second-best, or feasible, program. Therefore, if we are interested in a feasible optimal bankruptcy system, the public assistance and taxation literatures do not promise an easy answer. They do offer, however, valuable lessons and remind us that whatever system we do choose must be designed to operate in a non-optimal world.