Is the Personal System Bankrupt?

BY LORETTA J. MESTER

ver the past few years, several attempts CURRENT PERSONAL BANKRUPTCY SYSTEM have been made to reform the U.S. IN THE U.S. O bankruptcy system, to help stem perceived As Joseph Pomykala discusses in his interesting article, the word abuses of the system. In this article, “bankruptcy” has two roots. “Banca Loretta Mester outlines the components of reform rotta” is Latin for broken board. In medieval Italy, “creditors would proposals. She then looks at the empirical research break the workbenches of defaulting on bankruptcy to evaluate the rationale for merchants over the merchants’ heads.” “Banqueroute” is French for debtors reforming the system and the effectiveness of on the lam (route), as bankruptcy proposed changes. was considered an act of debtor fraud. As Pomykala points out, before the mid-18th century, bankruptcy was considered a crime, and in England, Neither a borrower nor a lender be; bills were passed in the House (HR 333) certain bankrupt debtors were subject For oft loses both itself and friend, and in the Senate (S 420) in March to capital punishment. The U.S. And borrowing dulls the edge of husbandry. 2001, but Congress adjourned before modified English law to be less harsh. Polonius reconciliation of the bills could be For example, the Pennsylvania Act 1, Scene 3 completed. Legislation is again being Bankruptcy Act of 1785 allowed for Hamlet by William Shakespeare considered this year, but regardless of those convicted of bankruptcy to be the outcome, the debate about the flogged while nailed by the ear to a U.S. personal bankruptcy system is pillory, after which the ear would be Over the past several years, unlikely to be resolved anytime soon. cut off. (Of course, how much more Congress has attempted to pass Indeed, a number of studies provide lenient this was is clearly debatable.) legislation to resolve perceived problems evidence on the rationale for changing Bankruptcy protection has in the personal bankruptcy system in the current system, on whether reform been part of U.S. federal law since 1898. the U.S. Although it has not proposed is necessary, and on whether the Indeed, Article I, Section 8 of the U.S. anything as drastic as Polonius proposed revisions will have the Constitution authorizes Congress to recommended, Congress has proposed intended effect. enact “uniform Laws on the subject several significant changes to the After reviewing the current of ” (see the article by current system. In the latest try, separate personal bankruptcy system and the Leonidas Mecham). The structure of proposed changes, we’ll discuss some the current bankruptcy system was of the findings of these recent studies. established in the Bankruptcy Reform These studies do shed some light on the Act of 1978. The idea is to allow a “fresh Loretta Mester is debate and cast some doubt that the start” (within limits) for honest people senior vice president who, often through unfortunate and director of proposed changes will yield benefits research at the as significant as intended. They also circumstances beyond their control, Philadelphia Fed. suggest further research is necessary have gotten into trouble with debt and to resolve all the issues. to allow for creditors to be repaid in an orderly fashion with the debtor’s available assets. www.phil.frb.org Business Review Q1 2002 31 The bankruptcy provisions the bankruptcy act.1 Chapter 7, extend to five years). The debtor must allow for a sharing of risk between sometimes called straight bankruptcy, repay creditors at least as much as they borrowers and creditors, offering some is liquidation — a filer hands over his would have received under Chapter 7, insurance to borrowers if they find assets (with some exemptions) to the and claims entitled to priority must be themselves unable to repay their debts. trustee, who then sells the assets and paid in full. The debtor cannot take The insurance gives consumers whose uses the proceeds to repay the debtor’s on any new debt without the trustee’s income may be low today but is creditors. The remaining debts (with approval, and the debtor’s secured expected to rise in the future the some exceptions) are then discharged and must be less than confidence to borrow now to pay for — that is, wiped clean — and the certain specified limits to allow him consumption. This raises consumers’ debtor retains control of his or her or her to file under Chapter 13. In economic well-being. If things go as future income.2 In many cases there exchange, debtors retain more of their planned, they will repay their debts. are few assets available to repay assets than they would under Chapter If some adverse event, like a job loss, creditors, and most unsecured debt, 7. At the end of the repayment prevents them from repaying, they such as , is not repaid period, any remaining unpaid debt is can file for bankruptcy and protect in bankruptcy (see the CBO study). discharged. Generally, someone can their future income from creditors. In other cases, a debtor may want to file for bankruptcy under Chapter 13 Bankruptcy procedures better enable keep control of an asset, like a car, as often as he or she wants (except consumers to smooth their consumption that is pledged as collateral against that, in some cases, the filing cannot over time, thereby increasing economic a loan. In this case, the debtor can be within 180 days of dismissal of a efficiency. The bankruptcy system also “reaffirm” the debt — the debtor and previous case). provides a joint debt-collection system creditor agree that the debtor will pay As part of the filing under for a debtor’s creditors (see the CBO the creditor all or part of the debt, either chapter, the debtor submits a list study for a nice review of personal even though the debtor has filed for of his assets, income, liabilities, creditors, bankruptcy fundamentals). A bankruptcy, and the creditor will not and debts. After a debtor files under bankruptcy system that is too harsh repossess the property. A person can file either chapter, an automatic stay stops would lower economic welfare by for bankruptcy under Chapter 7 every creditors from collecting on unpaid discouraging borrowing and risk sharing. six years. debts. This prohibits creditors from However, a system that is too lenient Chapter 13 involves adjustment filing lawsuits against the debtor for and allows debtors to escape from of debts of an individual with regular repayment, trying to garner the debtor’s their commitments too easily can hurt income. Under this chapter, some debts wages, or making telephone calls economic efficiency by causing creditors are reduced, but then debtors and demanding repayment (see Under- to restrict the supply of credit and raise creditors devise a plan by which the standing the Federal Courts). its cost to creditworthy borrowers. debtors repay their remaining obligations The federal bankruptcy courts out of their future incomes. Repayment WHY THE CALL TO REFORM? administer the bankruptcy system. All is made in installments over a three- Bankruptcy protection is bankruptcy cases are filed in these year period (which the court can designed to help people get out from federal courts. There are 94 bankruptcy under the burden of excessive debt and districts in the U.S. and its territories, to get a fresh start. Some people, though, each with a court. Pennsylvania is point to abuses of the system: Debtors broken into three districts: eastern, 1 They may also file under Chapter 11, but who actually have the ability to repay central, and western, while New Jersey this is rare. Chapter 12, which is similar to sometimes escape their obligations. Chapter 13, applies only to family farmers Knowing this escape route exists may and Delaware are each separate and in financial distress. See Report 106-49 on single districts. The courts in all three S. 625, U.S. Senate, and Mecham for fuller give borrowers an incentive to take on states (along with the District of the descriptions of the U.S. personal bankruptcy more debt, to the extent that lenders are system. Virgin Islands) are part of the third willing to supply them credit. Of course, 2 federal circuit. A bankruptcy case is According to Mecham, 18 categories of lenders should respond to a lenient debt cannot be discharged under Chapter 11; often overseen by a court-appointed there are fewer restrictions under Chapter 13. bankruptcy law that permits many trustee. Certain types of tax claims, debts for spousal debts to be discharged by restricting Under current law, individuals or child support or alimony, and debts for credit or raising the cost of credit so that willful and malicious injuries to person or considering bankruptcy can file property are examples of nondischargeable only the better credit risks could borrow. under Chapter 7 or Chapter 13 of debts. A bankruptcy system that is too

32 Q1 2002 Business Review www.phil.frb.org lenient would lead to economic inefficiencies whereby the supply of FIGURE 1 credit was too restricted and its cost too high. Annual Number of Bankruptcy Filings Over the years, changes have been made to the system to try to limit (12-month periods ending in June) the scope for abuse. For example, in 1984, judges were permitted to dismiss Chapter 7 cases if they thought granting relief would constitute a “substantial abuse” of the bankruptcy code. But the term “substantial abuse” was not defined, and creditors and trustees were not allowed to present evidence to the judge in a particular case on whether relief should be viewed as substantial abuse (see Report 106-49, Senate). Filings Have Increased in Recent Years. One factor pointed out in the most recent round of legislative consideration of the bankruptcy system Note: Shaded areas represent economic recessions. was the substantial increase in the Source: Bankruptcy data from Administrative Office of the U.S. Courts. number of filings that began in the 1980s. Total bankruptcy filings, including personal and business filings, FIGURE 2 hit a record 1.43 million in the year ended June 1998, and they have been Number of Personal Bankruptcy Filings, very high since then, with 1.39 million Total and by Chapter filings in the year ended June 2001 (12-month periods ending in June) (Figure 1).3 Indeed, the 400,000 filings in the second quarter of 2001 was the most ever for a three-month period. Total Personal Filings Chapter 7 Personal Filings Annual number of filings in thousands Over 97 percent of these filings are Chapter 13 Personal Filings 1600 personal as opposed to business, and 70 1379 1352 1349 1400 percent of personal filings per year are 1240 typically Chapter 7 filings (Figure 2).4 1200 985 969 951 1000 864 800

600 394 382 375 398 3 The number of personal filings before 400 and after 1979 are not directly comparable 200 because the Bankruptcy Reform Act of 1978 allowed for spouses to file a joint petition for 0 bankruptcy protection. See the CBO study 1998 1999 2000 2001 for further discussion.

4 Note that some of the personal filings Source: Administrative Office of the U.S. Courts. could actually represent business failures because some small businesses are funded by the personal credit lines of their owners. See Appendix A of the CBO study for further discussion of the data on personal bankruptcy filings.

www.phil.frb.org Business Review Q1 2002 33 Figure 3 shows the number of personal bankruptcies per thousand households FIGURE 3 (the bankruptcy rate) in the three states in the Third Federal Reserve Personal Bankruptcy Rate, Total and by Chapter District (Delaware, New Jersey, and Pennsylvania) and in the U.S.5 As (12-month periods ending in June) you can see, these numbers ranged between nine and 13 bankruptcies per 1000 households in 2001. In other words, in 2001, 0.9 to 1.3 percent of households filed for bankruptcy in the nation and in our three states. (Note that there is much wider variation in the personal bankruptcy rate across the other states in the U.S. In 2001, Tennessee, which had 24.5 filings per thousand households, had the highest rate; Iowa, which had 4.1 filings per thousand households, had the lowest rate. See the Table.) Note: Sum of Chapter 7 and Chapter 13 does not equal total because there are a few personal business filings under Chapter 11. Personal bankruptcy rate is the number of The number of personal personal bankruptcy filings per thousand households. bankruptcy filings began accelerating Sources: Administrative Office of the U.S. Courts and U.S. Census. in the 1980s, with an especially large increase between 1995-98.6 For example, from 1961 to 1980, the FIGURE 4 personal bankruptcy rate rose, on average, about 3 percent per year. Personal Bankruptcy Rate and Since 1980, the average increase in Growth in Personal Bankruptcy Rate the personal bankruptcy rate has been (12-month periods ending in June) almost 8 percent per year, with an especially sharp increase of 14 percent per year between 1995 and 1998 (Figure 4). We can look at the increase another way: In 1980, there was one personal bankruptcy filing for every 336 house- holds in the U.S.; in 2001, there was one personal bankruptcy filing for every 78 households. Some of the rise in bank- ruptcies in the 1980s can be attributed to bad economic times: There was a short recession from January 1980 to July

Note: Shaded areas represent economic recessions. 5 Delaware has about 300,000 households, Personal bankruptcy rate is number of personal filings per year per 1000 households. New Jersey about 3.1 million households, Sources: Bankruptcy data from Administrative Office of the U.S. Courts; household Pennsylvania about 4.8 million households, data from the Bureau of Census, www.census.gov. and the U.S. about 106 million households.

6 In contrast, business filings, which increased in the early 1980s, fell back in the latter half of the 1980s and in the 1990s.

34 Q1 2002 Business Review www.phil.frb.org 1980 and a long one from July 1981 to TABLE November 1982 (recessions are shown by shaded bars in the figures). But the Personal Bankruptcy Rate by State in rapid rise in the bankruptcy rate in the Year Ended June 2001 1990s is more difficult to understand, since this was a period of very good (Number of Personal Bankruptcies per 1000) economic conditions — economic Personal growth averaged 3.2 percent per year Total No. of Bankruptcy Rate in the 1990s, and the unemployment rate Nonbusiness Households (Nonbus. Filings State Filings (2000 Census) per 1000 Hhs) had fallen to 4 percent by the end of the decade. Even this is not unprecedented: Tennessee 54,730 2,232,905 24.51 Utah 16,915 701,281 24.12 The rate of bankruptcy filings rose Georgia 63,800 3,006,369 21.22 rapidly in the mid-1980s in the midst Nevada 15,833 751,165 21.08 of an economic expansion, and it has Alabama 36,116 1,737,080 20.79 Mississippi 20,561 1,046,434 19.65 risen in other periods of economic Arkansas 19,466 1,042,696 18.67 expansion as well. (Note that the rise Indiana 42,537 2,336,306 18.21 is not necessarily a bad thing, as it Maryland 32,956 1,980,859 16.64 accompanied an increase in credit Idaho 7,578 469,645 16.14 Oklahoma 20,892 1,342,283 15.56 availability to households.) Washington 34,087 2,271,398 15.01 Households did increase their Louisiana 24,730 1,656,053 14.93 borrowing in the 1990s, and the rise in Kentucky 23,609 1,590,647 14.84 Oregon 19,667 1,333,723 14.75 household debt-service burdens — that Illinois 66,817 4,591,779 14.55 is, required payments on mortgages Virginia 38,361 2,699,173 14.21 and other as a percentage Ohio 61,906 4,445,773 13.92 of disposable income — in that decade West Virginia 9,630 736,481 13.08 New Jersey 39,575 3,064,645 12.91 may explain part of the rise in the Missouri 27,989 2,194,594 12.75 bankruptcy rate (Figure 5, see next California 143,174 11,502,870 12.45 page). Yet the debt-service burden was Florida 78,702 6,337,929 12.42 Wyoming 2,343 193,608 12.10 at comparable levels in the mid-1980s, Kansas 12,554 1,037,891 12.10 and the bankruptcy rate was much Hawaii 4,712 403,240 11.69 lower.7 So factors other than debt- Arizona 22,036 1,901,327 11.59 service burdens appear to play some Rhode Island 4,690 408,424 11.48 New Mexico 7,420 677,971 10.94 role in the decision to file. Michigan 41,251 3,785,661 10.90 The most recent increase in Colorado 16,921 1,658,238 10.20 filings in the first half of 2001 might Pennsylvania 47,708 4,777,003 9.99 Montana 3,578 358,667 9.98 be the result of pending legislation Washington, DC 2,410 248,338 9.70 to change the bankruptcy system, as North Carolina 30,215 3,132,013 9.65 people contemplating bankruptcy may Nebraska 6,382 666,184 9.58 have accelerated their filings to get in Wisconsin 19,624 2,084,544 9.41 Delaware 2,730 298,736 9.14 under the old rules. Still, the fact that New York 63,642 7,056,860 9.02 the bankruptcy rate rose in the late Texas 66,290 7,393,354 8.97 1990s during good economic times and Connecticut 11,144 1,301,670 8.56 South Carolina 12,868 1,533,854 8.39 remains high is taken by many as an Maine 4,198 518,200 8.10 indication that the system is being Minnesota 15,216 1,895,127 8.03 abused by people who actually have North Dakota 2,021 257,152 7.86 South Dakota 2,228 290,245 7.68 New Hampshire 3,545 474,606 7.47 Massachusetts 16,676 2,443,580 6.82 7 Vermont 1,580 240,634 6.57 There have been periods — for example, Alaska 1,361 221,600 6.14 between 1988 and 1991 — when debt-service Iowa 9,662 2,336,306 4.14 burden and filings moved in opposite directions. www.phil.frb.org Business Review Q1 2002 35 the wherewithal to repay their debts. This view has led many observers to FIGURE 5 believe that the bankruptcy system needs to be revamped and has led Personal Bankruptcy Rate and to proposed legislation to change the system. Whether changes to the Debt-Service Burden bankruptcy system will have much of an effect on the rate of filings depends both on what changes will be enacted and whether the bankruptcy system itself has encouraged filings.

BANKRUPTCY REFORM LEGISLATION Although there have been many attempts to pass legislation over the past several years, bankruptcy reform legislation has yet to be signed into law. In March 2001, the House and Senate passed their own versions of Note: Shaded areas represent economic recessions. bankruptcy reform legislation (HR 333 Source: Bankruptcy data from Administrative Office of the U.S. Courts; debt-service and S 420); similar bills were passed the burden data from the Federal Reserve Board. Debt-service burden is household required payments on mortgage and consumer debt as a percentage of disposable personal income. previous year by the 106th Congress, Quarterly personal bankruptcy rate is quarterly filings per thousand households. Note, and hearings were held in 1997 by the sum of quarterly filings equal annual filings. Correlation between debt-service burden 105th Congress. Last year, Congress and quarterly personal bankruptcy rate was 0.84 from 1980 to 1987, –0.71 from 1988 to 1991, and 0.88 from 1992 to 2000. adjourned before reconciliation of the bills could be completed, but bankruptcy reform legislation is again being A means test would be applied to IRS uses to figure out living expenses considered this year. While the determine which debtors would be for people owing back taxes, but the versions passed by the House and the forced into Chapter 13 and how much bill specified the use of actual costs of Senate in 2001 differed in some ways, debt would have to be repaid over a other necessities (for example, child those differences have narrowed as five-year period. Debate has centered care, union dues, and so forth). Some legislation has worked its way through on whether means testing is necessary. income, such as Social Security and several sessions of Congress. In last Creditors favor such testing, saying war crimes compensation, would be year’s bills, there was general agreement that some debtors have abused the excluded from the calculations. Those on basic aspects of reform. Here I’ll current bankruptcy system. Consumer earning less than the median income review eight proposed changes to groups say only 3 percent to 5 percent in the applicable state would qualify the bankruptcy system. The first five of Chapter 7 filers would have to for Chapter 7 regardless of their ability of these reforms favor creditors by repay some of their debt under the to repay. limiting the benefits to debtors from proposed means tests, and they argue Under current law, if a debtor declaring bankruptcy. The last three that reform is not necessary, since the files under Chapter 13, the repayment reforms might be considered debtor rate of filings dropped in 1999 after period is three years unless the court, for protections. the record rate in 1998. As discussed cause, extends it to a maximum of five (1) Chapter 7 Means below, arguments for and against years. This provision would remain the Testing. The bankruptcy system would means testing are in dispute. same for debtors whose family income is be changed to what proponents of the The 2001 bills barred Chapter less than the median family income in bills call a “needs-based” system. If a 7 filing if, after living expenses and the applicable state. However, for debtor has sufficient income to repay a the cost of other necessities, the debtor families with higher income (who large part of his or her debts, he or she could afford to repay at least $100 per would be forced to file under Chapter could not pursue Chapter 7 liquidation month over a five-year period. The 13 by the means test), the repayment but only a Chapter 13 repayment plan. bills generally used the standards the plan would be extended to five years.

36 Q1 2002 Business Review www.phil.frb.org (2) Nondischargeable put a federal cap of $125,000 on the downs. The House bill barred strip- Debts. Bankruptcy courts currently exemption for a primary residence. downs for a motor vehicle acquired by presume that if a debtor bought more This cap would apply to all states. the debtor within five years prior to filing than $1000 in luxury goods or services The House bill maintained states’ bankruptcy and for other personal or took $1000 in cash advances in ability to opt out of the federal limit property bought within one year prior to an open-ended credit plan within 60 and reestablish an unlimited or other filing. The Senate bill differed from the days before filing for bankruptcy, these exemption by passing legislation.9 House bill in setting the period for motor debts were fraudulently incurred, Both bills lengthened the time from vehicles at three years prior to filing. and so they are not dischargeable. six months to two years that a debtor (5) Repeat Filings. Currently, Any debt incurred to pay an existing must live in a state before being able to debtors can file for Chapter 13 bank- nondischargeable debt is non- claim that state’s exemption. ruptcy at any time (except, in some dischargeable if it was incurred with (4) Lien Strip-Down. cases, within 180 days of a prior the intent of not repaying. Non- Currently, debts are secured only to dismissal). Debtors can file for Chapter dischargeable debts include certain the value of the collateral, with any 7 bankruptcy six years after a discharge taxes, family support obligations, and remainder treated as unsecured debt. in bankruptcy. The legislation would debts arising from fraud. The proposed Thus, a debtor could pay the amount have limited repeat filings. Both bills legislation would make more debts of the collateral’s current market value would have barred a Chapter 7 filing nondischargeable in bankruptcy. and keep the collateral — this is a strip- within eight years of prior discharge. The 2001 bills extend the down. For example, a debtor could The Senate bill would have barred a definition of fraudulently incurred purchase a car, file for bankruptcy, and Chapter 13 discharge within three years (and, therefore, nondischargeable) keep the car by paying off the value of of a prior discharge under a Chapter 7, debt. In the House bill, the threshold for the car, even though this is less than the 11, or 12 filing, and within two years of presumption of fraudulent purchases of amount he contracted to pay in the a prior discharge under Chapter 13. The luxury goods was lowered to $250 within loan. Debtors who bought furniture, House bill was harsher, disallowing a 90 days of filing for bankruptcy, and the which has little resale value, are able to Chapter 13 discharge within five years threshold for cash advances was lowered keep the furniture by repaying the low of any prior discharge. to $750 within 70 days of filing. The resale price rather than paying off the (6) Reaffirmation of Debts. Senate bill agreed with the House, much larger debt. Consumer groups Consumer advocates say creditors are except that the threshold for luxury argue that changing the rules to allow pressuring debtors into reaffirming debts goods was lowered to only $750. creditors to repossess the collateral — in other words, pressuring them into (3) Homestead Exemption. would force reaffirmations, wherein saying that they will pay the debt and Many states allow a debtor to keep debtors agree to pay certain debts and not discharge it in bankruptcy. Such possession of his or her residence (up to not discharge them in bankruptcy. The reaffirmations are supposed to be filed some limit) when filing for bankruptcy; 2001 legislation sought to limit strip- with and approved by the bankruptcy the residence would not be available to court. But Sears was recently held pay off creditors. Five states (Florida, liable in a class action based on

Iowa, Kansas, South Dakota, and 9 Prior to 1978, there was no federal reaffirmations of debt that were not Texas) put no limit on the exemption exemption; the states controlled exemptions. filed with bankruptcy courts. The case for a primary residence; other states The Bankruptcy Reform Act of 1978 set involved Sears’ pressuring debtors into federal exemptions for certain assets, are quite restrictive (for example, including personal goods, tools of trade, reaffirming their debts with Sears New Jersey allows no homestead autos, and homesteads. But individual states rather then having them discharged in 8 were allowed to opt out and set their own exemption). One of the major limits, and by 1983, all of them had. The bankruptcy. Sears admitted to criminal differences between the House Bankruptcy Reform Act of 1994 raised the fraud and paid a fine of $60 million. and Senate bills passed last year was federal exemption level and tied it to the The bills passed in 2001 Consumer Price Index starting in 1998. As their treatment of the homestead of January 1, 2000, the federal exemption was sought to stem abusive reaffirmation exemption. The Senate version would $16,150 per debtor; 35 states had set their agreements by mandating that certain own exemption levels and did not allow the federal exemption; and the remaining states specific disclosures be made in writing allowed debtors to choose between their state to the debtor, explaining the terms exemption or the federal exemption when filing 8 of the underlying credit agreement These data are as of January 1, 2000. See for bankruptcy. See footnote 13, page 490 of footnote 13, page 490 of Report 107-3, House Report 107-3 Part 1, House of Representatives and the reaffirmation. The bills of Representatives. and the study by Jon Nelson. asked the attorney general to www.phil.frb.org Business Review Q1 2002 37 enforce prohibitions against abusive to negotiate a repayment plan with evidence on the relationship between reaffirmations. his creditors on his own. However, the social forces (stigma) and the number (7) Consumer Credit requirement might also delay filings of bankruptcy filings. A lessening in Disclosures. Some argue that creditors until debtors are unable to come up the effectiveness of these social forces have led consumers into bankruptcy by with a repayment plan in Chapter 13. would help support arguments for misleading them about the true cost of Both bills required the debtor to have reform. Next I’ll discuss evidence credit. The legislation sought to make undergone within 180 on the extent to which the current credit card companies disclose more days before filing under Chapter 7 or bankruptcy system is being abused. about the consequences to the borrower Chapter 13. High levels of abuse favor the of making only the minimum monthly reformers. Finally, I’ll discuss evidence payment. The implications of concerning the efficacy of proposed The bills would have amended reforms. Even if one believes the the Truth-in-Lending Act to require the empirical work current system needs reform, it is not disclosures on credit-card bills. Credit [on bankruptcy] to clear that the proposals will yield the card issuers would have had to provide desired effect. generic examples of the consequences date is mixed. (1) Market Forces. In a study of making only minimum payments on done in 2000, Lawrence Ausubel argued bills in terms of how long it would take As suggested by this overview, that market forces have tempered some to pay off the debt. Issuers also would while the House and Senate bills passed of the recent acceleration in personal have had to give a toll-free number in 2001 differed in some of their details, bankruptcy filings and that legislation where holders could get specific there was general agreement on major is unnecessary. As the bankruptcy information about repayment scenarios items of reform. The real question is rate increased, lenders responded by for their own accounts. Enhanced whether these reforms are necessary tightening their credit standards, thus disclosures for home-equity , and, if so, whether they will be effective. leading to the decrease in the number introductory loan rates, and late of bankruptcies between 1998 and 1999. payment deadlines and penalties EVIDENCE RELEVANT TO In Ausubel’s view, if there ever was a would have been required. Both bills BANKRUPTCY REFORM “bankruptcy crisis,” it is self-correcting. would have barred the termination of There have been several In congressional testimony in a credit card just because a customer studies of personal bankruptcy. The 1998, Ausubel argued against the means- hadn’t incurred a finance charge. implications of the empirical work to test approach because, in his view, the In testimony before the date is mixed as to whether the proposed immediate cause of the record number House in March 1999, Federal Reserve bankruptcy reform is needed and of bankruptcies is the high level of Governor Edward Gramlich said the whether it will have the intended effect. household debt, which he attributes in Board of Governors wondered whether Partly at issue is the extent to which part to aggressive lending tactics. In a repeated disclosures to consumers might borrowers respond to the incentives 1999 study, Ausubel found that in create “information overload.” He said provided by the bankruptcy law — randomized trials on preapproved credit- the Board does not generally favor borrowing more than they otherwise card solicitations conducted by a major laws that restrict a creditor’s discretion would and declaring bankruptcy even U.S. issuer of credit cards, offers that to determine which accounts or though they could eventually repay their included higher interest rates and fees transactions it deems as economically debts — or, alternatively, whether they tended to attract riskier borrowers with viable. The Board’s view is that if declare bankruptcy when they face an higher delinquency, charge-off, and creditors are terminating accounts of unexpected hardship that makes it bankruptcy rates than offers with better customers who use their credit cards impossible for them to repay their debts. terms. In other words, issuers face a only for transactions purposes, they are There are four main issues. so-called adverse selection problem.10 doing so because they consider these First, I’ll discuss empirical evidence accounts unprofitable. on the source of the recent rise in bankruptcies and the extent to which (8) Credit Counseling. 10 Adverse selection in the credit-card market Provisions are included to address market forces place limitations on the has also been documented in my paper with the issue of unsophisticated borrowers. number of bankruptcies. If these forces Paul Calem. The fact that credit-card rates are very sticky and don’t tend to come down These provisions are intended to ensure are effective, reforms are less needed. when other rates do is partly attributable to that the debtor made a good-faith effort Similarly, I’ll discuss empirical this adverse selection problem.

38 Q1 2002 Business Review www.phil.frb.org Ausubel places some of the blame for The relationship between was removed from the report, the more higher bankruptcies on the creditors debt levels and bankruptcies is more creditworthy past filers initiated new who have issued the debt. He favors complicated than the studies by credit relationships, especially high- the approach of earlier proposed Ausubel and Stavins might suggest. limit credit cards, at a much faster rate legislation that would restrict the There is no doubt a strong correlation than normal — evidence that the flag claims of lenders who caused a debtor’s between debt burdens (measured by was a constraint on their getting ratio of unsecured debt to income to the debt-to-income ratio, debt-to- credit.11 Nevertheless, some of the exceed 40 percent. He also favors a assets ratio, or debt-service burden) negative effects from filing may have time priority in bankruptcy: unsecured and number of bankruptcies, since a declined over time. Filing for lenders would be repaid in the order higher debt burden means a negative bankruptcy may be more accepted in which they lent, with the earliest shock can have a more severe effect these days, since it has become more lender repaid first and the latest on a household. However, we also common. There are other costs lender repaid last, giving the later know that in the aggregate, debt and associated with filing that may have lenders more incentive to monitor debt-to-assets seem to increase after fallen as the number of filings has the borrower’s credit position. households see their incomes rise and risen. For example, it is easier to find Joanna Stavins reviewed expect their future incomes to rise. information on how to file (the forms studies that have shown riskier borrowers Debt seems to facilitate growth rather and the information are readily have gotten better access to credit- than inhibit it. Debt levels have available on the Internet); more card loans over time, noting that the been expanding not only because of people have experienced bankruptcy, rise in credit-card borrowing in the aggressive lending but also because of so there are more people who can give mid-1990s has coincided with the increase in bankruptcy filings. Using data from the Terms of Credit Card Another factor that may have contributed to Plans, a survey of about 200 of the the increase in the bankruptcy rate is the largest bank credit-card issuers conducted twice a year by the Federal decreased social stigma associated with Reserve Board, Stavins provided declaring bankruptcy. empirical evidence that credit-card issuers that offer higher rates and the expanding economy and because advice; and there are more bankruptcy fees in order to compensate for higher technological advances have allowed lawyers competing for business. risk do tend to experience higher creditors to offer loans to more In an interesting study, David delinquency rates (measured by the borrowers at a lower cost (see my Gross and Nicholas Souleles assembled fraction of outstanding credit-card 1997 article on credit scoring). a panel of over 25,000 individual loans 60 days or more overdue), a (2) Stigma. Another factor credit-card accounts, chosen to be finding similar to Ausubel’s. But unlike that may have contributed to the representative of all open accounts in Ausubel, Stavins found that these increase in the bankruptcy rate is June 1995. They studied the behavior issuers did not seem to have higher the decreased social stigma associated of these accounts for the next 24 charge-off rates (the fraction of with declaring bankruptcy. Certainly, months or until they first defaulted outstanding credit card loans that bankruptcy continues to have a are written off). Indeed, her empirical negative connotation. It can harm a results showed that banks that charged person’s reputation, and it can make it 11 Similarly, Stavins presented some data from higher rates and fees earned higher net more difficult to gain access to credit the Survey of Consumer Finances for 1998 income from credit cards than banks in the future. Federal law allows indicating that among those who have ever filed that charged lower fees. This implies credit bureaus to continue to report for bankruptcy (8.51 percent of respondents), the average level of credit-card debt is largest that at least over the period covered a bankruptcy filing in the person’s for those who filed nine or more years ago. But (1990-99), when the economy was in credit report for up to 10 years after she also found that the average credit-card debt for someone who filed one or two years ago was good shape, it was profitable for issuers the filing. A study by David Musto higher than for those who filed three to nine to extend credit to riskier borrowers. showed that this does restrict the years ago. Stavins posited that this might be Whether that would continue to be person’s access to credit. His work because once someone files under Chapter 7, he or she cannot file again for six years, so true in an economic downturn remains using credit-file data from 1994-97 issuers might feel relatively safe lending in the to be seen. showed that when the bankruptcy flag initial period after a filing. www.phil.frb.org Business Review Q1 2002 39 or were closed in good standing, in is consistent with the stigma nationally representative sample, an attempt to see whether the recent hypothesis, that is, that a decline in and a review by the General increase in bankruptcies is better the cost of declaring bankruptcy — Accounting Office disputed some explained by supply effects or demand either the social, legal, or information- of the study’s findings on a number effects. That is, did lenders increase gathering costs — is largely responsible of methodological grounds. For one the supply of credit to less credit- for the increased level of filings.13 worthy borrowers, who account for (3) Abuse. A basic premise the increase in bankruptcy filings? of bankruptcy reform legislation is that Increases in credit Or, even after researchers control for the current system is being abused by limits and other changes creditworthiness, have people become people who declare bankruptcy when more willing to default over time? they can still afford to repay their in risk-composition Has their demand for bankruptcy debts. Here the evidence is very explain only a small increased? According to Gross and mixed, and the results depend on Souleles’ estimates, riskier borrowers the various assumptions made about part of the significant — for example, those with lower credit the type of means test that would be increase in default scores, larger credit card balances, and enacted and the way different types of rates between 1995 smaller monthly payments — are debt would be handled. A 1997 study much more likely to default. Default by John Barron and Michael Staten and 1997. rates were also higher for people living published by the Credit Research in states where unemployment was Center at Georgetown University thing, the study used the information higher, house prices were lower, and estimated that if all secured debt debtors provided at the time of filing fewer residents had health insurance.12 was reaffirmed, about 32 percent about their income, expenses, and The authors documented that of Chapter 7 debtors in their sample debts without verification and assumed there was an increase in credit to riskier could repay about 31 percent of their that the filer’s ratio of income to borrowers. But increases in credit limits nonhousing, nonpriority debts. This expenses remained constant over and other changes in risk-composition would be an average payment per filing the five-year repayment period. explain only a small part of the of $3570 over five years. The study Visa and MasterCard have significant increase in default rates was based on a sample of 3798 families funded several studies, including three between 1995 and 1997. They found who filed for bankruptcy in 13 major by Ernst &Young (by Tom Neubig and that all accounts, even those with the U.S. cities during the spring and co-authors). The latest of these same risk characteristics, age, and other summer of 1996. It was not a studies, published in March 1999 and economic fundamentals, became more based on a nationally representative likely to default over the sample period. sample of 1997 filings, estimated that And this increase in the probability of 10 percent of Chapter 7 debtors would 13 There is a debate in the legal literature bankruptcy — about 0.06 percentage about whether an economic modeling be affected by a means test for ability point per month between the start of approach or a sociological approach is to pay either 25 percent or more of the appropriate methodology for studying the sample period in June 1995 to its bankruptcy decisions, especially when stigma unsecured debts or $5000 over five end in June 1997 — is comparable to is the factor being investigated (see Michelle years. This would yield $3 billion in that which would occur if the credit White’s 1997 article for an interesting debt recovery over five years. But discussion). The economic modeling approach, score of every account in the sample favored, for example, by White (1997), assumes these results assumed that the debtors were reduced by one standard deviation, that consumers act to maximize their welfare remained in payment plans for the and, therefore, may act strategically when it which would be a very large increase comes to filing for bankruptcy, as we’ll discuss full five years and that their incomes in the overall riskiness of the sample. below. The sociological approach, favored, for rose as fast as their expenses and debt While not conclusive, this evidence example, by Teresa Sullivan, Elizabeth Warren, (Report 106-49, Senate, p. 88). and Jay Westbrook (1989) and Rafael Efrat (1998), assumes that people file for bankruptcy Marianne Culhane and when their financial problems become severe Michaela White got significantly enough that they can no longer handle their debt; it is not something they anticipate or different results from the Ernst & Young 12 Gross and Souleles also documented plan for, and they do not act strategically studies. Their results were based on a a seasoning effect: They found that the regarding filing for bankruptcy. My training different sample of bankruptcy filings probability of delinquency rises from the puts me in the economic modeling camp; time the account is opened until it is about hence, the studies I review here largely (from 1995) and different assumptions two years old, then the probability falls. follow that approach. about, among other things, how long

40 Q1 2002 Business Review www.phil.frb.org debtors remain in their payment plans Wenli Li also developed a a home in New Jersey but an unlimited and debtors’ automobile expenses. general equilibrium model of bank- exemption in Texas. So there is an They estimated that 3.6 percent of ruptcy chapter choice and showed that economic incentive to move to a Chapter 7 debtors could afford to individuals with fewer assets but higher state with a higher exemption before repay some of their debts, with total income were more likely to choose declaring bankruptcy. recoveries of $450 million over five Chapter 7, while those with more In their provocative 1999 years. Because of their different sample, assets and lower income were more study, Ronel Elul and Narayanan even when Culhane and White likely to choose Chapter 13 and they Subramanian, using data from the changed their assumptions to those work less.15 The author’s theoretical Panel Study of Income Dynamics used by Ernst &Young, they still analysis of proposed reforms indicated (PSID), estimated that about 3 percent estimated significantly lower recoveries they will affect borrowers differently, of all moves to states with higher — $930 million — than Ernst & Young. depending on their level of wealth and exemptions are driven by bankruptcy One of the drawbacks of income. For example, a means test that considerations. This percentage doubles many of these types of studies is that shifts filers into Chapter 13 would hurt for moves made by households “at they assume lender behavior would the ones with few assets and medium risk” for bankruptcy (that is, with remain the same after bankruptcy income levels. Anticipating this, those an estimated probability of filing for reform. But if lenders lend even more filers with a higher probability of bankruptcy equal to the average filer). aggressively, the number (and cost) filing for Chapter 13 will reduce their Using data from the University of bankruptcies might increase after borrowing but also not work as hard. of Michigan’s Panel Study on Income reform. The studies also assume that This has important implications for Dynamics (PSID) for 1984-1995, Scott borrower behavior would remain the trying to assess the economic benefits Fay, Erik Hurst, and Michelle White same, but models suggest that this of bankruptcy reform or the amount found evidence that borrowers respond need not be the case.14 of debt that borrowers would be able to the incentives to file for bankruptcy. For example, one potential to repay in Chapter 13. They measured these benefits as the drawback of the means test as proposed in the legislation is that shifting A corollary of the premise that people are debtors with incomes above a certain threshold from Chapter 7 to Chapter 13 abusing the bankruptcy system by filing when essentially imposes a high tax on future they can repay is that people act strategically earnings, since Chapter 13 requires debtors to use some portion of their when filing for bankruptcy. future earnings to repay their debt. Michelle White’s 1999 study pointed Strategic behavior. A corollary amount of debt that is dischargeable out that this sets up perverse incentives, of the premise that people are abusing under bankruptcy less any assets over reducing debtors’ willingness to work the bankruptcy system by filing when the exemption level, which would and even giving them an incentive to they can repay is that people act have to be given up under bankruptcy. quit their jobs to avoid the tax. She strategically when filing for bankruptcy. The authors found that for each $1000 and Hung-Jen Wang have proposed But the empirical evidence on just how increase in benefits, the probability of combining Chapters 7 and 13 so strategically people are behaving when a household’s filing rises, on average, that debtors filing for bankruptcy they file is mixed. “Forum shopping” is by 0.021 percentage point, which would would have to use their assets and one strategy. The exemption levels for imply a 7 percent increase in filings per their future earnings, after certain personal bankruptcy vary widely across year.16 To see what this effect means, exemptions, to repay their debts. Their states. For example, as discussed earlier, simulations suggest that this system a single filer receives no exemption for would not have deleterious effects on 16 debtors’ incentives to work. One drawback of the study is that only 254 households included in the PSID had filed for bankruptcy. The rate of filings in the PSID 15 In an empirical study, Ian Domowitz and was only half of the national rate, suggesting Robert Sartain found that Chapter 13 filers that PSID households underreported their 14 In addition, these studies do not account more often tend to be married and employed bankruptcy filings. See Fay, Hurst, and White for the administrative expenses of imposing a and have higher income and higher equity-to- and the CBO study for further discussion of means test. debt ratios than Chapter 7 filers. this point. www.phil.frb.org Business Review Q1 2002 41 consider that there were about the fact that sometimes creditors do SUMMARY 580,000 personal filings in the year not take legal action against borrowers Congress continues to try ended June 1989, about the middle who default, so borrowers have less to pass legislation to reform the point of the period covered in the incentive to file for bankruptcy bankruptcy system. While the rate of study. A 7 percent increase in filings protection, and that borrowers may bankruptcy filings has risen over the would have meant 40,000 more filings want to preserve the option to file in past several years, the reason for that in 1989. If the size of the effect were the future, so they refrain from filing increase is still debatable and that the same in 2001, an increase of $1000 immediately. Another possibility is that means the rationale for reform is in benefits would have added 94,000 there is indeed stigma associated with debatable too. Some proponents of filings to the 1.35 million personal filing, which deters households from reform argue that people with the filings in 2001. declaring bankruptcy. wherewithal to repay their debts are Culhane and White also (4) Efficacy of Repayment taking advantage of the system and studied strategies and found that Plans. The reform proposals assume that significant cost savings would be sophisticated debtors could avoid that Chapter 13 repayment plans forthcoming from reform. Others being classified as having the ability to work. But a 1994 study by the argue that the real reason bankruptcies repay under a means test by taking on Administrative Office of the U.S. have increased is that the level of debt more debt or increasing charitable Courts found that 36 percent of has increased, perhaps because lenders contributions. Note that if such strategic debtors who voluntarily entered have encouraged risky borrowers to behavior is the rule, then estimates of Chapter 13 repayment plans between take on excess levels of debt. cost savings under means testing that 1980 and 1988 completed their plans. The empirical work on do not account for these reactions will The study did not indicate why 64 the causes and incentives to file for be overstated. percent of the plans failed. Only 14 bankruptcy and whether the proposed Other studies reviewed by percent of all Chapter 13 cases were bankruptcy reform will have the Michelle White (1998b) did not find converted to Chapter 7. This finding desired effect is decidedly mixed. that personal bankruptcy rates are on the efficacy of repayment plans While this means that proponents on significantly related to the incentives is important and contrary to the each side in the debate can find the to file.17 Indeed, the real conundrum assumption in many of the studies that ammunition they need by choosing the might not be why the bankruptcy rate found that bankruptcy reform would right study, it also means that more increased so much in the 1990s, but lead to significant cost savings, since research is necessary in order to fully why it didn’t increase more18 and why the studies assume that debtors remain understand the bankruptcy more people haven’t moved to Texas in the repayment plans the full five phenomenon. BR and Florida where the homestead years. exemption is unlimited. Two reasons suggested by White (1998a) include

17 White (1998b) and “Notes” review the literature on whether debtors behave strategically regarding bankruptcy decisions. See also the CBO study.

18 Fay, Hurst, and White found that about 18 percent of households would have benefitted from filing for bankruptcy over the 1984-94 period, and White (1998a) found that 15 percent of households would have benefitted from filing in 1992. Yet, on average, fewer than 1 percent of households filed over these periods.

42 Q1 2002 Business Review www.phil.frb.org REFERENCES

Ausubel, Lawrence. “Testimony Before Domowitz, Ian, and Robert L. Sartain. Mester, Loretta J. “What’s the Point of the Subcommittee on Commercial and “Determinants of the Consumer Bankruptcy Credit Scoring?” Federal Reserve Bank of Administrative Law of the Committee on Decision,” Journal of Finance 54 (February Philadelphia Business Review (September/ the Judiciary of the U.S. House of Represen- 1999), pp. 403-20. October 1997), pp. 3-16. tative,” hearing on Consumer Bankruptcy Issues, March 10, 1998. Efrat, Rafael. “The Moral Appeal of Musto, David K. “The Reacquisition Personal Bankruptcy,” Whittier Law Review, of Credit Following Chapter 7 Personal Ausubel, Lawrence. “Adverse Selection in 20 (1998), pp. 141-67. Bankruptcy,” Wharton Financial Institu- the Credit Card Market,” Working Paper, tions Center Working Paper 99-22, University of Maryland, 1999. Elul, Ronel, and Narayanan Subramanian. June 1999. “Forum-Shopping and Personal Bank- Ausubel, Lawrence. “Personal Bankruptcies ruptcy,” manuscript, Wharton School, Nelson, Jon P. “Consumer Bankruptcies Begin Sharp Decline: Millennium Data University of Pennsylvania, July 1999. and the Bankruptcy Reform Act: A Time- Update,” manuscript, University of Series Intervention Analysis, 1960-1997,” Maryland, January 18, 2000. Fay, Scott, Erik Hurst, and Michelle J. Journal of Financial Services Research White. “The Household Bankruptcy (September 2000). Barron, John M., and Michael E. Staten. Decision,” American Economic Review, “Personal Bankruptcy: A Report on forthcoming. Neubig, Tom, Gautam Jaggi, and Robin Petitioners’ Ability to Pay,” Credit Research Lee. “Chapter 7 Bankruptcy Petitioners’ Center Monograph #33, Georgetown Gross, David B., and Nicholas S. Souleles. Repayment Ability Under H.R. 833: University, October 1997. “An Empirical Analysis of Personal The National Perspective,” Ernst & Bankruptcy and Delinquency,” Review Young, March 1999, Calem, Paul, and Loretta J. Mester. of Financial Studies, 15, 2002, pp. 319-47. www.ey.com/global/vault.nsf/US/ “Consumer Behavior and the Stickiness Chapter_7_Bankruptcy_Petitioners_ of Credit Card Interest Rates,” American Li, Wenli. “To Forgive or Not to Forgive: Repayment_Ability_Under_H.R._833:_ Economic Review, 85 (December 1995), An Analysis of U.S. Consumer Bank- the_National_Perspective/$file/ pp. 1327-36. ruptcy Choices,” Federal Reserve Bank Report99.pdf. of Richmond Economic Quarterly, 87 Congressional Budget Office (CBO). (Spring 2001), pp. 1-22. Neubig, Tom, Fritz Scheuren, Gautam Jaggi, “Personal Bankruptcy: A Literature and Robin Lee. “Chapter 7 Bankruptcy Review,” September 2000. Mecham, Leonidas Ralph. Bankruptcy Petitioners’ Ability to Repay: Additional Basics. Administrative Office of the United Evidence from Bankruptcy Petition Files,” Culhane, Marianne B., and Michaela M. States Courts, revised second edition, Ernst & Young, February 1998, White. “Taking the New Consumer June 2000. www.ey.com/global/vault.nsf/US/ Bankruptcy Model for a Test Drive: Means- Chapter_7_Bankruptcy_Petitioners_ Testing Real Chapter 7 Debtors,” American Ability_to_Repay:_Additional_ Bankruptcy Institute Law Review (1999), Evidence_from_Bankruptcy_Petition_Files/ pp. 27-77. $file/feb98_report.pdf. Continued on next page

www.phil.frb.org Business Review Q1 2002 43 REFERENCES

Continued from previous page Report 107-3, Part 1, of the Committee White, Michelle J. “Economic Versus on the Judiciary, House of Representatives Sociological Approaches to Legal Re- Neubig, Tom, Fritz Scheuren, Gautam 107th Congress, 1st Session, To Accom- search: The Case of Bankruptcy,” Law Jaggi, and Robin Lee. “Chapter 7 Bank- pany H.R. 333, Bankruptcy Abuse and Society Review, 25 (1997), pp. 685-709. ruptcy Petitioners’ Ability to Repay: Prevention and Consumer Protection Act The National Perspective, 1997,” of 2001, Together with Dissenting Views, White, Michelle J. “Why Don’t More Ernst & Young, March 1998, February 26, 2001. Households File for Bankruptcy?” Journal www.ey.com/global/vault.nsf/US/ of Law, Economics, and Organization 14 Chapter_7_Bankruptcy_Petitioners_ Stavins, Joanna. “Credit Card Borrowing, (1998a), pp. 205-31. Ability_to_Repay:_the_ National_ Delinquency, and Personal Bankruptcy,” Perspective,_1997/$file/Report97Fin.pdf. New England Economic Review (July/ White, Michelle J. “Why It Pays to File August 2000), pp. 15-30. for Bankruptcy: A Critical Look at “Notes: A Reformed Model of Consumer Incentives Under U.S. Bankruptcy Laws Bankruptcy,” Harvard Law Review, 109 Sullivan, Teresa A., Elizabeth Warren, and a Proposal for Change,” University of (April 1996), pp. 1338-56. and Jay L. Westbrook. As We Forgive Our Chicago Law Review, 65 (Summer 1998b), Debtors: Bankruptcy and Consumer Credit pp. 685-732. Pomykala, Joseph. “Bankruptcy Reform,” in America. New York: Oxford University Regulation (Fall 1997), pp. 41-78. Press, 1989. White, Michelle J. “Viewpoints: Too Many Incentives to Go Bankrupt,” American Report 106-49 of the Committee on Understanding the Federal Courts. Banker, November 19, 1999. the Judiciary, Senate, 106th Congress, The Federal Judiciary, 1999, 1st Session, To Accompany S. 625, www.uscourts.gov/UFC99.pdf. Bankruptcy Reform Act of 1999, Together with Additional and Minority Views, Wang, Hung-Jen, and Michelle J. White. May 11, 1999. “An Optimal Personal Bankruptcy Procedure and Proposed Reforms,” Journal of Legal Studies 29 (January 2000), pp. 255-86.

44 Q1 2002 Business Review www.phil.frb.org