A tale of two debtors: bankruptcy disparities by race.
Legal policy has long struggled with the issue of official neutrality in the face of racially disparate results. While the days of laws explicitly discriminating against people of color may be gone, the legal system as a whole has not attained perfect race neutrality. Scholars have offered evidence of this tension in disparate spheres such as criminal justice, employment, and education. This paper adds to that history by offering evidence of racial difference in the court system in an area in which such differences had not been posited before: bankruptcy filings. Such an addition to the debate is particularly timely given the current credit turmoil and heightened prominence of bankruptcy as a societal actor.
When it amended the Bankruptcy Code ("the Code") in 2005, Congress sought to curb perceived debtor "abuse" of bankruptcy laws by pushing more debtors out of Chapter 7 and into Chapter 13. (1) The amendments thus deny some debtors Chapter 7's immediate and almost automatic (2) cancellation of debts, and instead thrust them into a Chapter 13 that requires the debtor to make exacting payments to creditors over a period of up to five years. (3) In so doing, Congress may have exacerbated racial disparity in bankruptcy relief.
The data from this paper suggest that minorities who enter bankruptcy are far less likely than whites to receive a bankruptcy discharge. Part of this is simply because of the choice that debtors make. Black debtors, for example, are three times more likely to choose Chapter 13 than are white debtors. (4) Because the overall relief rate was only 23% for Chapter 13, (5) this means that blacks are disproportionately denied relief based on the bankruptcy chapter they choose.
More worrisome is that the empirical data in this paper suggest that once minorities enter Chapter 13, they obtain bankruptcy relief far less often than do whites--the odds of a discharge are 40% lower for black or Hispanic debtors as compared to white ones, even after controlling for income, education, and employment. (6) In other words, Congress's recent amendments (7) have made it so that some minority debtors will no longer have the option of an immediate Chapter 7 discharge in which all races fare the same, (8) and must instead enter a long-term payment Chapter 13 in which their race may be a determining factor in whether they ever get a successful discharge.
A numbers-based discussion of minority debtors' likelihood of relief is new to bankruptcy scholarship, and fills in the middle part of the three-part story of race in bankruptcy law. (9) Scholars have already shown that black and Hispanic families are far more likely to enter bankruptcy than are white families. (10) At least one critical factor in this seems to be predatory lending practices: even residents in high-income, predominately black neighborhoods are more than twice as likely to get subprime mortgages as are residents in low-income white neighborhoods. (11) Scholars have also posited that the type of relief offered by bankruptcy laws favors white debtors over black debtors, since whites disproportionately own the type of assets that bankruptcy protects, and blacks disproportionately have the types of debts that bankruptcy does not relieve. (12) This would leave minority debtors who obtain relief worse off than white debtors who obtain relief. Thus, the literature offers a picture of different races before and after bankruptcy. The data presented in this paper begin to tell the story of what happens to minority debtors while they are in bankruptcy--which chapter they choose and what happens to them while they are pursuing a discharge of their debts in Chapter 13.
This paper thus informs the relationship between bankruptcy and race and, as such, fleshes out some larger issues surrounding race and the law. Until now, that debate lacked empirical information about what happens to different races once in bankruptcy. It also lacked any clear assertion that race played a role in whether a debtor received a discharge. Indeed, much of the criticism of "raced" bankruptcy laws seemed understandably premised on the assumption of equal availability of a discharge. (13) The data offered in this paper refute that assumption.
II. METHODOLOGY AND CORE FINDINGS
The data for this paper comes from the 2001 Consumer Bankruptcy Project, whose final completion rates for Chapter 13 cases only recently became available because a significant portion of the multi-year Chapter 13 repayment plans were not completed until 2006 and 2007.
The Chapter 13 database used in this paper has a core sample of 978 debtors, while the Chapter 7 database has 801 debtors. (14) The information about these debtors came from two principal information sources. First, debtors received questionnaires at their mandatory meetings with creditors. (15) These questionnaires asked for demographic information such as race, level of education, and whether the debtor owned a home and was employed. (16) Then, coders collected data from the corresponding public court records. (17) These records supplied additional information about income, bankruptcy outcome, and motions made against the debtor in court.
Overall, 69.1% of blacks who entered bankruptcy in 2001 eventually obtained a discharge, compared to 87.5% of whites. (18) Part of this is explained by the fact that blacks chose Chapter 13 with much higher frequency than did whites and Hispanics: of those debtors choosing Chapter 7 and Chapter 13 bankruptcy, 61.8% of blacks chose Chapter 13, compared to 29.4% of Hispanics and 20.5% of whites. (19)
Yet an analysis of this data also revealed a discrepancy among debtors only entering Chapter 13. Significantly fewer blacks and Hispanics who entered Chapter 13 left with a successful discharge of debts, even controlling for the influence of income, education, and employment. Whereas, 28.3% of all whites entering bankruptcy obtained a discharge, only 19.8% of blacks and 19.4% of Hispanics did. (20) In other words, merely being black lowers the odds of getting a discharge by 40%, and being Hispanic lowers the odds by 43%. (21)
[FIGURE 1 OMITTED]
Even after a judge approves the debtor's proposed repayment plan, and thus gives the court's seal of approval for feasibility, the disparity still exists between blacks and whites. Of whites who had their plans confirmed, 39% wound up receiving a discharge, compared to 28.6% of blacks, again even after controlling for income, homeownership, education, and employment. (22) Hispanics who had their plan confirmed had a completion rate of 27.7%. (23)
There are a number of possible contributors to the difference in completion rates. The discussion below focuses on the difference in completion rates within Chapter 13, rather than the choice between chapters, because the Consumer Bankruptcy Database offered a broader explanation for the Chapter 13 differences.
As far as the differences in completion rates within Chapter 13 are concerned, data and common-sense point to three principal explanatory avenues. The first is that actors in the bankruptcy system--judges, attorneys, or otherwise--are treating whites and minorities differently. A second potential explanation is a racial bias built into the Chapter 13 laws. Finally, there is the possibility that social factors, independent of debtors' socioeconomic status, offer less support to minorities, such as through inadequate representation.
A. Bankruptcy Actors
One explanation for the disparate completion rates is that actors in the system--such as trustees, judges, or attorneys--are treating minorities differently. This treatment is difficult to gauge because we cannot know with certainty what is going on in the minds of these actors. Thus, the discussion below is limited to those actors for whom the numbers demonstrated a difference in involvement: trustees and attorneys. (24)
One way treatment of debtors can be measured is through the number of adversarial moves made against them by other parties. In a bankruptcy case, the strongest adversarial move to make is a motion to dismiss, which can be made by the debtor, creditors, or the trustee at any time during the bankruptcy proceedings. (25) The data on motions to dismiss support the assertion that some of the race differences in completion rates might be explained by differences in the number of motions to dismiss made by third parties. (26) The principal party making those motions was the trustee.
A significantly higher percentage of discharged black debtors (30.3%) than white debtors (17.8%) (27) were subjected to motions to dismiss. (28) This means that on average, even successful black debtors had 12.5% more motions to dismiss made against them. Although the specific details of the motions are unknown, the difference is striking when one considers the comparable socioeconomic demographics of the groups and that such motions were made against ultimately successful debtors. Not only should these racial groups theoretically have had comparable motions to dismiss made against them, but it would also appear that any such motions to dismiss were ultimately off base because the debtors completed the plans.
Nor can these motions to dismiss be defended on the ground that they were made to encourage the debtor to make payments. The above motions to dismiss were motions that the court actually ruled on. Moving parties can--and often do--withdraw the motions to dismiss on the day of the hearing if their intent is to prod the debtor into payment. The fact that judges actually ruled on these motions, however, suggests the trustees made the motions with intent to remove the debtor from Chapter 13.
Because motions to dismiss can be made by many parties--mortgage holders, credit card companies, or the trustee (29)--it helps to take a closer look at who made these motions in the data set. The motions were spread out enough among different types of creditors so that no single category of creditors had enough of a sample size to draw any conclusion. One party had, however, made enough motions to dismiss to provide a significant enough sample size: trustees. Trustees made at least one motion to dismiss against 22.2% of discharged black debtors, but against only 11.1% of white debtors. (30) Trustees thus were twice as likely to make a motion to dismiss against a black debtor who ultimately completed her plan than against a similarly-positioned white debtor.
[FIGURE 2 OMITTED]
Considering the amount of discretion a trustee has, one reading of the above data is that trustees handle black debtors more aggressively than they do white debtors. Trustees' responsibilities in Chapter 13 create many opportunities to influence the outcome of cases. The Chapter 13 trustee recommends approval or denial of the plan, and then is charged with assuring that the debtor relinquishes the required amount of income during the repayment period. (31) If the debtor falls behind, the trustee will usually take steps such as asking the court for a wage garnishment or making a motion to dismiss. (32)
Bankruptcy courts and the Code give trustees a great amount of discretion in decisions such as whether or not a motion to dismiss should be filed. Case law has established that "the United States trustee has complete discretion in determining whether and how to act to advance the goals of case management." (33) The trustee also has "prosecutorial discretion" not to file the motion to dismiss. (34) The Code's language contributes to the trustee's discretion by saying that the trustee "shall ... if advisable, oppose the discharge of the debtor." (35) Yet what is meant by "advisable" is unclear.
Other explanations for these data are less convincing. The strongest alternative explanation is that independent factors caused black debtors to miss payments more frequently than white debtors. These missed payments, in turn, caused the trustees to move to dismiss. Missed payments could have been caused by external shocks to the debtors' lives, such as loss of employment--one could postulate that such events happen more often to blacks than to whites due to social prejudice or some other reason.
Alternatively, the aforementioned treatment of expenses by some courts--by not allowing for support of nonlegal dependents--could cause blacks to miss payments more often. That is, perhaps a greater percentage of black debtors were supporting nonlegal dependents, thus making their actual expenses higher than both their allowed expenses and whites' actual expenses. If that treatment caused even ultimately successful black debtors to have greater hardships during their repayment plans, it could have led to subsequent motions to dismiss.
Yet these hypotheses of trustees' disparate motion rates are unlikely for several reasons. First, debtors and trustees can account for external shocks on the debtor's life, such as a job loss or demotion, by making a motion to modify the repayment plan. (36) If external turbulence happened more frequently in discharged black debtors' lives, then one would expect to see more motions to modify made by black debtors seeking breathing room. The data do not, however, support this assertion. Among discharged debtors, there was no difference between the races in the frequency of motions to modify, although the numbers are small so this fact is merely suggestive. (37)
Moreover, if a debtor were making late payments, there would normally be some communication between the debtor and trustee, as the trustee appraises the situation. If the debtor were struggling to make the payments because of a need to support non-dependent family members, a reasonable trustee would probably consider this situation before attempting to jettison the debtor's chances of a fresh start. Combined with the fact that these debtors had comparable socioeconomic demographics and ultimately completed their plans, the information about motions to modify makes it highly unlikely that differences in frequencies of external shocks on white and black debtors' lives explains the difference in motions to dismiss.
There also does not appear to be any one geographic area accounting for the difference in trustee motions to dismiss. The trustees' motions to dismiss were spread throughout four of the five states studied in the data set. (38) While a larger data set would be necessary to study the full effects of geography on trustees' rates of motions to dismiss, the data in the current study suggest the disparity in trustee motions is not limited to any one geographic area.
[FIGURE 3 OMITTED]
Although the data are consistent with trustee bias, any such conclusion would be premature and should await further studies on trustees and race in bankruptcy. The data are also consistent with other interpretations, such as blacks living disproportionately in districts where the trustees are more likely to make motions to dismiss.
Attorneys wield considerable influence over the outcome of a case. When a debtor falls behind, lawyers can often work with the trustee to prevent a motion to dismiss the case. (39) And studies have consistently established that "the key actor in each consumer bankruptcy is the debtor's lawyer." (40) Lawyers principally contribute to the difference in the completion rates by whether they choose to represent and how well they represent minorities.
Lack of legal representation likely explains some of the disparity in completion rates. The numbers for Chapter 13 support the inference that black and Hispanic debtors were less often represented and that this lack of representation contributed to lower completion rates. Whereas 98.4% of dismissed white debtors (41) had attorneys, only 94.8% of blacks and 94% of Hispanics had attorneys. (42) These rates of representation differ starkly from successful debtors: every single one of the 207 black, Hispanic, and white debtors who obtained a discharge were represented by an attorney. (43)
[FIGURE 4 OMITTED]
A reasonable inference from these numbers is that a lack of representation contributed to the failure to obtain a discharge. Besides improving a debtor's chances during the bankruptcy proceedings, attorneys play a key role in deciding whether the debtor even files for bankruptcy in the first place. Although the debtor "has the ultimate right to decide, [because of] the complexity of the bankruptcy system, ... the lawyer has an overwhelming influence over the decisions of most clients to file or not and which chapter to choose." (44) Thus, it could be that some of the black and Hispanic debtors who filed and did not have their plans confirmed would have been wisely advised not to file.
Because the difference in representation is only about 4%, (45) this factor explains only part of the difference in completion rates. Even if only represented debtors are considered, white debtors still have a significantly higher completion rate (28.7%) compared to their black counterparts (20.6%). (46) Lack of attorney representation thus must be combined with other factors--such as quality of representation, trustee bias, or the structure of bankruptcy laws--to obtain a full picture of why black and Hispanic debtors fare worse in Chapter 13.
The other major way in which attorney representation could contribute to the disparate completion rates is through the quality of representation. Yet the data from the current study regarding quality of representation are inconclusive. Probably the most objective factor that would allow for a comparison of the quality level of the attorneys is their experience level. At least one governmental study has linked the experience level of attorneys to the chance of a debtor having her Chapter 13 plan confirmed and attaining a discharge. (47)
Although experience level was not known for the attorneys in the current study, one potential proxy for experience level is how much debtors paid for their attorney. The Model Rules of Professional Conduct requires attorneys to charge only reasonable fees, and one measure of reasonableness is attorney experience. (48) Interestingly enough, those black, Hispanic, and white debtors who were represented paid comparable amounts to their attorneys: blacks on average paid $1,495, whites paid $1,465, and Hispanics paid $1,446. (49) Thus, to the extent the price reflects the quality of the attorney, there is no support for an inference that the attorneys representing minority debtors are any less qualified than those representing whites. Alternatively, it could turn out that whites are receiving a higher quality of representation through more experienced attorneys, and that Hispanics and blacks are simply paying too much for their representation. But absent more data, there is no reason to suspect this.
B. Social Factors
A third potential source of causes of the disparity in completion rates--though the least convincing--is social factors. That is, perhaps something external to bankruptcy happens to blacks and Hispanics more frequently than to whites and makes it more difficult for them to complete their repayment plans. This explanation is tempting because the main avenue for dismissal of a case is through a motion to dismiss for failure to make the payments promised in the original plan. (50) Thus, one could argue that blacks and Hispanics simply fall behind more frequently on their payments. Both a statistical analysis of the data set and a legal analysis of Chapter 13, however, provide strong arguments why standard socioeconomic factors--income, education level, employment, and homeownership--do not explain the difference in completion rates.
1. Statistical Makeup of the Debtors
From a statistical standpoint, debtors entering Chapter 13 are mostly middle class Americans who have run into income problems. (51) Most own homes, have middle class jobs--as gauged by occupational prestige score--and have some college education. (52) Moreover, income and indices of class for Chapter 13 debtors are comparable for blacks, Hispanics, and whites. Most notably, the groups have essentially no difference in terms of income: whites ($2,273) have a slightly higher average monthly total income than Hispanics ($2,204) and blacks ($2,171), but these differences are not statistically significant. (53) Using the median income makes Hispanics the highest income group (higher than whites by about $70). Income by itself does not predict whether a debtor will be successful in Chapter 13.
Among debtors entering Chapter 13, homeownership was high: 92.5% of Hispanics, 89.6% of blacks, and 84.2% of whites were homeowners. (54) Education level followed a similar trend: out of a high of eight (advanced degree) and a low of one (no schooling), black debtors were the best-educated (5.5 average), followed by whites (5.2) and Hispanics (4.7). (55) This meant that the average black and white debtor had some college education, whereas the average Hispanic had a high school education. Hispanics, on the other hand, were the group most likely to be employed (90.3%), followed by whites (83.9%) and blacks (83.1%). (56)
These data suggest that there is no reason why socioeconomic factors would favor whites over blacks and Hispanics in Chapter 13. If anything, the data suggest that blacks and Hispanics had higher socioeconomic status, since they have comparable incomes but blacks have higher education levels and both Hispanics and blacks are more likely to own homes and have a job. Unsurprisingly, given the similarity of these numbers, none of the above factors contributed statistically to the difference in completion rates.
2. Bankruptcy's Level Field
Another reason why it would make sense that factors such as income would not lead to different completion rates is that Chapter 13 requires debtors to pay all "disposable income" to the repayment plan. (57) Disposable income is income that is not "reasonably necessary" to support the debtor or the debtor's dependants. (58) While there is plenty of leeway for what counts as reasonably necessary, there are also basic limits on what can and cannot be counted as expenses and a basic idea of what an average person would need to pay for support. The effect of comparing these limits to a debtor's income is that it doesn't matter whether the debtor earns $50,000 per month or $3,000 per month; the reasonably necessary expenses will take up a portion and then the rest will go toward plan payments. Consequently, each debtor would theoretically have the same amount of breathing room. (59) Chapter 13 plans thus, theoretically, level the playing field regardless of income, and this would explain why even debtors of different income levels should be expected to have similar completion rates.
C. Structure of the Laws
One possible explanation for the difference in completion rates is that something in the way Chapter 13 laws are drafted puts minorities at a disadvantage. Scholars have posited race bias in what would otherwise appear to be neutral laws--such as the Internal Revenue Code. (60) Mechele Dickerson has applied this line of thought to the Bankruptcy Code and found many elements of the Code to be problematic. (61)
The structure of Chapter 13 may disadvantage minority debtors because of the conflicting interpretations of "disposable income" among some bankruptcy courts. (62) Some courts refuse to exclude expenses for "nonlegal dependents" such as "parents, adult children, grandchildren, [and] domestic partners." (63) This interpretation could disproportionately hurt blacks and Hispanics more than other groups because "[t]he rearing or informal adoption of children by members of their extended family" is more prevalent than among other racial groups. (64) Thus, more Chapter 13 black debtors would have hidden costs not accounted for by the repayment plan. The debtor would then be faced with a difficult choice between providing for these informal children and meeting the expected payments. To the extent that this happens, it would be understandable why the debtor would choose to continue providing for nonlegal dependents and thus be unable to meet their Chapter 13 payments--essentially, the court would have overestimated the debtor's truly "disposable income." (65)
While raced bankruptcy laws is one viable explanation, more empirical support would be needed before drawing any conclusions about its contribution to completion rates. It may be true that blacks and Hispanics in general provide more support for nonlegal dependents, but that does not mean this is the case for Chapter 13 debtors, given that the socioeconomic makeup of Chapter 13 debtors is different in relevant ways from the population as a whole. For example, in 2005, the U.S. Census calculated the median income of blacks as $30,858, Hispanics as $35,967, and whites as $48,554. (66) Yet, as mentioned above, differences in incomes of Chapter 13 debtors were not statistically significant. Thus, before concluding that the laws themselves explain a significant portion of the difference in completion rates, it should be established that the above generalities about minorities in the studies on raced bankruptcy laws (that they support more nonlegal dependent relatives) hold true for Chapter 13 debtors.
D. Effects on Choice of Chapter
Although the data are more limited in explaining the racial differences in chapter choice, they do offer some further information relating to potential causes. One possible explanation is that minority homeowners place greater importance on holding onto their homes, and thus seek to enter Chapter 13 to do so. While 64% of white homeowners chose Chapter 13, 88% of black homeowners and 81% of Hispanic homeowners chose Chapter 13. (67)
Another possibility is that black debtors are not receiving competent legal representation, and thus might not be as informed about how difficult it actually is to obtain a discharge and hold onto one's home in Chapter 13. To do justice to this question, further studies would be needed to examine the competence of bankruptcy attorneys and related services or advertisements that would inform communities about bankruptcy. The data do reveal that black and Hispanic debtors choosing Chapter 7 are less likely to be represented by attorneys, at about 85%, compared to about 93% of white Chapter 7 debtors being represented by attorneys. (68) This discrepancy suggests that legal representation and the information that comes along with such representation could contribute to the difference in choice.
Another hypothesis would be that perhaps economic considerations help determine chapter choice. The numbers do not support such an inference; the differences in chapter choice are still significant independent of income level. (69) Nor do debtors of one economic subgroup, such as low-income debtors who own homes, account for the different rates of choosing Chapter 13. Although blacks and whites of higher incomes choose Chapter 13 more often, the differences between low-income blacks and low-income whites (and between high-income whites and high-income blacks) is the same as the difference between blacks and white debtors as a whole. (70)
There are three broad implications of these findings. First, they add a new sector of the court system to the longstanding debate about official legal neutrality in the face of disparate racial results. Second, they suggest that independent actors such as the U.S. trustee should have a modified role in which they weigh in on behalf of unrepresented debtors. Finally, they underscore the importance of shining more light on race in the area of bankruptcy, such as through annual reporting of racial data by district.
A. Official neutrality in the face of disparate racial results
The above results add to the ongoing tension between official neutrality and disparate racial results in the legal system. Bankruptcy laws, like their counterparts elsewhere in the legal system, are officially race-neutral, yet the above numbers paint a bleak picture of race in bankruptcy. Black, Hispanic, and white debtors enter bankruptcy with similar measures of class. They are, contrary to stereotypes, "overwhelmingly middle class." (71) Yet far more blacks choose a bankruptcy chapter that offers low probabilities of bankruptcy relief. (72) Then, even just looking at those that end up in Chapter 13, a far lower percentage of black and Hispanic debtors than white debtors leave with improved prospects of maintaining a middle-class lifestyle by holding onto their assets and freeing themselves from their debts. (73)
While the U.S. has moved beyond the days when public institutions banned people of color and laws explicitly discriminated against them, it has been clear that the replacement system has not achieved racially neutral results. The evidence of a previously unconsidered portion of the court system--bankruptcy filings--to the list of racially disparate results strengthens the argument for the breadth of the problem. These disparities parallel the larger concerns about race as a factor that affects who comes into the legal system and how that person is treated once they are there.
B. Discretion, Rules, and New Roles
The data also underscore the complications of the dichotomy between rules and discretion in the law. Discretion creates space for racially charged treatment. Yet rules can do the same even if written in an apparently race-neutral manner. (74) Thus, increasing or decreasing discretion may not solve the problem. Instead, independent actors such as the U.S. trustee should play a more active role in evaluating the quality of representation and weighing in on behalf of unrepresented debtors.
One potential way to alleviate actor bias against minorities--either if it were determined that there was prejudice by bankruptcy actors or if society wanted to guard against any potential prejudice--would be to include more concrete rules in Chapter 13. Bankruptcy operates on often fuzzy and ambiguous standards; for example, the laws do not define what "substantial abuse" means, which is a ground for dismissing a bankruptcy petition. (75) Such fuzzy rules could be seen as providing too much discretion for judges to dismiss petitions from "truly deserving debtors." (76) Bankruptcy scholars have postulated, without considering race, that "this lack of clarity increase[s] the risk that ... otherwise similarly situated debtors might receive disparate treatment." (77)
Rules, however, can also create opportunities for racial disparity. For example, the 2005 amendments to the Bankruptcy Code (78) made it so that educational expenses for dependent children are reasonable (and thus allowable in the debtor's budget) only if the child is less than eighteen years of age. (79) On its face, this change to the Code seems race-neutral, however, it likely makes it disproportionately more difficult for minorities to complete Chapter 13 repayment plans. Black and Hispanic families are more likely to have a child eighteen or older attending high school. (80) Consequently, they are more likely to need to make such expenses at some point during the repayment plan. Since these expenses will not be "allowable," the expenses will need to come out of disposable income that will be earmarked for creditors, thus increasing the possibility of missed payments and a motion to dismiss the case.
One improvement that avoids the complications of the discretion/rules tension would be to require independent actors to advocate on behalf of unrepresented debtors and to evaluate the quality of representation. For example, trustees could be required to exercise particular vigilance when the debtor is unrepresented. Trustees are already expected to assist the debtor in complying with the repayment plan, (81) and are "charged with objecting to improper creditor claims." (82) Yet the Code specifically says that trustees should not advise on legal matters. (83) While it would be problematic for trustees to give legal advice, in light of the above difference in frequency of motions to dismiss, in pro se cases, trustees should be charged with weighing in on behalf of debtors and investigating more thoroughly the debtors' ability to pay before making a motion to dismiss. When a party other than the trustee makes a motion to dismiss, the trustee should be required to make a statement stating the case for and against the motion, thus ensuring that at least one informed party will present the debtor's case to the court. (84)
C. Further Attention
More information about racial disparities and accompanying attention is a key step toward remedying the problem, not just in bankruptcy, but in other areas of the law. If more attorneys knew of the racial differences, they might put more resources toward pro bono representation or attorney education. (85) If more judges were aware of the differences, they might be quicker to conscript attorneys to help the unrepresented or to discipline poor attorneys. (86) If trustees knew of the differences, they might intervene on behalf of debtors more often.
One way to generate more information and thus attention about the issue would be to require districts to collect and annually report racial data. In bankruptcy, this would mean reporting the number of filings, which chapter those filings were in, how many motions of what kind were made, and each plan's outcome.
This data would make it possible to conduct further studies on key areas such as the causes of motions. For example, one helpful source of information would be how many days late the payments were before the trustee made the motion to dismiss. If the motions are for payments late by an average of forty days for whites and twenty-five days for blacks, it would suggest actor bias. If, on the other hand, the days are comparable, it would suggest independent social factors are playing a larger role.
Moreover, because a few trustees are involved in many cases, future studies should look at the role of individual trustees. Studies could examine whether some trustees consistently make motions with the same frequency regardless of the debtor's race, whereas other trustees consistently make more motions only for certain races. An investigation into certain trustees who consistently demonstrate more aggressive pursuit of minorities than whites would not only provide information, but could also spotlight any problematic actors in the bankruptcy machinery.
It would also be helpful to investigate why blacks and Hispanics are less likely to be represented, and if there are any differences in the quality of attorney representation between the races. One cause of the disparate percentages of representation could be that bankruptcy attorneys do not advertise to minority communities as often as they do to white communities. Advertising has been found to play a critical role, especially for certain segments of the population, in access to legal services. (87)
This paper has put forth data about outcomes for debtors of different races, and possible explanations for such data. It would be irresponsible to charge racism of any bankruptcy institution or actors at this point. What the above numbers do suggest, however, is that the problem of racial disparity in the court system extends further than was evidenced before. Furthermore, there is a great need to look more deeply into racial disparity in bankruptcy, especially in Chapter 13. Chapter 13 has long received less attention from scholars than Chapter 7 or Chapter 11 and is poorly understood even by many experts in the field. Minority debtors are far less likely to obtain a Chapter 13 discharge even when controlling for factors such as income, employment, homeownership, and education. BAPCPA will only exacerbate this problem by removing for many debtors the option of entering a clearly race-neutral Chapter 7, and by introducing arguably minority-unfriendly definitions of allowable expenses.
But there is also good news that comes with awareness of this particular danger. Greater awareness of the problem is the first step toward the attention necessary to remedy it. In bankruptcy, solving the problem would also serve a broad set of interests. Creditors would benefit from a system that does not reject what could otherwise be successful minority debtors from Chapter 13, since those debtors would otherwise pay back a larger portion of their debts over the repayment period. Judges, lawyers, and other bankruptcy actors would benefit from changes that would maintain the viability and integrity of the system. Finally, and most importantly, the changes would make a difference in the lives of many minorities by ensuring them equal opportunity for a fresh start when their lives are turned upside down by unpredictable events--an opportunity vital to their full participation in the risk-laden American economy.
It is also good news that there is no evidence that any of this is intentional; Congress, for example, surely wrote BAPCPA thinking it was race-neutral. Nor are the various actors--whether bankruptcy lawyers in their choice of advertising, or trustees in their motions--necessarily aware of the racially disparate impact of bankruptcy laws. Before the data from this study was available, those parties had an understandable excuse of a veil of darkness. Consequently, as this veil is lifted, there is reason to believe that decision-makers will take steps to ensure that the ladder out of indebtedness is not being unfairly knocked from underneath minority debtors.
What veils of darkness remain elsewhere in the legal system are unknown. But given the feasibility of collecting information on race in other areas of the law and the light that such information sheds on race in bankruptcy, there is reason to look for other areas of the law in need of spotlights. And just as the racially charged images following Hurricane Katrina raised awareness and directed resources toward lessening racial challenges facing communities such as New Orleans, such spotlights could similarly direct resources toward lessening racial differences in the legal system as a whole.
(1) See 151 CONG. REC. S1842 (daily ed. March 1, 2005) (statement of Sen. Hatch); U.S. Senate Republican Policy Committee, S. 256--The Bankruptcy Abuse Prevention and Consumer Protection Act of 2005, LEGISLATIVE NOTICE NO. 3, Feb. 28, 2005, at 1.
(2) Although bankruptcy judges have discretion to dismiss a Chapter 7 debtor for abuse, in practice this rarely happens. Less than 1% of Chapter 7 cases are dismissed. 2001 CONSUMER BANKRUPTCY PROJECT (on file with author) [hereinafter 2001 CONSUMER BANKRUPTCY PROJECT]. Additionally, there was no statistically significant difference among races in Chapter 7 dismissals. Id. The 2001 Consumer Bankruptcy Project is a large-scale longitudinal study led by professors at universities across the nation. The project was made possible through funding from the Ford Foundation, as well as grants from Harvard Law School and New York University Law School. The statistical calculations referenced in this article were calculated from data found in the 2001 Consumer Bankruptcy Project. These calculations were retained by the author and may be obtained from the author upon written request. For further discussion of the 2001 Consumer Bankruptcy Project see ELIZABETH WARREN & AMELIA WARREN TYAGI, THE TWO-INCOME TRAP (2004).
(3) See ELIZABETH WARREN & JAY LAWRENCE WESTBROOK, THE LAW OF DEBTORS AND CREDITORS 281-82, 326 (5th ed. 2006) (explaining that Chapter 13 bankruptcy requires debtors to make payments based on "projections of future income and living expenses," which often fail to reflect real life occurrences).
(4) P < .01. P is a measure of statistical significance. If P < .05 then it is considered statistically significant.
(5) This number comes from the percentage of all debtors who filed for Chapter 13 and eventually obtained a discharge. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(6) After controlling for income, education, homeownership, and employment, this odds ratio held statistical significance (P < .01 for blacks and P < .1 for Hispanics). See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(7) Bankruptcy Abuse Prevention and Consumer Protection Act of 2005 (BAPCPA), Pub. L. No. 109-8, 119 Stat. 23 (2005) (codified as amended in scattered sections of 11 U.S.C.).
(8) See 2001 CONSUMER BANKRUPTCY PROJECT. supra note 2.
(9) A. Mechele Dickerson, Race Matters in Bankruptcy, 61 WASH. & LEE L. REV. 1725, 1726-27 (2004) [hereinafter Dickerson, Race Matters].
(10) Elizabeth Warren, The Economics of Race: When Making It to the Middle Is Not Enough, 61 WASH. & LEE L. REV. 1777, 1779 (2004) [hereinafter Warren, Economics of Race].
(11) Id. at 1779, 1795. Professor Warren mentions several potential contributors to this greater vulnerability to financial collapse. Blacks and Hispanics may have "more pervasive job difficulties and more trouble financing medical care." Id. at 1779. The problem is also likely intertwined with the fact that racial minorities are "singled out for predatory loans and other subprime credit that drain billions of dollars out of the pockets of these families and push them into financial collapse." Id.
(12) See A. Mechele Dickerson, Race Matters in Bankruptcy Reform, 71 MO. L. REV. 919, 937, 956 (2006) [hereinafter Dickerson, Bankruptcy Reform]; Dickerson, Race Matters, supra note 8, at 1726.
(13) See Dickerson, Bankruptcy Reform, supra note 11, at 925, 937 (stating that "Everyone who files for bankruptcy generally receives some debt relief...." and "[t]here is ... no reason to believe that ... courts interpreted the pre-BAPCPA Code in a way that favored whites").
(14) See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(15) See Katherine Porter & Deborah Thorne, The Failure of Bankruptcy's Fresh Start, 92 CORNELL L. REV. 67, 81 (2006).
(16) See id. To view the questionnaire form used in the 2001 Consumer Bankruptcy Project see Elizabeth Warren, Bankrupt Children, 86 MINN. L. REV. 1003, 1028-32 (2002).
(17) See Porter & Thorne, supra note 13, at 81.
(18) P < .01 for the difference between blacks and whites, but data were not statistically significant for the difference between Hispanics and whites. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2. These numbers were obtained by normalizing to account for the reality that there were nearly three times as many people who entered Chapter 7 as Chapter 13 in 2001. See Ed Flynn et al., A Tale of Two Chapters: Financial Data, AM. BANKR. INST. J., Oct. 2002, at 20.
(19) See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(20) Using a Z test (a comparison of sample and population means to determine whether significant differences exist), this data was statistically very significant for blacks, with P = .0025, and was significant for Hispanics, with P = .042. Running a logistic regression model (a model used for prediction of the probability of occurrence of an event by fitting data into a logistic curve) in Stata (a type of statistical analysis software program) and removing biracial pairings, this data was statistically very significant for blacks, with P = .003, and was moderately significant for Hispanics, with P = .07. These significance levels held even after controlling for income, education level, homeownership, and employment. Although the sample size for Hispanics was still moderately significant upon running the logistic regression, its diminished p-value likely reflects the fact that the total sample size for Hispanics was small (93, compared to 318 whites and 500 blacks). That small size made the group of Hispanics more sensitive to adjusting for biracial attributions and controlling for socioeconomic factors. See id.
(21) The decreased odds of receiving a discharge were calculated in Stata using a logistic regression model, which yielded an odds ratio of .604 for blacks and .57 for Hispanics. See id.
(22) (P = .01). This significance level was obtained by running a logistic regression model in Stata. See id.
(23) (p = .1). Because of the smaller sample size for Hispanics, shrinking the sample size by eliminating those who had their plan dismissed prior to confirmation made the data for Hispanics only slightly significant. This significance level was obtained by running a logistic regression model in Stata. See id.
(24) No statistical significance was found for differential measures of judicial participation in the bankruptcy process, even though the data provided information on numbers, such as the percentage of motions granted. See id.
(25) See WARREN & WESTBROOK, supra note 3, at 282; 11 U.S.C. [section] 1307(c) (2006).
(26) Coders for the 2001 Consumer Bankruptcy Project did not count all motions made, but rather whether one of a given type of motion was made. Thus, if three motions to dismiss were made by trustees against a particular debtor, such would be coded as a motion to dismiss having been made by a trustee, but would not be counted as three separate motions. Also, only motions that were ruled on were included. Frequently, trustees or other parties made motions to dismiss and then, at the hearing, withdrew such motions because of something the judge or more likely the debtor said. It would be informative, in a future study, to record the withdrawn motions since these likely have a coercive effect on the debtors. Moreover, repeated motions by certain players would suggest a heightened level of aggression. The categories of motions to dismiss counted in the current database were those made by the trustee, the mortgage lender, the car lender, a child support/alimony creditor, other creditor, debtor, and unknown movant. None of the discharged debtors had a motion to dismiss made by the debtor, so the results discussed in the text reflect motions to dismiss made by third parties. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(27) p = .025 (N = 30 for discharged black debtors subjected to a motion, and N = 16 for whites. "N" represents the data/sample size). This significance value was attained using a one-sided Z test. See id.
(28) The examination of motions to dismiss was limited to discharged debtors because almost every dismissed debtor had a motion to dismiss made against him--that is how the debtor's case was dismissed. The data supports this logical inference: 99% of dismissed black and white debtors had at least one motion to dismiss made against them and 100% of dismissed Hispanic debtors had at least one motion to dismiss made against them. One unfortunate consequence of focusing on discharged debtors is that the sample size for Hispanic discharged debtors, eighteen, was too small for a significant statistical analysis. See id.
(29) See 11 U.S.C. [section] 1307(c) (2006) (giving any party in interest or the trustee the authority to request dismissal "for cause").
(30) p = .02 (N = 22 for discharged black debtors subjected to a motion, and N = 10 for whites). This significance value was attained using a one-sided Z test. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(31) 11 U.S.C. [section] 1302(b)(2)(B), (b)(3), (b)(5) (2006).
(32) WARREN & WESTBROOK, supra note 3, at 282.
(33) Steven W. Rhodes, Eight Statutory Causes of Delay and Expense in Chapter 11 Bankruptcy Cases, 67 AM. BANKR. L.J. 287, 308 (1993); see, e.g., In re Erwin, 376 B.R. 897, 902 (Bankr. C. D. Ill. 2007) (indicating that Chapter 13 trustees have broad discretion in plan administration); In re Passis, 235 B.R. 562, 564 (Bankr. D. N.J. 1999) (stating that the decision to file a motion to dismiss, in a Chapter 7 case, is within the discretion of the United States trustee).
(34) See In re Duffus, 339 B.R. 746, 748 (Bankr. D. Or. 2006).
(35) 11 U.S.C. [section] 704(a)(6) (2006).
(36) See WARREN & WESTBROOK, supra note 3, at 326 (citing 11 U.S.C. [section] 1329(a) (2006)).
(37) 8.1% of blacks (N = 8) and 13.3% of whites (N = 12) made a motion to modify their plans. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(38) See Figure 3. The 2001 Consumer Bankruptcy Project studied debtors from one district in each of five states. The total motions against black debtors in each of these states was spread out: Pennsylvania (8), Texas (6), Illinois (5), and Tennessee (3). In California, the only state not mentioned, none of the four discharged white debtors or two discharged black debtors was subjected to a motion to dismiss by a trustee. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(39) See, e.g., Michael J. Macco, Chapter 13 Trustee, http://macco.lawoffice.com/CustomPage.shtml#25 (last visited Sept. 29, 2008); Avoiding Stress in Court, www.atlantabankruptcy-attorney.com/html/ avoiding_stress_in_court.html (last visited Sept. 29, 2008); John E. Smith, What Happens Next?, http://johnesmithlawoffice.com/what_happens_next (last visited Sept. 29, 2008).
(40) WARREN & WESTBROOK, supra note 3, at 355.
(41) Dismissed debtors include those who had their cases dismissed either before or after confirmation.
(42) This difference in representation was significant for blacks (P = .02, N = 349) and Hispanics (P = .03, N = 63). These significance values were calculated using a one-sided Z test and the corresponding table for normal distribution. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(43) See id.
(44) WARREN & WESTBROOK, supra note 3, at 355.
(45) See Figure 4; see also 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(46) P = .005. This significance value was attained using a one-sided Z test by plugging the numbers into the Z test formula, which yielded a 2.55. This number was then looked up on a normal Z test table to attain the p value. See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(47) HULYA K. ERASLAN ET AL., THE REALITIES OF PERSONAL BANKRUPTCY UNDER CHAPTER 13 19 (2006), http://www.fdic.gov/bank/analytical/cfr/ 2006/oct/eraslan_li_sarte_pres.pdf.
(48) MODEL RULES OF PROF'L CONDUCT R. 1.5(a)(7) (2008).
(49) See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(50) See WARREN & WESTBROOK, supra note 3, at 326.
(51) See Warren, Economics of Race, supra note 9, at 1781 (describing middle class families as the majority filing for bankruptcy when looking at factors such as "education, occupation, and homeownership").
(52) Among families filing for bankruptcy, 70% held jobs in the fiftieth percentile of occupational prestige. See Elizabeth Warren, Financial Collapse and Class Status: Who Goes Bankrupt?, 41 OSGOODE HALL L.J. 115, 134 n.39, 135 fig.6 (2003) [hereinafter Warren, Financial Collapse].
(53) See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2. A T test--a test that determines whether differences between samples reflect actual differences between the populations from which they were drawn, or alternatively, merely random differences suggestive of similar populations--in Stata revealed that the difference is not significant between whites and blacks (P = .24), and between whites and Hispanics (P = .71). See id. These figures were adjusted for outliers because one white debtor's income was either entered mistakenly or the group contained a millionaire debtor, either of which would throw off group measures of central tendency.
(54) See id.
(55) See id.
(56) See id.
(57) 11 U.S.C. [section] 1322(a)(4) (2006); see Dickerson, Race Matters, supra note 8, at 1728.
(58) 11 U.S.C. [section] 1325(b)(2) (2006); see Dickerson, Race Matters, supra note 8, at 1728.
(59) Which is to say, very little breathing room; that may be one big reason why completion rates are so low.
(60) See generally Beverly I. Moran & William Whitford, A Black Critique of the Internal Revenue Code, 1996 WIS. L. REV. 751 (1996) (examining the Internal Revenue Code through the lens of critical race theory and concluding that members of the black community receive, on average, fewer of the tax benefits received by the average member of the white community).
(61) See generally Dickerson, Race Matters, supra note 8, at 1743-71 (examining racial bias within the Bankruptcy Code).
(62) See id. at 1728.
(63) Id. at 1728-29; see also id. at 1729 n.13 (citing numerous case law examples).
(64) Id. at 1746.
(65) 11 U.S.C. 1325(b)(2) (2006); see also Dickerson, Race Matters, supra note 8, at 1728.
(66) CARMEN DENAVAS-WALT ET AL., U.S. CENSUS BUREAU, INCOME, POVERTY AND HEALTH INSURANCE COVERAGE IN THE UNITED STATES: 2005, at 6 (2006), available at http://www.census.gov/prod/2006pubs/p60-231.pdf.
(67) See 2001 CONSUMER BANKRUPTCY PROJECT, supra note 2.
(68) P < .01. See id.
(69) See id.
(70) 81.5% of blacks making greater than $2,000 per month choose Chapter 13, compared to only 59.5% making less than $2,000 per month. 52.5% of whites making greater than $2,000 per month chose Chapter 13, compared to 33.8% making less than $2,000 per month (P < .01). See id.
(71) Warren, Economics of Race, supra note 9, at 1781.
(72) See supra notes 4-5 and accompanying text.
(73) See supra notes 6-7 and accompanying text.
(74) See supra notes 60-66 and accompanying text.
(75) R. Wilson Freyermuth, Crystals, Mud, BAPCPA, and the Structure of Bankruptcy Decisionmaking, 71 MO. L. REV. 1069, 1070 (2006).
(76) Id. at 1071.
(78) Bankruptcy Abuse Prevention and Consumer Protection Act of 2005 (BAPCPA), Pub. L. No. 109-8, 119 Stat. 23 (2005) (codified as amended in scattered sections of 11 U.S.C.).
(79) 11 U.S.C. [section] 707(b)(2)(A)(ii)(IV) (2006).
(80) See Shuk Wa Wong & Jan N. Hughes, Ethnicity and Language Contributions to Dimensions of Parent Involvement, 35 SCH. PSYCHOL. REV. 645 (2006) (mentioning that among kindergarten through twelfth-grade students, 18% of black and 13% of Hispanic students had repeated at least one grade, compared to 9% of whites).
(81) See 11 U.S.C. [section]1302(b)(4) (2006); WARREN & WESTBROOK, supra note 3, at 282.
(82) See 11 U.S.C. [section]1302(b)(1) (2006); WARREN & WESTBROOK, supra note 3, at 282.
(83) See 11 U.S.C. [section]1302(b)(4) (2006).
(84) Another example of independent actors affecting racial disparity would be making sure someone communicates to unrepresented debtors how important it is to have an attorney, such as by pointing out those studies that have shown that representation increases the chance of receiving a discharge. Congress could require delivery of the attorney-importance message at the mandatory credit counseling session or at some other point in conjunction with the court filings. Rather than communicating to all debtors, a court clerk could simply check for those few debtors without representation and accordingly send a message.
(85) Pro bono efforts by individual lawyers and bar associations have fallen short overall. See SUSAN R. MARTYN & LAWRENCE J. FOX, TRAVERSING THE ETHICAL MINEFIELD: PROBLEMS, LAW, AND PROFESSIONAL RESPONSIBILITY 70 (2004). Funding for legal aid services has been drastically cut over the past two decades. Id. at 70-71. Though ideally this trend would be reversed, at least for bankruptcy, members of the current Congress have only shown an inclination to cut further. See id. at 63. Philanthropic support for legal aid could help if more donors were made aware of the problem; after all, philanthropy in recent years has increasingly looked to fund entrepreneurial endeavors. See, e.g., Laura Koss-Feder, An Investment with Meaning, TIME, Apr. 9, 2007, available at http://www.time.com/time/magazine/article/0,9171,1604868,00.html. And historically bankruptcy relief has been linked to entrepreneurship because it enables risk-taking. See, e.g., John M. Czarnetzky, The Individual and Failure: A Theory of the Bankruptcy Discharge, 32 ARIZ. ST. L.J. 393, 436-38 (2000).
(86) Because the same lawyers come before the same judges time after time, bankruptcy judges are in a particularly good position to conscript attorneys. Moreover, the realities of Chapter 13 meet the general American judicial standards for when judges should compel a lawyer to represent someone in a non-criminal suit. Judges' "power to conscript lawyers ... exists for two primary purposes: (1) to ensure a 'fair and just' adjudicative process ... and (2) to maintain the integrity and viability of the ... entire civil justice system." See MARTYN & FOX, supra note 85, at 62. Though either of these is alone enough to justify conscription, both apply in the current situation. Disparate racial impact due to lack of representation clearly offends the adversarial system's ability to meet the standard of "just and fair." See id. at 63. Furthermore, one factor in determining whether this standard is met is "equal access to adequate legal assistance." See id. Minority debtors arguably have unequal access compared to two groups: creditors, who are likely to have attorneys because of their wealth, and whites, because attorneys tend to be less available in minority communities. Given that equal representation would help close the gap in Chapter 13 completion rates and thus increase the bankruptcy system's fairness, conscription of lawyers would serve this second goal of preserving the judicial system.
(87) See Richard J. Cebula, Historical and Economic Perspectives on Lawyer Advertising and Lawyer Image, 15 GA. ST. U. L. REV. 315, 328 (1998) (noting that "advertising has become an important way for certain segments of the population, especially the poor, less educated, and younger segments, to obtain access to legal services").
Rory Van Loo, Consultant, McKinsey & Co. J.D., 2007, Harvard Law School. Editor, Harvard Law Review, 2005-2007. I thank Professor Elizabeth Warren, Leo Gottlieb Professor of Law at Harvard Law School, for providing the initial vision for this empirical study and for her persistent guidance and feedback thereafter. I am also grateful to Jennifer Dillon, Biostatistician, Georgetown Medical Center, for early help with the statistical analysis.
|Printer friendly Cite/link Email Feedback|
|Author:||Van Loo, Rory|
|Publication:||Albany Law Review|
|Date:||Jan 1, 2009|
|Previous Article:||The judges v. the state: obtaining adequate judicial compensation and New York's current constitutional crisis.|
|Next Article:||Making peace and making money: economic analysis of the market for mediators in private practice.|