BANKRUPTCY BY THE NUMBERS

Exploring the Demographics of Consumer Chapter Choice

Contributing Editor: Gordon Bermant
Executive Office for U.S. Trustees (1)

     This column explores the demographics of consumer chapter choice by examining some basic bankruptcy statistics across the states. The results provide an introduction or refresher to some of the tough issues facing us when we try to understand how and why debtors choose between chapters 7 and 13.

The Significance of Chapter Choice

     Chapter choice in consumer bankruptcy is important for both policy and administrative reasons. On the policy side, there is an argument that some filers belong in chapter 13 rather than chapter 7 because their likely future income relative to their pre-filing unsecured debt gives them a good chance of making meaningful repayments over a 3 to 5 year period.. The argument continues that the rapid fresh start of chapter 7 should be reserved for debtors who are unable to make meaningful repayments.

     On the administrative side, the organization and burdens of managing a chapter 13 case differ from those attaching to a chapter 7 case.(2) Any significant shift in the relative numbers of filings in the two chapters will change the amount and nature of work for the bankruptcy courts, private trustees, U.S. Trustees, and attorneys representing debtors and creditors.

    The current statutory limitations on a debtor's initial choice of chapter 7 are essentially nil (11 U.S.C. 109(b)). Pending legislation may change this, by placing repayment and/or income tests at the threshold of a debtor's eligibility for chapter 7.(3)     

     For these reasons, it is important to attempt to understand the current environment of chapter choice. Our systematic understanding of the causes of chapter choice is weak, even though the opinions expressed by experts from every corner are very strong. As is frequently the case with empirical research, we are better at saying what is false than what is true. For example, in a highly quantified analysis of 1,529 consumer cases in Texas, Illinois, and Pennsylvania (all filed in 1981), Sullivan et al. tested chapter choice against a range of potential causes of the choice. They concluded that many of the plausible and intentional determinants of choice (e.g. ability to pay, assets to protect, unsecured debt levels, state exemption levels) had little if any causal effect. What mattered more, but still not a lot, were such factors as a recent move within the state or choice of a specialist attorney.(4)

    The authors noted that the district of filing is far and away the most powerful predictor of chapter choice, which is as true today as it was in 1981. Consider, for example, that during 1998 there were 22,840 non-business filings in the Western District of Tennessee, of which 74% were chapter 13 filings. During that period there were 25,011 non-business filings in the Western District of Washington, of which 17% were chapter 13 filings.(5) Such differences persist over time, revealing the potent but amorphous factor of "local legal culture." There are no simple ideas or models that will completely account for the large variations in chapter choice-or at least no one has found them yet. What follows is a brief description of a promising lead.

Chapter Choice in Relation to Filings per 1000 Households

    The percentage of chapter 13 filings in a state tend to vary directly with the numbers of filings per 1000 households in the state.(6) Where the total consumer rate per household becomes very high, it is often (but not always) due to large numbers of chapter 13 filings. The degree of association between chapter 13 filing rate and overall consumer rate per household is large enough to be substantively interesting. It is approximately as strong, for example, as the positive relationship between choice of chapter 7 and a recent intrastate move, as reported by Sullivan et al.(7).

      The relationship is shown in two forms in Tables 1 and 2. For Table1, the states (including the District of Columbia) were first sorted on percentage of chapter 13 filings. The highest 10 and lowest 10 states were then compared for the numbers of filings per 1000 households. To explore the possible relationship of these factors to the overall amount of consumer bankruptcy activity in the groups, the total numbers of filings were also included. In Table 2, the states were first sorted on numbers of filings per 1000 households, and the averages of the other factors were then calculated.

Table 1

States with Highest and Lowest Percentages of Chapter 13
(Values are averages for the states in the group)

GROUP PERCENTAGE CHAPTER 13 FILINGS PER 1000 HOUSEHOLDS ALL CONSUMER FILINGS
HIGHEST* 50% 17 16,668
LOWEST** 6% 10 4,140

*In decreasing order: GA, AL, NC, TN, SC, LA, TX, UT, MS, AR
** In increasing order: ND, SD, WV, RI, NH, IA, VT, ME, AK

Table 2
States with Highest and Lowest Rates of Filing per 1000 Households
(Values are averages for the states in the group)

GROUP PERCENTAGE CHAPTER 13 FILINGS PER 1000 HOUSEHOLDS ALL CONSUMER FILINGS
HIGHEST* 39% 21 45,171
LOWEST** 17% 9 8,633

*In decreasing order: NV, TN, GA, UT, AL, CA, MS, MD, ID, AR
** In increasing order: AK, SD, SC, IA, VT, ND, ME, MA, NC, NE

     The tables reveal both regularity and disparity. As has been observed repeatedly, the states with highest chapter 13 filing rates occupy the southern tier of the country, with the exception of Utah. The states with the lowest rates are uniformly northern except for West Virginia. And, most obviously the states with the lowest rates of chapter 13 tend also to have small overall filing rates both absolutely and relative to the number of households. When the data are first sorted on the number of filings per 1000 households, the positive association with chapter 13 rates is again observed, but there are some striking anomalies: North and South Carolina, which are in the top 10 states in percentage of chapter 13 filings, are in the bottom 10 states in terms of filings per 1000 households. And the geographic regularities of Table 2 disappear.

    An important problem to be solved in interpreting the relationship between percentage of 13 filings in a state and the rate of filings per 1000 households is to establish the direction of causality if there is one. Does a local legal culture that channels filers into chapter 13 thereby encourage more total filings than would otherwise arise? Is the population of chapter 13 filers in high percentage chapter 13 states composed of people who might not file at all in other states? Alternatively, do local conditions that produce high potential for bankruptcy produce a response in the business and legal communities that channels these filings into chapter 13?. A third possibility is that the two factors are not causally related, but are simply two consequences of other factors, like exemption limitations or wage garnishment laws.(8) It is beyond doubt that strong chapter 13 environments can thrive both where bankruptcy per 1000 households is very frequent (Alabama, Georgia and Tennessee, for example) and infrequent (North Carolina and South Carolina, for example). Nevertheless, for the nation overall, the direct relationship between the two factors is strong. The regularity and the important exceptions imply strongly that there is more than one set of causal factors leading to strong tendencies to file in chapter 13.

Conclusion: Why does this matter?

    The importance of understanding the roots of very strong (or very weak) support for chapter 13 filings in a community, a district, or a state, goes to the heart of our conception of consumer bankruptcy generally and the choice between chapter 7 and chapter 13 specifically. Economists have been attempting to use economic modelling to discern whether there has been a change in the stigma associated with filing bankruptcy.(9) The fact is that stigma, or shame, is a social and emotional factor, and we are extremely ignorant of the social and emotional contexts in which ordinary people operate at the time of filing and the months preceding it. Without an independent basis for asserting that increased rates of filing are caused by a decreased sense of personal responsibility, stigma, or shame in the population, there is no foundation for inferring such a relationship from economic factors alone. We need to look at the social contexts of the bankruptcy environment directly.

     The general prevalence of bankruptcy in a state, as measured by filings per 1000 households, is one example of such a measure-albeit not as precise or refined as we need. Sullivan et al. searched for other determinants in their work. There is support at many levels for the proposition that debtors should be at least encouraged to file chapter 13 if they can in fact repay a meaningful portion of their unsecured debt over the period of the plan. If we can understand how current high-percentage chapter 13 environments operate, and export that which is exportable into other environments, we might find a socially satisfying way to improve the consumer bankruptcy process.


End Notes:

* This article originally appeared in the American Bankruptcy Institute Journal, Vol. XVIII, No. 4, May 1999. It is reprinted with permission of the American Bankruptcy Institute.

1. All views expressed in this article are those of the author and do not necessarily represent the views of the Executive Office for U.S. Trustees.

2. For the courts, see Bermant, Gordon, Lombard, Patricia A., and Wiggins, Elizabeth C., A Day in the Life: The Federal Judicial Center's 1988-1989 Bankruptcy Court Time Study. 63 Amer. Bnkry. L. J. 491 (1991).

3. See, e.g., H.R. 833, The Bankruptcy Reform Act of 1999, and S.625, the Bankruptcy Reform Act of 1999, introduced in February and March, 1999, respectively.

4. Sullivan, Teresa A., Warren, Elizabeth, and Westbrook, Jay Lawrence, As We Forgive Our Debtors: Bankruptcy and Consumer Credit in America (1989). Pages 230-270.

5. Data compiled by the Administrative Office of the U.S. Courts.

6. The product-moment correlation coefficient for this relationship is +.41, which is statistically significant.

7. Sullivan et al., supra note 5, at 246.

8. Sullivan et al reported that exemption limitations did not influence chapter choice across the states they studied (Texas, Illinois, Pennsylvania), but there has been no study using information from all states. Nothing has been published about the effectiveness of wage garnishment laws.

9. Reference to be supplied.