It was estimated that at least 17 percent of all U.S. households would benefit financially from filing for personal bankruptcy (White, 1998)! However, 1.4 million households filed for bankruptcy in 2001, which represents only 1.3 percent of the 104 million U.S. households. White (1998) obtains her estimate of the percentage that could benefit by calculating the number of households that had a positive financial benefit to filing, which is the standard measure used in the bankruptcy literature.
Since the standard model predicts a much higher percentage of households filing than is actually the case, the standard model appears insufficient and the incentives to file do not appear to be entirely understood. As more households file for personal bankruptcy, it is becoming more important to understand these determinants. Further, since Congress is considering altering the personal bankruptcy laws, economists need to understand these determinants to be able to add to the policy debate.
One plausible explanation for White’s (1998) findings is that the benefit to filing is not large enough for some households. White (1998, p. 703) calculates the median benefit to filing for those with a positive benefit equals $1,650. It may be that some households do not file when the benefit is relatively small. On the other hand, for those with a relatively large benefit, it is harder to explain why they do not file. One possibility is that these households may not file in the first year that they have this large financial benefit. Rather, some households wait to file in hopes of avoiding bankruptcy altogether. After several years with a positive financial benefit, they may eventually file for bankruptcy.
This paper addresses this possibility by estimating whether the probability of filing is changing over time for a household, holding the financial benefit and other relevant factors constant. To accomplish this, I estimate a hazard function using the personal bankruptcy data in the Panel Study of Income Dynamics (PSID). While this is not the first to use the PSID to examine the personal bankruptcy decision, this represents the first to use the PSID applying duration analysis.
The results indicate that there is duration dependence in bankruptcy. Further, I find that the conditional probability of filing is largest around the third and fourth spell-years, where spell equals the number of years the household has been in its current marital status. These findings indicate that the probability of filing for a household that has been married (divorced) for one year is different than the probability of filing for a household that has been married (divorced) for ten years, ceteris paribus. One implication of these findings is that some households may be avoiding bankruptcy for several years, even when they have a positive financial benefit to filing. This can help explain White’s (1998) findings.
The main policy implication from these findings is that assistance early on may help some households avoid bankruptcy altogether. It is hypothesized that some households do not want to file for bankruptcy even when they have a financial benefit to file. However, after several years of financial difficulty, they revise their estimates regarding the costs and benefits of bankruptcy, which leads them to file. If policies were aimed at helping households early on in their financial difficulty, then these households may never reach the point where bankruptcy becomes optimal.
If the goal of policy is to decrease the number of bankruptcy filings, these results suggest two policies, one of which is supported directly in this paper. First, an increase in the average Unemployment Insurance (UI) benefits is found to decrease the hazard rate. Second, Sullivan and Worden (1990) found that increases in credit-counseling services decrease the number of filings. If households in financial trouble could receive credit-counseling fairly soon after their financial problems arise, these households may be able to avoid bankruptcy.
Download
PDF Ebook Duration Dependence in Personal Bankruptcy
