Credit Card Borrowing, Delinquency, and Personal Bankruptcy - Statistical Data Included

New England Economic Review, July-August, 2000 by Joanna Stavins

Filing for personal bankruptcy under Chapter 7 prevents future filing for several years. Therefore credit card issuers face relatively low risk by extending credit to those who filed very recently. The table at the bottom of Table 1 shows that those who filed for bankruptcy a year or two before the survey was conducted had, on average, a high unpaid credit card debt, probably because they receive many credit card offers during that initial post-bankruptcy period. The average debt is lower for those who filed between three and eight years ago, and increases again for those who filed even earlier, perhaps because a bankruptcy flag that could limit access to credit lasts no more than 10 years.

Domowitz and Sartain (1999) found that while the ratio of credit card debt to income was the largest contribution to bankruptcy at the margin, health problems leading to medical debt were the most important factor in a household's decision whether or not to declare bankruptcy. The self-rated health profile in our sample (not shown in table) is consistent with their finding: 21 percent of filers rated their health as excellent, compared to 36 percent of non-filers, and 24 percent of filers rated their health as fair or poor, compared to 19 percent of non-filers. On the other hand, our data showed no notable difference between the two groups in the fractions with health insurance.

Although the data do not allow us to test for causality, the above comparison shows that credit card borrowing and bankruptcy filing are correlated. Aggregate statistics also indicate that credit card default rates and personal bankruptcy rates are correlated over time (Figure 4), but other factors need to be controlled for. We present regression analysis below.

Regional Data

The fraction of people who file for bankruptcy or default on their loans varies across regions. Some of the variance is likely due to differences in laws and social values that do not change substantially over time, but some may be due to differences in debt and credit card borrowing patterns. We examined the data to test whether the fraction of people who had delinquent loans or had filed for bankruptcy is correlated with the average household credit card debt in each region. Table 2 shows regional averages and correlation coefficients based on the data for the nine Census divisions. (4) The regions with the highest fractions of households with delinquent loans are Mountain, West South Central, and East South Central, while the highest proportions of bankruptcy filers were in the Mountain, East South Central, and West North Central divisions. Recall that in order to be classified as being delinquent on a loan, a person must have been late on his payments for at least two months during the past year. Approxi mately one-fourth of the sample had no loans at all; they are not included in the denominator when the statistic is computed.

To examine which measures of debt are most strongly associated with bankruptcy and delinquency, we calculated correlation coefficients of the two variables with total debt and credit card debt. As shown in the bottom two rows of Table 2, credit card debt is more closely correlated with delinquency and bankruptcy rates in a region than is total debt. In fact, the correlation coefficients for total debt are negative. Because the divisions with the highest bankruptcy rates also have the lowest average household incomes, the ratio of credit card debt to household income or to household net worth is even more closely correlated with regional bankruptcy rates, with correlation coefficients of 0.71 and 0.76, respectively. The correlations indicate that regions with high credit card debt relative to income typically have high rates of bankruptcy filing.