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

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

Households with higher unpaid credit card balances today are more likely to have filed for bankruptcy in the past than those with lower balances. Note that the coefficient on credit card debt is higher than the coefficient on other debt. (6) Although both coefficients are small in magnitude, the result is notable because past filers are likely to have discharged their credit card debt at the time of filing and may still have binding constraints on their credit card limits (Musto 1999). Recall that the bankruptcy signal remains in credit reports for up to 10 years, whereas the median number of years since filing in our sample is eight.

Homeowners, married or older individuals, and those with higher incomes are less likely to have filed for bankruptcy, while persons having a large family are more likely. Unlike the case of the delinquency regression above, spouse's unemployment during the previous 12 months is more likely to be associated with a person having filed for bankruptcy in the past than the filer's own unemployment. Surprisingly, those with health insurance are more likely to have filed, although they may have acquired health insurance after the filing. Regional effects are found to be strong: New England residents (the omitted region) are significantly less likely to have filed for bankruptcy than residents of other regions, holding the demographic attributes constant. The strong regional effects indicate that factors other than individual characteristics play a role in bankruptcy decisions. They include differences in state bankruptcy exemption levels. (7)

We found that having filed for bankruptcy in the past or having been unemployed during the past year significantly raised the probability of delinquency, controlling for other characteristics. Households that had filed for bankruptcy in the past were found to carry higher unpaid credit card balances and have significantly higher ratios of credit card debt to income than those who had not filed. People who had filed for bankruptcy in the past were also more likely to default on their payments, even after controlling for their income and credit card debt.

III. Credit Card Lending--the Supply Side

Based on pre-approved credit card solicitations mailed out by a credit card issuer, Ausubel (1999) found that credit card issuers face adverse selection. Consumers who accept credit card offers are worse credit risks than consumers who do not, and consumers who accept inferior credit card offers (that is, those with higher interest rates) are worse credit risks than consumers who accept superior offers--they turn out to have higher delinquency, charge-off, and bankruptcy rates. Ausubel concludes that an "inferior offer yields inferior customers."

We test the adverse selection hypothesis using detailed data on the terms of credit card plans offered by many bank issuers over a long period of time: from January 1990 to July 1999. Both delinquency and charge-off rates increased substantially in the mid 1990s (Figures 3 and 4). Because our data include the period of steep increase in both rates but are not limited to it, using panel data allows us to isolate intertemporal shifts from individual issuers' lending strategies.