Does bankruptcy hurt an individual’s ability to be hired or borrow money?
Credit reports, a gauge of our dependability as borrowers, are commonly used in a wide range of lending decisions, helping lenders judge whether we should be trusted with a credit card, an auto loan, or a mortgage. But the use of credit reports has expanded to other, nonlending areas. Increasingly, these reports are considered by landlords in rental-housing decisions and by employers in hiring decisions. It is often speculated how a major credit report blemish, such as a bankruptcy, could hurt one’s chances to get a loan or a job, but research that uses empirical evidence to probe such speculations has been thin. In “Bad credit, no problem? Credit and labor market consequences of bad credit reports” (National Bureau of Economic Research working paper no. 22711, October 2016), economists Will Dobbie, Paul Goldsmith-Pinkham, Neale Mahoney, and Jae Song offer a more systematic examination of this question.
The authors’ empirical strategy is to determine whether the removal of a bankruptcy “flag” from a credit report—an event signifying a potentially major improvement in one’s credit standing—would have a causal effect on individuals’ financial and labor market outcomes. When filing for personal bankruptcy, most people choose to do so under chapter 13 or chapter 7 of the U.S. Bankruptcy Code, and the flags for each chapter are statutorily delisted at different times after filing (7 years and 10 years, respectively). The authors use this difference in timing to develop a model whereby outcomes of chapter 13 filers (the “treatment” group) are compared with those of chapter 7 filers (the “control” group). The model assumes that, before flag removal at the 7-year mark, outcomes are identical for both groups.
Two datasets are used in the analysis. To estimate financial outcomes—namely, credit scores and credit card and mortgage borrowing—the authors draw upon data from the Equifax Consumer Credit Panel, focusing on flag removals implemented during the 2002–11 period. To study labor market outcomes, they match Equifax data with Social Security Administration tax records, which contain information on individuals’ employment status and earnings.
The study’s regression results reveal large and significant effects of flag removal on financial outcomes but not on labor market outcomes. Credit scores jump contemporaneously with flag removal, increasing sharply by 10 points in the quarter of change, and subside thereafter. This impact, which the authors call a “first stage” effect, quickly propagates into patterns of borrowing. Both credit card borrowing and mortgage borrowing trend up with flag removal: in the former case, the increase is linear over the 3-year period after the change; in the latter, it is concentrated in the first year.
Perhaps surprisingly, these strong effects on consumers’ financial experiences are not mirrored by similar responses in labor market outcomes. The authors find no statistically or economically significant impacts of flag removal on employment (job-to-job) transitions or worker earnings—results reported with 95-percent confidence intervals and extending to the self-employed workforce. The zero effect persists in heterogeneity tests accounting for differences in individuals’ credit card use before flag removal, demographic characteristics, and business cycle conditions, among other factors.
But why would an improved credit report matter for financial outcomes but not for labor market outcomes? The authors conjecture that the answer probably lies with differences in decisionmaking priorities. Whereas lenders likely put greater weight on credit report data in making their lending decisions, employers may be more inclined to rely on a broader set of screening inputs in making their hiring decisions. The authors also point to survey data indicating that a substantial percentage of hiring managers (about 30 percent) conduct background credit checks mainly to comply with state legislative requirements and to minimize their legal liability.