What accounts for racial and ethnic differences in credit use?

Racial and ethnic differences across U.S. households in use of bank credit (e.g., credit cards) and nonbank credit (e.g., payday loans) are striking. We examine whether household characteristics and residential location can explain these differences. We use a novel dataset with information on previously unexplored factors, including income volatility, subjective attitudes about banks, geographic proximity to financial providers, and neighborhood population characteristics. We find that much of the raw disparities in credit use are attributable to observable household characteristics. Accounting for neighborhood population characteristics meaningfully reduces the disparities further. However, the residual racial and ethnic disparities remain large in magnitude. We show these disparities are not likely attributable to unobserved differences in households' family background, financial literacy, subjective attitudes, or credit scores. Instead, they are most likely attributable to unobserved supply‐side factors, such as racial and ethnic differences in households' exposure to marketing. We conclude with implications for policy.