Gender Discrimination in US Mortgage Lending: A Temporal & Spatial Analysis

Abstract

By analyzing data from 102 million mortgage loan applications in the US between 2004-2012 (about 90% of all applications) across 3,141 counties and 8,000 financial institutions, I investigate whether applicants’ gender matters in the bank's lending decision. I find evidence that female applicants face a higher rejection rate than males, holding all else constant. Similar results are obtained for Hispanic applicants compared to non-Hispanics, and minorities compared to non-minorities. Moreover, I document that this discrimination is pro-cyclical (i.e. peaks in recessions) and that it varies substantially across states. Most importantly, I show that non-economic factors help explain the variation in discrimination across states while economic factors do not. Specifically, I find that gender discrimination in lending is higher in states with (i) high level of conservatism, (ii) low support for gay rights, and (iii) low female representation in the state legislature.(JEL E32, G21, J16)