Detecting Double Discrimination in the Credit Market

Abstract

The design of tests for discrimination in the credit market is controversial. This paper revisits the issue and pays special attention to double discrimination, which consists in targeting loan applicants who belong simultaneously to two vulnerable groups. Double discrimination may take different forms, some of which contrast with the addition of two one-dimensional biases. Our novel testing methodology, based on recovery rates, rationalizes the fact that double discrimination can exist irrespectively of the presence of simple discrimination. We illustrate our suggested approach using a microcredit case study involving positive discrimination.

Keywords: Discrimination in Lending, Credit Market, Double Discrimination, Gender, Race, Ethnicity, Recovery Rate

JEL Classification: C12, C44, J15, J16, G21, D63