Abstract
This paper tests for bias in consumer lending decisions using administrative data from a high-cost lender in the United Kingdom. We motivate our analysis using a simple model of bias in lending, which predicts that profits should be identical for loan applicants from different groups at the margin if loan examiners are unbiased. We identify the profitability of marginal loan applicants by exploiting variation from the quasi-random assignment of loan examiners. We find significant bias against both immigrant and older loan applicants when using the firm’s preferred measure of long-run profits. In contrast, there is no evidence of bias when using a short-run measure used to evaluate examiner performance, suggesting that the bias in our setting is due to the misalignment of firm and examiner incentives. We conclude by showing that a decision rule based on machine learning predictions of long-run profitability can simultaneously increase profits and eliminate bias.
Highlights
There are large disparities in the availability and cost of credit across different demographic groups within many developed countries
Marginal female and male applicants yield statistically identical profits, suggesting no bias against female applicants. We show that these results cannot be explained by other ethnic or age-related differences in baseline characteristics, differences in the level of systematic risk across groups, or the way that the instrumental variable (IV) estimator averages the level of bias across different examiners
Immigrant and older applicants are more likely to default in the short run compared to native-born and younger applicants with the same level of expected long-run profits, while no such differences exist for female and male applicants
Summary
There are large disparities in the availability and cost of credit across different demographic groups within many developed countries. In the context of consumer lending, the outcome test is based on the idea that long-run profits should be identical for marginal applicants from all groups if loan examiners are unbiased and the disparities across groups are solely due to omitted variables or statistical discrimination. Immigrant and older applicants are more likely to default in the short run compared to native-born and younger applicants with the same level of expected long-run profits, while no such differences exist for female and male applicants Taken together, these three results suggest that examiners are equalizing the private returns of lending across groups at the margin, just as predicted by our incentive-based model of bias. The Supplementary Appendix provides additional results, details on two alternative models of bias, and information on the outcomes used in our analysis
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have