statistical discrimination

  • 详情 Attention Constraints and Financial Inclusion
    We show that attention constraints of decision makers create barriers to financial inclusion. Using administrative data on retail loan screening processes, we find that attention-constrained loan officers exert less effort reviewing applicants from lower social or economic status (SES) backgrounds and reject them more frequently. More importantly, when externally imposed increases in workload tighten attention constraints, loan officers are even more prone to quickly rejecting low SES applicants but quickly accepting very high SES applicants without careful review, further widening the approval rate gap between high and low SES applicants — a unique prediction of the attention-based mechanism. Our findings suggest that decision makers’ attention constraints could amplify taste-based and statistical discrimination, which further exacerbates financial inclusion gaps, and that financial technologies that reduce information-processing costs may promote more balanced financial access.
  • 详情 Attention Discrimination in Retail Lending
    Using proprietary loan-screening data, we document that loan o!cers exhibit statistical discrimination due to how they allocate attention. Discrimination happens at the early stage of information acquisition: officers exert less effort reviewing ex-ante unfavorable applicants and thus end up rejecting them at a higher rate than justified by credit quality. In other words, ex-ante unfavorable applicants are more likely to be rejected without careful review. Further, when loan o!cers are distracted by heavier workloads, the degree of discrimination increases. Relative to the lowest workload decile, the approval rate for ex-ante unfavorable applicants drops from 14.7% to 8.8% when o!cers are in the top workload decile. Overall, the results show that attention constraints magnify discrimination. Our findings suggest that reducing decision-maker attention constraint can help reduce statistical discrimination.