Black Loans Matter: Distributionally Robust Fairness for Fighting Subgroup Discrimination

TL;DR

This paper introduces a distributionally robust fairness approach for lending algorithms to combat subgroup discrimination, addressing limitations of current group fairness metrics and leveraging recent individual fairness methods.

Contribution

It proposes a novel framework combining distributionally robust optimization with individual fairness techniques to better detect and mitigate subgroup discrimination in lending.

Findings

Improves fairness across subgroups without sacrificing overall accuracy

Addresses legal and reputational risks associated with subgroup discrimination

Enhances existing fairness protocols with robust, data-driven methods

Abstract

Algorithmic fairness in lending today relies on group fairness metrics for monitoring statistical parity across protected groups. This approach is vulnerable to subgroup discrimination by proxy, carrying significant risks of legal and reputational damage for lenders and blatantly unfair outcomes for borrowers. Practical challenges arise from the many possible combinations and subsets of protected groups. We motivate this problem against the backdrop of historical and residual racism in the United States polluting all available training data and raising public sensitivity to algorithimic bias. We review the current regulatory compliance protocols for fairness in lending and discuss their limitations relative to the contributions state-of-the-art fairness methods may afford. We propose a solution for addressing subgroup discrimination, while adhering to existing group fairness requirements, from recent developments in individual fairness methods and corresponding fair metric learning algorithms.