Marginal Fairness: Fair Decision-Making under Risk Measures

TL;DR

This paper proposes marginal fairness, a new fairness criterion for decision-making under risk measures that ensures decisions are insensitive to protected attributes, applicable across various data types and regulatory contexts.

Contribution

It introduces marginal fairness for equitable risk-based decisions, modeling a two-stage process and extending fairness to complex dependencies via cascade sensitivity.

Findings

Framework applied to auto insurance data demonstrating practical fairness enforcement

Ensures decisions are insensitive to protected attributes under generalized risk measures

Applicable across diverse data types and regulatory environments

Abstract

This paper introduces marginal fairness, a new individual fairness notion for equitable decision-making in the presence of protected attributes such as gender, race, and religion. This criterion ensures that decisions based on generalized distortion risk measures are insensitive to distributional perturbations in protected attributes, regardless of whether these attributes are continuous, discrete, categorical, univariate, or multivariate. To operationalize this notion and reflect real-world regulatory environments (such as the EU gender-neutral pricing regulation), we model business decision-making in highly regulated industries (such as insurance and finance) as a two-step process: (i) a predictive modeling stage, in which a prediction function for the target variable (e.g., insurance losses) is estimated based on both protected and non-protected covariates; and (ii) a decision-making stage, in which a generalized distortion risk measure is applied to the target variable, conditional only on non-protected covariates, to determine the decision. In this second step, we modify the risk measure such that the decision becomes insensitive to the protected attribute, thus enforcing fairness to ensure equitable outcomes under risk-sensitive, regulatory constraints. Furthermore, by utilizing the concept of cascade sensitivity, we extend the marginal fairness framework to capture how dependencies between covariates propagate the influence of protected attributes through the modeling pipeline. A numerical study and an empirical implementation using an auto insurance dataset demonstrate how the framework can be applied in practice.