The Sensitivity of Counterfactual Fairness to Unmeasured Confounding

TL;DR

This paper develops tools to evaluate how unmeasured confounding affects the fairness of causal models, especially in non-linear additive noise models, providing practical methods for sensitivity analysis in real-world fairness applications.

Contribution

The authors introduce a novel sensitivity analysis framework for assessing the impact of unmeasured confounding on counterfactual fairness in causal models, including efficient algorithms for bivariate and multivariate cases.

Findings

Sensitivity analysis tools reveal the extent of bias due to confounding.

Efficient algorithms enable practical assessment in real-world scenarios.

The methods help identify when fairness measures are robust or compromised.

Abstract

Causal approaches to fairness have seen substantial recent interest, both from the machine learning community and from wider parties interested in ethical prediction algorithms. In no small part, this has been due to the fact that causal models allow one to simultaneously leverage data and expert knowledge to remove discriminatory effects from predictions. However, one of the primary assumptions in causal modeling is that you know the causal graph. This introduces a new opportunity for bias, caused by misspecifying the causal model. One common way for misspecification to occur is via unmeasured confounding: the true causal effect between variables is partially described by unobserved quantities. In this work we design tools to assess the sensitivity of fairness measures to this confounding for the popular class of non-linear additive noise models (ANMs). Specifically, we give a procedure for computing the maximum difference between two counterfactually fair predictors, where one has become biased due to confounding. For the case of bivariate confounding our technique can be swiftly computed via a sequence of closed-form updates. For multivariate confounding we give an algorithm that can be efficiently solved via automatic differentiation. We demonstrate our new sensitivity analysis tools in real-world fairness scenarios to assess the bias arising from confounding.