Fairness through Causal Awareness

TLDR

Biased historical datasets embed prejudices that cause classification algorithms to replicate unfair treatment patterns. The authors propose a causal modeling framework to learn from such biased data, linking fair classification to intervention strategies. They construct a causal model in which the sensitive attribute confounds both treatment and outcome, and train its parameters from observational data using deep learning and generative modeling techniques even when unobserved confounders exist. Experiments show that this fairness‑aware causal approach yields more accurate causal effect estimates and enables the derivation of policies that are simultaneously more accurate and fair on historically biased datasets.

Abstract

How do we learn from biased data? Historical datasets often reflect historical prejudices; sensitive or protected attributes may affect the observed treatments and outcomes. Classification algorithms tasked with predicting outcomes accurately from these datasets tend to replicate these biases. We advocate a causal modeling approach to learning from biased data, exploring the relationship between fair classification and intervention. We propose a causal model in which the sensitive attribute confounds both the treatment and the outcome. Building on prior work in deep learning and generative modeling, we describe how to learn the parameters of this causal model from observational data alone, even in the presence of unobserved confounders. We show experimentally that fairness-aware causal modeling provides better estimates of the causal effects between the sensitive attribute, the treatment, and the outcome. We further present evidence that estimating these causal effects can help learn policies that are both more accurate and fair, when presented with a historically biased dataset.