Causal Mediation Analysis with Hidden Confounders

TL;DR

This paper introduces a deep learning-based method for causal mediation analysis that accounts for hidden confounders, overcoming limitations of traditional approaches that require all confounders to be measured.

Contribution

It proposes a novel framework combining proxy variables and deep learning to uncover hidden confounders and estimate causal effects without assuming sequential ignorability.

Findings

Effective in synthetic and semi-synthetic datasets

Improves causal effect estimation accuracy

Potential applications in causal fairness analysis

Abstract

An important problem in causal inference is to break down the total effect of a treatment on an outcome into different causal pathways and to quantify the causal effect in each pathway. For instance, in causal fairness, the total effect of being a male employee (i.e., treatment) constitutes its direct effect on annual income (i.e., outcome) and the indirect effect via the employee's occupation (i.e., mediator). Causal mediation analysis (CMA) is a formal statistical framework commonly used to reveal such underlying causal mechanisms. One major challenge of CMA in observational studies is handling confounders, variables that cause spurious causal relationships among treatment, mediator, and outcome. Conventional methods assume sequential ignorability that implies all confounders can be measured, which is often unverifiable in practice. This work aims to circumvent the stringent sequential ignorability assumptions and consider hidden confounders. Drawing upon proxy strategies and recent advances in deep learning, we propose to simultaneously uncover the latent variables that characterize hidden confounders and estimate the causal effects. Empirical evaluations using both synthetic and semi-synthetic datasets validate the effectiveness of the proposed method. We further show the potentials of our approach for causal fairness analysis.