Estimating Aleatoric Uncertainty in the Causal Treatment Effect

TL;DR

This paper introduces measures for aleatoric uncertainty in individual treatment effects, proposing kernel-based estimators and demonstrating their effectiveness through extensive experiments.

Contribution

It defines VTE and CVTE as measures of aleatoric uncertainty, proves their identifiability, and develops nonparametric estimators with theoretical guarantees.

Findings

Estimators achieve convergence as shown in theory

Method performs well on synthetic datasets

Outperforms naive baselines in experiments

Abstract

Previous work on causal inference has primarily focused on averages and conditional averages of treatment effects, with significantly less attention on variability and uncertainty in individual treatment responses. In this paper, we introduce the variance of the treatment effect (VTE) and conditional variance of treatment effect (CVTE) as the natural measure of aleatoric uncertainty inherent in treatment responses, and we demonstrate that these quantities are identifiable from observed data under mild assumptions, even in the presence of unobserved confounders. We further propose nonparametric kernel-based estimators for VTE and CVTE, and our theoretical analysis establishes their convergence. We also test the performance of our method through extensive empirical experiments on both synthetic and semi-simulated datasets, where it demonstrates superior or comparable performance to naive baselines.