Estimation and Inference of Heterogeneous Treatment Effects using Random Forests

TL;DR

This paper introduces causal forests, a non-parametric extension of random forests, enabling consistent estimation and valid inference of heterogeneous treatment effects in various scientific fields.

Contribution

It develops a theoretical framework for causal forests, proving their consistency and asymptotic normality, and provides practical methods for confidence interval construction.

Findings

Causal forests outperform classical matching methods in power.

They are valid for any type of random forest, including classification and regression.

The method is effective even with irrelevant covariates.

Abstract

Many scientific and engineering challenges -- ranging from personalized medicine to customized marketing recommendations -- require an understanding of treatment effect heterogeneity. In this paper, we develop a non-parametric causal forest for estimating heterogeneous treatment effects that extends Breiman's widely used random forest algorithm. In the potential outcomes framework with unconfoundedness, we show that causal forests are pointwise consistent for the true treatment effect, and have an asymptotically Gaussian and centered sampling distribution. We also discuss a practical method for constructing asymptotic confidence intervals for the true treatment effect that are centered at the causal forest estimates. Our theoretical results rely on a generic Gaussian theory for a large family of random forest algorithms. To our knowledge, this is the first set of results that allows any type of random forest, including classification and regression forests, to be used for provably valid statistical inference. In experiments, we find causal forests to be substantially more powerful than classical methods based on nearest-neighbor matching, especially in the presence of irrelevant covariates.