Nonparametric estimation of causal heterogeneity under high-dimensional confounding

Abstract

This paper considers the practically important case of nonparametrically estimating heterogeneous average treatment effects that vary with a limited number of discrete and continuous covariates in a selection-on-observables framework where the number of possible confounders is very large. We propose a two-step estimator for which the first step is estimated by machine learning. We show that this estimator has desirable statistical properties like consistency, asymptotic normality and rate double robustness. In particular, we derive the coupled convergence conditions between the nonparametric and the machine learning steps. We also show that estimating population average treatment effects by averaging the estimated heterogeneous effects is semi-parametrically efficient. The new estimator is an empirical example of the effects of mothers' smoking during pregnancy on the resulting birth weight.

References in corpus (2)

• Least squares after model selection in high-dimensional sparse models
• Adaptive Concentration of Regression Trees, with Application to Random Forests

Cited by in corpus (7)

• Double Machine Learning based Program Evaluation under Unconfoundedness
• Analysing a built-in advantage in asymmetric darts contests using causal machine learning
• Kernel Methods for Unobserved Confounding: Negative Controls, Proxies, and Instruments
• Cross-Fitting and Averaging for Machine Learning Estimation of Heterogeneous Treatment Effects
• Regularized Orthogonal Machine Learning for Nonlinear Semiparametric Models
• Group Average Treatment Effects for Observational Studies
• Doubly robust estimation for conditional treatment effect: a study on asymptotics