Evaluating (weighted) dynamic treatment effects by double machine learning

TL;DR

This paper develops a robust double machine learning approach to evaluate dynamic treatment effects over multiple periods, accounting for high-dimensional covariates and enabling subgroup analysis.

Contribution

It introduces Neyman-orthogonal score functions for dynamic treatments, ensuring robustness and consistency in effect estimation with high-dimensional data.

Findings

Estimators are asymptotically normal and $\\sqrt{n}$-consistent.

Method performs well in finite samples as shown in simulations.

Applied to Job Corps data to assess training program sequences.

Abstract

We consider evaluating the causal effects of dynamic treatments, i.e. of multiple treatment sequences in various periods, based on double machine learning to control for observed, time-varying covariates in a data-driven way under a selection-on-observables assumption. To this end, we make use of so-called Neyman-orthogonal score functions, which imply the robustness of treatment effect estimation to moderate (local) misspecifications of the dynamic outcome and treatment models. This robustness property permits approximating outcome and treatment models by double machine learning even under high dimensional covariates and is combined with data splitting to prevent overfitting. In addition to effect estimation for the total population, we consider weighted estimation that permits assessing dynamic treatment effects in specific subgroups, e.g. among those treated in the first treatment period. We demonstrate that the estimators are asymptotically normal and $\sqrt{n}$-consistent under specific regularity conditions and investigate their finite sample properties in a simulation study. Finally, we apply the methods to the Job Corps study in order to assess different sequences of training programs under a large set of covariates.