On Double Robustness in Double Machine Learning

Simon Valentin, Gianluca Detommaso, Yikuan Li, Manfred Opper, Michael Brückner

Keywords: Causal Inference, Machine Learning, Double Machine Learning, Doubly Robustness

TL;DR: This paper clarifies common misconceptions about the doubly robustness of Double Machine Learning, characterizes under which conditions doubly robustness holds and proposes a new maximum likelihood estimator that achieves it.

Abstract: Double Machine Learning (DML) is widely used for causal estimation from observational data and is often assumed to be doubly robust. While this holds for the Z-estimator proposed by Chernozhukov et al., many practical implementations rely on the Robinson estimator, which crucially depends on correct treatment model specification. This misunderstanding has important implications, as many practitioners incorrectly assume robustness to misspecification. We provide analyses clarifying when double robustness holds for popular DML estimators. Based on these insights, we develop a maximum likelihood estimator that achieves double robustness, providing a likelihood-based alternative to the Z-estimator.

Submission Number: 10