A Meta-learner for Heterogeneous Effects in Difference-in-Differences

We address the problem of estimating heterogeneous treatment effects in panel data, adopting the popular Difference-in-Differences (DiD) framework under the conditional parallel trends assumption. We propose a novel doubly robust meta-learner for the Conditional Average Treatment Effect on the Treated (CATT), reducing the estimation to a convex risk minimization problem involving a set of auxiliary models. Our framework allows for the flexible estimation of the CATT, when conditioning on any subset of variables of interest using generic machine learning. Leveraging Neyman orthogonality, our proposed approach is robust to estimation errors in the auxiliary models. As a generalization to our main result, we develop a meta-learning approach for the estimation of general conditional functionals under covariate shift. We also provide an extension to the instrumented DiD setting with non-compliance. Empirical results demonstrate the superiority of our approach over existing baselines.

Understanding how treatment effects vary across different population is essential in policy evaluation, especially when using observational data. We address this by developing a new method for estimating heterogeneous treatment effects using panel data, where we have repeated outcome observations across time. At the core of our approach is a meta-learning framework, which transforms the causal task into structured prediction problems, which can be solved using any machine learning approach. Importantly, our method is doubly robust, meaning it remains reliable even if some parts of the model are inaccurate. Moreover, it can also extend to more complex settings like treatment non-compliance and shifting data distributions. Empirical results show that our method consistently outperforms existing alternatives. This makes it easier for researchers to understand not just whether a treatment works, but for whom and under what conditions.