De-biased Machine Learning in Instrumental Variable Models for Treatment Effects

We introduce a de-biased machine learning (DML) approach to estimating complier parameters with high-dimensional data. Complier parameters include local average treatment effect, average complier characteristics, and complier counterfactual outcome distributions. In our approach, the de-biasing is itself performed by machine learning, a variant called automatic de-biased machine learning (Auto-DML). By regularizing the balancing weights, it does not require ad hoc trimming or censoring. We prove our estimator is consistent, asymptotically normal, and semi-parametrically efficient. We use the new approach to estimate the effect of 401(k) participation on the distribution of net financial assets.