Doubly Robust Estimation of Local Average Treatment Effects Using Inverse Probability Weighted Regression Adjustment

Abstract

We revisit the problem of estimating the local average treatment effect (LATE) and the local average treatment effect on the treated (LATT) when control variables are available, either to render the instrumental variable (IV) suitably exogenous or to improve precision. Unlike previous approaches, our doubly robust (DR) estimation procedures use quasi-likelihood methods weighted by the inverse of the IV propensity score - so-called inverse probability weighted regression adjustment (IPWRA) estimators. By properly choosing models for the propensity score and outcome models, fitted values are ensured to be in the logical range determined by the response variable, producing DR estimators of LATE and LATT with appealing small sample properties. Inference is relatively straightforward both analytically and using the nonparametric bootstrap. Our DR LATE and DR LATT estimators work well in simulations. We also propose a DR version of the Hausman test that can be used to assess the unconfoundedness assumption through a comparison of different estimates of the average treatment effect on the treated (ATT) under one-sided noncompliance. Unlike the usual test that compares OLS and IV estimates, this procedure is robust to treatment effect heterogeneity.