A0909
Title: Quantifying omitted variable bias in nonlinear instrumental variable estimators
Authors: Yu-Min Yen - National Chengchi University (Taiwan) [presenting]
Abstract: We develop a framework for quantifying omitted variable bias (OVB) in a class of nonlinear instrumental variable (IV) estimators, including the local average treatment effect (LATE), the LATE on the treated (LATT) and the partially linear IV model (PLIVM). We derive bias decompositions for these parameters, establish partial identification bounds for the structural estimand, and construct statistical inference procedures that yield OVB-adjusted confidence intervals. We propose to estimate the OVB bounds and conduct relevant statistical inferences using the double machine learning (DML) with the median method. An empirical application to the U.S. Job Training Partnership Act (JTPA) experiment illustrates usefulness of the method.
