A1312
Title: Nonparametric estimation of local treatment effects with continuous instruments
Authors: Luke Keele - University of Pennsylvania (United States) [presenting]
Abstract: Instrumental variable methods are widely used to address unmeasured confounding, yet much of the existing literature has focused on the binary instrument setting. Extensions to continuous instruments often impose strong parametric assumptions for identification and estimation, which can be difficult to justify and may limit their applicability in complex real-world settings. Theory and methods for nonparametric estimation of treatment effects with a continuous instrumental variable are developed. An estimand is introduced that, under a monotonicity assumption, quantifies the treatment effect among the maximal complier class, generalizing the local average treatment effect framework to continuous instruments. Considering this estimand and the local instrumental variable curve, connections to the dose-response function and its derivative are drawn, and doubly robust estimation methods are proposed. Convergence rates and conditions for asymptotic normality are established, providing valuable insights into the role of nuisance function estimation when the instrument is continuous. Additionally, practical procedures for bandwidth selection and variance estimation are presented. Through extensive simulations, the advantages of the proposed nonparametric estimators are demonstrated. Finally, the methods are applied to data where excess travel time is an instrument for patients likelihood of receiving care at specialized health care facilities.
