Semiparametric instrumental variable estimation of treatment response models

This article introduces a new class of instrumental variable (IV) estimators for linear and nonlinear treatment response models with covariates. The rationale for focusing on nonlinear models is that, if the dependent variable is binary or limited, or if the e(ect of the treatment varies with covariates, a nonlinear model is appropriate. In the spirit of Roehrig (Econometrica 56 (1988) 433), identi.cation is attained nonparametrically and does not depend on the choice of the parametric speci.cation for the response function of interest. One virtue of this approach is that it allows the researcher to construct estimators that can be interpreted as the parameters of a well-de.ned approximation to a treatment response function under functional form misspeci.cation. In contrast to some usual IV models, heterogeneity of treatment e(ects is not restricted by the identi.cation conditions. The ideas and estimators in this article are illustrated using IV to estimate the e(ects of 401(k) retirement programs on savings. c © 2002 Elsevier Science B.V. All rights reserved. JEL classi&cation: C13; C14; H31