Nonparametric Estimation of Models with Generated Regressors.

This paper considers nonparametric kernel estimation of models with generated regressors and derives the asymptotic distribution of the resulting estimators. It is also shown how generated regressors may be used to reduce the dimensionality of certain nonparametric models. A labor supply equation is estimated to illustrate the technique. Copyright 1996 by Economics Department of the University of Pennsylvania and the Osaka University Institute of Social and Economic Research Association.