Title: Nonparametric instrumental estimation of additive models
Authors: Samuele Centorrino - Stony Brook University (United States) [presenting]
Sorawoot Srisuma - University of Surrey (United Kingdom)
Abstract: A two-step estimator is proposed for nonparametric additive regression functions with multiple endogenous and exogenous conditioning variables. In the first step we construct a sieve nonparametric instrumental variable estimator that achieves the optimal rate of convergence in a minimax sense. We smooth this over in the second step using kernel methods. The subsequent estimator has an asymptotic normal distribution and has an oracle property. In particular, the asymptotic distribution of each additive component is the same as it would be if all the other components were known.