Undergraduate Honors Theses

Thesis Defended

Spring 2015

Document Type


Type of Thesis

Departmental Honors



First Advisor

Carlos Martins-Filho

Second Advisor

Martin Boileau

Third Advisor

Jem Corcoran


In this paper we discuss the derivation, and use a Monte Carlo study to examine the finite sample performances of select estimators put forth in Martins-Filho et al. (2015) for partially linear semiparametric models under nonparametric endogeneity. We find that the selected estimators sufficiently account for the explicit nonparametric endogeneity of the underlying model in finite samples, and conclude that the one step back fitting and unweighted pilot estimators are more efficient than their counterparts in estimating m(x) and β respectively. Our findings support the assertion that both estimators for m(x) are oracle efficient.

Included in

Econometrics Commons