B0732
Title: Testing the separability condition in two-stage nonparametric models of production
Authors: Cinzia Daraio - University of Rome La Sapienza (Italy) [presenting]
Leopold Simar - Universite Catholique de Louvain (Belgium)
Paul Wilson - Clemson University (United States)
Abstract: A statistical model was previously provided that can rationalize two-stage estimation of technical efficiency in non-parametric settings. Two-stage estimation has been widely used, but requires a strong separability' assumption: the second-stage environmental variables cannot affect the support of the input and output variables in the first stage. We provide a fully non-parametric test of this assumption, deriving test statistic having asymptotic normal distribution under the null. We obtain this for both the DEA and the FDH first stage estimators. Our simulation results indicate that our tests (including some bootstrap alternatives) perform well both in terms of size and power. We present a real-world empirical example by updating a previous analysis on U.S. commercial banks. Our tests easily reject the assumption required for two-stage estimation, calling into question results that appear in hundreds of papers that have been published in recent years.
