In a Monte Carlo experiment we show that using a small probability of Type I error may lead to reduced pretest estimator MSE when a Hausman pretest is used to choose between least squares and instrumental variables estimators.

The Hausman test is used in applied economic work as a test of misspecification. It is most commonly thought of (wrongly some would say) as a test of whether one or more explanatory variables in a regression model is endogenous. There are several versions of the test available with modern...