A Monte Carlo Comparison of Various Semiparametric Type-3 Tobit Estimators 

Insik Min

Sheng jang Sheu

and

Zijun Wang

This paper compares recently developed semiparametric estimators of Type-3 Tobit model using Monte Carlo imulations. In particular, we examine the finite sample performance of the recently proposed method by Li and Wooldridge and compare it to some alternative semiparametric estimators. Simulation results indicated that Li and Wooldridge (2002) estimator under the independence restriction compares well relative to other alternative estimators, especially when the sample size is small or the error distribution has a thick tail.