Abstract
We propose a class of rank tests for testing the randomness of technology parameters in a stochastic frontier regression model. The asymptotic distribution of the test statistic is derived using the weak convergence results of empirical and rank processes. Since the distribution is quite complex and involves the unknown distribution of the error term, a bootstrap procedure is suggested and its validity is established. Simulation studies indicate that the test works well for most common choices of inefficiency distributions. The proposed test procedure is applied to a data set on electricity utility companies.
Acknowledgments
The authors would like to thank the Editor and referees for their comments and suggestions. Chanchala Ghadge would like to acknowledge the University Grants Commission (UGC) of India for a financial support in the form of a Research Fellowship.