Maximum likelihood estimation of the revenue function system with output-specific technical efficiency

In this paper we propose a non-radial and output-specific measure of technical efficiency which is new in a stochastic frontier model. We consider a multi-output multi-input transformation function formulation that incorporates output-specific technical efficiency (OSTE) in a revenue maximizing framework. Starting from the dual revenue function with OSTE, we develop the maximum likelihood method to estimate the parameters of the translog revenue-share system as well as predict OSTE components using some novel results from the closed skew-normal distribution.