Estimation of a Four Component Semiparametric Stochastic Production Frontier Model with Endogenous Regressors and Determinants of Inefficiency

Abstract

This paper proposes a semiparametric stochastic production frontier model where the technology parameters are unknown smooth functions of environmental variables, and inputs are allowed to be endogenous. There are four components in the error term of this stochastic frontier model, where two of them are the noise components including the time-invariant and time-varying noises, and the other two of them are the inefficiency components including the time-invariant (i.e., persistent) and time-varying (i.e. transient) inefficiencies. The transient inefficiency is allowed to be a function of the environmental variables as well. We apply the proposed methodology to the Norwegian salmon production data, and analyze the estimated smooth coefficients (i.e., input elasticities), marginal effects of farm age, and persistent, transient, and overall technical efficiency scores.

About the presenter’s visit

Dr Kai Sun will be visiting the School of Economics from 8-11 April 2019  While here he will be sharing a room with his host, Dr David Du (606 Colin Clark Building).  If you would like to meet with him or have lunch or dinner with him please contact Dr Du.  Dr Du can be contacted on k.du1@uq.edu.au.