Technical Efficiency in Rice Production at Farm Level in Northern Thailand: A Stochastic Frontier with Maximum Entropy Approach

Kittawit Autchariyapanitkul, Jirakom Srisirisakulchai, Kasem Kunasri, Apiwat Ayusuk

Abstract

In this paper, we apply a maximum entropy estimation method that is used as substitute for maximum likelihood method approach. This methodology was applied to sticky rice production data from Northern area, Chiang Mai province, Thailand. The mean square error (MSE) and standard error of parameters confirm that the method of GME in stochastic frontier model is more accurate than the conventional stochastic frontier model. In particular, the standard stochastic frontier model underestimate the technical inefficiency scores for the lower rankings and overestimate for the higher rankings. In this study, 95\% of the sticky rice farmers were found to have high inefficiencies and  most of them had technical inefficiencies between 0.6 and 0.8. These findings suggest that a considerable amount of productivity is lost due to inefficiency.