Technical efficiency analysis of milkfish (Chanos chanos) production in Taiwan—an application of the stochastic frontier production function

Abstract

Milkfish has been farmed in Taiwan for over 300 years. Faced with limited land resources, a labor shortage, decreasing demand, and growth of imported fisheries products, the industry is looking at the problem of how to maintain a sustainable and efficient production. This study specifies a stochastic production frontier function to estimate potential milkfish farm output and efficiency by using 1997–1999 data from a survey of 433 aquaculture milkfish farms. Both Translog and Cobb-Douglas frontier production models are estimated using the maximum likelihood estimation method. Empirical results show that the Translog stochastic production function model fits the data better and that milkfish farming in Taiwan exhibits diminishing returns to scale. We also compare estimated maximum potential milkfish production per hectare under various pond conditions to provide managers with information about how to boost efficiency. In addition, this study estimates substitution elasticities and complementarity of input factors for milkfish farms to provide helpful information for milkfish farmers on how to reallocate input resources and help raise milkfish productivity through improvements in technical efficiency.