A Technical Efficiency Analysis of Taiwan’s Barramundi Production - An Application of the Stochastic Frontier Production Function

• Main Authors: Kun-Shan Huang, 黃崑山
• Other Authors: Fu-Sung Chiang
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/55565125832863774117

碩士 === 國立臺灣海洋大學 === 應用經濟研究所 === 96 === Taiwan was the leading producing country of barramundi in the world between 1990 and 1998. Because of the domestic market could not be effectively developed, barramundi production has been decreasing since 1998, and slumped to the lowest point in 2002. Since 2002, export demand for barramundi has increased and the industry has expanded. In 2005, the production of barramundi in Taiwan has ranked second in the world. This research examines the relationship between input and output in barramundi production using accounting information provided by barramundi farmers. We conducted interviews with the managers of barramundi farms between October and December in 2007. The information on the inputs and output for the barramundi from fry to harvest were collected in Chiayi, Kaohsiung and Pingtung counties. The data period ranged from 5 to 19 months for a total of 80 harvests from 50 farms. We make use of the concept of production frontier suggested by Farrell (1957). The model developed by Battese and Coelli (1995) was used to estimate the stochastic production frontier; this production frontier function was used to analyze the impact the manager's ability on productivity and the sources of inefficiency. Our result shows that the Translog production function has a better goodness of fit in terms of the likelihood ratio than the Cobb-Douglas production function. The estimated value for the return to scale is 0.855, and the average value of technical efficiency is 0.917, with 93.25% of the farms having a value exceeding 0.8. For all farms, results of this study suggest that 99% of technical inefficiency is due to human errors. Among all factors, culture density, water source, depth of the ponds, size of the pond, workshop attendance, land ownership, and export of the production have significant impacts on production efficiency. Based on our results, the best farming density is between 50,000 and 100,000 fries per hectare. Use of either pure freshwater or sea water should be avoided. A mixture of fresh and sea water is most appropriate. The depth of pond is recommended to be higher than 1.5 meters. The best size of the pond is less than 0.5 hectares due to the decreasing return to scale found in this study. We recommend managers to attend related training workshops to obtain more information on the safe level of drug usage and ways to prevent fish diseases. On land ownership, this study finds that farm managers who own the land tend to be more efficient in producing barramundi than those who rent. We recommend that if managers want to run their business long term, they would be better off purchasing the land for production. Finally, for the export markets, the fish size should have to be bigger than those sold in domestic markets and hence fish will have to be fed longer. As the rate of body growth decreases with time, it will be less efficient in production. However, export is a safe way to avoid risk. If prices are not good in domestic market, then a better choice is to extend the feeding time for export. The result on the maximum output per hectare of fish farms after correcting for technical inefficiency factors will provide the fishery authority and related fishery groups information to help increase productive efficiency and achieving production goals in the future. In addition, the estimated technical efficiency values will also serve as the basis for comparison and exchange of experience among fish farms, which in turn will provide less efficient farms directions for improvements; and thereby increase the competitiveness of the barramundi industry in Taiwan.