Data envelopment analysis for analyzing technical efficiency in aquaculture: The bootstrap methods

The paper adopts the bootstrap procedures for data envelopment analysis to analyze the technical efficiency in the Vietnam's intensive white-leg shrimp farming. The results demonstrate that the null hypothesis of constant returns-to-scale is rejected in favor of variable returns-to-scale for the production technology. Moreover, the potential improvement is greater using bootstrapping than that achieved by using the conventional data envelopment analysis, which has been widely adopted in the aquaculture literature for technical efficiency estimation. By adopting the double bootstrap proposed by Simar and Wilson, the bias-corrected technical efficiency is 0.69 and at the 95% confidence interval, with the lower limit of 0.65 and the upper limit of 0.75. In addition, statistically positive influences on technical efficiency in this type of farming are larger farm size and having access to formal credit, with cultivation period in the production year being a negative influence.