Efficiency analysis of fisheries using stock proxies

•We perform technical efficiency analyses with proxies of fish stock estimates.•The DEA measures provide more robust estimates of production elasticities.•The CPUE measures can provide robust estimates of efficiency scores.•The CPUE index can be a good empirical approximation for stock size changes.•Both the CPUE and DEA measures indicate decrease in stock abundances. This study proposes three methods for technical efficiency (TE) estimation using three fish stock proxy measures while applying the stochastic production frontier (SPF) approach. We apply these methods to two Vietnamese offshore fisheries, gillnet and hand-line, for which measures of stock abundance are unavailable. Based on the assumption of unitary elasticity for both effort and the stock index, our results show that using data envelopment analysis (DEA) is more robust than catch-per-unit-effort (CPUE) measures in deriving a composite stock index to account for differences in stock conditions between periods. The SPF model using the DEA estimate of the stock index is free of production-related assumptions and is not subject to a distortion in the measures of production elasticities. Based on the consistency conditions of the efficiency estimates, we find no difference between efficiency scores based on CPUE or DEA measures. When the average characteristics of the vessels over periods are similar, the CPUE measures are not subject to a distortion in the measures of TE and can provide robust efficiency estimates. We also find that the CPUE index can be a good empirical approximation for stock size changes in fisheries with limited information. The empirical results indicate a decrease in stock abundances - probably due to overfishing of offshore resources.