Alternative Approaches to Evaluating Multispecies Fishery Efficiency

In this paper, the estimate of the productive efficiency for some Italian fleet segments is proposed. The estimate is based on a Stochastic Frontier Production (SFP) function and a Data Envelopment Analysis frontier (DEA) using micro data of fishing vessels. The essential random nature of the fishing processes under consideration suggests that stochastic method might be preferable to non-parametric DEA method for efficiency estimation. In other words, the risk connected with DEA is that random noise, probably due to exogenous shocks or measurement error, may be considered as inefficient. On the other hand, DEA does not identify an explicit functional form for the production function combined with a distributional form for the inefficiency term. Therefore, DEA should be less prone to mis-specification than Stochastic Frontier Production. In fishery literature, several empirical comparisons between DEA and SFP have been conducted in order to assess the level of efficiency of fishing vessels and fleet segments. This paper attempts to apply both techniques, simultaneously, to the Mediterranean area which is characterised by multi-species and artisanal-type fishery