Technical and scale efficiency in the Italian Citrus Farming: A comparison between Stochastic Frontier Analysis (SFA) and Data Envelopment Analysis(DEA) Models

This paper aims to estimate technical and scale efficiency in the Italian citrus farming. Estimation was carried out from two different approach: a non parametric and a parametric approach using a Data Envelopment Analysis (DEA) model and a Stochastic Frontier Analysis (SFA) model, respectively. Several studies have compared technical efficiency estimates derived from parametric and non parametric approaches, while a very small number of studies have aimed to compare scale efficiency obtained from different methodological approaches. This is one of the first attempts that aims to put on evidence possible difference in scale efficiency estimations in farming due to methods used. Empirical findings suggest that the greater portion of overall inefficiency in the sample might depend on producing below the production frontier than on operating under an inefficient scale. Furthermore, we found that the estimated technical efficiency from the SFA model is substantially at the same level of this estimated from DEA model, while the scale efficiency arisen from SFA is larger than this obtained from DEA analysis.