Spatial nonstationarity in the stochastic frontier model: An application to the Italian wine industry

This research estimates the efficiency of a representative sample of Italian wine producers from the Italian FADN survey following a recent spatial stochastic frontier framework that allows to isolate the spatial dependence among units and to evaluate the role of intangible local factors in influencing the economic performance of firms. The empirical exercise shows that the specific territorial patterns in the data cannot be merely explained using a standard set of contextual factors. This intangible component can be interpreted as the role of the local business climate: in most localities, the presence of an embedded community stimulates a process of local learning that generates the diffusion of tacit knowledge through continuous interaction among the local actors. This effect is found to be different across firm size, with a larger impact on small firms.