Multilateral comparison of total factor productivity and convergence in Italian agriculture (1951-2002)

This article uses the index number approach to estimate the regional differences of agricultural output, input and total factor productivity (TFP) in Italy. Data cover the period 1951 to 2002 and are taken from Agrefit, which is a new database of Italian agriculture at regional level. Comparisons across space and time are constructed in two steps: first, bilateral Fisher indexes, which are not transitive, are spatially chained to obtain transitivity, using the EKS method, then, multilateral indexes are linked chronologically over time by means of the so-called TFGG method, which satisfies temporal fixity. The second focus of the paper is to test for convergence in agricultural productivity. We consider the problems of heterogeneity and stability of cross-section growth regressions using a hierarchical Bayesian method based on panel data.