• O130;
  • O180;
  • Q110
  • Agricultural productivity;
  • TFP growth;
  • Convergence;
  • Spillovers;
  • Panel Data


This article analyzes long-term agricultural Total Factor Productivity (TFP) growth at regional level by testing its time-series properties and identifying factors associated with divergence as opposed to convergence. The empirical application concerns Italian regions over the 1951–2002 time period. TFP growth decomposition ultimately attributes the observed productivity growth performance to these contrasting (convergence vs. divergence) forces. We find that technological spillovers are the key convergence force regardless of how the spillover effects are computed. At the same time, forces favoring convergence are almost offset by divergence forces (mainly scale or learning effects). This decomposition may explain the persistence of TFP growth rate differences in Italian agriculture, and could be applicable elsewhere.