Economic structure, technology diffusion and convergence - the case of the Italian regions

Over the years, Italy’s regional disparities have been the object of much investigation in the literature. Recent evidence points to a revival of the convergence process, which had come to a halt in the mid-1970s, and renews the interest in the subject. This paper aims at contributing to the debate. Endorsing the critiques of the neoclassical assumption of technology as a public good, we investigate the link between economies’ structural characteristics and their growth performance. Specifically, treating technology as sector-specific and modelling technological spillovers as a positive function of the degree of similarity between economies’ sectoral features, a modified version of the Solow model is put forward and used to derive an “extended” convergence equation. The latter is then estimated by means of Panel Data procedures and data on the Italian regions over the 1970-1995 period. The results bring empirical support to our approach. From a theoretical viewpoint, our model suggests that the effects of technology diffusion on the convergence process are twofold. Firstly, if technological progress is partly dependent on external innovations, the temporal evolution of each economy’s productivity level, and its speed of convergence to the steady state value, cannot be ascribed solely to the existence of diminishing returns to capital, as suggested by Neoclassical Growth Theory, but is affected by technology diffusion as well. The difficulty in disentangling the effects of the two factors on the convergence rate remains, but the “extended” convergence equation arrived at reveals that the size of technological spillovers will have a level effect on productivity. Secondly, treating technological progress as sector dependent, our model implies potential steady-state growth rate heterogeneity. Thus, the estimated convergence rate can be ascribed to the concept of “Weak Conditional Convergence” [Islam (2003)], with each economy converging to its own steady state growth rate, which is more likely to be different from the others the more diverse the steady state production structures. These arguments lend support to the concept of “Club Convergence”.