This paper compares the long run prediction of convergence clubs introduced by Quah (1996 and 1997) with the actual observed dynamics of the Italian regions during the period 1970-2004. Economic dynamics is described by the evolution per capita GDP and different notions of distance are introduced to compare the trajectories of the regions. In addition, by means of hierarchical clustering methodologies the set of economies are segmented. By using the average distance, the study identifies two main performance clubs resembling the long run prediction of two converge clubs. On the other hand, the distance correlation shows different co-movements between members of the same cluster, indicating a variety of responses to external shocks. In particular the average distance identifies a clear division between a high performance club consisting of regions from the Center North, and a low performance club composed by regions from the South and islands. The presence of a cluster composed by center north regions is substantially confirmed by the distance correlation analysis, suggesting an homogeneous response to external shock. By contrast Southern regions display the same dynamical evolution but difference in co-movements.
Our analysis provides hints about the fundamentals that link the regions in their process of divergence. In fact the performance clubs pattern we discovered reflects the distribution of economic activities as well as the structural attributes of the regional economies.