European regional convergence in a human capital augmented Solow model

In this paper, the process of productivity convergence is investigated for the enlarged European Union using regional (NUTS-2) data. The Solow model extended by human capital is employed as a workhorse. Alternative strategies are proposed to control for spatial effects. All specifications confirm the presence of convergence with an annual speed between 3 and 3.5 percent towards regional steady states. Furthermore, a geographically weighted regression approach indicates a wide variation in the speed of convergence across the regions, where a higher speed is striking in particular in France and the UK. Clusters of convergence can be identified, where regions with high convergence also have high initial income levels. -- Solow model ; regional convergence ; spatial lags ; spatial filtering