Analysing Growth and Distribution Dynamics - Isolating Divergence Factors

This paper analyses the factors driving convergence and divergence processes in the growth dynamics of European urban regions over the period 1978 to 1994. To achieve this, we develop a two-stage procedure. First, viewing growth of real GDP per capita as a multivariate process, a fully specified growth model is estimated assessing growth dynamics and the role of a rich set of variables. In particular, as noted elsewhere (Cheshire and Magrini, 2005), evidence is found which is supportive of a spatial adaptation of the endogenous growth model, with human capital related variables having a highly significant role in explaining growth differences. The second stage then uses a ‘distribution dynamics’ approach to evaluate the contribution of individual growth factors to convergence dynamics. More specifically, regression results are used to simulate alternative end period incomes which, via the estimation of stochastic kernels, enable us to isolate the role of selected variables in shaping the dynamics of the cross-sectional distribution of per capita income. This analysis suggests that the most important factor determining the form of FUR growth and convergence dynamics observed from 1978 to 1994 was the differing distribution of human capital. These results are compared with the results using the Markov Chain approach and it is found both techniques lead to similar conclusions.