Search

This paper examines the growth rate volatility of European regions' per capita GDP from 1992 to 2008. We measure the regional volatility using a new methodology based on Markov matrices, and investigate its main determinants. Volatility displays a geographical pattern and a significant spatial dependence. Output composition appears one of the main drivers of volatility; among the other determinants we find a negative impact of the size of regional economies and of labour market flexibility, and a positive impact of sectoral concentration, financialization of the economy, and, occasionally, of participation in EMU. [ABSTRACT FROM AUTHOR]

At the end of 1990s, Danny Quah devoted several papers to the analysis of polarization and stratification in the convergence processes of economies, creating the image of the 'convergence clubs' and suggesting the importance of studying the distribution dynamics of the macroeconomic variables. As for the labour markets, Overman and Puga (2002) showed that a progressive polarization of unemployment was in fact occurring among the European regions in 1986-1996, causing a phenomenon of cross-border clusterization. Here we propose to analyse the evolution of the unemployment rates of the EU 27 regions in the last two decades assuming that the unemployment rates evolve according to a Gompertz stochastic process. The estimated parameters of the process - intrinsic growth rate, deceleration factor, volatility - represent the evolutionary path of the unemployment rate and allow for estimating the steady state of the process. A cluster analysis is performed on the steady state values of the unemployment rates. The analysis confirms the emergence of several 'convergence clubs' among the European regional labour markets, which are compared to the clusters resulting from the more traditional clusterization on the current unemployment rates. [ABSTRACT FROM AUTHOR]

This paper extends the seminal Blanchard and Katz regional labour market model to include interaction effects using a dynamic spatial panel data approach. Three key contributions of this extended model are: ( i) the unrealistic assumption that regions are independent of one another no longer has to be made, ( ii) the magnitude and significance of so-called spillover effects can be empirically assessed, and ( iii) both the temporal and spatial propagation of labour demand shocks can be investigated. Using annual data from 1986-2010 for 112 regions across eight EU countries, both the non-spatial and spatial models are estimated. It is found that the majority of the spillover effects are highly significant. Consistent with economic theory, the impact of a region-specific demand shock is largest in the region where the shock instigates. The shock also propagates to other regions, especially impacting the first and second-order neighbours. [ABSTRACT FROM AUTHOR]