Cracolici, Maria Francesca Cuffaro, Miranda Nijkamp, Peter
Year of Publication:
Tinbergen Institute Discussion Paper 07-065/3
Unemployment rates appear to vary widely at a subregional (e.g., local or provincial) level. Using spatial econometric models for spatial autocorrelation, this paper focuses attention on the spatial structure of regional unemployment disparities of Italian provinces. On the basis of findings from the economic literature and of the available socio-economic data, various model specifications including different explanatory variables are tested to investigate the geographical distribution of unemployment in the 103 provinces of Italy for the years 1998 and 2003. The results suggest that there is a clear explanation of unemployment differentials in terms of spatial equilibrium and disequilibrium factors and a significant degree of spatial dependence among labour markets at the provincial level in Italy. Provinces marked by high unemployment, as well as those characterized by low unemployment, tend to be spatially clustered, demonstrating the presence of unemployment ‘persistency’ in space and time regimes.
unemployment spatial lag model spatial autocorrelation Italian provinces