Diversity of convergence and spatial dependence between European metros and non-metros. An application of the two-regime spatial autoregressive panel model

Abstract

The aim of the paper is to analyse the difference in the speed of convergence and instability of spatial dependence between the metro and non-metro subregions. The tworegime spatial autoregressive panel data model estimated in the Bayesian approach using a Markov Chain Monte Carlo technique was applied for both purposes. The model was evaluated in the small Monte Carlo simulation. The main analysis was performed at the NUTS3 level, which allowed delimitating the metropolitan areas according to the Larger Urban Zone definition. A balanced panel consisted of data for 31 European countries for the years 2000–2010. A significant catching up process was found for both metros and nonmetros. However, our evidence confirms the lower speed of convergence for metros than for non-metros. The significant parameter for spatial interactions was found only for the nonmetros. The results indicate the crucial role of national factors for identifying inequalities in the convergence and spatial dependence.