Giuseppe Arbia, Faculty of Economics, University "G. d'Annunzio" of Pescara-Chieti, Italy, Pescara, Italy, Julie Le Gallo, IERSO (IFReDE-GRES)
Université Montesquieu-Bordeaux IV
, Pessac, France, Gianfranco Piras, Faculty of Economics, Tor Vergata University of Rome, Italy, Rome, Italy
A meta-analysis of regional economic convergence of the NUTS-2 European regions, 1977-2002 (assigned to theme
The study of the geographical distribution of income and wealth and its evolution in the European Regions has attracted the interest of many researchers in recent time and a growing attention has been devoted to this topic in applied econometric literature (for a review see Islam, 2003). Perhaps the most popular theoretical framework used to quantify the speed of economic convergence among countries is the one proposed by Barro and Sala-i-Martin (1995) (referred to as $\beta$-convergence model) in which a primary role in the explanation of the income growth is played by the initial level of the per-capital GDP. Due to the nature of the empirical data used, some of this literature pays the right attention to the spatial econometric methodology needed for a sound statistical inference on regional data (see e.g. Fingleton, 2003). On the other hand various approaches have been proposed as alternatives to the popular beta convergence model in order to better capture the dynamics of per-capita income. These approaches range from panel data models (Arbia and Piras, 2004), to continuous time stochastic equations models (Arbia and Paelinck, 2003a, 2003b), to Bayesian spatial autoregressive model (Ertur et al., 2003), to the idea of unit roots and stochastic convergence (Arbia and Costantini, 2005) to the concept of convergence clubs (Durlauf and Johnson, 1995; Baumont et al., 2003). Each of these models presents relative advantages and drawbacks. In some sense, a trade-off between simplicity and methodological satisfaction has been experienced. Furthermore the results deriving from these different models are not clearly comparable with those obtained using the simple beta-convergence conceptual framework. The aim of this paper is to provide a key for the interpretation of the various results obtained in the literature. In order to achieve this aim we estimate different models on the same dataset and we make the comparison of the various results obtained trying to provide a picture of the distribution of European income and wealth and its evolution in the last 25 years. The data used are the per-capita GDP (expressed in millions of euro 1995) drawn from the Cambridge Econometrics European Regional database including 125 NUTS2 regions belonging to 10 European Countries (namely Belgium, Denmark, France, Germany, Luxembourg, Italy, Netherlands, Portugal, Spain and UK) for the years ranging from 1977 to 2002.
 | Conference organized through conf-vienna
(copyright Gunther Maier) |