Convergence Process in the European Region - Cluster Analysis

Abstract

The primary aim of our research was to assess the character of the convergence process in member countries of the European Union during the period 2002–2012. For this purpose, we applied cluster analysis, a multivariate statistical technique. The results of the cluster analysis allowed us to classify countries into several groups with similar features. In this context, we compared cluster analysis results for single years from period 2002–2012. This comparison allows us to measure whether certain changes exist in defined clusters because changes in particular clusters can be interpreted as divergence or convergence tendencies. We analyzed whether the group (which represents the Eurozone core) was defined more broadly during the analyzed period or not. A more broadly defined core suggests that certain countries converge to the Eurozone core. Data for the convergence criteria for individual countries were obtained from the Eurostat, OECD, and IMF databases, or from the statistical offices of the countries under analysis. An important contribution to this research area was written by Artis and Zhan (International Journal of Finance and Economics, 1997), which classified a sample of 18 developed countries around the globe into five groups. Three clusters were identified in the European context: core (France, the Netherlands, Belgium and Austria), northern periphery (Denmark, Ireland, Sweden, Norway, Finland and the UK) and southern periphery (Italy, Spain, Portugal and Greece). Cluster analysis was also used to explore the convergence process by Krcilkova (Agrarian Perspectives XV, 2006), who clustered the countries in the European Union using growth in the gross domestic product (GDP), the inflation rate derived from the GDP deflator, the unemployment level, real interest rates and the public budget deficit versus GDP as convergence criteria. In her conclusions, she provides evidence for significant convergence in the economies Int Adv Econ Res (2014) 20:459–460 DOI 10.1007/s11294-014-9488-6