摘要：This article tests the hypothesis that the geographical location plays a major role in dif- ferentiated economic integration of central and eastern regions at different geograph- ical scales. We implement a local measure of Gini index allowing to measure the eco- nomic catching-up of regions (across the period 1995-2007 at the NUTS 2-3 level). The findings indicate that a convergence of the CEEC toward the EU-15 seems to take place but at the cost of widening regional inequalities within each state. Our results suggest also the existence of a west-east gradient of regional economic integration. var currentpos,timer; function initialize() { timer=setInterval("scrollwindow()",10);} function sc(){clearInterval(timer); }function scrollwindow() { currentpos=document.body.scrollTop; window.scroll(0,++currentpos); if (currentpos != document.body.scrollTop) sc();} document.onmousedown=scdocument.ondblclick=initialize 54 Bourdin Convergence and economic integration Reproduced with permission of the copyright holder. Further reproduction prohibited. ground. According to the works by Rey (2003, 2004), we hypothesize that the emphasis placed on a global measurement of regional inequality may hide different regional trajecto- ries that may have explicit spatial rep- resentations reflecting pockets of poverty, convergence clubs and other forms of geographical grouping. The first part of this article pre- sents the theoretical foundations and conceptual framework of our analysis. We then present the methodology and data used. The third part of the ar- ticle sets out the results of this ap- proach which aims at demonstrating that the process of convergence and regional integration in Central and Eastern Europe is geographical by na- ture. Theoretical foundations and concep- tual framework The neoclassical growth theory devel- oped by Solow (1956) stating that economic convergence between countries with uneven development is possible, is at the origin of work on economic convergence. He works on the assumption that there are regional disparities at the beginning and that these disparities tend to decrease over time. Each region converges towards a long-run income per capita growth rate called .steady-state rate. 1 . This explains how a country's economy tends to converge more rapidly if it is below its steady state. The beta and sigma convergence econometric and statistical tests performed on a wide sample of countries, over a period of nearly thirty years, lead us to reject the catch-up hypothesis. 2 In other words, the growth of the countries that are initially the least developed is not systematically quicker than that of developed countries. Subsequent to these studies, new theories on economic convergence were developed. They are based on Myrdal's research ...... which states that growth is a spatially cumulative process likely to increase ineq ualities. This saw the emergence of new en- dogenous growth theories resulting from the introductory work of Romer (1986) and Lucas (1988). These models underline the lack of .diminishing re- turns to capital., which is related to the endogenous nature of .produ c- tion technology.. The level of human capital, the amount of investment in R&D or even .knowledge spillovers. are explanations for divergent growth trends. Another theoretical movement developed following the lead by Paul Krugman (1991): the New Geographic Economy (NGE). It acknowledges that economic activities are focused on clusters and this polarization modifies the spatial distribution of wealth be- tween regions (Fujita & Thisse, 1996). Williamson (1965) is at the origin of the emergence of this economic trend as the author was already highlighting the role of space to explain regional growth at that time. The role of space in this economic theory is essential, as it helps explain the phenomena of economic growth. The geographical distribution of regional growth phe- nomena is hardly ever random: on the contrary, the economic performance of adjacent regions is often similar. Studies on convergence resulting from the NGE show that some coun- tries manage to take more advantage of growth, while others fail to do so. We can thus observe two phenomena: economic convergence for some countries, and divergence for others in a synchronous movement. The result of this process is not predetermined, and the new theoretical models that highlight the role of geographic space show that there can be both economic convergence and divergence. As a re- sult, the effects of economic conver- gence on regional integration can be both positive and negative. It there- fore depends a lot on the initial situa- tion, on the ab ility of the regions to adapt and on the neighbourhood ef- fects. This article proposes the meas- urement of local convergence in this framework, to assess the simultane- ous phenomena of economic conver- gence and divergence, using the breakdown of inequalities on a local scale. The neoclassical theory implies that uneven development between countries tends to decrease over time. Having said this, it does not give a sat- isfactory explanation for the long-run factors of economic growth. This is why we have favoured an ap proach based on the arguments of the en- dogenous growth theory (Romer, 1986; Lucas, 1988) and geographic economy in our article, as they show that the economic situation of a re- gion depends on its interrelations with its neighbours. Measuring economic convergence at local level is of interest as it enables us to understand the origin of the per- sistence or reduction in regional dis- parities in some parts of the EU. Local convergence can be defined as a situa- tion in which convergence rates in terms of economic growth are similar for observations located nearby in space .Ertur, Gallo, & LeSage, ...... )n other words, there could be a spatial clustering of regions with similar sig- ma convergence coefficients, thus confirming the geographical nature of the regional catch -up process. We can therefore estimate a local conver- gence index for each region in our sample, then investigate whether these estimations provide empirical confirmation of our concept of local convergence. The interest is therefore to consider both economic conditions and geographic proximity as being po- tential influences on economic con- vergence. For instance, the lack of catch-up .even divergence. of a lag- ging region could be explained by the fact that it is surrounded by other lag- ging regions with similar convergence rates. Local convergence measures the reduction (or increase) in economic disparities within a group of adjacent regions. In other words, there is local convergence when the GDP per capita of the regions included in the deline- ated area tend towards the average level of the GDP per capita of the area in question. Methodology and data During the last decade, the empirical research on economic convergence has rapidly developed and often pro- duces contradictory results (Quah, 1996; Islam, 2003; Abreu, de Groot, & Florax, ....; Dall'erba & Le Gallo, 2008; Le Pen, 2011). The origin of the
