Determinants of Growth and Business Cycles: Theory, Empirical Evidence and Policy Implications: INFER Annual Conference 2003

Regional Convergence in Central and Eastern Europe: Evidence from a Decade of Transition (p. 165-166)

Bernhard Herz, University of Bayreuth, Germany Lukas Vogel, University of Bayreuth, Germany

1 Introduction

Over the past ten years, the Central and Eastern European transition economies have experienced considerable rates of economic growth, exceeding those of Western European countries. After this period of economic transition and restructuring, a group of eight transition countries (together with Malta and Cyprus) is going to join the European Union in 2004. The Eastern enlargement of the EU is widely perceived as a big challenge not only for the accession economies but also for the European Union. One dimension is the regional economic development in an EU-25.

Over the past decades, regional development and, especially, the development of economically lagging regions within the European Union have been priorities of the European Commission’s economic policy. With the Eastern enlargement the regional disparities will widen, thus reinforcing this policy issue. The question then is whether regional disparities in Central and Eastern Europe are going to diminish over time, and whether the Eastern European regions are going to catch up with the Western European ones.

In this context, the paper investigates the regional development in Central and Eastern Europe. In particular, we are interested in the prospect for regional economic catch-up and convergence of the accession countries. We consider regions as small open economies. Given the dynamics induced by the liberalisation of trade and factor flows since the early 1990s, we ask whether there are discernible patterns of economic development, in particular whether lagging regions tend to catch-up with richer ones. By estimating empirical growth models, we search for structural determinants of regional growth performances.

Most empirical studies on growth and convergence in transition economies have focused on the country level (e.g. Campos (2001), Eichengreen / Ghironi (2001), Fidrmuc (2000)). While these contributions analysed cross-sections of countries, we focus on CEEC regions. Methodologically, the paper differs from the above-mentioned contributions in that we do not only report cross-section estimates of growth equations for the period 1991-2002, but also exploit the time-series information in our data. The empirical approach is comparable to a recent contribution by Tondl / Vuksic (2003), who investigate regional development in a sample of 36 Eastern European regions over the period 1995-2000. However, whereas Tondl / Vuksic (2003) use a growth-accounting framework and focus especially on the role of economic geography, the present paper relies on the "convergence methodology". To give an account of the regional economic growth during economic transition, we compare the results from a cross-section of regions for the period 1991-2002 with results from two sub-periods (1991-1996, 1996-2002) and with panel estimates. We investigate the development of regional gross-value added per capita and per employee and find a pronounced reduction in regional disparities during the first years of transition. Thereafter, regional dispersion has remained stable. Our empirical findings reject unconditional convergence.

However, we find evidence for conditional convergence, i.e. poorer regions conditionally grow faster than richer ones. Our results further suggest that structural factors, e.g. the rate of labour participation, the share of agriculture and manufacturing in total employment, and the level of education, are relevant for the regional growth performance. We also find a strong influence of country characteristics. As a consequence, regional disparities between countries have diminished whereas, on average, they remained stable within countries. The results from splitting the sample in two sub-periods also suggest that conditional convergence is a phenomenon of the first half of the 1990s, but not of the second half. The ability of our control variables to explain the variation in income or productivity growth substantially weakens in the period 1996-2002.