Economic Convergence in the EU: A Complexity Approach

Abstract

This paper examines the effect of economic complexity on economic growth in the context of the EU countries. Our results suggest that a group of countries in the EU whose economic complexity exceeds a certain threshold tend to converge to levels of income corresponding to their measured complexity. On the other hand, the interaction of the current account deficit with economic complexity has important effects on growth for a second group of countries with lower levels of complexity. Since we also find that income convergence is faster within the first group, we argue that convergence is much faster for countries whose economic complexity exceeds a certain threshold.

Keywords:

• economic complexity
• European Union
• growth
• heterogeneity
• income convergence

JEL Classification:

• O11
• O52
• F43

ACKNOWLEDGMENTS

The authors would like to thank Alper Duman and participants of the Seventh Conference on Economic Challenges in Enlarged Europe, held on June 14–16, 2015, in Tallinn, Estonia, for their comments and discussions.

Notes

1. Lower-income new members grew faster than the core members of the EU between 1995 and 2011, except for the years of global crisis starting in 2008, and there were growth disparities among them.

2. Beta-convergence can be unconditional (absolute) or conditional, the latter meaning controlling for other country-specific variables than initial income per capita. Empirical applications of the hypothesis originate with Barro and Sala-i-Martin (1992), Baumol (1986), and Mankiw, Romer, and Weil (1992). We focus on articles that examine real income per capita convergence in the enlarged EU.

3. Employing an econometric analysis adapted from a neoclassical growth model augmented with endogenous technological progress, which varies across countries and over time, Borsi and Metiu (2015) find no evidence of real income convergence within the EU for either 1995–2010 or 1970–2010. Yet they identify a separation between new EU members and the old member countries in the long run. Our article is mostly related to previous literature as we follow the neoclassical growth theory with exogenous technological process.

4. We exclude Luxembourg, Cyprus, and Croatia from our analysis due to data limitations.

5. Hidalgo and Hausmann (2009) worked with the SITC rev. 4 (772 products, 129 countries), the HS at the 4-digit level (1,241 products, 103 countries), and the NAICS at the 6-digit level (318 products, 150 countries). The trade data are from the UN Commodity Trade Statistics.

6. We opted to divide the countries into two groups based on the complexity threshold of −0.25. Separating the two groups at the 0 threshold puts Hungary and Poland in group 2. Our main findings are robust to this particular choice of threshold. The results for this alternative grouping are available from the authors upon request.

Additional information

Notes on contributors

Gül Ertan Özgüzer

Gül Ertan Özgüzer is a Professor in the Department of Economics at Izmir University of Economics, Izmir, Turkey. Ayla Oğuş Binatlı is an Assistant Professor in the Department of Economics at Izmir University of Economics, Izmir, Turkey.