to a lower log GDP per capita in the non-EU samples. More trade openness leads to a higher GDP per capita in all samples, but more human capital leads to higher log GDP values in the full sample only. A higher population density leads to a higher log GDP per capita in the full sample and the EU samples, but it leads to a lower log GDP per capita in the non-EU samples. However, this effect is only weakly significant or not significant at all in the non-EU samples.
When we compare the results of GMM and FE models, we observe significant β-convergence in all cases. Moreover, the convergence speed is always larger in the EU countries than in the non-EU countries when we look at the point estimates. However, the convergence of the non-EU countries seems to be slightly stronger in the FE models, and the convergence of the non-EU countries seems to be slightly stronger in the GMM models. This leads to a larger difference in convergence speed between the EU and non-EU countries in the FE models compared to the GMM models. This could be caused by the fact that we used much more control variables in the FE models compared to the GMM models. When we perform an unconditional FE estimation with only the lagged value of log GDP per capita growth as an independent variable, we get opposite results in which EU coun-tries converge slower than non-EU councoun-tries. However, when we start adding control variables, the convergence speed of the EU countries becomes larger compared to the non-EU countries.
So it seems that unconditional β-convergence is stronger in the non-EU countries and conditional β-convergence is stronger in the EU countries.
Figure 6: Standard deviation
(a) EU and non-EU (b) EU with and non-EU without Schengen
deviation decreases approximately by the same percentage in both regions, so in 2018 it is still much higher in the non-EU countries than in the EU countries.
In Figure 6b, we treat all Schengen countries as EU countries. Now, the initial standard deviation of the non-EU countries reduces heavily and is even lower than the standard deviation of the EU countries. This can be explained by the fact that the non-EU Schengen countries have a relatively high GDP per capita, such that the values of the remaining countries in the non-EU sample are much closer to each other. The standard deviations of the EU countries barely changed, as the non-EU Schengen countries have log GDP per capita values comparable to the values of the EU countries. The standard deviation of the non-EU and EU countries decreases respectively by 38.7 and 25.8 per cent in the years 1994-2018. This means that σ−convergence in terms of standard deviation seems to be stronger in the non-EU countries if the Schengen countries are included in the EU sample.
Figure 7a shows that the initial coefficient of variation is also much higher in the non-EU countries, comparable with Figure 6a. The coefficient of variation of the non-non-EU and non-EU countries reduces respectively by 32.1 and 30.7 per cent in 1994-2018. So we do not observe a large difference in the decrease of the coefficient of variation.
Figure 7: Coefficient of variation
(a) EU and non-EU (b) EU with and non-EU without Schengen
In Figure 7b, the initial coefficient of variation of the non-EU countries is slightly above the value of the EU countries. The coefficient of variation of the non-EU countries decreases by 45.9 per cent in 1994-2018, which is much more than the 29.8 per cent decrease of the EU countries.
So also in terms of coefficient of variation, we see that σ-convergence is stronger in the non-EU countries if the Schengen countries are included in the EU sample.
5. CONCLUSION
In this paper, we examined the convergence of the log GDP per capita of the European countries. We compared the EU and non-EU countries to investigate whether convergence differs in these regions. In this section, we answer the research questions and make suggestions for further research.
Hypothesis 1 claims that we can divide the European countries into convergence clubs and that the EU and non-EU countries form separate convergence clubs. Using data in the period 1999-2018, we find five convergence clubs and one diverging country. When we treat all Schengen countries as EU countries, the first two convergence clubs only consist of EU countries, and the last convergence club only consists of non-EU countries. The only non-EU country in the third convergence club is Turkey, and the only EU countries in the fourth convergence club are Bulgaria
and Croatia. From these convergence clubs, we can conclude that EU and non-EU countries almost form separate convergence clubs, such that hypothesis 1 is true.
Hypotheses 2A and 2B mention that the economies of respectively the EU and the non-EU countries converged after the start of the EU. We look at different convergence models to answer this question, using European country data from 1999-2018. In all models for β−convergence, we see that higher previous log GDP per capita values lead to less log GDP per capita growth.
Therefore, both EU and non-EU countries converge in terms of β−convergence. We find that the standard deviation and coefficient of variation decrease over time in all models. This means that the disparities between the EU and the non-EU countries reduce after the start of the EU, so we observe σ−convergence as well. Therefore, we can confirm hypotheses 2A and 2B.
Hypothesis 3 states that the convergence speed is higher for countries in the EU than for the non-EU countries in Europe. The speed of unconditional β−convergence is higher in the non-EU countries, but the speed of conditional β−convergence is higher in the EU countries in both GMM and FE models. This holds in the standard case, but also when we treat all Schengen countries as EU countries. However, the 95 per cent confidence intervals are overlapping, which indicates that we are not highly certain about these results due to the sampling uncertainty. We observe a comparable speed of σ−convergence for the EU and non-EU countries in the standard case, but if we treat all Schengen countries as EU countries, the speed of σ−convergence is higher in the non-EU countries. σ−convergence does not take the explanatory variables into account, such that it strengthens the argument that unconditional convergence is stronger in the non-EU countries.
Therefore, hypothesis 3 only seems to hold for conditional β−convergence.
In this research, we found quite some differences in the convergence between the EU and non-EU countries. However, it is still unclear whether the EU caused these differences. It could be the case that we would observe similar differences if the EU was never formed and that other factors can explain these convergence differences. Therefore, the role that the EU plays in convergence should be further investigated. Also, regional data could be used to compare the convergence in EU and non-EU regions. Besides, other indicators can be used to measure the convergence of
the economies instead of the log GDP per capita. For example, Kilinc et al. (2017) use different banking and stock market measures to indicate convergence.
REFERENCES
Arbia, G., Gallo, J., & Piras, G. (2008). Does Evidence on Regional Economic Convergence Depend on the Estimation Strategy? Outcomes from Analysis of a set of NUTS2 EU Regions. Spatial Economic Analysis, 3(2), 209-224.
Arellano, M., & Bond, S. (1991). Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The review of Economic Studies, 58(4), 1029-1054.
Arnold, N., & Dagher, J. (2015). Growth opportunities, strategic savings, and the dot-com boom and bust. Economic Inquiry, 53(4), 1850-1871.
Azariadis, C., & Drazen, A. (1990). Threshold externalities in economic development. Quarterly Journal of Economic Development, 105(2), 501-526.
Barro, R. (2015). Convergence and Modernisation. The Economic Journal, 125, 911-942.
Bartkowska, M., & Riedl, A. (2012). Regional convergence clubs in Europe: Identification and conditioning factors.
Economic Moddeling, 29, 22-31.
Bond, S. (2002). Dynamic panel data models: a guide to micro data methods and practice. Portuguese economic journal, 1(2), 241-162.
Bond, S., Hoeffler, A., & Temple, J. (2001). GMM Estimation of Empirical Growth Models. CEPR Discussion Paper(3048).
Bouayad-Agha, S., & Verdrine, L. (2010). Estimation Strategies for a Spatial Dynamic Panel using GMM. A New Approach to the Convergence Issue of European Regions. Spatial Economic Analysis, 5(2), 205-227.
Chen, C., & Hsu, T. (2004). A Novel Approach to Great Circle Sailings: The Great Circle Equation. The journal of navigation, 57, 311-325.
Corrado, L., Martin, R., & Weeks, M. (2005). Identifying and interpreting regional convergence clusters across Europe. The Economic Journal, 115(502), C133-C160.
Dalgaard, C., & Vastrup, J. (2001). On the measurement of σ-convergence. Economics Letters, 283-287.
Den Haan, W., & Levin, A. (1998). Vector autoregressive covariance matrix estimation. University of California, San Diego.
Dey, P. S., & Neogi, D. (2015). Testing Sigma and Unconditional Beta Convergence of GDP for SAARC Countries:
Can Inclusion of China further Consolidate the Convergence? Global business review, 16(5), 845-855.
Dheane, G., & Jochmans, K. (2015). Split-panel Jackknife Estimation of Fixed-effect Models. The Review of economic studies, 82(3), 991-1030.
Durlauf, S., Johnson, P., & Temple, J. (2006). Growth econometrics. Amsterdam, North Holland.
Everaert, G., & Pozzi, L. (2007). Bootstrap-based bias correction for dynamic panels. Journal of Economic Dynamics and Control, 31(4), 1160-1184.
Fiaschi, D., Lavezzi, A., & Parenti, A. (2018). Does EU cohesion policy work? Theory and evidence. Journal of regional science, 58(2), 501-526.
Galor, O. (1996). Convergence? Inferences from theoretical models. The Economic Journal, 106(437), 1056-1069.
Hansen, L. (1982). Large sample properties of generalized method of moments estimators. Econometrica, 50(4), 1029-1054.
Hembram, S., & Haldar, S. (2019). Beta, sigma and club convergence: Indian experience from 1980 to 2015. Economic Review, 54(2), 343-366.
Hurwicz, L. (1950). Least-squares bias in time series. Statistical Inference in Dynamic Economic Models, 365-383.
Islam, N. (2003). What Have We Learnt from the Convergence Debate? Journal of Economic Surveys, 17(3), 309-362.
Kilinc, D., Seven, U., & Yetkiner, H. (2017). Financial development convergence: New evidence for the EU. Central Bank Review, 17, 47-54.
Kiviet, J. (1995). On bias, inconsistency, and efficiency of various estimators in dynamic panel data models. Journal of Econometrics, 68(1), 53-78.
Lyncker, K., & Thoennessen. (2017). Regional club convergence in the EU: evidence from a panel data analysis.
Empirical economics, 52(2), 525-553.
McKelvey, R., & Zavoina, W. (1975). A statistical model for the analysis of ordinal level dependent variables.
Mathematical Sociology, 4, 103-120.
Mileva, E. (2007). Using Arellano – Bond Dynamic Panel GMM Estimators in Stata. Economics Department, Fordham University.
Monfort, P. (2008). Convergence of EU regions Measures and evolution. Working Paper.
Mora, T. (2008). Factors conditioning the formation of European regional convergence clubs. The Annals of Regional Science, 42(4), 911-927.
Neal, L. (2007). The economics of europe and the european union. Cambridge University.
Newey, W., & West, K. (1987). A Simple, Positive Semi-Definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix. Econometrica, 5(3), 703-708.
Nickell, S. (1981). Biases in Dynamic Models with Fixed Effects. The Econometric Society, 49(6), 1417-1426.
Phillips, P., & Sul, D. (2007). Transition Modelling and Econometric Convergence Tests. Econometrica, 75(6), 1771-1855.
Phillips, P., & Sul, D. (2009). Economic transition and growth. Journal of applied econometrics, 24(7), 1153-1185.
Phillips, P., Sun, Y., & Jin, S. (2007). Long run variance estimation and robust regression testing using sharp origin kernels with no truncation. Journal of Statistical Planning and Inference, 137(3), 985-1023.
Pietrzykowski, M. (2019). Convergence in GDP per capita across the EU regions— spatial effects. Economics and Business review, 5(19), 64-85.
Quah, D. (1993). Galton’s Fallacy and Tests of the Convergence Hypothesis. The Scandinavian journal of economics,, 95(4), 427-443.
Quah, D., & Barro, R. J. (1992). Convergence. Journal of Political Economy, 100, 223-251.
Ram, R. (2018). Comparison of cross-country measures of sigma-convergence in per-capita income, 1960–2010.
Applied Economics Letters, 25(14), 1010-1014.
Ramajo, J., Marquez, M., Hewings, G., & Salinas, M. (2007). Spatial heterogeneity and interregional spillovers in the European Union: Do cohesion policies encourage convergence across regions? European economic review, 52(3), 551-567.
Roodman, D. (2009). How to do xtabond2: An introduction to difference and system GMM in Stata. The Stata Journal, 9(1), 86-136.
Simionescu, M. (2014). Testing σ-convergence across EU-28. Economics Sociology, 7(1), 48-60.
Sun, Y., & Dhaene, G. (2019). xtspj: A command for split-panel jackknife estimation. The Stata Journal, 19(2), 335-374.
Tselios, V. (2009). Growth and Convergence in Income Per Capita and Income Inequality in the Regions of the EU.
Spatial Economic Analysis, 4(3), 1742-1772.
Windmeijer, F. (2005). A finite sample correction for the variance of linear e!cient two-step GMM estimators. Journal of Econometrics, 126(1), 25-51.
Young, A., Higgins, M., & Levy, D. (2018). Sigma Convergence versus Beta Convergence: Evidence from U.S.
County-Level Data. Journal of money, credit and banking, 40(5), 1083-1093.
APPENDIX