Linkage between international trade and economic growth in GCC countries: Empirical evidence from PMG estimation approach

Abstract

This paper explores the empirical evidence of the links between economic growth and openness to international trade by controlling for auxiliary variables in the model for the six Gulf Cooperation Council (GCC) countries over the annual sample period 1980–2010. After testing for cointegration based on a recent bootstrap panel test, we employ the Pooled Mean Group (PMG) estimation technique of M.H. Pesaran, Y. Shin, and R. Smith (1999. “Pooled Mean Group Estimation of Dynamic Heterogeneous Panels”. Journal of the American Statistical Association 94: 621–634) that is appropriate for drawing sharper conclusions in dynamic heterogeneous panels by considering long-run equilibrium relations. The results show evidence of cointegration relationship between the variables of interest, and reveal that economic growth responds positively to trade openness over both the short run and long run. The evidence is robust to using various trade openness measures and to alternative model specifications, suggesting thus the non-fragility of the linkage between economic growth and openness to international trade for the GCC region. Our findings are then promising and support the view that economic growth is directly and robustly linked to trade openness for the GCC countries.

Keywords:

• economic growth
• trade openness
• dynamic heterogeneous panels
• PMG approach
• GCC region

JEL Classifications:

• C23
• F10
• F14
• F43
• O53

Acknowledgements

The author is grateful to the editor and the anonymous referees for their helpful comments and suggestions, which have significantly improved the paper. He would like to thank the Deanship of Scientific Research at King Saud University represented by the research center at CBA for supporting this research financially.

Notes

1. See, for example, Mamun and Nath (2005), Reppas and Christopoulos (2005), Shirazi and Manap (2005), Uğur (2008), Gurgul and Lach (2010), Sakyi (2011) and Ulaşan (2012).

2. Trade openness may be considered as source of imported inflation especially for countries that are highly dependent from abroad.

3. This index measures the extent to which the real exchange rate is distorted away from its free trade level by the trade regime.

4. The sample period is considered as sufficiently recent to account for the latest economic events.

5. During these years, the GCC region has been considered as a free trade area.

6. Note that more than 80% of the GCC's total revenue comes from the oil sector.

7. Rivera-Batiz (1985) outlines that imports could be risen due to the increase of economic activity, as high real income improves consumption.

8. Various openness variables are used, namely direct trade policy indicators (tariff rates and non-tariff barriers), the ratio of exports to GDP, the ratio of imports to GDP and the ratio of trade volume (exports plus imports) to GDP. For the last variable, measures from the World Bank and the Penn World Tables are employed.

9. These approaches embody the dynamic econometric panel model, the panel Fully Modified OLS (FMOLS) method, the panel Seemingly Unrelated Regression (SUR) model, the panel cointegration techniques, the panel VAR process and Granger non-causality tests, and others.

10. The readers are referred to Breitung and Pesaran (2008) for a literature overview.

11. The number of breaks is determined based on the Bayesian information criterion, and the break locations are estimated by the OLS method as suggested by Bai and Perron (1998, 2003).

12. The LM statistic is normally distributed under the null hypothesis.

13. The residuals can be obtained by estimating the model by the Fully Modified OLS method proposed by Phillips and Hansen (1990).

14. Pesaran, Shin, and Smith (1999) argue that the obtained PMG estimators are consistent and normally distributed asymptotically.

15. According to Pesaran, Shin, and Smith (1999), the PMG approach can be adapted to allow only a subset of the long-run parameters to be identical while the others differ across groups.

16. The readers are referred to Pesaran, Shin, and Smith (1999) for more details on the MG and DFE methods.

17. Real measures are preferable on theoretical grounds to nominal measures largely and conventionally advocated in the related literature.

18. The nominal variables are taken in current values, while the real variables are measured at constant prices (2005) and constant exchange rates (2005).

19. Throughout the paper, some empirical results are not reported to preserve space, but they are available upon request from the author.

20. The nominal GDP of GCC countries is evaluated at US$1.4 trillion in 2011.

21. Unlike the other GCC countries, the UAE are less dependent on exports and rely on services to boost the GDP.

22. Oil price registers a contraction in 2009 since it decreased by 33% because of the global financial crisis and economic slowdown observed at the end of 2008, causing thus a fall by about 20% in the nominal GDP in the GCC region.

23. These insights show the crucial role of the GCC region in the global economy.

24. A large portion of imports is destined for re-export.

25. Saudi Arabia monopolizes 45% of the labor force in the GCC region.

26. Within the GCC region, the ratio of government expenditures to GDP ranges from 27% in Qatar to 43% in Kuwait.

27. We have applied tests developed in the time series framework to test for unit root in all level and first-differences series for each country. The results (not reported here) outline that we generally conclude in favor of variables integrated of order one.

28. We have applied other panel cointegration tests and found long-run linkage between the variables.

29. For the UAE, the break date detected for all models is located in 1988.

30. The PMG short-run estimates are the means of the coefficients obtained by averaging across countries.

31. We have also estimated the models with only the openness measures as explanatory variables as they are the main variables of interest. The results (not reported here) confirm the significant positive linkage between economic growth and trade openness.

32. This finding is in line with several empirical works in the literature mentioning that investment is among traditional determinants that positively affect economic growth.

33. The growth models consider FDI as factor of economic growth through the role played in technological diffusion. Faras and Ghali (2009) opt for the ARDL approach, and find that the contribution of FDI inflows to economic growth differs across GCC countries.

34. The variables are taken in logarithm except of FDI that contains negative values.

35. The LM cointegration test presented in this paper is not applied when adding the new determinants because the number of explanatory variables exceeds the number of groups in this case. Therefore, we have employed other tests developed in the panel framework to test for cointegration between variables.

36. The unit root and cointegration test results are not reported, but they are available upon request.

37. The Hausman test concludes in favor of common long-run coefficients as shown in Table 11.