Is education important in assessing the impact of institutions on economic growth in emerging economies?

ABSTRACT

In this article, we empirically revise the hypothesis that institutions cause economic growth for emerging countries starting from a theoretical model. Our sample consists of 21 countries covering different zones: European Emerging, Asia Pacific Emerging, Latin America, Middle-East and Africa while the status advanced versus secondary emerging countries based on FTSE (Financial Times Stock Exchange) classification is accounted for. The period analysed is 1995–2014. The methodology is based on System GMM estimator of Arellano-Bover and Blundell-Bond for dynamic panel data. Empirical findings suggest that only variables such as voice and accountability and government effectiveness have a significant positive impact on economic growth rates of the analysed countries. In the presence of control variables, i.e. trade and government final consumption, results are robust. Results remain robust for countries that have a high level of government expenditure on tertiary education which proves the role of education in assessing the impact of institutions on economic growth.

KEYWORDS:

• Institutions
• economic growth
• System GMM
• dynamic panel
• emerging economies

JEL CLASSIFICATION:

• F43
• C23
• O57

Acknowledgements

We would like to thank participants at the EEFS Conference Amsterdam 2016 and INFER Annual Conference Reus 2016, for helpful comments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Brazil, Chile, China, Colombia, Czech Republic, Egypt, Hungary, India, Indonesia, Malaysia, Mexico, Morocco, Pakistan, Peru, Philippines, Poland, Russia, South Africa, Thailand, Turkey and United Arab Emirates.

² Graphs with stylized facts’ evolution can be provided upon request.

³ We have also tried various interactions between institutional variables. Results can be provided upon request. In many cases, we did not obtain significant results. One explanation is that the variables are not continuous in time and their variability is low, case that remains the same even when interactions are considered. This fact, technically, could be a reason of obtaining low and insignificant coefficients.

⁴ Tables could be provided upon request.

⁵ Tables with results can be provided upon request.

Additional information

Funding

This work was supported by a grant of the Romanian National Authority for Scientific Research and Innovation, CNCS – UEFISCDI, project number PN-II-RU-TE-2014-4-0443.