Assessing the Performance of National Innovation Systems with a Helix-Based Composite Indicator: Evidence from 24 European Countries

ABSTRACT

We employ the conceptual framework of “N-tuple helices” and devise a composite indicator, namely SFIGA (Society-Finance-Industry-Government-Academia), that aims at measuring the innovation performance of 24 OECD European countries from 2008 to 2017. We also investigate the effects of actors-helices on GDP per capita. The preliminary results from panel data regressions (OLS, quantile, and GMM) indicate their decisive economic importance as key actors of national innovation systems. However, the negative impact of academia on economic growth may explain a highly required resource allocation for basic research in the “knowledge economy” era. Its potential positive economic effects probably require time and patience.

KEYWORDS:

• N-tuple helices
• composite indicator
• innovation performance
• economic growth
• Europe

JEL CLASSIFICATION:

• C23
• O32
• O47
• O52

Acknowledgments

The author would like to thank the three anonymous reviewers for helpful critical feedback in an earlier version of the paper. The author is also grateful to Stavroula Alexiou (University of Thessaly) for her helpful remarks which tackle issues in this paper. The author would also like to thank the participants of the 19^th Triple Helix International Conference, 16-18 June 2021, that took place at the University of São Paulo for freely sharing their thoughts, and suggestions, for an earlier version of this paper that has been presented. The usual disclaimer applies.

Disclosure Statement

No potential conflict of interest was reported by the author(s).

Notes

1. In all indicators selected, the higher values represent better performance, except for the indicator titled “Human Flight and Brain Drain” which, however, has been rescaled.

2. The indicators drawn from World Governance Indicators were initially on a scale of −2.5 (worst) to 2.5 (best). However, we calculated them on a scale of 0 to 5 to avoid negative values.

3. Demand-side indicators have been added as well in the helices of industry and government such as sales, exports, market size, and government procurement of advanced tech products.

4. We must mention that missing data accounted for < 0.27% of the data.

5. Composite indicators are formed when thematic indicators are compiled into a synthetic index and presented as a single composite measure (Freudenberg 2003).

6. The indicator is derived from the database of World Bank and is measured in constant prices ($ 2010).

7. The Sargan test is used to determine whether the econometric models are valid or not, and whether the instruments are correctly specified or not (Ullah, Akhtar, and Zaefarian 2018).

8. To examine the validity of a strong exogeneity assumption, the Arellano-Bond test for no auto-correlation (or no serial correlation) is used under the null hypothesis that the error terms of two different time periods are uncorrelated (Ullah, Akhtar, and Zaefarian 2018).

9. We have also calculated Cronbach’s Alpha (C-alpha) coefficient of reliability for each of the latent constructs-helices. C-alpha in each case is above the 0.70 thresholds of acceptable reliability (Nunnally 1978). More specifically C-alpha is 0.905 for Industry, 0.881 for Academia, 0.836 for Government, 0.875 for Society, and 0.757 for Finance, respectively. So, they are highly likely to share common factors as outlined by our framework and it is evidence that the indicators are measuring the same underlying construct (Nardo et al. 2008).

10. A statistically significant correlation between the composite and another indicator can be interpreted as either: (i) there may be a cause-and-effect relationship, although the direction is not known without other information, (ii) at a minimum, the two factors do not interfere with each other, or (iii) it may just be a spurious relationship, i.e. both indicators are driven by a third one (Freudenberg 2003).

11. Both the SFIGA index and its components (helices) allow for comparisons between (groups of) countries over time. The historical analytical scores of five synthetic indicators are not presented for the economy of space reasons. They are available upon request.

12. It is worth noting that Norway ranked higher and Slovenia lower in our composite indicator than in SII.

13. For an equal weight of helices (λ = 20%) on the construction of the SFIGA index (see Equation 3(3) $SFIG{A}_{index}=\sum _{i=1}^5{\lambda }_i{H}_i$(3) ), the Pearson correlation coefficient between LGDP and SFIGA is 0.928.

14. The authors examine the simultaneous effects of three components (synthetic indices) of technology upgrading on total factor productivity for 16 economies during the 2002–16 period.

15. The diagnostic tests indicate strong autocorrelation, heteroskedasticity, and functional form misspecification, in the models of industry, government, and finance at the 1% statistical significance level.

16. The coefficient magnitude of society is similar in quantiles 10th and 75th, respectively.

17. For instance, Lee and Kim (2009) found that secondary education and political institutions are important for low-income countries, whereas policies facilitating technological development and higher education seem to be highly effective in generating growth for upper middle- and high-income countries.

18. The usual graphical illustration of panel quantile regression results is not shown for the sake of brevity but is available upon request.

19. If the FE model is selected, then the effect of the independent variable is assumed to be identical across all the groups and the regression merely reports the average within-countries effect over time. Alternatively, if the RE model holds then the effect of the independent variable varies randomly within the countries.

Additional information

Funding

This work was supported by the Hellenic Foundation for Research and Innovation (HFRI) under the HFRI PhD Fellowship Grant (Fellowship Number: 455).

Notes on contributors

Apostolos Vetsikas

Apostolos Vetsikas is a Ph.D. Candidate at the Department of Economics, University of Thessaly in Greece. He holds a B.Sc. in Economics, a M.Sc. in Applied Economics, and a M.Sc. in New Entrepreneurship, Innovation and Development (University of Thessaly). His main research interests include Innovation Systems, Applied Economics, and Applied Econometrics.