The dual role of R&D expenditures in European Union’s member states: short- and long-term prospects

Abstract

The objective of this article is to assess a potential dual role of public expenditures in R&D upon economic growth and employment, using these dimensions as partial representations of the socioeconomic state of affairs in European Union’s Member States. First, we look into direct, short-term impacts arising from R&D expenditures, much in the sense of a multiplier effect. Second, we analyse impacts from the stage of development of National Innovation Systems (NIS) upon the macroeconomic conditions of interest, assuming that current stages of development are products of previous commitment to innovation, that is, a structural, long-term outcome of innovation-oriented investments. In order to empirically test our propositions, we have analysed 28 EU Member States (1990–2013) through three sets of econometric (static and dynamic panel data) models. Results highlight that EU countries’ governmental commitment to their respective innovation systems catalyses current and prospective economic growth and employment levels, suggesting a complementarity between Neo-Schumpeterian and Neo-Keynesian perceptions over governmental R&D involvement. This can bring innovation efforts closer to the macroeconomic debate on monetary and fiscal policies and function as a criticism to austerity measures, as this may not only affect the present socioeconomic situation, but also generate the cornerstone for a perennial state of divergence among EU Member States.

Keywords:

• Innovation policy
• European Union
• R&D expenditures
• national innovation systems

JEL:

• O38
• B52
• E61

Acknowledgements

Authors’ would like to acknowledge valuable contributions from anonymous reviewers and from the Editor, Dr Ronald Pohoryles.

Disclosure statement

Authors do not report any conflict of interest involved in this research.

Notes

1. The OECD (2009) proposes four key areas for governmental intervention concerning the recovery from the 2008 crisis: (i) infrastructure, (ii) science, R&D and innovation, (iii) education and (iv) green technologies. The potential for overlaps between different dimensions is properly addressed by the OECD.

2. The Framework Programme became known ad Horizon 2020 after 2013.

3. These divergence trends among EU Member States are not restricted to the STI environment, causing peripheral countries to reduce governmental involvement with social issues (Censolo and Colombo 2016), negatively affecting welfare states (Guillén, González-Begega, and Balbona 2016).

4. Even though these dimensions do not comprehend the entire reality behind the dynamics of socioeconomic systems, they provide a robust starting point for this discussion.

5. The proxy used for “Institutions” identify the institutional quality of countries. This is included as a dimension of innovation systems as there is an expectation that institutions drive the efficiency in the relationship between innovation inputs and outputs.

6. In this case, the penetration of internet among the population is used as a proxy for the “knowledge society” and infrastructure development.

7. The factor analysis reduces the set of existing variables to a set of non-observable hypothetical or theoretical factors which summarize most of the information contained in the original set of variables. A similar procedure is undertaken by Fagerberg, Srholec, and Knell (2007) and Fagerberg and Srholec (2016) for estimation of technology and capacity competitiveness.

8. We cannot rule out the potential endogeneity of this variable, that is, the possibility of GDP growth attracting FDI rather than FDI influencing growth. However, this discussion lies outside of the scope of this article. For further discussions on this issue see Nair-Reichert and Weinhold (2001).

9. Interestingly, while inflationary levels seem to reduce unemployment, they negatively related to GDP growth. This conflicting outcome deserves further attention in future analyses dealing with monetary policy in the EU context.

10. We did not apply longer timeframes because of missing data issues. Also for this reason, Luxembourg and Malta are not represented in this analysis.

11. We did not apply longer timeframes because of missing data issues. Also for this reason, Luxembourg and Malta are not represented in this analysis.

Additional information

Funding

This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo [grant number 2016/17801-4].