Skip to main content
Log in

R&D tax incentives in EU countries: does the impact vary with firm size?

  • Published:
Small Business Economics Aims and scope Submit manuscript

Abstract

This paper studies the effect of R&D tax incentives on the research activity of manufacturing firms based in France, Italy, Spain and the UK, over the period 2007–2009. Using a matching procedure, we show that, in all the examined countries but Spain, R&D tax incentives induced a statistically significant increase in the intensity of R&D expenses over sales. However, this effect is driven only by the behaviour of small firms. By assessing the benefit-cost ratio of R&D tax policies, we find evidence of substantial additional effects in Italy and the UK.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1

Similar content being viewed by others

Notes

  1. Industry-level evidence for the USA indicates that fiscal incentives to R&D have an impact on aggregate output growth well beyond their positive effect on R&D engagement (Minniti and Venturini 2017a). Minniti and Venturini (2017b) document that, in the USA, this impact is channelled by a faster rate of TFP growth, induced by the R&D tax policy, which is stronger for firms far from the technology frontier.

  2. Assessment of the “output additionality” of R&D consists in checking whether tax incentives do stimulate innovation outputs, such as patents and product innovations. Examples of these studies are Czarnitzki et al. (2011), Cappelen et al. (2012), Bozio et al. (2014), Bodas Freitas et al. (2017), Bösenberg and Egger (2017).

  3. Albeit earlier works based on structural models indicate that the demand for R&D is not very sensitive to changes in R&D tax price, the latest evidence seems more supportive. Parsons and Phillips (2007) and CPB et al. (2014) survey the wide micro-econometric literature concluding that the demand elasticity of R&D to its tax price ranges from 1 and 2% in the long run and between 0.20–0.30% in the short run. Similar findings are obtained by Bloom and Griffith (2002) using country-level data.

  4. For the sake of simplicity, we neglect the presence of other special fiscal allowances for R&D outlays.

  5. The database was collected within the European Firms in a Global Economy (EFIGE) project, supported by the Directorate General Research of the European Commission through its 7th Framework Programme. The original sample was identified along three dimensions of stratification: industries (11 NACE-CLIO industry codes), regions (at the NUTS-1 level of aggregation) and size classes (10–19; 20–49; 50–250; more than 250 employees).

  6. In spite of the well-known fact that small firms do not carry out (or do not report) formal R&D activities, the share of small companies on total R&D performing firms ranges from 62 to 63% in France and the UK and to 73–74% in Spain and Italy. SMEs represent a large majority of the total sample, ranging between 88.6% in France to 93% in the UK.

  7. As a robustness check, we also employed a nearest neighbour matching (NNM) procedure, with NN = 1 (with replacement) and the imposition of a common support. The results, available from the authors upon request, are consistent with those achieved with the kernel algorithm. However, according to the usual covariate balancing tests, the quality of the matching turns out to be lower than that yielded by the kernel procedure (see Table 13 in the “Appendix” section).

  8. Large firms represent a handful of companies in the working sample and hence are grouped together with medium-sized firms.

  9. Table 13 in the “Appendix” section reports common joint significant tests of the balancing properties of the covariates (i.e. the firms’ characteristics included in the probit regressions) arising from the kernel matching. For all country and size sub-sample, all the joint tests indicate a good quality of the matching. In fact, after the matching, the mean bias in the covariates (i.e. the mean of the absolute differences between the treated and untreated sub-samples) is remarkably lower; the likelihood ratio (LR) test rejects the hypothesis of joint significance of the covariates. Similarly, the pseudo-R2 approaching to zero suggests a successful matching.

  10. Moreover, testing the additionality of R&D fiscal incentives for Spanish firms seems redundant because the estimated effect of such incentives is not statistically significant (cf. Table 5).

References

Download references

Acknowledgements

With the usual disclaimers, we wish to thank two anonymous reviewers for their helpful comments and suggestions.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Alessandro Sterlacchini.

Appendix

Appendix

Table 13 Covariate balancing tests for kernel matching

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sterlacchini, A., Venturini, F. R&D tax incentives in EU countries: does the impact vary with firm size?. Small Bus Econ 53, 687–708 (2019). https://doi.org/10.1007/s11187-018-0074-9

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11187-018-0074-9

Keywords

JEL classifications

Navigation