Page 31 - IDEA Studie 07 2023 TACR

ARE SUBSIDIES TO BUSINESS R&D EFFECTIVE? REGRESSION
DISCONTINUITY EVIDENCE FROM THE TA CR ALFA PROGRAMME IDEA 2023
Conclusion
This study provides comprehensive evidence of causal effects of the TA CR ALFA subsidy programme. The results of RD estimators, based on micro data for business sector participants near the cutoff, indicate that the programme had strong and persistent additionality effects on R&D inputs, but only in SMEs, and not in large firms. In addition, the results are inconclusive for additionality effects on R&D outputs and economic performance irrespective of the size of firms and either during or after the project.
The results raise a number of questions about the design of policy interventions in this area, which are of great importance for evidence-based decision making. First, the results indicate that similar programmes in the Czech Republic, including the follow-up programmes to ALFA, could become more efficient by reallocating money from funding large firms to SMEs. It is noteworthy that some of the most prominent programmes abroad, such as R&D subsidies in the Small Business Innovation Research (SBIR) program in the United States and the Small and Medium Enterprise Instrument (SMEI) of the European Commission, target not only small but specifically young innovative firms and start-ups, whereas their Czech counterparts support relatively established firms. It is striking that the median age of subsidy recipients in both programmes above is 5 years, whereas in ALFA, it is 19 years. Because small and young firms are also more likely to be cash and credit constrained, policymakers should seriously consider shifting the focus of the support to these groups of firms.
Second, if this type of subsidy continues to be small compared to the size of their recipients, and are thus relatively evenly spread across a large population of firms, we might never be able to find out whether the programmes deliver impacts on economic performance and competitiveness of the economy, on the grounds of which the subsidies are primarily justified. Admittedly, this should be acknowledged ex-ante when funding for these programmes is considered by the government.
Third, interestingly, the inputs additionality effect appears to be negatively associated with the extent of co-funding requirement by the recipient from non-public sources – the maximum proportion of eligible costs of the project that could be covered by the subsidy – because the latter was significantly higher in large firms than in SMEs, while the oppo- site holds for the former. In other words, large firms were required to co-fund a larger share of project costs but the subsidy actually turned out to have a smaller input- additionality effect on them. Hence, the co-funding requirement must not be confused
29