Page 16 - IDEA Studie 07 2023 TACR

ARE SUBSIDIES TO BUSINESS R&D EFFECTIVE? REGRESSION
DISCONTINUITY EVIDENCE FROM THE TA CR ALFA PROGRAMME IDEA 2023
The database provides a broad range of dependent variables that are relevant for testing the additionality effects, which fits well with the fact that the ALFA programme not only provided subsidies for business R&D, but also mandated the recipients to produce specific R&D outputs for the project to be deemed successfully completed, including patents as one output. As noted earlier, at the general level, the programme was justified by increasing competitiveness, and thereby, probably increasing growth and productivity. We therefore consider effects of the treatment on the following:
1. R&D inputs – R&D expenditures, not only total, but also by the source of funding (private vs. public) and the type of R&D costs (current vs. capital).
2. R&D outputs – patent applications filed in the Industrial Property Office of the Czech Republic.
3. Economic performance – employment (full-time equivalent), sales and labour productivity (value added per employee).
In addition, we use a number of other variables as covariates and to test the underlying assumptions of the RD design, including time since incorporation, whether the firm is foreign-owned, incorporated as a stockholding company, classified in the manufacturing sector, and located in the capital city of Prague, and proposal-level characteristics with regards to the number of partners in the consortium and whether the consortium included a research organization. For more detailed definitions of the variables, see Appendix Table A1.
Throughout the analysis we make a distinction between small and medium size enterprises (SMEs) versus large firms, defined as firms with 250 or fewer employees versus firms with more employees in the pre-treatment year (t). First, this is because the maximum subsidy of 3 million EUR is relatively small and with data at the firm-level it is difficult to trace the impact of the subsidy in large firms, with typically much larger overall cash flows. Admittedly, for large firms, receiving such a subsidy often amounts to a mere accounting operation with little to no impact on actual allocation of resources. Without more fine-grained data for sub-units, which is not available to us, the estimates we present could prove to be rather imprecise for large firms. Second, existing studies on additional effects of R&D subsidies, as we review in the introduction, confirm that there are noteworthy differences between SMEs and large firms. For the sake of transparency, we report results for both groups of firms, but the estimated coefficients for large firms should be interpreted with caution.
14