Page 9 - IDEA Studie 07 2023 TACR

ARE SUBSIDIES TO BUSINESS R&D EFFECTIVE? REGRESSION
DISCONTINUITY EVIDENCE FROM THE TA CR ALFA PROGRAMME IDEA 2023
The prime reason, up to now, why no study has been able to apply the RD design toestimate the effects of subsidies on R&D inputs is probably that the requisite administrative microdata from the evaluation of project proposals, such as project scores or rankings, is not only difficult to obtain from funding providers, but is also difficult to merge with information from R&D surveys. This is because R&D surveys are typically collected by national statistical offices and are only available for research purposes at the firm level in an anonymized form, making it difficult to merge them with other sources. We have been able to overcome these challenges thanks to research cooperation with the Czech Statistical Office in the framework of the OECD microBeRD project (OECD, 2020), which allowed us to merge the official R&D microdata with project data from TA CR.
In the Czech Republic, econometric estimates of the impact of R&D subsidies on business enterprises remain scant in general, despite repeated calls for systematic assessment of their impact based on principles of what has been dubbed “counterfactual evaluation” in the evaluation community (Arnold and Mahieu, 2011, Office of the Government of the Czech Republic, 2015, Srholec, 2015, and Horák, 2016). Previous studies have been limited to using matching estimators to infer causality based on observables without information on R&D activities (Sidorkin and Srholec, 2017, 2021; TA CR, 2019; Ratinger et al.; 2020). The only exception is an analysis by Palguta and Srholec (2016), a pilot attempt to use the RD design to examine input additionality of R&D subsidies that have been awarded in only one out of four calls of the ALFA programme. This paper follows up with significantly more complete and extensive examination. Hence, for the first time in the Czech context, this line of work brings much-needed evidence on the effectiveness of business R&D subsidies based on a research design that allows for relatively reliable causal interpretation of the estimated effects.
7