C31 - Multiple or Simultaneous Equation Models: Cross-Sectional Models; Spatial Models; Treatment Effect Models; Quantile Regressions; Social Interaction ModelsReturn

Evaluation of the economic impact of public research and development (R&D) support has become an integral part of the policymaking process. The paper examines the economic effects of the Czech programme TIP supporting R&D in private companies in the recent decade and aspires to contribute to the debate on quantitative methods for assessing outputs additionality of public R&D programmes. The analysis is based on the counter-factual econometric approach, because an appropriate control group could not be found among the unsupported companies, a so-called generalized propensity score matching (GPSM) was used. Two data sources were used: monitoring data of the programme TIP and economic data from the database Bisnode-MagnusWeb. The impact of R&D support on the firms' performance was measured by output indicators: gross value added, profit and productivity. Treatment effects are diverse depending on firm sizes. Generally, the R&D support effects are positive and more substantial in small and medium firms (SMEs) while there are no such effects confirmed for large firms. The support could only impact positively on the output (profit, productivity and gross value added) of small firms only after reaching a certain minimum threshold of the support whereas output diminishes for medium firms after a point of saturation.
Implications for Central European audience: This paper results could be useful for policymakers and R&D funding providers because it provides evidence about the microeconomic impacts of programmes, namely it shows the differentiated effects of public interventions toby the size categories of beneficiaries. The analysis confirms that GPSM approach provides deliver logical outcomes and if the investigation is situated to the right period after the programme is accomplished, it has a potential to bring more insight in the effects than the standard dichotomous model.