R&D grants and R&D tax credits in Belgium : evidence on the policy mix

Drawing on a longitudinal database of Belgian firms over the years 2014-2020, this study investigates the joint effect of R&D grants and R&D tax credits on R&D inputs and innovation outputs. We estimate Conditional Difference-in-Difference (CDiD) models and apply both treatment effects estimators that account for heterogeneous, staggered treatments as well as standard two-way fixed effects DiD estimators. We find positive treatment effects for both grants and tax credits on R&D employment, R&D employment intensity, and total R&D expenditures. R&D tax credits have a significant positive impact on the share of sales of new or improved products. By comparing the results obtained by the two econometric methods, we also find that the standard two-way fixed effects models may lead partially to potentially wrong conclusions about the impacts of such policies, as the traditional estimators may not sufficiently account for the complexity of how the policy instrument affect firm-level outcomes.