By Otto Toivanen,
We estimate the welfare effects of government support to private R&D using comparable R&D project level data from Belgium, Finland, and Germany. In a counterfactual analysis we evaluate the existing policies against alternative policies, including first best. There is considerable heterogeneity in R&D investments, R&D participation rates, spillovers, and profits across firms. Socially optimal R&D participation rates are only marginally higher than those observed in the data, suggesting that most of the benefits from activist policies come from increasing R&D in firms already doing R&D rather than from enticing new firms to start R&D. We find that activist policies increase r&D substantially, but have essentially no effect on welfare. We also find that the gap between laissez-faire and first- and second-best policies is narrow at 3-4 per cent. EU-wide innovation policy is clearly more effective than national ones.