6. Conclusions and discussion
The literature on R & D subsidies largely agrees on their usefulness as a policy tool to trigger additional R & D in the private sector (e.g. González and Pazó, 2008; Carboni, 2001,b; Takalo et al., 2013a,b). Despite the broad range of settings in which R & D grants have been evaluated, we still knew little about possible differences in the responsiveness of research versus development activities to this policy instrument. Possible differences between general R & D subsidy schemes and schemes dedicated to research versus development projects remained unexplored. With higher outgoing spillovers, higher risk and constrained access to external financing, gaps between social and private returns as well as market failures tend be more severe for research than for development projects. Research subsidies may therefore yield higher additionality effects than development subsidies. At the same time, research and development are complementary activities, with investment in one increasing the productivity of the other (Cassiman et al., 2002). Targeted schemes are therefore likely to generate cross-scheme effects, with research grants having knock-on effects on development expenditures and vice versa.
This study contributes to the literature on R & D subsidies by estimating the direct and cross scheme average treatment effects on both research and development investments in recipient firms. Making use of a policy design that explicitly distinguishes between research projects, development projects and mixed R & D projects, matching estimations confirm that, on average, grants increase net R & D spending. The results thus reject total as well as partial crowding out. More importantly, when decomposing the type of grant and the type of investment, the findings show that research grants yield higher average direct effects than development grants. In addition, there are significant cross effects from research grants on development expenditures and from development grants on research expenditures. The results also show that mixed grants, which support both research and development activities, turn out to trigger additional research, but not development intensity. Furthermore, as the funding agency gradually moved from mixed schemes towards targeted schemes over time, overall R & D additionality increased.
