The temporal effects of R&D subsidies on R&D, innovation and innovation behaviour: evidence from UK

This study contributes to the understanding of the temporal dimension of the effects of R&D subsidies on R&D inputs, outputs and innovation behaviour. We argue that while the time dimension of the effects on R&D inputs (i.e. input additionality; IA) tightly relates to public support programmes’ requirement of private contribution in R&D spending, the effects on innovation (i.e. output additionality; OA) and innovation behaviour (i.e. behavioural additionality; BA) are temporally delayed, linked to the explorative nature of learning associated with publicly funded projects, and obstacles to changes in the innovation behaviour of firms. We empirically test our theoretical framework for a sample of highly R&D intensive, UK firms to find that while IA occurs in the year following receipt of R&D subsidies, BA takes three years after receipt of R&D subsidies to occur. We also find some evidence that OA is mainly driven by process innovation and it takes (at least) three years to occur. In measuring BA, we employ for, the first time, a ‘classical’ treatment and control group evaluation without relying on firms’ perceptions of how the receipt of public subsidies affected their innovation behaviour. Our findings have significant implications for both policy and evaluation.