Governments have a long history of subsidising private research and development (R&D), but there is still little evidence regarding which mechanisms actually produce innovation outcomes and how policies interact. This paper develops a quasi-experimental method that combines difference-in-differences and regression-discontinuity research designs to study how interacting R&D support schemes and their generosity impact innovation. Importantly, the approach allows for the effects of indirect funding (tax incentives) and direct funding (grants) to be disentangled. I build a database of firm performance and innovation measures by linking several micro-level datasets of UK firms. Preliminary results show that tax credits and grants are complementary for producing some types of innovations, but not all. Continued work explores why this is the case.