The study utilizes a dataset with enterprise, RDI subsidy, and innovation information containing more than 2.5 million firm-year observations. It examines the association between various forms of subsidy ‘treatments’ and firms’ innovation output using matching control group.