This paper empirically investigates the potential determinants of business-sector R&D intensity using a panel of OECD countries for the period of 1970-2002 with data measured as five-year year averages. Estimates using a system GMM estimator controlling for endogeneity show a high degree of persistence in business-sector R&D. Tax incentives for R&D have a significant and positive impact on business R&D spending regardless of specification and estimation techniques. Furthermore, we find that expenditures on R&D performed by universities are significantly positively related to the business enterprise sector expenditures on R&D indicating that public sector R&D and private R&D are complements. Direct R&D subsidies and the hightech export share are significantly positively related to business-sector intensity but these effects are only significant using the first-differenced GMM estimator. Static fixed effects results show that countries characterised by a low level mark-up ratio appear to have higher R&D intensities but this effect disappears after controlling for lagged R&D intensity. Similarly, the Ginarte-Park index of patent rights is significantly positively related to business-sector R&D intensity in the static panel data model but is no longer significant in the dynamic panel data model.
Innovation R&D Government support Industry structure