In recent years, a large body of empirical research has investigated whether the predictions of secondgeneration growth models are consistent with actual data. This strand of literature has focused on the longrun properties of these models by using productivity and innovation data but has not directly assessed the effectiveness of R&D policy in promoting innovation and economic growth. In the present paper, we fill this gap in the literature by providing a unified growth setting that is empirically tested with US manufacturing industry data. Our analysis shows that R&D policy has a persistent, if not permanent, impact on the rate of economic growth and that the economy rapidly adjusts to policy changes. The impact of R&D tax credits on economic growth appears to be long lasting and statistically robust. Conversely, more generous R&D subsidies are associated with an increase in the rate of economic growth in the short run only, indicating that, at best, this policy instrument has only temporary effects. Overall, the evidence regarding the effectiveness of R&D policy provides more support for fully endogenous growth theory than for semi-endogenous growth theory.