An empirical note on international R&D spillovers

Whether international R&D spillovers are global and trade-related is still a debated issue. By adopting two specifications that nest models previously estimated in the literature, we test the hypothesis that international R&D spillovers are global and trade-unrelated for a sample of OECD countries over the period 1971-2004. In particular, via a randomization exercise, we reject the null hypothesis of a "global pool of technology" and show that there are partitions of countries associated with relatively strong/weak knowledge spillovers. Then, we estimate a nonlinear specification that includes simultaneously geographical distance and international trade among the determinants of domestic TFP. We find robust evidence that both factors affect how foreign knowledge impacts on the domestic productivity of each recipient country.