(Coorganized with LUISS) Multinationality, R&D and Productivity. Evidence from the top R&D investors worldwide

The paper provides new conceptualization and empirical evidence into the relationship between multinationality and productivity. First, we argue that this is a composite relationship, which can be split into a direct and an indirect one. In particular, we uncover an often neglected channel of multinationality on productivity, channelled through higher incentives to invest in R&D. Second, this composite effect differs on the intensive margin (higher share of multinational activities) and the extensive margin (greater geographical dispersion of multinational activities). Using a large sample of top R&D investors in the world, we propose and econometric model based on both an R&D and a productivity equation, which are both allowed to depend on multinationality. With this model we can disentangle the direct and indirect effects of multinationality on productivity appropriately. We find that: i) a positive direct impact of multinational intensity on productivity, while the geographical dispersion of multinationality is negatively associated with productivity; ii) multinationality (along both margins) has a positive indirect impact through higher investments in R&D iii) this positive indirect effect is however not large enough to compensate the negative direct one at the extensive margin. Results are largely consistent with a theoretical approach that combines transaction cost theory explaining the direct effects with an economic analysis of how incentives to invest in R&D depend on multinationality explaining the indirect effect