International R&D Spillovers and Unobserved Common Shocks

This paper investigates whether returns to domestic R&D and international R&D spillovers should be estimated without considering the heterogeneous impact of unobserved common shocks, as has been done by the literature in this area. Using a panel of 50 economies from 1970-2011, I find that when unobserved common shocks are disregarded, estimates of domestic R&D and foreign R&D weighted by bilateral imports might be biased and inconsistent. Once unobserved common factors are accounted for, by allowing for heterogeneous technology coefficients, significant estimates become more sizable, consistent and not seriously biased in most cases. However, these estimates might be capturing not only returns to domestic R&D and trade-related knowledge spillovers, but also unobserved common spillovers and other effects. This indicates that knowledge spillovers and effects of unknown form cannot be easily separated. Therefore, unobserved common shocks should not be ignored when estimating returns to domestic R&D and international R&D spillovers.