Do spillovers matter when estimating private returns to R&D?

Abstract

A large body of literature estimates private returns to R&D adopting the Griliches knowledge production framework which ignores the potential impact of spillovers on consistent estimation. Using a panel of 12 manufacturing industries across ten OECD economies, we investigate whether ignoring spillovers leads to bias in the estimated private returns to R&D. We compare results from a common factor framework, which accounts for spillovers and other unobserved shocks, to those from a standard Griliches approach. Our findings confirm that conventional estimates conflate own-R&D and spillover effects, implying that spillovers cannot be ignored even when the interest lies exclusively in evaluating private returns to R&D.