Productivity Spillovers of R&D in Sweden

Although Sweden is one of the most R&D-intensive OECD-countries, the importance of R&D spillovers in the country has not been systematically analyzed. This paper employs a cross-sectional dataset of 264 R&D-performing Swedish firms from 1996-97. With this set, knowledge production functions are estimated, where industry groups are treated as subsamples. In addition, 160,614 non R&D-performing firms are used to examine the effects of R&D spillovers also among non R&D-performers. The estimations use three different weight methods for R&D that spills over from other industries: two input-output measures and a technology flow matrix in the spirit of Jaffe (1986). The results indicate that R&D-performing firms gain in Total Factor Productivity from their own R&D. In two of the three weighing matrices spillovers from R&D result in higher Total Factor Productivity among R&D-performers. Among non R&D-performers, the Total Factor Productivity effect of R&D-spillovers is robustly positive and significant across specifications. Examination of the social returns of R&D from specific industries, one at a time, on other industries does not reveal substantial social effects beyond the effect on the own firm. It is reasoned that the most likely reason for the small size of R&D-spillovers rests in the Swedish corporate structure, with most R&D being conducted by large multinationals.