Addressing Keller's Critique: More on the Identification of Productive Technology Spillovers

25  March 2010    4:00 pm

Bart Los, University of Groningen


wiiw, Rahlgasse 3, 1060 Vienna, lecture hall (entrance from the ground floor)


R&D activities by one industry can have positive effects on the productivity performance of other industries, as a consequence of technology spillovers. As revealed by Wolfgang Keller in the late 1990s, severe multicollinearity problems have precluded the identification of industries that have generated the most productive technology spillovers, despite a number of sophisticated attempts. Consequently, we do not know which industries should be targeted for technology policy. This paper proposes an alternative estimation approach (Minimum Cross Entropy econometrics) that can overcome the identification problem. For 18 manufacturing industries, elasticities of productivity to R&D expenditures by the industry itself and by other industries are estimated on a bilateral basis. Furthermore, productivity effects of spillovers from the foreign counterparts of the industry are estimated. The analysis is done for twelve OECD countries in the period 1982-2003

Keywords: technology spillovers, sectoral productivity growth, entropy econometrics

JEL classification: C29; O33; O47