University-Industry partnerships in the smart specialisation era.

Abstract

• Technological diversification is associated with more opportunity discovery • Low technological proximity and diversification is associated with the lowest performance in terms of licenses generating income • High technological proximity and low diversification is associated with the low numbeers of licenses generating income • High technological proximity and diversification is associated with more opportunity discovery but also more spillover • Low technological proximity and high technological diversification is associated with the best performance in temrs of licenses generating income The effect of diversification versus specialisation, as well as the technological proximity of R&D partners have been at the heart of innovation studies. Articles in this field either take a regional or company point of view. During the last few decades, studies on the relatedness of knowledge and its importance for innovation and commercialisation have pushed policy makers towards clustering strategies such as smart specialisation in the EU and the Innovation Supercluster Initiative in Canada. Interestingly, universities as the source of new knowledge and technologies have been absent from this literature as the focal point. This paper aims at filling one of the missing links between the literature on technological relatedness and university research commercialisation. We use patent and licence data from the USPTO and the AUTM survey to study the effect of patent portfolio composition on university research commercialisation. We use Shannon’s entropy index to differentiate between the effects of related and unrelated diversification on the number of licences generating income. Our results show a positive association of related diversification with the number of licences, but none for unrelated diversification. Furthermore, technological proximity follows an inverted-U shaped association with the number of licences generating income. However, the effect is observed only for smaller universities. We conclude that the curvilinear association is the result of cognitive distance and the absence of boundary spanners. Our findings indicate that regional policy makers intending to use universities as an engine for innovation and regional economic growth should consider policies and initiatives aimed at bridging the cognitive gap between university and industry by either increasing technological proximity or reducing cognitive distances by financing boundary spanning organisations.