Self sustaining R&D networks

This paper contributes to the knowledge-incentive based explanation of R&D networks. It argues that knowledge overlap and novelty are complementary inputs of any R&D alliance, and the complementarity coefficients depend on the incremental or radical nature of the research activity. The relation between the specialization of knowledge endowments and the structural properties (clustering, components, asymmetry of link distribution) of the resulting incremental, radical and mixed research networks are investigated. The paper addresses the structural conditions enabling the persistence of a research network through time.