The structure and dynamics of knowledge networks: a proximity approach

Local knowledge networks are often held responsible for the competitiveness and innovativeness of geographical clusters. However, the literature on spatial clustering tends to assume that firms in clusters have equal access to the knowledge that circulates in those networks and that this knowledge is inaccessible for firms beyond the clusters boundaries. In addition, the majority of studies on spatial clusters neglect to analyse the structure of these networks over time. This study aims to account for these shortcomings. Through the application of social network analysis this dissertation provides an in-depth study on the structure and dynamics of knowledge networks within and across clusters. More specifically, it analyses how geographical, social and cognitive proximity can explain the evolution of knowledge networks. It appears that the effect of geographical proximity on network formation is not constant over time. On the basis of a longitudinal analysis of the collaboration network among German biotechnology inventors the research shows that geographical proximity is most relevant for network formation for young emerging industries where knowledge is predominantly tacit. In many other contexts, geographical proximity between firms does not result in the formation of local knowledge networks within clusters. In particular, social proximity appears to be much more important for local knowledge networks to emerge. The research demonstrates that a lack of social proximity might prevent the emergence of networks of local collective learning in clusters despite geographical proximity. Even in the business park of Sophia-Antipolis – often conceived to be the European “Silicon Valley” – a rather cohesive local knowledge network has emerged over its 30-year history only in one of its two main sectors. Differences in the way spatial clustering in these two sectors have unfolded historically underlie these divergent patterns of network dynamics, demonstrating that the processes of cluster dynamics and network dynamics are strongly interrelated