Page - (000019) - in Knowledge and Networks

7 Space as a Condition for Network Formation One dominant line of research has focused on the geographical constraints on the very process of network formation4 and its evolution. This approach is based on findings that geographical proximity tends to facilitate network formation; it increases the likelihood for social encounter, the exchange of information, and the creation of social relations (Allen, 1977; Zipf, 1949). Empirical research on net- work evolution confirms the plausible expectation that new relations are more likely to emerge in geographical proximity than over large distance (Broekel & Boschma, 2012; Glückler, 2010; Powell, White, Koput, & Owen-Smith, 2005). Recent research in evolutionary economic geography has underscored the association between geographical proximity and tie formation in networks by empirically con- trolling for other forms of proximity (Balland, 2012). These and other accounts of geographical constraints of network formation and evolution form part of what Borgatti and Halgin (2011) would classify as geographical theories of networks. Consequently, networks are located in space, and the creation of new linkages seems to be somewhat affected by this geography. Conversely, the creation of new ties in a network is an important strategy for bridging physical separation and enable com- munication and exchange over large distance (Glückler, 2005). The mutual condi- tionality between space and networks is thus a fascinating and still unexplored area of research. More complicated, however, is the question of how space and networks interact in their combined effect on the creation and reproduction of knowledge. Space as a Moderator of Network Effects on Knowledge The first interactive linkage between space and networks points to the moderating effect of space on the impact of network on innovation. In an interesting research design, Owen-Smith and Powell (2004) analyzed Boston-based biotech firms and their global alliance network. They found that although actor centrality was an important factor of innovativeness in the global network, it was insignificant within the regional network. Defying the intuition of network theorists, firms were equally likely to innovate independent of their position’s centrality in the network as long as they were connected to the local alliance network. Geographical proximity thus moderated the effect of network centrality on innovation. One explanation of this finding is that proximity allows for knowledge spillovers within the entire network of alliances and beyond the dyadic alliances. In a subsequent analysis of U.S. bio- technology, Whittington et al. (2009) demonstrated that the innovativeness of bio- tech firms benefits from geographical proximity and network centrality in ways that 4 See Adams et al. (2012), in particular Butts, Acton, Hipp, and Nagle (2012); Daraganova et al. (2012); Doreian and Conti (2012); Lomi and Pallotti (2012); Viry (2012); Preciado, Snijders, Burk, Stattin, and Kerr (2012); Sailer and McCulloh (2012); Takhteyev, Gruzd, and Wellman (2012); Verdery, Entwisle, Faust, and Rindfuss (2012). 1 Introduction