Page - (000340) - in Knowledge and Networks

338 same partners always cooperate or whether they switch partners over time. Increasing trust, experience, and common understanding tend to contribute to the continuation of the partnership because they increase the efficiency of knowledge exchange and sharing. Conversely, the declining difference between knowledge stocks of continuously cooperating partners—that is, an increase in their cognitive proximity (the degree of similarity of their knowledge bases)—indicates that oppor- tunities to exchange and share knowledge have been exploited by them and should therefore lead to partner-switching. Hence, the relation between certain proximity dimensions and continuation of collaboration is by no means unidirectional (Ter Wal & Boschma, 2011). In fact, individual characteristics (e.g., technological capabilities), and thus the proximity to others, coevolve with continuous collaboration (Balland, Boschma, & Frenken, 2015; Ter Wal & Boschma, 2011). These dynamics have undergone little empirical analysis (Balland et al., 2015). Although the coevolution of factors driving collaboration choice and the evolution of ties can be explored only with a dynamic approach, most of the studies on the relation between proximity and cooperation have been rather static (e.g., Cantner & Meder, 2007; Paier & Scherngell, 2011; Wuyts et al., 2005). In this chapter we want to contribute to the field of dynamic approaches and analyze the interplay between cognitive proximity, knowledge exchange, and col- laboration. We focus our analysis on ties within innovator networks defined as an ensemble of direct and indirect connections, with the direct ones being research collaborations intended to produce innovations (Cantner & Graf, 2006). Tracking the individual actors and their collaborations over time, we pursue the following core research question: To what extent do knowledge dynamics between two coop- erating actors determine the continuation of their innovative ties? Accordingly, we concentrate mainly on the dynamics of partners’ cognitive proximity. In addition, we analyze the other two dimensions, trust and competencies, as further important covariates. Our descriptive analysis suggests that firms are generally prone to switching their cooperation partner rather than to repeating the collaboration with that partner. We thus find that the knowledge transfer and cooperation that partners have experi- enced with each other have no significant effect on the likelihood that they will repeat their cooperation. Our empirical analysis also shows that cooperation is pro- moted by several factors: an overlap between the firms’ knowledge bases, an uneven distribution of the reciprocal potential for knowledge exchange, general collabora- tion experience of the partners, and similarity in the degree of popularity of the collaboration partners. We also find that firms prefer to cooperate with partners that are different in organizational nature and age. We begin by providing a general overview of basic concepts and principle argu- ments that describe the relation between similarity in knowledge, experience, and their effect on tie formation. After characterizing how these relations dynamically coevolve with ongoing collaboration, we present our hypotheses. In the second sec- tion we explain our methodological approach, including descriptions of the data and variables. The third section presents the final results and our discussion of them. We conclude with suggestions for further research. U. Cantner et al.