Page - (000345) - in Knowledge and Networks

343 more likely and sustainable (Cantner, Meder, & Wolf, 2011). Furthermore, the bal- ance between partners’ invested effort and reciprocated learning determines how well the alliance functions and how long it endures. Unilateral learning or an imbal- ance of resources might result in asymmetric bargaining power and dependency (Hamel, 1991; Khanna et al., 1998). Firms (actors) find that their attractiveness in terms of resources and efforts is reciprocated in collaborations with others similarly endowed. In sociological studies on the relations of individuals, the attractiveness of similarity has been termed homophily (McPherson et al., 2001; Rogers & Bhowmik, 1970). In the context of R&D collaborations, homophily might be driven by the search for reciprocity. If so, then actors similar in experience and competence will exhibit higher reciprocal potential than will dissimilar actors and will thus have mutual incentive to associate with each other (Cantner & Meder, 2007). The Dynamics of Tie Formation Although much work has been done to identify factors that lead to the formation of innovative alliances, little is known about the factors that determine the continua- tion3 of these alliances (Dahlander & McFarland, 2013). Because comprehensive longitudinal data on collaboration is difficult to find, most studies on innovation networks have relied on static analyses. Conceptual frameworks, too, such as Boschma’s proximity approach, are basically static in nature (Balland et al., 2015). In addition, the relation between the competence, proximity, and collaboration of a firm is characterized by strong interconnectedness. The embeddedness of firms also feeds back into the proximity to other actors, influencing their attractiveness as potential partners and future collaboration opportunities (Balland et al., 2015). The proximity of the partners changes throughout their bilateral collaboration as well, a shift that has consequences for its continuation. Both the underexplored coevolution of these factors and the evidence of the paradoxical effects of proximity and embed- dedness make it unclear whether collaboration alliances are finite (develop toward a specific date of expiration) and whether one can use an alliance’s continuation or termination to indicate an R&D alliance’s success. These coevolutionary processes can be captured only by dynamic approaches. Advances in this direction have been recently made mainly in the research on networks by scholars such as Balland, de Vaan, & Boschma (2013), Broekel (2015), and Ter Wal (2014). They have developed frameworks for empirically analyzing the parallel development of proximity, structural embeddedness, and the overall linkage distribution. One of this literature’s foremost contributions has been the inclusion of endogenous network forces (the feedback effects of structural position in the net- work) as an explanation for the probability of link formation other than relational 3 In this chapter the continuation of a linkage is synonymous with its persistence, recurrence, or repetition. It is defined technically as the reappearance of a link over multiple years in our time frame of observations. 16 Coevolution of Innovative Ties, Proximity, and Competencies