The bidirectional causality of tie stability and innovation performance

Scholars increasingly focus on the tie stability of firms' alliance networks and regard innovation as its consequence. However, they overlook that a firm's innovation performance may shape its tie stability. This potential simultaneous causality is poorly accommodated in the empirical literature, which impedes progress in innovation and network research. From the relational embeddedness and behavioral perspectives, we propose a mutually reinforcing relationship between tie stability and innovation performance. A stochastic actor-oriented SIENA model enables us to investigate bidirectional causality. Longitudinal data on 876 firms from 2003 to 2014 were analyzed to find evidence of a self-reinforcing force; this relationship was moderated by collaborative purpose diversity. These findings support the bidirectional causality between tie stability and innovation and the contingency effect of alliance relationships, providing new insights into network research, the innovation literature, and practice.