Nodal Enterprises (NEs) (Dyer & Nobeoka, 2000) create value by exploiting their position at the center of networks of supplier firms. In managing these networks of organizations, however, NEs face a dilemma: how much knowledge must be shared with partners in the value-chain? In this paper, we propose that knowledge sharing must be managed dynamically.

Building on selected applications of social network theory to clusters’ analysis (Kogut, 2000), we assume that NEs manage a combination of two types of rents when they share knowledge among their networked firms: “Coleman-type” and “Burt-type” rents. How the combination between the two dynamically unfolds is an unexplored issue. Using a computer simulation model, we explore robustness of alternative strategies of knowledge sharing among networked firms. Learning from our simulation experiments, we suggest a desired inter-temporal pattern of knowledge sharing and we show how the steady-state of knowledge-sharing depends counterintuitively on the speed at which knowledge leaks among the firms that are connected in the network.