Advances in Strategic Network Formation: Preferences, Centrality, and Externalities

The importance of social and economic networks has meanwhile been widely acknowledged throughout social sciences. Network positions are supposed to be important for both, consumers and suppliers. Examples are R&D collaborations, strategic alliances, knowledge management within organizations, formal and informal leadership, personal (business) contacts, information about job openings, bargaining power etc. Given these beneficial aspects of networks, the stability is in question: How do networks change when agents follow incentives for profitable network positions? This question serves as a leitmotif for this thesis. First, it is addressed how to formally model situations where individuals purposefully alter the network structure. Major advancement was achieved in recent years (see Jackson and Wolinsky, 1996, Bala and Goyal, 2000, and Bloch and Jackson 2006). However some issues remain, especially concerning the structure on the utility function needed to characterize certain equilibrium outcomes. For instance, given specific benefits of linking, we illustrate how the stable networks depend on the assumptions of increasing and decreasing marginal returns. Second, I want to have a closer look at the individual incentives that drive network formation. Typically, there are two kinds of aspects which play a major role: One is the gain of access to information and support by having many other actors in close reach (e.g. Bala and Goyal, 2000); the other one is the bargaining position that is attained by being a broker for others (e.g. Burt, 1992, Goyal and Vega-Redondo, 2008). I propose a model in which agents strive for these two types of benefits measured by closeness and betweenness centrality. The stable and emerging networks in this model are studied by three complementary methods. We contrast the emerging network structures in a scenario with either betweenness or closeness incentives and a scenario in which closeness and betweenness incentives are combined. In any of those scenarios, we find low local density ("clustering coefficient") in the emerging networks. After analyzing the stable networks, I examine the efficient networks for different parameter combinations. Surprisingly, incentives for intermediation rents (i.e. betweenness) frequently lead to networks that have lower welfare than the starting networks. Finally, one of the focal topics of strategic network formation is the "tension of stability and efficiency" (see Jackson and Wolinsky, 1996), meaning that there is a general conflict between individual interest and collective outcome. What has not been studied systematically are the sources of inefficiency. I approach this omission by analyzing the role of positive and negative externalities of link formation. This yields general results that relate situations of positive externalities with stable networks that cannot be "too dense" in a well-defined sense, while situations with negative externalities tend to induce "too dense" networks. Those result enable us to better characterize many of the models known in the literature on strategic network formation. Moreover, the results provide a clear signal about how to reduce inefficiency.