Power-use in cooperative competition: A power-dependence model and an empirical test of network structure and geographic mobility

•We propose a power-dependence model of network formation and geographic mobility.•We analyze the dynamics of power-use in a community of cooperatively competing amateur sports teams.•Poisson regression as well as binary and valued exponential random graph models are used.•We find that power-outcomes are governed by status and geographic distance.•Results suggest that network formation and geographic mobility are intertwined with power and dependence.

Although the social exchange paradigm has produced a vibrant research program, the theoretical tradition is rarely used to model the structure of social networks outside of experiments and simulations. To address this limitation, we derive power-dependence predictions about network structure and geographic mobility—the outcomes of power-use—and test these predictions using complete data on competition networks and travel schedules among amateur sports teams. Poisson regression and exponential random graph models provide strong support for our predictions. The findings illustrate exchange dynamics in which status resources desired by teams, coupled with the availability of geographically proximal alternatives, create power and dependence that dictate where and with whom teams compete. Although evidence supports Georg Simmel’s classic proposition that networks form on the basis of values and propinquity, we show that this complex dynamic is conditional on power and dependence. We conclude by discussing implications and directions for future research.