【光华讲坛】Structural Interventions of Social Ties
主题：Structural Interventions of Social Ties
This paper examines the impacts of structural intervention on individual's equilibrium behavior in the framework proposed by Ballester et al. (2006). We use minimal information, centrality measures of the players involved in the intervention, to characterize individuals' behavioral perturbations caused by a structural intervention. This characterization heavily relies on the fact that any structural intervention is outcome equivalent to a corresponding changing on the characteristics of the players involved in the structural intervention. Compared with recomputing the equilibrium of the intervened network game, this characterization simplifies the calculation by reducing the dimension of calculation especially when the structural intervention is sparse".
As an application of this result, we investigate the behavioral perturbation caused by building up a bridge between isolated networks. We construct an exact index to identify the key bridge players, linking up whom results in increasing aggregate activity the most. We illustrate that, by an example, the key bridge players may consist of neither each group's key player nor central player because both centrality and self-loop of bridge players contribute to aggregate effort.
We then investigate the network formation process in which a planner arranges links in the population sequentially and we show that the efficient networks (as well as the most active networks) are nested split networks.
Finally, we consider the policies aiming at optimally removing a group in a network. As a generalization of Ballester et al. (2006)'s key player result, we use the indexes in the original network to construct a measure to rank the impacts of the removal of a group on aggregate activity.