Optimal targeting strategies in a network under complementarities

The paper analyzes the optimal targeting strategies of a planner (a governmental agency, a firm) who aims to increase the aggregate action of a population. The agents interact through a social network and react to their exposure to neighbors' actions. The reaction function describes, for example, the best response in a strategic game, a mechanical influence in a contagion disease or a mimetic behavior. The reaction is assumed to be increasing in exposure, resulting in complementarity in actions. When it is linear, the optimal planner's strategies are explicit, characterized by well-known centralities indices computed on the bilateral impacts. When the reaction function is concave or convex, the optimal strategies depend not only on the impacts but also on the pattern of agents' attentions. The value of information on the interaction structure is shown to be (almost always) positive and related to some form of heterogeneity between agents.