Targeting in social networks with anonymized information

This paper studies whether a planner who only has information about the network topology can discriminate among agents according to their network position. The planner proposes a simple menu of contracts, one for each location, in order to maximize total welfare, and agents choose among the menu. This mechanism is immune to deviations by single agents, and to deviations by groups of agents of sizes 2, 3 and 4 if side-payments are ruled out. However, if compensations are allowed, groups of agents may have an incentive to jointly deviate from the optimal contract in order to exploit other agents. We identify network topologies for which the optimal contract is group incentive compatible with transfers: undirected networks and regular oriented trees, and network topologies for which the planner must assign uniform quantities: single root and nested neighborhoods directed networks.