Agents with Other-Regarding Preferences in the Commons

We present a unified approach to study the problem of the commons for agents with other-regarding preferences. This situation is modeled as a game with vector-valued utilities. Several types of agents are characterized depending on the importance assigned to the components of their utility functions. We obtain the set of equilibria of the game with two types of agents, pro-social and pro-self, and some refinements of this set for conservative agents. The most relevant result is that only a pro-social agent is required to avoid the tragedy of the commons, regardless of the behavior of the rest of the agents.