Cooperation among Self-Regarding Individuals

Cooperation among unrelated self-regarding individuals is generally ex-plained by economists and biologists by noting that when interactions are repeated, non-cooperative acts may be punished by withdrawal of cooper-ation in future periods. We review these models, from Trivers ’ reciprocal altruism to more recent models of indirect reciprocity and the Folk Theorem, asking if they provide an adequate explanation of the structure and evolution of cooperation among unrelated individuals. We find the models inadequate on several grounds. First, they rest on implausible assumptions concerning both the degree of future orientation of humans and the quality of information available to them. Second, they apply at best only to cooperation in very small groups. Finally, the cooperative equilibria supported by these models, when they exist, are evolutionarily inaccessible and dynamically unstable under realistic assumptions concerning the quality of information