Cooperation in the Short and in the Long Run

The emergence of cooperative behavior in a competitive world poses something of a puzzle for classical theories of competition, since it appears to be inconsistent with the pursuit of self-interest by individuals. Of course, cooperation among closely related individuals makes sense if it increases the fitness of a specific gene (Hamilton, 1964). But this leaves open the question of why cooperation is often observed among unrelated individuals when “cheating” would yield a higher payoff for any one of them. An ingenious game-theoretic explanation for this case has been put forward by Axelrod and Hamilton (1981) and Axelrod (1984). Imagine a large population of individuals who engage in pairwise interactions. Every time that two individuals meet, they play a “game” whose outcome affects the number of offspring that each of them leaves in the next period. The fittest strategy is the one with the highest reproductive success rate. Even though a strategy is temporarily successful, however, it may eventually become less fit as the frequency of the other strategies in the population changes. In particular, a necessary condition for a strategy’s con-