The Evolution of Cooperation.

This is an important book, with an elegant research design. For Axelrod, the fundamental problem is to understand how cooperation can emerge among self-seeking actors who lack centralized authority. This formulation has long been the crux of international relations theory, particularly "Realism," but Axelrod's approach breaks new ground and produces powerful answers to basic questions about international cooperation. Axelrod relies on mathematical game theory, which he employs with great sophistication. Yet this book is fully accessible to nonspecialists. Axelrod writes clearly, in plain English, and always remains focused on the problem of how cooperation can emerge among egoists. He starts by exploring a well-known game, the Prisoners' Dilemma, which represents many (but not all) of the obstacles to cooperation. In this game of strategic choice, there are only two players, each facing a single decision, to "cooperate" or to "defect." Each player must choose without knowing what the other will do. The payoffs for each player depend on the combination of both players' choices. The dilemma lies in the pattern of payoffs. In this game, if both players cooperate, each will do better than if they both defect. But what if only one chooses to cooperate while the other chooses to defect? In that case, the Prisoners' Dilemma rewards the defector with the highest possible payoff (even higher than the reward for mutual cooperation). The disappointed cooperator suffers the worst payoff, even worse than he would get if both players had defected. In colloquial terms, he is a "sucker." Cooperation still pays—but only if both players cooperate. This payoff structure creates powerful incentives to defect. In fact, for each player, defecting offers the best individual payoff no matter what the other player does.