The Evolution of Trust and Cooperation between Strangers: A Computational Model
Michael W. Macy, John Skvoretz
Social and economic exchanges often occur between strangers who cannot rely on past behavior or the prospect of future interactions to establish mutual trust. Game theorists formalize this problem as a one-shot prisoner's dilemma and predict mutual noncooperation. Recent studies, however, challenge this conclusion. If the game provides an option to exit (or to refuse to play), strategies based on projection (of a player's intentions) and detection (of the intentions of a stranger) can confer a cooperator's advantage. Yet previous research has not found a way for these strategies to evolve from a random start or to recover from invasion by aggressive strategies that feign trustworthiness. The authors use computer simulation to show how trust and cooperation between strangers can evolve without formal or informal social controls. The outcome decisively depends, however, on two structural…