It is well known in evolutionary game theory that population clustering in Prisoner's Dilemma games allows some cooperative strategies to invade populations of stable defecting strategies. We adapt this idea of population clustering to a two-person trust game. Without knowing it, players are typed based on their recent track record as to whether or not they are trusting (Players 1) and whether or not they are trustworthy (Players 2). They are then paired according to those types: trustors with trustworthy types, and similarly non-trustors with untrustworthy types. In the control comparisons, Players 1 are randomly repaired with Players 2 without regard to type. We ask: are there natural tendencies for people to cooperate more frequently in environments in which they experience more cooperation in comparison with controls?
Available at: http://works.bepress.com/vernon_smith/26/