We consider a modified Prisoners' Dilemma game in which each agent can offe
r to pay the other agent to cooperate. The subgame perfect equilibrium of t
his two-stage game is Pareto efficient. We examine experimentally whether s
ubjects actually manage to achieve this efficient outcome. We find an encou
raging level of support for the mechanism, but also find some evidence that
subjects' tastes for cooperation and equity may have significant interacti
ons with the incentives provided by the mechanism.