Since the performance of a system of interacting peers depends strongly on their individual contributions, the ratio between their provided and consumed resources should be set in line with the social welfare improvement, without relying on a central coordinator. In this paper, we devise a self-organized mechanism for cooperation policy setting of the interacting peers based on decision-theoretic analysis. By extensive simulation experiments, we demonstrate that when the proposed mechanism is followed, a Pareto optimal equilibrium emerges in the system from fairly coordinated decisions of the rational peers on their cooperation policies.