Learning with whom to Interact: A Public Good Game on a Dynamic Network

We use a public good game with rewards, played on a dynamic network, to illustrate how self-organizing communities can achieve the provision of a public good without a central authority or privatization. Given that rewards are given to contributors and that the choice of whom to reward depends on social distance, free-riders will be excluded from rewards and the (almost efficient) provision of a public good becomes possible. We review the related experimental economics literature and illustrate how the model can be tested in the laboratory.