Abstract In this paper, we discuss several issues related to public goods provision. Unlike many Austrians, we do not think that the concept of public goods - or of collective action - is an inherently flawed idea, even though we reject the alleged welfare implications of
public goods theory, as proposed by orthodox public finance literature. We then argue that the structure of a generic public goods problem is more a game of chicken or an assurance game than a prisoner's dilemma and that this has important implications with regard to the plausibility of cooperative
outcomes. Namely, when the public goods problem has the weakest-link structure and can be represented as an assurance game, then the cooperative outcome will be self-enforcing. In many cases, the public goods problems can be transformed into weakest-link games or different mechanisms can be
found to ensure cooperation. We also discuss the difference between a public goods problem and collusion. We assert that, unlike public goods problems, collusive agreements have the structure of a prisoner's dilemma. Overall, our paper suggests that there are reasons to be optimistic about
stability and efficiency of stateless societal orders.