On the Notion of Social Institutions∗

We argue that it is natural to study social institutions within the framework of standard game theory (i.e., only by resorting to concepts like players, actions, strategies, information sets, payoff functions, and stochastic processes describing the moves of nature, which constitute a stochastic game when combined) — concepts like “social norms,” and “mechanisms” can be easily accommodated, as well as philosophical/sociological definitions of social institutions Focusing on strategies rather than on mechanisms have two advantages: First, focusing on strategies allows us to distinguish between those aspects that are behavioral in nature and are subject to alternative design, and those that are part of the environment. Second, considering strategies allows for a more detailed look into the way an outcome function is “genuinely implemented” (Hurwicz (1996, p. 123)). ∗This paper corresponds to Chapter 2 of my Ph.D. thesis. I wish to thank my advisors, Narayana Kocherlakota, and Marcel K. Richter, for their guidance, encouragement, and very helpful comments and suggestions. I am also very grateful to the other members of my dissertation committee, Leonid Hurwicz, and Max Jodeit. While they all improved the material in many ways, all remaining errors are mine. This paper tries to address some questions posed to me by Leo Hurwicz, who, without being in complete agreement with some of my answers, made invaluable comments. Financial support from the Subprograma Ciência e Tecnologia do 2 ̄ Quadro Comunitário de Apoio is greatufully acknowledged.