Kenneth A. Shepsle, “Studying institutions: Some lessons from the rational choice approach”, Journal of Theoretical Politics 1 (1989), pp. 131-147.

Summary: Develops the theory of structure-induced equilibrium and explores the two sides of the institutional coin – an examination of the consequences of institutional structure and an explanation of the development of institutional structure. Overall, he’s making an argument for using rational choice.
Important Insight: Structure-induced equilibrium. The concept of structure-induced equilibrium is based on the idea that an institutional process, described by its rules, can be graphed as an extensive form game… A structure-induced equilibrium may be defined as an alternative (a status quo ante) that is invulnerable in the sense that no other alternative, allowed by the rules of procedure, is preferred by all the individuals, structural units, and coalitions that possess distinctive veto or voting power. The idea of equilibrium contained in this concept is founded on the structural and procedural features of the process being modelled. Its focus is on institutional equilibrium – the equilibrium in outcomes produced by a specific institutional configuration.
—-
-key point: develops the theory of structure-induced equilibrium and explores the two sides of the institutional coin – an examination of the consequences of institutional structure and an explanation of the development of institutional structure. Overall, he’s making an argument for using rational choice.
-the concept of structure-induced equilibrium is based on the idea that an institutional process, described by its rules, can be graphed as an extensive form game… A structure-induced equilibrium may be defined as an alternative (a status quo ante) that is invulnerable in the sense that no other alternative, allowed by the rules of procedure, is preferred by all the individuals, structural units, and coalitions that possess distinctive veto or voting power. The idea of equilibrium contained in this concept is founded on the structural and procedural features of the process being modelled. Its focus is on institutional equilibrium – the equilibrium in outcomes produced by a specific institutional configuration.
-this raises two questions (re: a theory of equilibrium institutions, which is the challenge to rational choice): how are institutions selected, and how are they sustained? He argues that these should be modelled explicitly (but doesn’t do it here).
-in earlier equilibrium theories, equilibrium was preference driven. An outcome, x, was said to be an equilibrium if there existed no y preferred to it by a decisive coalition of agents (normally a simply majority of the set of agents). The unhappy conclusion of many of these theoretical stories, themselves generalizations and extensions of Arrow’s impossibility theorem, was that in politics there typically were no equilibria.
-Yet empirically, outcomes appeared to track the preferences of distinguished actors upon whom institutional structure and procedure conferred disproportionate agenda power. All the ‘action’, so to speak, was not in the straightforward aggregation of preferences, but rather was found in the sometimes subtle influence provided by control over structure and procedure.
-problems with traditional rat choice theories of institutions re: choosing institutions:
-what an institution facilitates may in fact be a by-product of what its founder intended. (critique against functionalism).
-most rat choice theories conceive of institutions as ex ante bargains – as the one-time-only choice of a game form. In reality, institutions undergo transformation. Not all of the rules of the game can be fixed in advance, since not everything can be anticipated.
-a theory of equilibrium institutions (the problem of robustness) must be modelled explicitly
-the transactions costs of change provide an institution with something of a cushion, giving it a stability it might not otherwise enjoy. Thus, even when institutional arrangements are not optimally suited to a given environment, they may nevertheless endure b/c prospective gains from change are more than outweighed by the costs of effecting them.