Jurisdictions and Structure-Induced Equilibria

Institutions are widely held to affect outcomes (e.g., neo-institutionalist theories in the social sciences). An important argument in this context concerns the stability of political interactions. This argument was developed in reaction to the “chaos theorems” of majority rule. The global cycling theorem of Richard McKelvey and Norman Schofield demonstrates that, in a multidimensional space with three or more voters who cast their vote in favor of the closest alternative and decide according to simple majority rule, a sequence of votes can (almost always) be constructed that leads to any outcome. Any proposal that attracts a majority of votes can in turn be beaten by another proposal preferred by a different majority. Two consequences follow from this result. First, an actor who controls the agenda effectively chooses the outcome. Second, majority rule is inherently unstable because any outcome can be overturned. In this interpretation, the result of a voting process is just caused by the (arbitrary) end of voting. An equilibrium only exists in a very rare preference configuration—for example, a median in all directions. Thus, we would expect to observe vote cycles according to the chaos theorem's interpretation of these formal results.

In reality, decision making in legislatures is characterized by stability. Neo-institutionalist scholars, most prominently Kenneth Shepsle, have argued that the observed stability of decision making results from the institutions that govern decision making. Thus, the equilibrium of decision making is induced by the structure of the process. One main application of the concept of structure-induced equilibrium has been to legislative organizations. Whereas majority rule in a multidimensional space is unstable, Duncan Black's median voter theorem tells us that majority voting on a unidimensional issue leads to a stable outcome at the position of the median voter. Legislators are faced with a multitude of public policy problems on which there is disagreement on various aspects. In other words, they act in a multidimensional policy space. As legislatures typically consist of more than three voters and often act by simple majority rule, the chaos theorem seems to apply. Shepsle asserts that the stability of legislative decision making in this situation can be explained by a structure-induced equilibrium. Institutions determine what issues are considered when and by whom. By shaping the behavior of legislators, these institutions can induce stability. One example is a strong committee system, which effectively reduces the multidimensional space, which is prone to instability, into a series of unidimensional spaces, in which the median voter theorem ensures stability by dividing the space into separate jurisdictions. Furthermore, institutions can require legislators to vote on one issue at a time within the committees. One objection to this view is that the structure of the process—that is, the institutions—is subject to debate as well. The Riker objection says that, if legislators were aware of the implications for policy of a given set of institutions, they would be tempted to change them to manipulate the outcome in their favor. Thus, the choice of institutions inherits the instability of the choice of policies.