In recent years important theoretical contributions have shown that majority rule is a very badly behaved collective choice mechanism. In the absence of artificial restraints on preferences majority rule processes are almost always in disequilibrium. Moreover, the extent of the disequilibrium is pervasive, as captured by the observation that "anything can happen". What are the implications of such nihilistic results for the study of democratic political processes? Some authors believe that the implications are major, that they in fact preclude the development of a science of politics. Other authors take a more sanguine view. This essay argues that equilibrium notions, as presently formulated, are neither necessary nor sufficient for the development of a scientific study of politics. The newly proved disequilibrium results do suggest a change in the research agenda facing political scientists. The broad outlines of that agenda, and a general strategy for proceeding are discussed.

This paper has been prepared for the Conference on Political Equilibrium in Honor of William H. Riker, Washington-Hilton, August 27, 1980. The authors acknowledge comments on an earlier draft, ranging from constructive hostility to benign neglect, from: Randall Calvert, John Ferejohn, Robert Parks, Charles Plott, Robert Salisbury, and Barry Weingast. Published as Fiorina, Morris P., and Kenneth A. Shepsle. "Equilibrium, disequilibrium, and the general possibility of a science of politics." Political equilibrium. Springer Netherlands, 1982. 49-64.

Submitted - sswp364.pdf