How Actual Governments Cope with the Paradoxes of Social Choice: Some Anglo-American Comparisons

Abstract

Political scientists have been paying increasing attention to the paradoxes, theorems, and problems of social choice theory. Since these theorems are formally valid, their conclusions are, in a strict sense, inescapable. Yet, in practice, a system of government that is to maintain or appear to maintain the principle of popular sovereignty has to operate as if the constraints implicit in these theorems did not exist. Systems of government have done this for generations, long before the theorems were formulated. Taking Arrow's General Possibility Theorem as a prototype of the social choice theorems, I suggest that the strategies used by different systems of government to escape the constraints implicit in Arrow's conditions of rational choice can be used as a method of classifying and revealing differences between them. I apply this notion to the British and American systems of government. I find, in this context, that the two most commonly used strategies are (1) to relax the condition of unrestricted choice by breaking down complex issues into a sequence of binary choices and (2) to relax the condition of irrelevance of independent alternatives by entrusting to elected or appointed officials the responsibility for judging the objective interests of the community, in effect making interpersonal comparisons of utility. Both the U.S. and the U.K. make use of both strategies, but there is a significant difference between the two systems in the weight placed on each. Moreover, while the former strategy is characteristic of issues resolved by partisan or adversarial political processes, the latter is more characteristic of administrative processes operating within the parameters of politically determined and agreed objectives. The greater weight placed by the American system on adversarial procedures results in issues becoming more politicized in the U. S., whereas in the U. K. there is more scope, particularly at the administrative level, for collaborative processes seeking to build consensus between the interests concerned within the framework of accepted policy objectives. Kenneth Arrow published the proof of his General Possibility Theorem in 1951.1 Political scientists appear at first to have taken remarkably little notice. Although an extensive literature subsequently developed, it was not until William Riker's Liberalism against Populism, published in 1982, that a major work of political science was devoted to showing how Arrow's Theorem, and other findings of social choice theory, effectively undermined one whole strand of conventional democratic theory, that which Riker calls populism.2 My suspicion is that the basic reason for the neglect, on which Riker comments, was the relative isolation at the time of the two