Positive Political Theory

Positive political theory is a variant of rational choice theory initiated by William H. Riker. In understanding the relationship between positive political theory and rational choice theory, it is sufficient to understand that both are committed to positive as opposed to normative political theory, and that both presuppose purposive action. However, not all the models of political phenomena studied by positive political theorists are developed in terms of rational choice theory. The hallmark feature of positive political theory is its adherence to the scientific method of building up descriptive generalizations, or universal laws, from minimalist intuitive assumptions. Often relying on the formal language of rational choice theory, these general laws are deduced from axiomatic systems that can be tested [Page 724]against actual human behavior. Positive political theory eschews making normative claims about how political agents or systems should function and instead emphasizes building explanatory and predictive models that serve as the basis for a progressive study of political phenomena. Positive political theorists study elections, legislative behavior, political institutions, and the formation of international alliances.

Riker coalesced positive political theory from what are now recognized to be canonical texts in the rational choice tradition of research: John Von Neumann and Oskar Morgenstern's Theory of Games and Economic Behavior (1944), Duncan Black's “On the Rationale of Group-Decision Making” (1948), Kenneth J. Arrow's Social Choice and Individual Values (1951/1963), and Anthony Downs's An Economic Theory of Democracy, Riker's Theory of Political Coalitions, along with James M. Buchanan and Gordon Tullock's Calculus of Consent (1962), and Mancur Olson, Jr.'s Logic of Collective Action complete the early rational choice canon that spans public choice and social choice theories as well. These results progressively build on each other, realizing the hope of positive political theorists that formal deductive models tested against empirical phenomena yield a growing body of knowledge. Von Neumann and Morgenstern's innovative treatment of game theory provided the basis for a new study of politics directly relating numerous individuals' strategically rational decisions to collective outcomes. Importantly, they developed the idea of equilibrium strategies that, when played against each other, yield predictable, law-like outcomes that exhibit stability.

Black's and Arrow's late 1940s research addressed electoral processes for reaching collective decisions. Both researchers used the eighteenthcentury Condorcet voting paradox as a point of departure for their investigations. According to the Marquis de Condorcet's analytic assessment of voting, it is possible for three individuals voting over three alternatives to reach deadlock if their preferences are structured such that, as an example, Alex prefers candidate Zeus to Athena to Hercules, Loren prefers Athena to Hercules to Zeus, and Pat prefers Hercules to Zeus to Athena. When the votes are tallied over pairs, the collective preference has the form that Zeus is preferred to Athena, who is preferred to Hercules, who is preferred to Zeus. This circular statement of a collective preference order was recognized by Condorcet, Black, and Arrow to be cyclical, and therefore irrational and unstable. Black contributed to the analytic study of election procedures by demonstrating that in the case in which voters' preferences over three outcomes are single peaked—that is, outcomes may be represented in a linear order with two distinct poles and no voter least prefers the middle outcome—then cyclic stalemate outcomes can be avoided. Arrow worked in the opposite direction of Black in developing his impossibility theorem. He proved that the cyclic instabilities that Condorcet alerted us to characterize voting procedures of n individuals selecting from n alternatives more generally. In the 1950s, Riker pursued these results to empirically investigate if actual voting procedures tend to result in unstable, cyclic outcomes. Riker's research indicated less cyclic outcomes than are anticipated by the theoretical models of voting.

...