In lieu of an abstract, here is a brief excerpt of the content:

CHAPTER 3 Ideology The decisions made by CEU maximizers are greatly influenced by the way they categorize themselves and other people. If they see the world through ethnic lenses, their decisions will not likely be the ones they would reach if they looked through religious lenses. Marxists certainly recognize that workers who unite on the basis ofclass must in some sense ignore or transcend both ofthese sources of intraclass division. Unfortunately, despite the importance of conceptual schemes for CEU theory generally, chapter 2 is quite casual about them. It tells us they are chosen based on their usefulness in making predictions . It does not tell us how these schemes are compared or revised. When the agents modeled in chapter 2 condition their choices on a category such as ethnicity, all their other beliefs are subsumed under the nondescript title ofadditional information. This suggests, quite unreasonably, that categories such as ethnicity exist in isolation as arbitrary elements of an agent's entire set of beliefs. A rational agent's beliefs presumably are coherent and systematic. Indeed in politics coherent sets ofbeliefs are so important that they have been given a special name. A system of beliefs constituting a coherent and systematic guide to action is typically termed an ideology. Yet calling the beliefs systems invoked in chapter 2 ideologies only alerts us to additional gaps in that chapter's discussion. For while a coherent belief system-an ideology-seems to be just the sort ofthing a rational agent would have, ideologues hardly fit anyone's common-sense view of rational agents. Moreover, a substantial literature to be discussed subsequently argues that only a fraction of modem mass publics has a working ideology. The majority is ideologically challenged. But if rationality and ideology are intimately related and most people do not have an ideology, what happens to the assumption of universal rationality? This chapter addresses these issues. We will see that while particular categories may be the innovations ofspecific rational agents, the existence ofa system ofcategories is an inherent part ofrational choice, not just a product of it. By the same token, there is no reason to suppose that rationality implies a shared system of categories and beliefs. Rationality plus exposure to new information does not necessarily lead to a convergence in belief systems, an 75 76 Choice-Free Rationality end to ideology, or even an end to ideological disagreement. Ideology and ideological difference are in the nature of rational behavior. Two conceptions of ideology-the "empiricist" and the "rationalist" (Stimson 1975)-predominate in contemporary political science. To Philip Converse (1964, 1975, 84-85), the leading advocate of the empiricist approach, an ideology is a logically and psychologically constrained belief system used to organize the substantial body of political information ideologues happen to have. Since individuals who think ideologically fit newer experiences into their ideological preconceptions (Converse I966a), their political beliefs generally will not converge, even when these individuals are exposed to large amounts of the same political information. Thus Converse is comfortable with the idea of agents who reason in a distinctly ideological way. He is also comfortable with the idea that the mass public is unwilling or unable to reason in this fashion and therefore displays a set of incoherent and unstable nonattitudes concerning a variety of political issues. Whereas Converse treats ideology as a belief system whose distinct constraints influence the use of information, the "rationalist" program initiated by Downs (1957) suggests the opposite. As clarified and revised by James Enelow and Melvin Hinich (1984) and Hinich and Michael Munger (1992, 1994), the Downsian approach conceives of ideology not as a psychological constraint on induction but as a means of avoiding the costs associated with induction. Because particular political parties and candidates are identified with particular systems of ideas, economizing voters can use ideological labels to determine their preferences instead of acquiring information specific to the electoral contest. If voters' political beliefs do not converge, this is not because they think in distinctly ideological ways but because they have not been subjected to the same political information. In short, voters who use ideologies do not process information differently than their more informed counterparts; they simply have less information to process. This chapter draws from both conceptions ofideology. On the one hand, it reverts to the empiricist approach by defining ideology as a psychological feature ofthe way political actors see the world and process information about it. On the other hand, it maintains the rational actor assumption. An ideology will...

Share