Belief system constraint and policy voting: A test of the unidimensional consistency model

Abstract

In this paper, the concepts, methods, and findings underlying disputes between traditionalist and revisionist schools of mass-belief-system research are examined. It is argued here that revisionist critiques merit important modifications and qualifications in traditionalist concepts and methods, but not wholesale revisions. The quantitative section seeks to demonstrate the continuing usefulness of the traditional approach for understanding public opinion and voting behavior. First, groups of individuals whose belief systems display a unidimensional or near-unidimensional structure are isolated from the general electorates of 1972 and 1976 using an individual level measure of belief consistency advanced by Barton and Parsons. Next a method of discerning voter projection of candidate issue stands developed by Bruner is used to compare the issue voting competence of consistent and inconsistent voters. Voters whose issue perferences are consistent with the liberal-conservative continuum are found to project less than their counterparts in both data sets examined.