Whose opinion counts? An analysis of the opinion-policy linkage in the U.S. states

Date of Completion

January 2007


Political Science, General|Sociology, General




A number of empirical studies have found that public opinion influences the policies implemented by the United States government. However, much of this work assumes that the opinions of all individuals have equal influence. Both theory and research suggest that some people may have more influence than others. This project uses data on opinions and policy in the American states to investigate four hypotheses: (1) the opinions of the educated will have greater influence than those of less educated; (2) the opinions of non-blacks will have greater influence than those of blacks; (3) the opinions of voters will have more influence than those of non-voters; and (4) the opinions of those with higher incomes will have more influence than the opinions of those with lower incomes. Thus, I examine whether or not social inequality leads to political inequality. The research measures the preferences of groups within states and examines which opinions are more closely related to the policies of those states, after controlling for group size. The measures of public opinion, which are based on an original dataset constructed from several public sources including the General Social Survey, Gallup Polls, and Harris Polls, consist of self-rated ideology, indexes based on a variety of attitudinal questions, and opinions on specific issues. Measures of policy include both general indexes and specific issue areas such as abortion, the death penalty, or welfare. The findings generally support the hypotheses, especially those regarding education and income.* ^ *A Doctoral Dissertation Research Improvement Grant from the National Science Foundation (SES-0602341) to David Weakliem and Casey Borch supported this project. ^