The Dynamics of Ambivalence


  • I thank Brian Gaines, Ben Highton, Howard Lavine, Beth Miller, Kenneth Locke, and the anonymous reviewers for their helpful comments and suggestions. The data analyzed in this article can be obtained from the National Election Studies website, Funding for ANES 2008–2009 Panel Study was provided by the National Science Foundation under grants SES-0535332 and SES-0535334, Stanford University, and the University of Michigan.

Thomas J. Rudolph is Professor of Political Science, University of Illinois at Urbana-Champaign, 240 Computer Applications Building, 605 E. Springfield Ave., Champaign, IL 61820 (


This article explores the dynamics of candidate ambivalence over the course of a presidential campaign. Candidate ambivalence tends to decrease as a campaign unfolds, although the rate of ambivalence decay is not constant across time or individuals. Two alternative theories of ambivalence change are considered and tested. Consistent with a motivational account, the results indicate that partisan reasoning contributes to the diminution of ambivalence over time. Consistent with an informational account, the results suggest that exposure to heterogeneous information heightens ambivalence. Ambivalence is least likely to decline among people who are exposed to cross-cutting information, politically sophisticated individuals with weak partisan attachments, and, during the general election phase of the campaign, those who live in homogenous areas with little political competition.