A Theory of Rigid Extremists and Flexible Moderates with an Application to the U.S. Congress

Diversity of opinion is an endemic feature of society. Different people can hold divergent opinions on a subject. They can also differ in the intensity with which they hold their opinions. Some people are vehement and uncompromising. Their views seem relatively impervious to contradictory evidence and the opinions of others. Then there are people who hold their opinions tentatively and show a willingness and inclination to modify them as new information comes to light. Heterogeneity in how people form new opinions in response to new facts and the opinions of others has been documented (for example, Walter Mischel and John Schopler, 1959). Psychologists refer to such a personality trait as lying along a rigidity-flexibility dimension. By way of example, consider an influential segment of society: politicians. As reflected in their espoused ideologies, politicians differ in their opinions concerning what is the ideal society and what are the appropriate means for achieving it. Empirically, there are vast ideological differences among politicians with some being extreme and some being mainstream. On top of this heterogeneity, we find that some politicians are notorious for being uncompromising while others, contrarily, have built a reputation for accommodation. Examples of recent vintage include Newt Gingrich and Ronald Reagan in the rigid category and George Bush and Bill Clinton in the flexible one. That there should be extensive differences among people in their opinions and their rigidity is hardly surprising in light of the diversity of backgrounds and the complexity of the environment that people must interpret and act upon. What is intriguing, however, is the apparent relationship between these two traits. Casual observation and what data are available suggest that individuals with relatively extreme views tend to be relatively rigid. As an example, a survey of enlisted soldiers were queried about their attitudes toward a variety of issues including conscription, officers, the Army, and a military career (Edward A. Suchman, 1950). They were asked a number of questions to which they were to respond with “positive” or “negative” and to report whether or not they strongly held this opinion. A soldier’s positive score equaled the number of positive responses while his intensity score equaled the number of opinions held “strongly.” The study concluded: “People on both ends of the content scale feel more strongly than people in the middle of the scale” (Suchman, 1950 p. 275). The first objective of this paper is to explain why rigidity and extremism might be related. Our theoretical analysis begins with a population for which traits relating to extremity of opinions and rigidity of thinking are * Blomberg: Department of Economics, Wellesley College, Wellesley, MA 02181 (e-mail: sblomberg@ wellesley.edu); Harrington: Department of Economics, Johns Hopkins University, Baltimore, MD 21218 (e-mail: joe.harrington@jhu.edu). The authors would like to thank Ralph Braid, Allen Goodman, Bruce Hamilton, Robert Moffitt, Sevi Rodriquez, Peyton Young, seminar participants at Johns Hopkins University, Universitat Pompeu Fabra, University of Cambridge, University of Oxford (Nuffield College), American University, and Wayne State University; two anonymous referees (in particular, one of them for suggesting how we could empirically test the theory) for their comments; and Tim Groseclose, Steven Levitt, and Jim Snyder for sharing their data. Discussions of this topic with Colin Camerer, Ken Kollman, Keith Poole, and Matt Rabin were also helpful. This research was completed while the second author was a Visiting Scholar at the Institut d’Analisi Economica-CSIC (Barcelona, Spain). The hospitality and financial support of the Spanish Ministry of Education are gratefully acknowledged. The second author also acknowledges the financial support of the National Science Foundation under Grant No. SES-9708910. 1 See, for example, Abraham S. Luchins and Edith Hirsch Luchins (1959), William A. Scott (1966), and David H. Jonassen and Barbara L. Grabowski (1993). 2 For a ranking of U.S. presidents in terms of flexibility, see Gary M. Maranell (1970).