Believing in Economic Theory: Sex, Lies, Evidence, Trust and Ideology

Abstract

People's beliefs about how well economic theory predicts outcomes may affect policy through democratic processes. Knowing what determines those beliefs is then important. We investigate how individual attitudes and characteristics correlate with those beliefs using a classroom Double Auction experiment, combined with a survey and ex-ante and ex-post elicitations of student beliefs. We find that Sex is a robust correlate of both ex-ante and ex-post beliefs: women are more skeptical than men in both instances. An index of socially desirable responding is positively related to prior belief in the theory: subjects who manage their image by telling Lies to please others also claim less skepticism on the ex-ante survey about the economic theory's predictive power. Subjects respond to Evidence in a minimally reasonable way: those who saw prediction errors in their experimental demonstration change their beliefs less between the ex-ante and ex-post surveys than those who saw none. While Trust-specifically, trust of authority-strongly correlates with Ideology, it is an insignificant predictor of beliefs. Finally, Ideology has complex effects on beliefs. As expected, the relatively liberal respondents are relatively more skeptical about economic theory in the ex-ante belief elicitation. Surprisingly, however, the relatively conservative respondents update beliefs in response to evidence much less strongly than their more liberal counterparts and, as a result, are actually relatively more skeptical than them in the ex-post belief elicitation. Acknowledgments. We have benefited from comments and help from Sarah Austin, Thilo Bodenstein, Dirk Engelmann, Jan Kmenta, Kathleen Knight, John Matsusaka, Andreas Ortmannand Christopher Wlezien, as well as seminar participants at the Public Choice/Economic Science Association meetings in San Antonio, Economic Science Association meetings in Barcelona and Tucson, the European Economic Association meetings in Venice, the California Institute of Technology, Technische Universität Chemnitz and the Max Planck Institut Strategic Interaction Group in Jena. Of course, none of these people are responsible for remaining errors or ambiguities.