A Tax and Redistribution Experiment with Subjects that Switch from Risk Aversion to Risk Preference

Benabou and Ok use expected utility to argue the prospect of upward mobility may cause majority rejection of redistributive taxes that benefit the majority. In paid computerized experiments subjects in a lower pay class exhibit risk preference but are risk averse in a higher class. This suggests behavior toward risk is not constant but depends on relative position. It is as if the reference point in prospect theory is socially determined. Such strong links between behavior toward risk and inequality are foreshadowed in the works of Bolton and Ockenfels (2000), Friedman and Savage (1948) and Thaler, Kahnemann and Tversky (1997). (JEL D63, C92) Beckman is a professor at the University of Colorado, Denver. Campus Box 181, POB 173364, Denver Colorado, 80217-3364. Phone: (303) 556 3048. Fax: (303) 556 3547. Email: [email protected] Professor Zheng made the original suggestion to do experiments on the POUM hypothesis. W. James Smith suggested using parameters from real societies. I am also grateful for the comments of Werner Guth, Stefan Traub, Rupert Sausgruber, Glenn Harrison, James Fowler, John Formby and Cecilia Rumi.