Testing the Predictions of Decision Theories in a Natural Experiment when Half a Million is at Stake

In the television show Deal or No Deal an individual faces a sequence of binary choices between a risky lottery with equiprobable prizes of up to half a million euros and a monetary amount for certain. The decisions of 348 contestants from Italian and British versions of the show are used to test the predictions of ten decision theories: risk neutrality, expected utility theory, the fanning-out hypothesis (weighted utility theory, transitive skew-symmetric bilinear utility theory), (cumulative) prospect theory, regret theory, rank-dependent expected utility theory, Yaari's dual model, prospective reference theory and disappointment aversion theory. Both Italian and British contestants violate assumptions of risk neutrality and loss aversion. There appears to be no evidence of nonlinear probability weighting or disappointment aversion. Observed decisions are generally consistent with the assumption of regret aversion and there is strong evidence for the fanning-out hypothesis. Contestants become more risk averse after foregoing several of the highest ranked prizes, which challenges expected utility framework (or prospective reference theory that gives identical prediction)