A quantitative and qualitative test of the Allais paradox using health outcomes

There have been many tests of the descriptive validity of the axioms of expected utility theory (EU) using money outcomes. Such tests are relatively uncommon with respect to health outcomes. This is unfortunate, because the standard gamble - considered by many health economists to be the gold standard for cardinal health state value assessment - is implied from the axioms of EU. In this paper, the classic Allais paradox, which predicts a systematic violation of the independence axiom, is tested in the context of health outcomes. Seventeen of 38 participants demonstrated strict violations of independence, with 14 of these violating in the direction predicted by Allais. The violations were thus significant and systematic. Moreover, the participants’ qualitative explanations for their behaviour show seemingly rational and not inconsistent reasoning for the violations. This evidence offers a further challenge to the descriptive validity of EU, and underlines the need to test alternative theories of risk and uncertainty in the context of health outcomes.