Is There an Allais Paradox? A Note on its Resolution

Abstract

If a decision maker uses J. M. Keynes' “conventional coefficient of risk and weight” in making decisions under conditions of risk and/or uncertainty, then a consistent pattern of ranked outcome results is not paradoxical as proposed by Allais. Similarly, the paradoxical aspects of the “reflection effect” in Kahneman and Tversky's Prospect Theory analysis of the Allais' paradox, also vanishes. The Allais Paradox can only exist when probabilities are used in a decision-making process without a measure of confidence. By incorporating an information base for calculated probabilities the Allais Paradox results are rendered nonparadoxical.