Belief consistency and invariant risk preferences

Karni and Viere (2013) introduce a novel theory characterizing the evolution of decision makers' beliefs in response to growing awareness of new conceivable contingencies. As awareness grows, the original subjective expected utility (SEU) preferences need to be reformulated. Under two key axioms linking the original and new preferences, Invariant Risk Preferences and Consistency, the authors derive representation theorems and updating rules for beliefs referred to as reverse Bayesianism. We prove that their Consistency axiom together with the SEU axioms implies Invariant Risk Preferences. In other words, we show that the theory of reverse Bayesianism can be behaviorally underpinned without directly imposing Invariant Risk Preferences. (C) 2018 Published by Elsevier B.V.