
Edi Karni
24
maker’s posterior beliefs, are obtained by updating the probabilities representing his/
her prior beliefs using Bayes’ rule. The critical aspect of Bayesian decision theory is,
therefore, the existence and uniqueness of subjective probabilities, prior and posterior,
representing the decision maker’s prior and posterior beliefs that abide by Bayes’ rule.
As was argued in Section 1.3.5, in decision theories invoking Savage’s analytical
framework, the unique representation of beliefs is predicated on a convention, that
constant acts are constant utility acts, that is not implied by the preference structures.
Consequently, these models ...