Worst- and Best-Case Expected Utility and Ordinal Metautility

We provide a theory of decision under ambiguity in which the worst-case and best-case expected utility representations are sufficient statistics for a decision maker's preferences; we refer to the new utility functions as meta-utilities. Our approach, which generalizes many of the commonly used models, provides an ordinal value of these worst- and best-case expected utilities, mirroring the familiar ordinal utility functions over commodity baskets. Ambiguity attitude is simply the marginal rate of substitution in the meta-utility between worst-and best-case expected utilities