A more robust definition of multiple priors

This paper provides a multiple-priors representation of ambiguous beliefs à la Ghirardato, Maccheroni, and Marinacci (2004) and Nehring (2002) for any preference that is (i) monotonic, (ii) Bernoullian, i.e. admits an affine utility representation when restricted to constant acts, and (iii) suitably continuous. Monotonicity is the main substantive assumption: we do not require either Certainty Independence or Uncertainty Aversion. We characterize the set of ambiguous beliefs in terms of Clarke-Rockafellar differentials. This allows us to provide an explicit calculation of the set of priors for several recent decision models: multiplier preferences, the smooth ambiguity model, the vector expected utility model, as well as confidence function, variational, general "uncertainty-averse" preferences, and mean-dispersion preferences.