Ever since the Harsanyi-Sen debate, it is controversial whether someone's welfare should be measured by her von-Neumann-Morgenstern (VNM) utility, for instance when analysing welfare intensity, social welfare, interpersonal welfare comparisons, or welfare inequality. We prove that natural working hypotheses lead to a different welfare measure. It addresses familiar concerns about VNM utility, by faithfully capturing non-ordinal welfare features such as welfare intensity, despite resting on purely ordinal evidence such as revealed preferences or self-reported welfare comparisons. Using this welfare measure instead (...) of VNM utility alters social welfare analysis -- for instance, Harsanyi's `utilitarian theorem' now effectively supports prioritarianism. VNM utility is shown to be a hybrid object, determined by an interplay of two factors: welfare and attitude to intrinsic risk, i.e., to risk in welfare rather than outcomes. (shrink)

An obituary of Philippe Mongin (1950-2020).

Following Gilboa and Schmeidler (Games Econ Behav 44:184–194, 2003) we consider a scenario where the decision maker holds, for every possible probabilistic belief about the states, a preference relation over his choices. For this setting, Gilboa and Schmeidler have offered conditions that allow for an expected utility representation. Their central condition is the diversity axiom which states that for every strict ordering of at most four choices there should be a belief at which it obtains. It turns out that this (...) axiom excludes many natural cases, even when there are no weakly dominated choices. We replace the diversity axiom by two new axioms—three choice and four choice linear preference intensity, which reflect the assumption that the preference intensity between two choices varies linearly with the belief. It is shown that in the absence of weakly dominated choices, the resulting set of axioms characterizes precisely those scenarios that admit an expected utility representation. In particular, our set of axioms covers a significantly broader class of scenarios than the Gilboa-Schmeidler axioms. (shrink)