Aggregate uncertainty and the structure of utility functions
Nabil Al-Najjar*, Luciano Pomatto
Last modified: 2014-04-05
Abstract
We consider the impact of aggregate uncertainty on a social planner's ranking of lotteries over finite population profiles. We show that the planner (1) uses expected utility to evaluate lotteries, and (2) has separable preference over deterministic profiles, if and only if he has an aggregative
utility that may be sensitive to aggregate uncertainty. The standard additive utilitarian criterion is the special case characterized by indifference to aggregate uncertainty. When the size of the population increases, idiosyncratic risk disappears but aggregate uncertainty remains.
We use aggregative utility to examine precautionary public policies, fractional treatment rules, strict preference for randomization, and the evolution of attitudes towards risk. In all these cases, observed behavior may be understood as an attempt to hedge against an aggregate uncertainty that cannot be removed even when the population is large.
utility that may be sensitive to aggregate uncertainty. The standard additive utilitarian criterion is the special case characterized by indifference to aggregate uncertainty. When the size of the population increases, idiosyncratic risk disappears but aggregate uncertainty remains.
We use aggregative utility to examine precautionary public policies, fractional treatment rules, strict preference for randomization, and the evolution of attitudes towards risk. In all these cases, observed behavior may be understood as an attempt to hedge against an aggregate uncertainty that cannot be removed even when the population is large.