ABSTRACT
The bulk of the literature on inequality measurement assumes that the income profiles do not involve uncertainty. It is natural to suppose that if uncertainty (or risk) is present, one may use the theory of decision under uncertainty to reduce the inequality problem to the case of certainty, say, by replacing each individual’s income distribution by its expected value or expected utility. Alternatively, it would appear that one may use the theory of inequality measurement to reduce the problem to a single decision-maker’s choice under uncertainty, say, to the choice amongdistributions over inequality indices.Weclaim, however, that neither of these reductions would result in a satisfactory approach to the measurement of inequality under uncertainty. Rather, inequality and uncertainty need to be analyzed in tandem. The following example illustrates.
