Bare preference and interpersonal utility comparisons

Abstract

From time to time it has been proposed to use just noticeable preference intervals as standard intervals upon which to base interpersonal utility comparisons. Such proposals have met with a number of difficulties due especially to the discovery that such intervals are difficult to take as representing equal differences in utility even for one individual.

In this paper I suggest conditions upon individual utility functions which, if satisfied, would make such ‘bare preference’ intervals useful as a basis for interpersonal utility comparisons even if the intervals are not interpreted as representing equal utility differences for any single individual.