How should income be divided? questionnaire evidence from the theory of “Impartial preferences”

Abstract

The idea of using as a basis for deriving social preferences about income distributions the choices that an individual would make in a hypothetical situation in which he had an equal chance of being put in the place of any member of society has a long tradition in economics. Harsanyi (1953, 1955), most notably, argued that the value judgements underlying that type of choice would be consistent with those of an impartial spectator in the narrower tradition of classical utilitarianism (David Hume 1740, Adam Smith 1759). We provide questionnaire evidence that, although there is a strong affinity between the two types of value judgement, neither is utilitarian, that is, consistent with an additively separable social welfare function. In fact, we found violations of utilitarianism qualitatively similar to those against the expected utility model for choice under risk, like those entailed in the Allais Paradox and other well known violations (including those against the Betweenness axiom). Our questionnaire also gives evidence of violations of some of the principles most used to measure and value inequality in a society, like the Pigou-Dalton Principle of Transfer and the Pareto unanimity rule. In a concluding paragraph we discuss certain similarities between the violations of the expected utility axioms and those of these more basic principles.