On the Income Distribution as a Public Good

Abstract

Lester Thurow's provocative paper,' which considers the circumstances in which attainment of a Pareto optimum requires redistribution, unfortunately does as much to obscure and confuse as to clarify the logic of distributional adjustments. As Thurow sees it, there are three reasons why redistribution may be necessary to achieve a Pareto optimum: (1) one man's may be an argument in the utility function of another; 2 (2) utility may derive from the act of giving per se, rather than its effect on the recipient; and (3) the distribution of -i.e., a measure of inequality may itself be an argument in the utility function. This third situation, Thurow argues, arise because of externalities (different crime rates, degrees of political stability, etc.) associated with alternative distributions, or (even if there are no externalities) because of individual tastes for equality or inequality. Though Thurow recognizes that (1) and (3) shade into one another, and that all three cases can lead to substantial transfers, he concentrates on (3), since it implies that income transfers take on a different characteristic than when they are generated by either of the other two motives. In this case, he argues, The distribution is a pure public (his emphasis,3 whereas it is not if transfers are a matter of deriving utility from the incomes of others and from giving gifts. 4 But the implication that a market failure problem exists only if the distribution itself enters individual functions is erroneous. This very same type of public good problem arises if utility derives from the incomes of others. If A's level YA affects