A CONSTITUTIONAL APPROACH TO TAXES AND TRANSFERS

What are the characteristics ofa “fair” systemoftaxes and transfers? All too often, contemporary political discussion oftaxes and transfers uses a concept of fairness that provides little basis for agreement. A proposed change in taxes or transfers is usually considered fair only when it benefits one’s own group or some other group that one favors, whether or not the existing taxes and transfers are fair by any standard. In this context, agreement is possible only if those who would pay higher taxes have some marginal benevolence toward those who would receive higher transfers (or lower taxes), given the existing distribution of income after taxes and transfers. Given the existing welfare state, this set of “Pareto optimal redistributions,” a concept first developed by Harold Hochrnan and James Rogers (1969), may be empty. In this context, without such marginal benevolence, any system of taxes and transfers is a negative-sum game, a form of legalized theft, reducing the total income of the community. The conventional focus on the distributional outcomes of this game, thus, is not a sufficient basis for determining whether these outcomes are the results of a fair game. This paper takes a “constitutional” or “contractarian” approach to taxes and transfers, and is based on a perspective first articulated by Frank Knight (1947), two applications first suggested by Richard Zeckhauser (1974), and a recent development of this perspective by James Buchanan (1985). As is so often the case, it is most appropriate to start with Knight. According to Knight (1947, p. 392):