Progress in the Theory of Social Choice and Distributive Justice

By definition, “consequentialist” behaviour in finite decision trees is explicable by its consequences. Both cost-benefit tests and “consequentialist” choices of economic policy necessarily require distributional judgements. These should emerge from a social welfare objective incorporating interpersonal comparisons. To accommodate them, Arrow’s IIA condition should be weakened to independence of ethically irrelevant alternatives. When consequences are risky, dynamically consistent consequentialist behaviour on an unrestricted domain of finite decision trees entails maximizing expected social welfare. Combined with an “ethical liberalism” condition, this leads to “fundamental” utilitarianism, which requires a further weakening of IIA to independence of ethically irrelevant mixed alternatives. Journal of Economic Literature classifications: D63, D71, D81.