Decision-Theoretic Analysis of Rawls’ Original Position

Abstract

Bruno de Finetti, in an article1 recently summarized in Theory and Decision, has compared the value judgments which should guide policy-making to the objective function of an optimization problem in Operations Research. He emphasizes that parochial features of current economic and social arrangements should not be presupposed in formulating the policy problem, and that the objective function is to depend not only on individual utilities but also on certain social interests concerning, e. g., equality and provision for future generations. Although de Finetti offers an attractive schema for approaching the policy problem that faces any society, he does not indicate, in the English summary anyway, how this ethical ‘objective function’ is to be arrived at. One approach compatible with de Finetti’s general decision-theoretic framework for ethics is that developed by J. C. Harsanyi. Harsanyi views ethics as a branch of the general theory of rational behavior.2 In particular, the ‘objective function’ for Harsanyian policy evaluation corresponds to the objectives that would be pursued by a fully rational, fully informed agent judging on the basis of ‘impersonal social considerations’ alone.3 Ethics as rational, impartial choice is a common but vague philosophical notion; Harsanyi’s accomplishment was to give content to ‘impersonality’ as a certain sort ofrisky choice: the agent chooses policy as if he had equal probability of being ‘put in the place of any individual in his society.4