Conditions for a Social Welfare Function

Abstract

_HE social welfare function, formulated first by Bergson in 1938, gave I promise of serving as a single, compact, formally elegant foundation for welfare economics by generalizing group value scales. With very little empirical content, the function could yield all the analytic conclusions of received welfare theory. With additional empirical content, the function might even help in the formulation of public policy. Recently, however, the usefulness of the social welfare function in the corpus of economic theory has been challenged. One attack is that the function is too abstract, too clumsy. It does not-and, indeed, cannot be made to-possess an empirical substance simple yet adequate enough to enable it to solve specific problems.3 Another attack is that posed by Arrow in his book Social Choice and Individual Values. His argument is that if we require a social welfare function to fulfil certain apparently reasonable conditions, then it can be proved that no such function exists.