Non - rationalizability of utilitarian consistent choice functions

Abstract

In much of the applied welfare economics, one finds the recurrent use of utilitarian objective functions, in arriving at social decisions. Apart from being completely insensitive to distributional issues, the utilitarian rule does not make single valued choices. The purpose of this paper is to show that choice functions which are utilitarian consistent (i.e. formed by selecting a point from the set of maximizers of a utilitarian objective function) cannot be rationalized by a continuous social welfare ordering. This would imply espousing kinds of objectives for capital budgeting problems other than the usual utilitarian one, if one desires to have a rational basis for investment planning decisions. A further result noted in the paper is that if a choice function is utilitarian consistent and symmetric then it cannot be rationalized by a social welfare function. This strengthens considerably the earlier result.