Pareto Versus Welfare

Abstract

Many normatively oriented economists, legal academics and other policy analysts appear to be "welfarist" and Paretian to at least moderate degree: They deem positive responsiveness to individual preferences, and satisfaction of one or more of the familiar Pareto criteria, to be reasonably undemanding and desirable attributes of any social welfare function (SWF) employed to formulate social evaluations. Some theorists and analysts go further than moderate welfarism or Paretianism, however: They argue that "the Pareto principle" requires the SWF be responsive to individual preferences alone - a position I label "strict" welfarism - and conclude that all social evaluation should in consequence be formulated along strictly welfarist lines. I show that no strictly welfarist social welfare function can give complete expression to a normative social evaluation, as distinguished from simply describing a social allocation. I show also that SWFs employed to formulate complete social evaluations must accordingly, in virtue of the entailment relations obtaining among strict welfarism and the several Pareto criteria when these latter are interpreted along preference-regarding lines, be constructed without regard to Weak, Strong or Indifferentist Pareto. The results derived here generalize Arrow's (1951), Sen's (1970), and cognate impossibility results.