Abstract Since Adam Smith's and David Hume's days there has never been so much interaction between philosophy and economics. A theory of distributive justice, from J.S. Mill to John Rawls, concerns the rationale of public decisions. Economists see distributive justice as
a branch of welfare economics; philosophers, after Rawls and Nozick, pay attention to the definition of ethical constraints within the range of collective decision-making. What comes out is a normative political theory, based on the intersection between «utility» and «rights».Philosophers
and social scientists often discussed this matter as an ordering (a function) of social alternatives. Before Rawls, there were two main research programs in this area, Harsanyi's and Arrow's. This paper considers just the second one, Arrow's, that concerns the construction of a social welfare
function from a set of individual orderings (or utilities) to a unique social ordering (or social utility index). One refers to Arrow's program as «collective choice*, and one makes an attempt to show how Rawlsian argument can interfer with the main result of collective choice theory,
that is as well-known an impossibility theorem.The discussion is based on the following assumptions:(i) Arrow's impossibility theorem captures fundamental features of a liberaldemocratic position;(ii) at least from the justification point of view, the Arrovian impossibility represents
a difficulty for liberal-democracy;(iii) a Rawlsian argument is shown to be able to overcome Arrovian impossibility;(iv) it follows that Rawls’ principles of justice imply a radical shift in our conception of liberal-democracy.