In the past decade, scholars such as Samuel Freeman, Martin O'Neill, Alan Thomas and others have argued that no matter how widely Rawls's theory of justice (TJ) was understood as a defence of the welfare state (WS), the socio-economic system Rawls defends and always defended is property-owing democracy (POD). In this article I present the argument that Rawls did not defend POD in TJ. However, while the claim that it was POD the socio-economic system implied by the principle of difference generated heated reactions by defenders of the WS, the argument I raise differs from most of them. In the article I maintain that, for Rawls, principles of justice underdetermine the choice of social systems, and that although this is true of both principles, it is especially true of the difference principle, whose ability to arbitrate between different social and economic policies is limited. I show how the attempt to uniquely determine the set of social and economic policies authorized by the difference principle clashes with what Rawls explicitly states in TJ and contrasts with central aspects of his sociological and political conception of socio-economic decision-making. Although my aim is to show that the difference principle itself does not discriminate among socio-economic systems, this does not preclude that additional reasons can help in narrowing the range of socio-economic systems allowed. With reference to the ideal of social union that Rawls expounds in the third part of TJ, I argue that TJ does, in fact, offer additional reasons to appraise the WS as preferable to POD. I maintain that the centrality Rawls attributes to the social unity of a just society provides the basis for an understanding of Rawls's position towards POD after TJ.
