1 External values may be rejected as an appropriate foundation for political philosophy either because such values are held not to exist or because, even if existent, they cannot be invoked for epistemological or moral reasons. We need not make a commitment to either of these positions, although, of course, the difference between them may be significant in other contexts.

2 Both of us have discussed such derivations elsewhere. See, Buchanan, James M., The Limits of Liberty (Chicago: University of Chicago Press, 1975)Google Scholar, and Lomasky, Loren, “Personal Projects as the Foundations for Basic Rights” Social Philosophy and Policy, Spring 1984, 1(2).CrossRefGoogle Scholar

3 Rawls, John, A Theory of Justice (Cambridge: Harvard University Press, 1971).Google Scholar

4 The problem of enforcement of any contractual agreement is, of course, one of intense importance in both the Hobbesian and the Rawlsian models of contract. We do not examine enforcement in this paper, although we recognize that this problem must occupy a crucial place in any comprehensive theory of justice.

5 With the notable exception of the Sovereign, who enjoys immeasurably greater liberty and power than anyone else. It is as if Hobbes believed that, by placing only one unequal in a position of power over a whole set of equals, he could dispel criticism that his outcome lacks formal justice. Critics historically have not been mollified.

6 Hobbes, Thomas, Leviathan, ed. Oakeshott, Michael (New York: Collier Macmillan, 1962), p. 120.Google Scholar

7 The whole construction of preferences, indicated by the set of indifference curves in Figure 1, could not be as depicted under the Hobbesian constraint. Under the latter, no positions would dominate I, for A and B, except those on the 45 degree line.

8 Proponents of a conception of positive liberty maintain that person P has the liberty to do x if and only if no one or no social institution constrains P from doing x, and if it is the case that P has the ability to do x (i.e., does not lack the means necessary for doing x). On this conception, if someone lacks the means to take a round-the-world cruise, then that person is not at liberty to take the cruise even though no individual or institution is constraining such travel. We believe this to represent a serious conceptual confusion. Not all morally valuable commodities are liberties and, in particular, power or ability should not be conflated with liberty. Perhaps more to the point, it seems impossible to read Rawls as putting forth a positive conception of liberty. He distinguishes basic liberties (afforded by the first principle of justice) from the other primary goods to be allocated via the difference principle. Since the latter are instrumental to persons' abilities to pursue successfully their chosen conceptions of the good, Rawls is clearly not identifying liberty with the ability to secure desired outcomes.

9 The two dimensions are not, of course, independent one from another. An increase in civil liberty of one person necessarily involves a reduction in the natural liberty of others since the duty to refrain from interfering constitutes a limit in natural liberty. That is why natural liberty and civil liberty are both to be regarded as components of liberty sans phrase.

10 Cf. Nozick's, Robert “Tale of the Slave”, notably stage nine. Anarchy, Stale, and Utopia (New York: Basic Books, 1974), pp. 290–292.Google Scholar

11 In this respect, the two-person models used to illustrate the analysis later may be misleading if not properly interpreted. In the strict two-person model, a single individual's vote is not without direct consequences for the collective outcome. Hence, it is necessary to keep in mind that the two-person example is designed to be illustrative of the more inclusive many-person setting.

12 Libertarian critics, in particular, may suggest that individuals will predictably enjoy lower levels of absolute liberty, even as defined in the Rawlsian sense, in the collectivist, highly regimented order described in a₁b₁ than they might expect to enjoy in genuine Hobbesian anarchy. There are two ways that we might counter such critics. We could point out that the illustrative descriptions in Figure 2 are designed to serve our expository purposes and are not meant to be actual settings, even as summarized. We might men redefine a₁b₁ to be that social order that dominates Hobbesian anarchy, in the sense discussed, but which embodies lower levels of absolute liberty than any other alternative that is considered to be within the Rawlsian choice set.

More straightforwardly, but less rigorously, we might simply postulate that the threshold values of the economic goods component in Hobbesian anarchy are so low that there is a presumptive, non-dominance argument for ruling this position out of account.

13 It may be charged that we have connived to tinker with the Rawlsian choice situation so as to conjure up a set of institutional arrangements for a political-legal order that seems to describe something like a Nozickean minimal state. We should argue that a careful carrying out of the Rawlsian contractual exercise does produce a political order that, in a formal sense, is far closer to the minimal state than most commentators on Rawls have seemed to recognize. The discussion in Section V advances this finding. However, this result is a consequence of taking seriously the conditions Rawls stipulates as characterizing his contractors. Why, then, does Rawls himself not recognize the quasi-libertarian structure of the institutions he recommends as just? Perhaps, the absence of clarity here arises because Rawls does not clearly define the precise nature and scope of liberty. Consequently, the implications for social arrangements of the first principle of justice remain murky.

The extent to which the formally defined “Rawlsian solution,” in a₄b₄, resembles in practical application, the Nozickean minimal state, as this slate is normally perceived, will depend on the range and scope of “interpersonal” externalities, broadly defined, as these are predicted behind the veil of ignorance.

14 See, A Theory of Justice, especially pp. 243–251 and 541–548.

15 One basis for arguing that such prospects may not emerge would be a presupposition that there are no “natural” differences among individuals. But this presupposition would be unsupportable on any formal grounds. If such natural differences are allowed as possibilities, even behind the veil of ignorance there may be predictions that institutional regimes that “treat individuals differently” with respect to allowable spheres of liberty may generate greater liberties for all parties.

Suppose, as an example, that there are “naturally” two types of persons, one of whom behaves morally toward others, the other of whom remains amoral. Behind the veil, ofcourse, the chooser does not know which of these types he or she may be. Maximum equal liberty may be attained only in a regime that imposes relatively severe constraints on individuals' freedom of action. It may prove possible, however, that everyone's freedom of action may be extended by a regime that allows differential treatment for the moral and the amoral persons, once these two types are fully identified in the institutional operational sequence.

16 A critic might demur, claiming that autonomy is simply a function of who decides, not whether decision is made in-period rather than at a constitutional level. However, the manner in which a decision is made may have a bearing on the autonomy of the one who decides. The conditions of constitutional choice are significantly different from those of in-period choice. The former is made anonymously and with no knowledge of particularities. One brings about an outcome that will be an outcome for oneself in one way or another, but what that way is remains unknown. The latter choice is made in full knowledge of who one is and what one wants. It is determinate; one is not merely accepting some outcome or other, in which one's own role is unknown, but this outcome. We maintain that this amounts to a significant difference in degree of autonomy exercised. One who knows what he is bringing about for himself and who acts intentionally to produce that result acts autonomously in a way no Rawlsian contractor can.

17 This conclusion requires that these collective institutions operate ideally. In any realizable context, of course, collective ventures aimed at promoting economic equality introduce inequalities of their own, both in liberties and in access to economic goods.

18 Nozick, , Anarchy, Slate, and Utopia, p. 160.Google Scholar