¹ An absolute or indefeasible right is one which generates a duty which may not be violated, whatever the consequences. It would be useful to define an “almost indefeas ible right”, where the excusing conditions are limited to “moral catastrophes”. On this, see Nozick, Robert, Anarchy, State, and Utopia (New York: Basic Books, 1974), 29n.Google Scholar

² I will suppose Nozick and other libertarian natural rights theorists to have established this claim.

³ As I do in “Natural Rights and Public Goods”, in Attig, Thomas, Callen, Donald, and Gray, John, eds., in The Restraint of Liberty, Bowling Green Studies in Applied Philosophy VII (Bowling Green, OH: Bowling Green State University Press, 1985), 102–117.Google Scholar

⁴ See Levin, Michael, “A Hobbesian Minimal State”, Philosophy & Public Affairs 11 (Fall 1982), 338–353Google Scholar; Narveson, Jan, The Libertarian Idea (Philadelphia, PA: Temnle University Press, 1989)Google Scholar; and Gauthier, David, “Rational Cooperation”, Nous 8 (1974), 62–63CrossRefGoogle Scholar. Also relevant are Narveson, , “Human Rights: Which, if Any, Are There?”, in Pennock, J. R. and Chapman, J. W., eds., Human Rights, Nomos XXIII (New York: New York University Press, 1981), 175–197Google Scholar; and “Contractarian Rights”, in Frey, R. G., ed., Utility and Rights (Minneapolis, MN: University of Minnesota Press, 1984), 161–174Google Scholar. Although David Gauthier would not endorse the minimal state exactly as it is characterized by Nozick and others, many of the conclusions he reaches in his recent Morals by Agreement (Oxford: Clarendon Press, 1986)Google Scholar are rather libertarian.

⁵ This follows both from Hobbes' defense of absolute sovereignty and from his claim that the sovereign has no moral or legal obligations. See his Leviathan, especially chapters 18, 26, 29.

⁶ I defend this reading in “Leviathan and the Minimal State: Hobbes' Theory of Government”, in Moyal, Georges and Tweyman, Stanley, eds., Early Modern Philosophy: Metaphysics, Epistemology, and Politics, Essays in Honour of Robert F. McRae (Delmar, NY: Caravan Books, 1986), 373–395Google Scholar. This reading of Hobbesian theory, as well as Levin's, are criticized by Kavka, Gregory in “Some Neglected Liberal Aspects of Hobbes' Philosophy”, Hobbes Studies 1 (1988), 89–108CrossRefGoogle Scholar. On issues of redistribution and of revolution, Kavka argues that Hobbes is on the side of contemporary liberals rather than conservatives, as these terms are used in the United States. I would want, however, to distinguish between Hobbesian theory and Hobbes' own writings. While I do not disagree with Kavka concerning the logic of the former, I would claim that the latter are extremely minimalist, surprisingly so given Hobbes' enthusiasm for absolute sovereignty and given the pre-Smithian character of his thought. (See also note 57.)

⁷ This is one of the conclusions of Hampton, Jean in Hobbes and the Social Contract Tradition (Cambridge: Cambridge University Press, 1986).Google Scholar

⁸ The assumption of self-interest, although standard in this tradition, is not essential to the Hobbesian approach, or so I would argue. For a discussion of these issues, see my “The Relation of Self-Interest and Justice in Contractarian Ethics”, Social Philosophy & Policy 5 (Spring 1988), 119–153.Google Scholar

⁹ The matter is very complicated. To forestall some possible objections, I shall assert, without argument, that hypothetical collective choice in Hobbesian theory should be understood only as a heuristic device. The more complex matter of justification comes up later and has nothing to do with consent or contract, contra standard interpretations of social contract theory. For some recent discussions of these issues, see Gauthier, , Morals by AgreementGoogle Scholar; Hampton, , “The Social Contract Explanation of the State” (manuscript, UCLA, 1986)Google Scholar, as well as Hobbes and the Social Contract Tradition, chap. 9; and Nelson, Alan, “Explanation and Justification in Political Philosophy”, Ethics 97 (10 1986), 154–176.CrossRefGoogle Scholar

¹⁰ Hobbes' “right of nature” is a mere (Hohfeldian) liberty, not a claim-right entailing obligations on the part of others. In the Hobbesian state of nature, there are no claim-rights, only liberties.

¹¹ The metaphor of a social contract should not be taken too literally or seriously, certainly not as representing any sort of promise or commitment, as Levin in fact supposes. See Levin, “A Hobbesian Minimal State”, 340, 352. This misinterpretation of the original agreement as a pledge or a binding agreement does not, however, affect his argument.

¹² Ibid., 343ff.

¹³ Provided, of course, that we can solve the problem of controlling the sovereign now that it has all of our swords. I shall ignore this important but seemingly intractable problem in this essay.

¹⁴ Levin, , “A Hobbesian Minimal State”, 345.Google Scholar

¹⁵ Ibid., 351.

¹⁶ This simplifying assumption will not have any effect on my argument.

¹⁷ Readers familiar with game theory will recognize this as similar to an Assurance Game. See Sen, A. K., “Isolation, Assurance, and the Social Rate of Discount” (1967)Google Scholar, reprinted in his Resources, Values and Development (Oxford: Basil Blackwell, 1984), 135–146. The preferences that give rise to Assurance Games are:

1. mutual disarmament

2. arm, disarm

3. arm, arm

4. disarm, arm.

This is represented in the 2 × 2 matrix below:

The difference between Sen's analysis and mine has to do with the second and third preferences. Since this difference will not significantly affect the reasoning of the parties, we may broaden Sen's notion and think of both as species of Assurance Games.

¹⁸ Levin, , “A Hobbesian Minimal State”, 344, 345Google Scholar. Levin's references to “Schelling effects” (349) reinforce my suspicion that he has confused coordination and similar problems with games of partial conflict.

¹⁹ Hillel Steiner has suggested to me that Levin's analysis, thus understood, is similar to that of the Campaign for Nuclear Disarmament (CND): if the only reason for the Soviets to possess nuclear arms is their fear of the West's nuclear arsenal, then the latter can unilaterally disarm without any fear.

²⁰ An outcome is Pareto-efficient if and only if there are no outcomes which are “Paretosuperior” to it. An outcome is Pareto-superior to another if and only if no one (strictly) prefers the second to the first and at least one person (strictly) prefers the first to the second. In the Prisoners' Dilemma, the equilibrium outcome is “Pareto-inferior” to another.

²¹ An outcome is stable in this sense if and only if no one has an incentive to act differently given the action of others. In genuine Prisoners' Dilemmas, the equilibrium outcome where the parties all adopt their utility-maximizing strategy possesses a particular kind of stability, that of being a Nash equilibrium.

²² There are problems with understanding the state of nature as a Prisoners' Dilemma, problems that are not apparent if we limit our exposition to two-person games and to 2 × 2 matrices. The game theoretical structure of such situations may not give rise to dominant strategies. See Hampton, Jean, “Free Rider Problems in the Production of Collective Goods”, Economics and Philosophy 3 (1987), 245–273CrossRefGoogle Scholar. These problems, however, will not affect my argument against Levin.

²³ Nozick, , Anarchy, ix, 26.Google Scholar

²⁴ Indivisibility should not be confused with rivalness: a good is rival to the extent that one person's enjoyment of the good affects another's enjoyment of the same good (e.g., a beach or park subject to crowding).

²⁵ My account of social order as a public good is indebted to Michael Taylor, especially his Community, Anarchy and Liberty (Cambridge: Cambridge University Press, 1982).Google Scholar

²⁶ This is one of the main theses of James Buchanan's important work, The Limits of Liberty (Chicago, IL: University of Chicago Press, 1975)Google Scholar. The provision of public goods, as well as fiscal and distributive decisions regarding their provision, are the concern of the “Productive State”. The latter Buchanan contrasts to the “Protective State” which has as its central concern the enforcement of law.

²⁷ The move from state provision of one public good, social order, to its provision of others may be blocked by pointing to asymmetries between goods that are the absence of invasion or harm (e.g., defense, pollution control, prevention of fraud) and goods that involve the provision of some benefit or positive service (e.g., public transportation). I am sceptical that this asymmetry can be made out; support for this line of objection might, however, be found in Gauthier, Morals by Agreement, chap. 7. I am indebted to Jan Narveson for this objection to my argument. It should, of course, be noted that efficient state provision of one public good is not an argument for inefficient attempts to provide another.

²⁸ See note 23.

²⁹ See virtually any introductory text in micro-economics or welfare economics.

³⁰ J. Roland Pennock interprets Levin as conceding to the Hobbesian state the establishment of the right to make contracts and, he presumes, that of eliminating fraud. Although Levin does not challenge this interpretation in his reply to Pennock, I think that the original article is unclear on these matters. See Pennock, , “Correspondence”, esp. 256–257Google Scholar, and Levin, , “Reply to Pennock”, both in Philosophy & Public Affairs 13 (1984), 255–262, 263–267.Google Scholar

³¹ Nozick, , AnarchyGoogle Scholar, Part 1.

³² Raz, Joseph, “The Functions of Law” (1973), reprinted in his The Authority of Law (Oxford: Clarendon Press, 1979), 170–171.Google Scholar

³³ At least initially. Gauthier's Hobbesian moral theory moralizes the state of nature prior to and independently of the state. See Gauthier, , Morals by Agreement, chap. 7.Google Scholar

³⁴ Contra Levin: “since the minimal Hobbesian state would protect people against invasion, we can identify property in such a state as everything an individual can secure without invading the sphere of another's activities, and this would effectively coincide with the fruits of one's labor and cooperative undertakings” (“Reply to Pennock”, 265).Google Scholar

³⁵ This is what Richard Musgrave has called the “distribution function”, contrasting it with allocation and stabilization functions. See Musgrave, Richard A. and Musgrave, Peggy B., Public Finance in Theory and Practice (4th ed.; New York: McGraw-Hill, 1984)Google Scholar, chap. 1.

³⁶ It should be pointed out that normative questions aside, it turns out to be very difficult to ascertain how much redistribution actually goes on in Western democracies and who, if anyone, is the net beneficiary. For some recent studies see Tullock, Gordon, Economics of Income Redistribution (Boston, MA: Kluwer-Nijhoff, 1983)CrossRefGoogle Scholar and Page, Benjamin I., Who Gets What From Government (Berkeley and Los Angeles, CA: University of California Press, 1983)Google Scholar. Although Tullock and Page are at different ends of the political spectrum, they appear to agree that the poor are not significant beneficiaries of redistribution in the United States compared to other groups.

³⁷ I should note that here and elsewhere when I talk of “benefits”, “interests”, and their cognates I do not mean “preferences”. The latter can be other-regarding and do not range only over human interests.

³⁸ Redistribution is possibly wasteful in other ways. One of these may be an unproductive use of resources to bring about the desired transfers—Arthur Okun's “leaky bucket”. Another may be the expenditure of resources by parties interested in securing redistributed resources for themselves. The latter activity has been called “rent-seeking” by some political economists. See Buchanan, James, Tollison, Robert, and Tullock, Gordon, eds., Toward A Theory of the Rent-Seeking Society (College Station, TX: Texas A & M University Press, 1980)Google Scholar. I shall not discuss complications resulting from these inefficiencies.

³⁹ Buchanan, James M. and Tullock, Gordon, The Calculus of Consent (Ann Arbor, MI: University of Michigan Press, 1962), chap. 13CrossRefGoogle Scholar; Zeckhauser, Richard, “Risk Spreading and Distribution”, in Hochman, Harold M. and Peterson, George E., eds., Redistribution through Public Choice (New York: Columbia University Press, 1974), 206–228Google Scholar; and Goodin, Robert, The Politics of Rational Man (London and New York: John Wiley & Sons, 1976), 115.Google Scholar

⁴⁰ See my “A Non-Egalitarian Defense of Redistribution”, in Bradie, Michael and Braybrooke, David, eds., Social Justice, Bowling Green Studies in Applied Philosophy IV (Bowling Green, OH: Bowling Green State University, 1982), 68–84Google Scholar, as well as the references therein.

⁴¹ See Frank, Robert H., Choosing the Right Pond: Human Behavior and the Quest for Status (New York: Oxford University Press, 1985)Google Scholar, especially chap. 6.

⁴² “Libertarian welfare state” comes from Robert Frank. See chap. 12 of Finding the Right Pond.

⁴³ This manner of making the point against Levin was suggested to me by Michael Robins.

⁴⁴ Perhaps this feature of property, that it facilitates private arrangements in ways that benefit all is a rationale for considering property law part of private law, along with contracts, business law, marriage law, and the like.

⁴⁵ Nozick, , Anarchy, 198.Google Scholar

⁴⁶ I owe this thought originally to Alan Donagan.

⁴⁷ This is the much discussed “collective charity” argument of Harold M. Hochman and John D. Rodgers. See their “Pareto Optimal Redistribution”, American Economic Review 59 (1969), 542–547Google Scholar, as well as the extensive literature that their idea provoked. See also Nozick, , Anarchy, 265–268Google Scholar, for some criticisms of this line of argument.

⁴⁸ Thurow, Lester, “The Income Distribution as a Pure Public Good”, Quarterly Journal of Economics 85 (1971), 327–336.CrossRefGoogle Scholar

⁴⁹ See Buchanan, James M., “The Political Economy of Franchise in the Welfare State”, in Seiden, R., ed., Capitalism and Freedom (Charlottesville, VA: University of Virginia Press, 1975), 52–77.Google Scholar

⁵⁰ It can sometimes be a difficult matter determining what legislation actually is pater nalistic. For instance, requiring hockey players to wear helmets may actually be non-paternalistic. See Schelling, Thomas, “Hockey Helmets, Daylight Saving, and Other Binary Choices”, in his Micromotives and Macrobehavior (New York: W. W. Norton, 1978), 211–243.Google Scholar

⁵¹ It is important to distinguish clearly the ways in which paternalistic and nonpaternalistic practices affect people's interests. Prohibitions of theft are not paternalistic, even though our interests are secured thereby. For it is, at most, the system of rules forbidding theft that is in my interests. More precisely, it is the state's prohibition of theft by others that is in my interests, just as the state's prohibition of theft on my part is in the interests of others. It is not in my interest to refrain from taking the possessions of others; this, however, I must do if I am to secure my possessions against others. By contrast, the state's requiring me to wear a helmet when I ride a motorcycle or to save part of my income for my retirement can be paternalistic if it is the case the state requires something of me for my interest.

⁵² Thomas Schwartz argues that “Cycles [that is, violations of transitivity, as well as of acyclicity] are not peculiar to collective as opposed to individual preference”. The Logic of Collective Choice (New York: Columbia University Press, 1986), 113–115.Google Scholar

⁵³ There is some disagreement about this. Members of the Austrian school in economics tend to understand the assumption of rationality as a priori. See also Davidson, Donald, Essays on Actions and Events (Oxford: Clarendon Press, 1980).Google Scholar

⁵⁴ Admitting that people are not always fully rational is not inconsistent with rational choice theory. See Elster, Jon, Ulysses and the Sirens (Cambridge: Cambridge University Press, 1979)Google Scholar, and Sour Grapes (Cambridge: Cambridge University Press, 1983).Google Scholar

⁵⁵ I recommend Taylor, Community, Anarchy and Liberty, to those who do not take seriously the possibility of anarchy as an alternative to the modern state.

⁵⁶ Levin's argument, however, was not meant to rest on largely empirical considerations. Not so of the Friedmans' defense of the minimal state. See Friedman, Milton and Friedman, Rose, Free to Choose (New York: Harcourt Brace Jovanovich, 1979)Google Scholar. For the more radical position of the family, see Friedman, David, The Machinery of Freedom (New Rochelle, NY: Arlington House, 1978).Google Scholar

⁵⁷ Leviathan, chap. 30. See my essay referred to in note 6.

⁵⁸ Barman, Gilbert, “Moral Relativism Defended”, Philosophical Review 84 (1975), 12–13Google Scholar. Also, “The fact that compromises are necessary among people of different powers and resources accounts for the fact that the duty to avoid harm to others is much stronger in our morality than is the duty to help those who need help …. Since everyone would benefit equally from a duty not to harm others, whereas the poor and weak would benefit much more from a duty of mutual aid than the rich and powerful would, the expected compromise is a fairly strong prohibition against harm and a weaker duty of mutual aid, which is what we have in our morality” (“Relativistic Ethics: Morality as Politics”, in French, Peter A. et al. eds., Midwest Studies in Philosophy, vol. 3 [Morris, MN: University of Minnesota, 1978], 114Google Scholar). See also “Moral Relativism as a Foundation for Natural Rights”, Journal of Libertarian Studies 3 (1979), 367–371.Google Scholar

⁵⁹ Here it is not important for my purposes to determine which of these standards is most appropriate.

⁶⁰ The moral framework that Karman and other Hobbesians employ is individualistic in the sense that it is the perspective of the agent that determines whether she has a reason for action. Thus, claims about equal benefit must be carefully formulated so as to avoid presupposing a standard for interpersonal comparisons of utility. It is not that interpersonal comparisons are not possible; for they are. Rather, it is that they do not necessarily generate reasons for action for individuals.

⁶¹ For such a measure of individual advantage, see Gauthier, , Morals by AgreementGoogle Scholar, chap. 5.

⁶² This is the argument of Buchanan, “A Hobbesian Interpretation of the Rawlsian Difference Principle”, in his Freedom in Constitutional Contract (College Station, TX: Texas A & M Press, 1977), 194–211.Google Scholar

⁶³ See Tullock, Gordon, “The Charity of the Uncharitable”, Western Economic Journal 9 (1971), 379–392Google Scholar, and “The Rhetoric and Reality of Redistribution”, Southern Economic Journal 9 (1979), 895–907.Google Scholar

⁶⁴ This explanation is, of course, speculative. Another explanation might be that our governments do not do what we wish them to do and that elected officials redistribute in ways that best serve them; citizens may vote, foreigners may not. We may not be able to infer very much from the pattern of government redistribution alone.

⁶⁵ Murray Rothbard and other contemporary American libertarians defend anarchism as being the only position compatible with the assumption of natural rights. I tend to think, contra Nozick, that they are right on this point. See Rothbard, Murray, For A New Liberty (rev. ed.; New York: Collier Books, 1978).Google Scholar