Published online by Cambridge University Press: 18 June 2009
Some ways of defending inequality against the charge that it is unjust require premises that egalitarians find easy to dismiss—statements, for example, about the contrasting deserts and/or entitlements of unequally placed people. But a defense of inequality suggested by John Rawls and elaborated by Brian Barry (who themselves reject the premises that egalitarians dismiss) has often proved irresistible even to people of egalitarian outlook. The persuasive power of this defense of inequality has helped to drive authentic egalitarianism, of an old-fashioned, uncompromising kind, out of contemporary political philosophy. The present essay is part of an attempt to bring it back in.
1 See Rawls, John, A Theory of Justice (Cambridge: Harvard University Press, 1971), sections 10–17Google Scholar; and Barry, Brian, Theories of Justice (London: Harvester Wheatsheaf, 1989), pp. 213–34.Google Scholar
2 Whether Rawls and Barry should themselves be styled “egalitarians” is not a matter that needs to be addressed in this essay.
3 Barry, , Thearies of Justice, p. 217.Google Scholar
4 Definitions: State A is strongly Pareto-superior to state B if everyone is better off in A than in B, and weakly Pareto-superior if at least one person is better off and no one is worse off. If state A is Pareto-superior to state B, then state B is Pareto-inferior to state A. State A is Pareto-inferior (tout court) if some state is Pareto-superior to A. State A is Pareto-optimal if no state is Pareto-superior to A: it is strongly Pareto-optimal if no state is weakly Pareto-superior to it, and weakly Pareto-optimal if no state is strongly Pareto-superior to it. States A and B are Pareto-incomparable if neither is (even weakly) Pareto-superior to the other. A change is a weak Pareto-improvement if it benefits some and harms none, and a strong Pareto-improtfmetit if it benefits everyone. The Pareto principle mandates a Pareto-improvement whenever one is feasible: the strong principle mandates (even) weak Pareto-improvements, 2nd the weak one only strong Pareto-improvements.
5 Rawls, , A Theory of Justice, p. 7.Google Scholar Cf. Rawls, John, Political liberalism (New York: Columbia University Press, 1993), p. 80.Google Scholar
6 See Barry, Theories of Justice, pp. 213–15; and see Rawls, A Theory of Justice, p. 104.
7 Barry, Theories of Justice, p. 224. That may seem to be a strange thing to say. In Section II, I try to make it seem less strange.
8 Ibid., p. 226.
9 The claim that the Pareto-improving unequalizing move might be accepted on grounds other than justice—that, indeed, Rawls's own case for it is not really one of justice—was made, persuasively, by Lyons, David in his “Nature and Soundness of the Contract and Coherence Arguments,” in Reading Rawls, ed. Daniels, Norman (Oxford: Blackwell), 1975), pp. 152–53.Google Scholar
10 Rawls, , A Theory of Justice, p. 65.Google Scholar
11 Although, as I have said, I am in sympathy with the first part of the argument, I also have reservations about it. Rawls's use of the motif of moral arbitrariness is subjected to (as yet) largely unanswered searching criticism by Nozick, Robert in his Anarchy, State, and Utopia (New York: Basic Books, 1974), pp. 213–27.Google Scholar
12 Rawls, , “The Basic Structure as Subject,” in Political Liberalism, p. 281Google Scholar, emphasis added; and see Rawls, , “A Well-Ordered Society,” in Philosophy, Politics, and Society, 5th series, ed. Laslett, Peter and Fishkin, James (New Haven: Yale University Press, 1979), pp. 15–16.Google Scholar
13 Cohen, G. A., “Incentives, Inequality, and Community,” in The Tanner Lectures on Human Values, vol. 13, ed. Peterson, Grethe B. (Salt Lake City: University of Utah Press, 1992).Google Scholar
14 Rawls, , A Theory of Justice, p. 72.Google Scholar
15 Ibid.
16 Thomas Nagel's defense of utilitarianism against Rawls is relevant here: see Nagel, , “Equality,” in his Mortal Questiotts (Cambridge: Cambridge University Press, 1979), p. 119.Google Scholar
17 As I indicate in note 20 below, I do not believe that Rawls construes libertarianism in the suggested strained fashion.
18 Barry, , Theories of Justice, p. 222.Google Scholar
19 Rawls, , A Theory of Justice, p. 64.Google Scholar
20 In the foregoing paragraphs I present the movement from natural liberty to equality as an entirely immanent one, each of its first two sub-stages yielding to its successor because the latter is more faithful to its predecessor sub-stage's rationale. I believe, without being certain, that Barry intends to ascribe precisely such an argument to Rawls, but I am certain that the structure of the argument Rawls himself presents (see A Theory of Justice, sections 12 and 13) is in one way different from the structure of the one that I have constructed, which is based on pp. 218–20 of Barry's Theories of Justice.
Rawls proceeds as follows. He cites the principles of “natural liberty,” “liberal equality,” and “democratic equality” as affording three readings of the ideal of equality of opportunity. Natural liberty is rejected because of its complacency about the influence of natural and social luck on shares. Because liberal equality aims to suppress social luck, it is commended, but it is then criticized, on the ground that there is no difference between social and natural luck that would justify exclusive preoccupation with the former.
Accordingly, the movement from liberal equality to democratic equality is indeed an immanent one, but the movement from natural liberty to liberal equality is not. Natural liberty is not rejected by Rawls because of any internal incoherence of the sort that Barry gestures at, but because it fails to resist the morally arbitrar).
21 Barry, , Theories of Justice, p. 224.Google Scholar
22 Different egalitarians believe that different things should be equalized, so they would endorse and condemn different outcomes, but that does not matter here: the phrase “outcome that egalitarians would endorse” can be treated here as a variable, since the argument we shall examine is for inequality no matter what conception of equality is embraced.
23 If, as Barry says, true equality of opportunity amounts to equality of outcome, then stage two of the Pareto argument, in asking us to abandon equality, also asks us to abandon true equality of opportunity. I doubt that Barry was fully alive to this aspect of his position, which implies that “the most just society” (Theorics of Justice, p. 217Google Scholar, my emphasis; and cf. p. 234) lacks equality of opportunity.
24 Strictly, of social primary goods, but, as Rawls himself does, I shall abbreviate to “primary goods” where misunderstanding is unlikely. Primary goods are “things that every rational man is presumed to want” since they “normally have a use whatever a person's rational plan of life.” The “social primary goods” are the primary goods “at the disposition of society,” to wit, “rights and liberties, powers and opportunities, income and wealth,” and the social bases of self-respect. “Other primary goods such as health and vigor, intelligence and imagination, are natural goods; although their possession is influenced by the basic structure, they are not so directly under its control” (Rawls, , A Theory of Justice, p. 62Google Scholar). Will Kymlicka argues that Rawls has no good reason for restricting his index of well-being to the social primary goods: see Kymlicka, , Contemporary Political Philosophy (Oxford: Oxford University Press, 1990), pp. 70–73.Google Scholar Thomas Pogge's defense of what he calls Rawls's semiconsequentialism-see Pogge, , Realizing Rawls (Ithaca: Cornell University Press, 1989)Google Scholar—supports Rawls's restriction of the purview of justice to social primary goods.
25 To, that is, the worst off: better-off people gain immediately from inequality. The idea is that we withhold from the poor in order to give more to the rich and, as a result, the poor end up better off, later.
26 Rawls, , A Theory of Justice, p. 151.Google Scholar
27 Ibid., p. 546.
28 Ibid., p. 62.
29 Ibid., p. 101; and see ibid, pp. 15, 75, 79, 102, 179; and, decisively in favor of the interpretation of Rawls given in the present paragraph, see Rawls, , “A Well-Ordered Society,” pp. 16–17.Google Scholar
30 Rawls, , Justice as Fairness: A Briefer Restatement (Harvard University typescript, 1989), p. 57Google Scholar; and see ibid., pp. 56, 89.
In my interpretation of the passages cited in this and the previous footnote, the constraint placed on how far ahead the talented may get depends on its being good for them that they have more talent. In a suggested counter-interpretation, Rawls says that talent endowments are morally arbitrary, so no one may benefit from them as a matter of desert, but, if the well endowed benefit from their talent in ways that also help the untalented, then the benefit redounding to the former is acceptable. In the counter-interpretation, it is immaterial to Rawls's argument that, as he mentions, talent is a good thing.
The counter-interpretation argument may be both coherent and Rawlsian (for there is more than one argument in Rawls), but, for two reasons, I rest with my own interpretation of the cited passages. First, the counter-interpretation does not account for the rhetorically strategic placement of the repeated reminder that talent is beneficial. Second, and perhaps more importantly, Rawls is in my interpretation readily seen to be addressing an intuitively grounded protest against superior emolument to the talented, a protest that such emolument follows the notion that “to them that hath shall be given”: see the reference to Mill at the beginning of Section III below. Rawls is naturally interpreted as replying: that is fine, provided that those who hath not also benefit.
As I have pointed out elsewhere (“Incentives, Inequality, and Community,” footnote 6), there is ambiguity in Rawls with respect to which inequalities are permitted. Many texts, such as those cited above, permit only inequalities that help the worst off. Other texts, those, for example, in which the leximin principle is affirmed, also permit inequalities that do not harm them (see, e.g., Rawls, , A Theory of Justice, p. 83Google Scholar). The counter-interpretation argument corresponds to the less-egalitarian leximin principle. Unlike my interpretation, it does not explain why maximin might take a restrictive form, and that is the form in which maximin appears in the cited passages.
31 John Stuart Mill so describes higher reward to the talented in section 4 of Chapter I of Book II of his Principles of Political Economy: see Collected Works of John Stuart Mill, ed. Robson, J. M. (Toronto: Toronto University Press, 1965), vol. 2, p. 210.Google Scholar For an adverse contrast Between Rawls and Mill on this issue, see the final page of my “Incentives, Inequality, and Community.”
32 Rawls, , A Theory of Justice, p. 62.Google Scholar
33 For related points, see Dworkin, Ronald, “Equality of Resources,” Philosophy and Public Affairs, vol. 10, no. 4 (Fall 1981), p. 343.Google Scholar
34 As Michael Lessnoff pointed out, it was inconsistent with Rawls's understanding of justice as a matter of “the proper distribution of the benefits and burdens of social cooperation” (A Theory of Justice, p. 5Google Scholar; emphasis added) that he omitted labor burdens from the index of social primary goods; see Lessnoff, , “Capitalism, Socialism, and Justice,” in Justice and Economic Distribution, ed. Shaw, William and Arthur, John (Englewood Cliffs: Prentice-Hall, 1978 edition), p. 143.Google Scholar
Later, Rawls contemplated including leisure time among primary goods: see his “Reply to Alexander and Musgrave,” Quarterly Journal of Economics, vol. 88 (11 1974), p. 654Google Scholar, and “The Priority of Right and Ideas of the Good,” Philosophy and Public Affairs, vol. 17, no. 4 (Fall 1988), p. 257.Google Scholar And Thomas Pogge is quite definite: “(T]he index must also include leisure time as a distinct social primary good. This good can be defined simply as the inverse of time worked, which is a burden of social cooperation” (Realizing Rawls, p. 198Google Scholar). This is not entirely satisfactory since, if “time worked” is a burden, it is one that varies importantly in size with the character of the job and the make-up of the person. (Cf. Baker, John, “An Egalitarian Case for Basic Income,” in Arguing for Basic Income, ed. Van Parijs, Philippe (London: Verso, 1992), p. 125Google Scholar, note 4.) It makes matters more difficult still that some people find their fulfillment in their work. Is it a burden of social cooperation for them too?
Nineteen years after the publication of “Reply to Alexander and Musgrave,” Rawls sustains his noncommitment on the matter of whether leisure should bo included among the primary goods (see Political Liberalism, pp. 181–82Google Scholar). I believe that this fence-sitting reflects the fact that labor is a burden (and, sometimes, benefit!) of social cooperation (so it should be a primary good) that fails the test of public checkability (see A Theory of Justice, p. 95Google Scholar) laid down for primary goods (so it should not be one).
Henceforth, when I speak of “labor burden,” I mean to refer not only to the quantity of a person's labor but also to its character in the sense of how burdensome or oppressive it is. Notice that, even if A Theory of Justice had included leisure time as a primary social good (or labor time as a primary social bad), we should still have been in the dark not only about the character but also about the mere quantity of people's labor inputs in D1.
35 Rawls, , A Theory of Justice, p. 62Google Scholar; see the text marked by note 28 above.
36 See note 4 above. D2 and D3 are Pareto-incomparable because in each some are better off than they are in the other.
37 For some readers, the following tabular presentation of the above array of comparisons may be helpful:
(The symbol “<” means “is less than,” “>” means “is greater than,” and “=” signifies equality.)
38 See my “Incentives, Inequality, and Community,” section VII, for an extended discussion of, in effect, this issue.
39 Which is to say that their labor supply curve is in the relevant region vertical. It could also be backward-bending, in which case they would produce more at We than they do at Wt.
40 It might therefore be suggested that they be paid less than others. But I shall not pursue that suggestion here: I am deconstructing an argument-for inequality, not sketching a complete egalitarian theory.
41 Rawls, , A Theory of Justice, p. 15.Google Scholar
42 Ibid., p. 75.
43 See, respectively, ibid., pp. 151, 78.
44 This holds whether or not D1 itself turns out, after further required specification, to be an equal distribution: that would partly depend on comparative labor burdens in D1, which were not specified in Section III above.
Note that, had we specified earned-income equality in D1 as wage per unit output (as opposed to, as we did, as wage per unit time), we should then have been unable to represent D1 as a state of equality without stipulating that the talented carry special labor burdens, since, for morally arbitrary reasons, they would then have been earning more per hour than the untalented. By stipulating equal wages per hour, we concealed, up to now, the labor-burden issue.
Ex hypothesi, the talented get the same primary goods as others at D1. If leisure is a primary good, then they work as many hours as others, therefore produce more, but get the same income. If leisure is not a primary good, and payment is per unit output, then they get more leisure and/or more primary goods, and not, in general, the same basket of primary goods, thus specified, but just the same rate of primary goods per unit output. It would be hollow to call that a society of equality, as Rawls's own (curiously unconsummated-see note 34 above) inclination to acknowledge leisure as a primary good betrays. If talented people should not get more just because they are talented, they should not get more leisure in the initial equality.
45 See Rawls, , A Theory of Justice, p. 62.Google Scholar
46 He does so in his affirmation of the Aristotelian Principle: “[Olther things equal, human beings enjoy the exercise of their realized capacities (their innate or trained abilities), and this enjoyment increases the more the capacity is realized, or the greater its complexity” (Ibid., pp. 426–27). It is the final phrase, much emphasized in the rest of the paragraph that this statement of the Principle introduces, that matters in our inquiry. (Anthony Skillen noted that the Aristotelian Principle is hard to combine with Rawls's endorsement of incentives for people of talent. Rawls “assumes … that the very minority who can enjoy stretching their capacities, the entrepreneurs, will need great wealth and prestige to motivate their activity”; see Skillen, , Ruling Illusions [Hassocks, Sussex: The Harvester Press, 1977], p. 47.)Google Scholar
47 Rawls, , A Theory of Justice, p. 105.Google Scholar
48 Which would mean, here, that the brothers should move to Chicago, where they could convert their benefit levels to 7.55 each.
49 Or, more strictly, such a D4 will be possible only in virtue of those few inequalities, at which I nodded at the beginning of Section IV above, which do not depend on the will. There will then be much less inequality than Rawls and Barry think can be justified.
50 Compare the defense of Barry's use of the phrase “equality of outcome” in Section II above. There could be a feasible set of collections of apples and oranges, the same for all takers, and we could think of income and leisure on the model of apples and oranges.
51 For David Gauthier's contractariantsm, see his Morals by Agreement (Oxford: Oxford University Press, 1986)Google Scholar, especially ch. 6; and for Thomas Scanlon's, see his “Contractualism and Utilitarianism,” in Utilitarianism and Beyond, ed. Sen, Amartya and Williams, Bernard (Cambridge: Cambridge University Press, 1982).CrossRefGoogle Scholar
52 See Barry, , Theories of Justice, pp. 7–8Google Scholar, and passim, using index entry on “impartiality.” In his “Constructing Justice,” Philosophy and Public Affairs, vol. 20, no. 3 (Summer 1991), pp. 266ff.Google Scholar, Allan Gibbard sketches a possible third conception, “justice as fair reciprocity.” which would be intermediate between Barry's two. Gibbard is uncertain that it is truly distinct and coherent. Rawls (Political Liberalism, p. 17Google Scholar) commends Gibbard for successfully identifying what he, Rawls, is after, but he does not address Gibbard's reservations about the notion, and, in Ibid., pp. 17 and 50, he mischaracterizes “mutual advantage” so as to lose contact with Barry's notion.
In my own view, Gibbard's “justice as fair reciprocity” is only superficially distinct: if we press for its rationale, its hybrid character becomes apparent, a fork opens, and we have to choose, again, between mutual advantage and impartiality.
53 See, e.g., Barry, , Theories of Justice, p. 213.Google Scholar
54 The curves are presented in Rawls, , A Theory of Justice, pp. 76–77Google Scholar, and in Rawls, , Justice as Fairness, p. 46AGoogle Scholar; and they are discussed by Barry, in Theories of Justice, pp. 229ff.Google Scholar
55 Barry, , Theories of Justice, p. 226; my emphases.Google Scholar
56 Rawls himself does not say that the initial equality is prima facie just. But he is seeking to justify a principle of justice, and he regards equality as the correct starting point in such an exercise, because citizens are “free and equal” (see Section I above). The obscurity surrounding “prima facie just” attaches, similarly, to the idea of the correct starting point.
57 One may distinguish (see Section I above) between justification by causes and justification by consequences, but why does the fact that causes come first in the causal order mean that they are at a more basic level in the order of justification?
58 Barry, , Theories of Justice, p. 232.Google Scholar
59 Ibid., p. 234.
60 Ibid.
61 Strictly, that is the wrong thing to say here, and for reasons that have nothing to do with the criticism of this kind of formulation that dominates the present essay. For, as Barry recognizes elsewhere, one may “not ascribe a fixed identity to the worst off group” (ibid., p. 216). Yet he does just that here; and again at p. 233, where he says: “The worst off gain as much as they possibly can from inequality, so they have no reasonable complaint”; and again at p. 242. where he says that “the reply to one who does badly is that if others did not do better he would be even worse off.” The right reply, in line with p. 216, is not that he, but that someone would be even worse off than he now is, be that someone he or somebody else.
The recurrent violation of the stricture laid down at p. 216 is a not uninteresting slip. In this slip the contractarian motif, which requires fixed identities, replaces the framework of impartiality, in which identities are dissolved.
For criticism of Rawls's significant equivocation across fixed and variable reference in his uses of the phrase “the worst off,” see Parfit, Derek, Reasons and Persons (Oxford: Oxford University Press, 1984), appendix HGoogle Scholar; and Cohen, G. A., “Self-Ownership, World-Ownership, and Equality,” in Social Justice Here and Now, ed. Lucash, Frank (Ithaca: Cornell University Press, 1986), pp. 133–34.Google Scholar
62 Barry, , Theories of Justice, p. 233; emphasis added.Google Scholar