Justice as Fairness and Reciprocity

Because of the prominence it gives to motivation by reciprocity, John Rawls's theory of "justice as fairness" has been criticized for being insufficiently demanding with respect to our duties to the disabled and to people in other countries. Rawls's defenders have responded by downplaying reciprocity, emphasizing instead equal respect for persons as moral subjects as the foundation of the contractualist point of view. Yet reciprocity played important roles in Rawls's account of stability and his account of the limits of duties of justice. This paper tries to reconcile reciprocity with a fundamentally "subject-centred" ethic by interpreting the reciprocity condition as a consequence of the fact that justice is in part a relational value. Duties of egalitarian distributive justice are not grounded on the duty to reciprocate benefits already received, but limited by a reasonable assurance of compliance on the part of those able to reciprocate, because their point is to constitute a valuable relationship, one of mutual recognition as equals. We have unconditional duty to help establish just global institutions, institutions which would allow us to share fairly in the burdens and benefits of global economic cooperation, but no unilateral duty to share fairly, where such institutions are not in place. Since non-contribution on the part of those unable to contribute involves no failure of recognition, the disabled do not fall outside the scope of distributive justice. Acknowledgements I have presented earlier earlier work on justice and reciprocity at the Queen's Political Philosophy Reading Group, the 2010 meeting of the Canadian Political Science Association, and the Manchester Political Theory Workshop. I presented earlier versions of this paper to the University of Zurich's Centre for Ethics and Balliol College. I am very grateful to all of these groups as well as to one anonymous referee for the comments and criticisms received. 0. Introduction The idea of reciprocity played a prominent role in John Rawls's conception of "justice as fairness." Rawls described our sense of justice as an extension of an innate tendency to "answer in kind." He assumed that people were capable of acting from a sense of justice 1. John Rawls, A Theory of Justice (Cambridge, Mass.: Harvard Univeristy Press, 1999), 433. At one point Rawls referred to his theory by the label "justice as reciprocity," the title of a 1959 essay that appeared in 1971; John Rawls, “Justice as Reciprocity,” in Collected Papers, ed. Samuel Freeman (Cambridge, MA: Harvard University Press, 1999). Rawls later called political liberalism's fundamental requirement of public justifiability a principle of reciprocity; John Rawls, Political Liberalism (New York: Columbia University Press, 1996), xlvi, li. This paper focuses on the role of reciprocity in "justice as fairness," touching on broader issues related to political liberalism only briefly with respect to the idea of overlapping consensus. Henceforward all references to A Theory of Justice will be made parenthetically in the text. All references are to the revised edition. against personal interest, but not willing to comply with principles of justice without a reasonable assurance that others would also comply. Thus the natural duty of justice, according to Rawls, was to help create and comply with just institutions, not to share fairly on a unilateral basis where institutions ensuring general compliance do not exist. The importance Rawls placed on reciprocity may seem a concession to feasibility and human nature. A series of prominent critics has argued that "justice as fairness" is insufficiently demanding with respect to our duties to persons with disabilities and persons in other countries. The severely disabled do not generally contribute to economic production. Foreigners benefit us by trading with us, but they do not benefit us by complying with the laws of our political society, or by contributing to essential public goods, and they can't necessarily be counted upon to share fairly with us, if we share fairly with them, because there is no global state to enforce a particular standard of distributive justice. Any view that derives duties of distributive justice from a more fundamental duty to reciprocate benefits or that limits duties of distributive justice by an expectation of compliance on the part of others will not see us as having stringent duties to non-contributors and outsiders. In response, Rawlsians have sought to downplay the importance of reciprocity in justice as fairness, emphasizing instead the principle of equal respect for persons as moral agents, independent of their capacities to threaten or benefit others. While respect for persons was undeniably central to Rawls's view, reciprocity also played important roles in the argument for stability and the account of the limits of duties of justice. This paper tries to reconcile the Kantian concern with respect for persons with the Humean concern with reci2. Brian Barry, Theories of Justice, vol. (Berkeley: University of California Press, 1989); Alan Buchanan, “Justice as Reciprocity Versus Subject-Centered Justice,” Philosophy and Public Affairs 19, no. 3 (1990): 227-52; Brian Barry, “John Rawls and the Search for Stability,” Ethics 105 (1995): 874-915; G. A. Cohen, SelfOwnership, Freedom, and Equality (Cambridge: Cambridge University Press, 1995); Martha Craven Nussbaum, Frontiers of Justice (Cambridge, MA: Harvard University Press, 2006). 2 procity by construing the reciprocity condition as a consequence of the fact that justice is in part a relational value. Duties of egalitarian distributive justice are not grounded on the duty to reciprocate benefits already received, but they are limited by a reasonable assurance of compliance on the part of those able to reciprocate, because their point is to constitute a valuable relationship, one of mutual recognition as equals. The basic idea is that some duties are required by people's status as moral agents, independent of their attitudes and conduct, while others are constitutive of a relationship that is appropriate to equals engaged in cooperation. The first section of the paper describes the core complaint of the disability critique of justice as fairness, and assesses the Rawlsian response. The second explains the roles reciprocity plays in Rawls's theory of justice, arguing that reciprocity cannot ground but only limit duties of justice. The third section interprets this reciprocity condition as an intrinsic part of distributive justice rather than merely a psychological constraint or a limitation of human nature. The main idea is that some duties are conditional on reciprocity because the point of the duties is to constitute a relationship. The fourth section sketches the view's implications for the issues of global justice and disability. 1. Justice as Fairness and the Mutual Advantage Tradition A Theory of Justice was a theory for a highly simplified setting: a single, isolated society of non-disabled persons. Rawls's assumption was that with a theory for the simple 3. The 'single isolated society' restriction of scope is explicit in Rawls's opening chapter (4, 7). The 'no disabilities' assumption only comes to the fore in Chapter 2; "I shall assume that everyone has physical needs and psychological capacities within the normal range, so that the questions of health care and mental capacity do not arise... The first problem of justice concerns the relations among those who in the everyday course of things are full and active participants in society and directly or indirectly associated together over the whole span of their life" (84). Following this terminology, Cynthia Stark calls the 'no disabilities' assumption the "fully cooperating assumption;" Cynthia Stark, “How to Include the Severely Disabled in a Contractarian Theory of Justice,” The Journal of Political Philosophy 15, no. 2 (2007), 129-30. Stark's interpretation of the FCA is that all mentally disabled persons and all severely physically disabled persons do not count as fully cooperating. The only problem with the "fully cooperating" terminology is that it might be taken to suggest that the severely disabled are excluded from the scope of those owed duties of justice because they are uncooperative, instead of simply being outside the set of subjects about which the theory makes claims. Adam Cureton offers a good account of what he calls Rawls's "limited project of justice;" Adam Cureton, “A Rawlsian Perspective on Justice for the Disabled,” Essays in Philosophy 9, no. 1 (2008). 3 case, we could more easily address the question of what justice requires in the more general case. As a result, justice as fairness makes no direct claims about what disabled persons are owed, as a matter of justice. However, in setting up the original position Rawls appealed to Hume's circumstances of justice: moderate scarcity, limited benevolence, and rough equality of capabilities (109-110). The equal capabilities condition led critics such as Brian Barry and Martha Nussbaum to assert that Rawls espouses, at least to some extent or in some form, the view of justice as mutual advantage. According to an old tradition, rules of justice are the result of an agreement between rational egoists roughly equal in capabilities, each of whom settles for not doing harm so as not to suffer it. On this view, people do others good by complying with rules of justice, but only so as to ensure that other people do the same with respect to them. Rational individuals are committed to reciprocity – returning good in proportion to the good received – but only for forward-looking and self-interested reasons. Brian Barry called this view "justice as mutual advantage"; Allen Buchanan called it "justice as self-interested reciprocity." Hume was explicit that if there were a species of rational creatures so inferior in strength that they were incapable of all resistance, we would owe them only humanity, not justice. Hume's frankness on this point led Barry to question 4. "A theory that fails for the fundamental [i.e. simpler] case is of no use at all," John Rawls, “Kantian Constructivism in Moral Theory,” Journal of Philosophy 77 (1980), 545-46; cited by Harry Brighouse, “Can Justice as Fairness Accommodate the Disabled?,” Social Theory and Practice 27, no. 4 (2001), 538. 5. Barry, Theories of Justice., Part II, "Hume and Rawls on Justice In Society"; see alsoBrian Barry, Justice as Impartiality (Oxford: Oxford University Press, 1995), 50; Nussbaum, Frontiers of Justice, 61-62. 6. For Rawls's description of this tradition, see John Rawls, “Justice as Fairness,” in Collected Papers, ed. Samuel Freeman (Cambridge, Mass.: Harvard University Press, 1999), 56. 7. Lawrence C. Becker, Reciprocity (London ; Boston: Routledge & Kegan Paul, 1986), 3. 8. Barry, Justice as Impartiality, 31. 9. Buchanan, “Justice as Reciprocity Versus Subject-Centered Justice,” 228-30. 10. The point of what Hume called "rules of justice " (rules of property and contract) was to stabilize possession. Where there is very great inequality of powers, there is no problem of instability for property and contract to solve; the strong simply take what they want; David Hume, The Philosophical Works of David Hume, vol. IV (Endinburgh: Adam Black and William Tait, 1826), 261-62. 4 whether the theory of mutual advantage is a theory of justice at all, since "justice is normally thought of not as ceasing to be relevant in conditions of extreme inequality but, rather, as being especially relevant to such conditions." Nussbaum says that she sees "no reason to think Rawls has diverged from Hume on this point," namely that cooperation must be mutually advantageous as opposed to the alternative of attempted domination, and that since this is not the case for very unequal parties, the severely disabled ought to be excluded from the scope of justice. Eve Kittay concurs that on Rawls's account, the severely disabled should not be owed duties of justice but merely charity. The attribution to Rawls of the mutual advantage view is uncharitable. More interesting is the line of criticism that assumes Rawls is committed to fair rather than self-interested reciprocity. Allan Gibbard criticized Barry for ignoring a possibility that lies between pure egoism and strict impartiality: the pursuit one's conception of the good within the con11. Barry, Theories of Justice, 163. Speaking of Justice as Impartiality, Nussbaum says that Barry's objections are "not precisely the same" as hers but "complementary," and that Barry "does not mention what seems to me fundamental, Rawls's acceptance of Humean Circumstances of Justice as his analogue for the state of nature." However, Barry had devoted a chapter of his 1989 book to Hume, Rawls, and the circumstances of justice. Both in this chapter and the first chapter of his 1995 book, his basic point is Nussbaum's, which is that if principles of justice solve the problem of dividing a cooperative surplus, they ought not apply to those we can expect not to make a net contribution to cooperation. 12. Nussbaum, Frontiers of Justice, 61-62. 13. Eva Feder Kittay, “Human Dependency and Rawlsian Equality,” in Feminists Rethink the Self, ed. Diana Tietjens Meyers (Boulder, CO: Westview Press, 1997), 232. 14. There is an innocent explanation for the presence of the equality condition in Theory. Prior to Rawls’ introduction of the veil of ignorance in Theory, it was the rule of having to make a once-and-for-all decision that was supposed to represent our intuitions about impartiality in the design of institutions. A particular principle might benefit me at the expense of others today, but if tomorrow I might find our positions reversed, I will be inclined to choose a principle that is fair for all. Uncertainty only generates impartiality, however, if our abilities are not too unequal, as Thucydides' Melians found in dealing with the mighty Athenians. In Theory, the parties to the original position still had roughly equal physical and mental powers, but they were now behind a veil of ignorance, denying them knowledge of their strength, intelligence and other abilities. Ignorant of their relative standing, the parties would be unable to capitalize on any superiority they might have, making the assumption of roughly equal capacities superfluous. Not surprisingly, equality of capacities was absent from Rawls’ account of the circumstances of justice in his final statement of ‘justice as fairness.’ John Rawls, and Erin Kelly, Justice as Fairness : A Restatement (Cambridge, Mass.: Harvard University Press, 2001), 84-85. I discuss the changing role of the equality condition in more detail in Andrew Lister, “Hume and Rawls on the Circumstances and Priority of Justice,” History of Political Thought 26 (2005), 683-84. 5 fines of a principled rather than a strategic commitment to treating well those who treat one well. Reciprocity in this sense is a moral duty. According to Gibbard, Rawls applied this idea to society as a whole. All those who have benefited from the rule-compliance of others are themselves motivated to benefit others by complying with social rules. The response from Barry and Buchanan was that even if it involves a principled commitment to reciprocity defined relative to a fair baseline, the theory is still based on returning like for like, and thus "allows for the exclusion of those who cannot provide benefits for others, such as the congenitally handicapped." At least twice Rawls describes principles of justice as rules for determining the proper division of the benefits of cooperation, in a manner that suggests he means the benefits over and above non-cooperation. Yet if the principles of justice are principles for dividing a cooperative surplus, it seems justice as fairness ought not apply to those who are not able to benefit others, even if (according to G.A. Cohen) Rawls was too decent a person to accept this conclusion. 15. "Reciprocity is a moral virtue. We ought to be disposed, as as a matter of general moral obligation, to return good in proportion to the good we receive, and to make reparation for the harm we have done;" Becker, Reciprocity, 3, 130-32. 16. Alan Gibbard, “Constructing Justice,” Philosophy and Public Affairs 20, no. 3 (1991), 266. 17. Barry, Justice as Impartiality, 50. Even if we think of reciprocity as "fair reciprocity" rather than selfinterested reciprocity, only those who have the necessary "strategic capacities" will be owed duties of justice; Buchanan, “Justice as Reciprocity Versus Subject-Centered Justice,” 228-31. 18. "[A]lthough a society is a cooperative venture for mutual advantage, it is typically marked by a conflict as well as by an identity of interests. There is an identity of interests since social cooperation makes possible a better life for all than any would have if each were to live solely by his own efforts. There is a conflict of interests since persons are not indifferent as to how the greater benefits produced by their collaboration are distributed, for in order to pursue their ends they each prefer a larger to a lesser share. A set of principles is required for choosing among the various social arrangements which determine this division of advantages and for underwriting an agreement on the proper distributive shares. These principles are the principles of social justice: they provide a way of assigning rights and duties in the basic institutions of society and they define the appropriate distribution of the benefits and burdens of social cooperation" (4, emphasis added; see also 109-110). It is true, as Jonathan Quong emphasizes, that the parties to the original position make no comparison with a state of nature; Jonathan Quong, “Contractualism, Reciprocity, and Egalitarian Justice,” Politics, Philosophy, Economics 6, no. 1 (2007), 80. However, in this passage Rawls suggests that we, the people using the original position to think about justice, construe the problem principles of justice solve as that of dividing a cooperative surplus. For further discussion of this point, see page 11. 19. Cohen, Self-Ownership, Freedom, and Equality, 224. 6 In response to this criticism, defenders of Rawls have downplayed the importance of reciprocity, in justice as fairness, emphasizing instead Rawls's commitment to equal respect for moral persons, or what Buchanan called "subject-centered justice." On Rawls's account, the scope of our duties of justice is set by possession of the two moral powers (the capacity for a sense of justice and a conception of the good). The parties to the original position may be mutually disinterested, but this thought experiment is just a heuristic device; the motivation of the parties is not the motivation of the citizens using the original position. Rawls thought he could make use of the idea of rational agreement between mutually disinterested parties to clarify our commitment to justice by building plausible constraints on arguments about justice into the conditions under which the parties would select principles. It is these conditions that reflect our respect for persons as equal moral subjects. For Rawls, reciprocity always meant mutual benefit from a fair baseline, an equal division of social primary goods, not mutual benefit relative to a non-agreement point, as in the mutual advantage tradition. The parties cannot make a comparison between how they would fare under a particular principle and how they would fare in a state of nature because they do not know their own attributes; knowledge of personal attributes would allow inequalities of power to influence the choice in the original position, undermining Rawls's aim of "modelling the fundamental equality of persons." Rawls rejected "bargaining contractarianism," insisting that "to each according to his threat advantage is not a conception of justice," because it does not settle conflicting claims based on morally relevant aspects of persons. Whereas "contractarians" 20. According to this view, "moral status depends upon features of the individual herself other than her power to affect others for good or ill;" Buchanan, “Justice as Reciprocity Versus Subject-Centered Justice,” 228-31. 21. Brighouse, “Can Justice as Fairness Accommodate the Disabled?”, 544. 22. Samuel Freeman, “Book Review: Frontiers of Justice,” Texas Law Review 85 (2006), 400-04. 23. Quong, “Contractualism, Reciprocity, and Egalitarian Justice,” 80. 24. Cynthia Stark, “Contractarianism and Cooperation,” Politics, Philosophy & Economics 8 (2009), 77; citing 7 trace the mechanics of self-interested bargaining, "contractualists" such as Rawls articulate the value of equal respect for persons. This Kantian ideal of the person annuls the Humean requirement of equal ability, which plays no necessary role in a theory based on mutual recognition and respect. Although this response is surely correct in highlighting the "subject-centred" aspects of justice as fairness, it does not resolve the tension between the Humean and Kantian elements of the theory. Perhaps we should not attempt to resolve these tensions, but simply abandon the demand for reciprocity. Henry Richardson suggests that the few passages in which the idea of reciprocity seems to serve as an argument for Rawls's principles are "vestigial survivals" of an earlier version of the argument. However, reciprocity played a more important role in the theory than the Rawlsian response to the disability critique allows. 2. The Roles of Reciprocity in Justice as Fairness Henry Sidgwick suggested that the principle of justice, in common sense morality, was that of "gratitude universalized." "[W]e have not only a natural impulse to requite benefits, but also a conviction that such requital is a duty." Universalizing this conviction yields the principle "that good deeds ought to be requited," and generalizing over all of the services upon which society depends, "we get the proposition that men ought to be rewarded A Theory of Justice, p.116 25. Christie Hartley, “Justice for the Disabled: A Contractualist Approach,” Journal of Social Philosophy 40, no. 1 (2009), 18. 26. Stark, “Contractarianism and Cooperation,” 74. 27. Hartley, “Justice for the Disabled,” 21-22. See also Cureton, “A Rawlsian Perspective on Justice for the Disabled.” 28. Henry S. Richardson, “Rawlsian Social-Contract Theory and the Severely Disabled,” The Journal of Ethics 10, no. 4 (2006), 428. 29. Henry Sidgwick, The Methods of Ethics (London: Macmillan and Co., 1874), 253. 8 in proportion to their deserts." Having made the transition from reciprocity to desert, Sidgwick went on to rehearse all of the difficulties involved in applying this idea, in particular whether rewards are to be apportioned according to effort or results, how services of different kinds were to be compared in value, and how to separate "that part of a man's achievement which is due strictly to his free volition from that part which is due to the original gift of nature and to favouring circumstances." Rawls followed Sidgwick's critique of distribution according to desert, but preserved the idea that social justice was a generalization of the duty of reciprocity. Rawls repeatedly described society as a cooperative scheme for "reciprocal advantage" (29), "mutual advantage" (4, 13, 79, 109, 456) or "mutual benefit" (402). He also claimed that the principle of utility was inconsistent with this ideal. Once the principles of justice are thought of as arising from an original agreement in a situation of equality, it is an open question whether the principle of utility would be acknowledged. Offhand it hardly seems likely that persons who view themselves as equals... would agree to a principle which may require lesser life prospects for some simply for the sake of a greater sum of advantages enjoyed by others... Thus it seems that the principle of utility is incompatible with the conception of social cooperation among equals for mutual advantage. It appears to be inconsistent with the idea of reciprocity implicit in the notion of a well-ordered society (13; see also 155). This appeal to reciprocity looks like an intuitive argument for the two principles, analogous to the considerations presented in Chapter 2, prior to the argument from the original position. The attempt to ground distributive justice in reciprocity persists in some recent treatments of global justice, where the demands of egalitarian distributive justice are thought to apply only where there exist the appropriate relationships of cooperation, because only where others have cooperated with us do we owe them a fair return. However, distributive justice cannot be derived from the moral duty to reciprocate benefits, because the duty of reciprocity will 30. Sidgwick, The Methods of Ethics, 253-54. 31. Sidgwick, The Methods of Ethics, 258-59. 32. Rawls, A Theory of Justice, 64, 89, 274. 33. Andrea Sangiovanni, “Global Justice, Reciprocity, and the State,” Philosophy and Public Affairs 35, no. 1 (2007): 3-38. 9 inevitably presuppose an account of justice in specifying what counts as a benefit and when the duty of reciprocity applies. The norm of returning benefits is found in virtually every society, Lawrence Becker points out, but it is articulated in very different ways, in different cultures. What counts as a benefit, and what counts as a fitting and proportionate return? Proportionality is ordinarily "gauged by entitlements," Ekkehart Schlicht argues. "An individual will perceive something he receives as a ‘favour’ only if it surpasses his entitlements. He will then reciprocate by giving more than the other party is entitled to obtain." Conversely, negative reciprocity (retribution) is called for in the case of rule violations, not in the case of just any conduct that makes me worse off than I would have been. The idea that norms of reciprocity are shaped by prior conceptions of right goes back to Aristotle. The Pythagoreans had claimed that simple requital is just, in that "what a man has done to another should be done to him." Yet a literal norm of tit-for-tat cannot be just, Aristotle argued. "[I]f an officer strike a man [in the course of duty], he ought not to be struck in return; and if a man strike an officer, he ought not merely to be struck, but to be punished." We can see how reciprocity presupposes justice in Rawls's own argument. Rawls's 34. Becker, Reciprocity, 73, 81. For a survey of the empirical literature, see pp.36-7, where Becker comments that "the empirical evidence for the existence of a social norm of reciprocity is overwhelming;" see also Ekkehart Schlicht, On Custom in the Economy (Oxford ; New York: Clarendon Press, 1998), 116. In economics, psyhology, and evolutionary biology, the principled commitment to reciprocity has been termed "strong reciprocity;" see, for instance, Ernst Fehr, et al., “Strong Reciprocity, Human Cooperation, and the Enforcement of Social Norms,” Human Nature 13 (2002): 1-25. For a recent overview of the formal theory and experimental evidence of strong reciprocity, see Gerald F. Gaus, The Order of Public Reason: A Theory of Freedom and Morality in a Diverse and Bounded World (Cambridge: Cambridge University Press, 2010), Ch. 7 "The Evolution of Rule-Following Punishers". 35. Becker argues that what is universal is only the "vague form" of reciprocity: "Good is to be returned for good, but not necessarily for evil." "Societies differ markedly in their norms about retaliation, for example... And there are striking differences as well in how various cultures have worked out matters of proportionality... voluntariness... and the type and directness of the required return" Becker, Reciprocity, 81. 36. Schlicht, On Custom in the Economy, 116. 37. Aristotle, The Nicomachean Ethics, trans. F. H. Peters, 5th ed. (London: Kegan Paul, Trench, Trubner & Co., 1893), 152, Book V Chapter 5. Thanks to Professor Schlicht for this reference. 10 invocation of reciprocity involves a quite specific account of the fair baseline: an equal division of socially produced goods. If people are asking themselves why they should agree to lesser prospects than others simply for the sake of maximizing aggregate benefits, they already accept that inequalities must benefit all. The difference principle is "a principle of reciprocity or mutuality" (xv) because everyone is benefited by the difference principle, in that "each man's position is improved with respect to the initial arrangement of equality" (69, emphasis added; see also 88-9). Why should we define "benefit" relative to an equal split of the social product, instead of relative to shares of resources in a state of nature, without any cooperation, or relative to what Robert Nozick called "partitioned cooperation?" Talented individuals may benefit from general cooperation as opposed to none, but not from general cooperation as opposed to less extensive cooperation, in which "better endowed" individuals cooperate only amongst themselves. The criterion of reciprocal benefit is arbitrarily weak if it simply requires that an individual be better off in society than if he withdraws alone, G. A. Cohen explains, but if the criterion requires benefit relative to withdrawal in coalitions, it will no longer justify the difference principle. The argument for Rawls's two principles from the idea of reciprocity presupposes a particular construal of the fair baseline; it is not an argument for construing the baseline in this way. Rawls's principles cannot be grounded in a duty of reciprocity, because the specification of any such duty depends on prior standards of justice. 38. Moreover, the principle of mutual benefit relative to this baseline must be applied incrementally, in Rawls's view, to the choice between more inequality and less, not just to the choice between some inequality and none, i.e. everyone must be better off or at least no worse off than under any more equal division, not just better off than under a perfectly equal division (89). 39. Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 193-94. 40. "The root of the difficulty is that Rawls' contractualist characterization of the problem of justice is inconsistent with the Rawlsian principles of justice (even when we set aside the problem of wholly infirm people). Rawls avoids the profoundly inegalitarian conclusions a consistent contractualist would reach only by being untrue to his own contractual starting point;" Cohen, Self-Ownership, Freedom, and Equality, 225-26. Here, Cohen is using the term "contractual" to mean what others would call contractarian. 11 Reciprocity plays a second and more substantial role in Theory, however, which is that of supporting the argument for the stability of justice as fairness. A conception of justice is stable when public recognition of its realization by the social system tends to bring about the corresponding sense of justice in citizens, with the result that if any deviations from justice occur, there is a tendency for justice to reassert itself (397-401). Rawls's argument for the stability of justice as fairness relies on the empirical premise that people are motivated by backward-looking, non-strategic reciprocity. It is "a deep psychological fact"(433), presumably rooted in natural selection (440), that we have a tendency to "answer in kind," loving those that love us, and showing friendship to those who have shown us friendship (433). Our sense of justice extends this basic human trait; we want to comply with the rules of institutions when we have benefited from others likewise complying with these rules (414-19, 429, 433). Rawls claimed that this deeply ingrained disposition can be harnessed as a stabilizing force more easily by justice as fairness than by utilitarianism. If social inequalities benefit all, then people will tend to see others' compliance with society's rules as benefits that ought to be reciprocated, whereas under utilitarianism, some people will not see any benefit to themselves in other people following the rules. As an empirical claim about human psychology, this argument about the greater stability of justice is not question-begging, as was the attempt to ground justice as fairness in the moral duty of reciprocity. However, its force is limited. Utilitarianism has a problem with stability because it concedes that everyone counts for one and none for more than one, yet maintains that slavery would be just if it maximized 41. "[T]he active sentiments of love and friendship, and even the sense of justice, arise from the manifest intention of other persons to act for our good. Because we recognize that they wish us well, we care for their well-being in return. Thus we acquire attachments to persons and institutions according to how we perceive our good to be affected by them. The basic idea is one of reciprocity, a tendency to answer in kind. Now this tendency is a deep psychological fact. Without it our nature would be very different and fruitful social cooperation fragile if not impossible... A capacity for a sense of justice built up by responses in kind would appear to be a condition of human sociability. The most stable conceptions of justice are presumably those for which the corresponding sense of justice is most firmly based on these tendencies (§76)" (433). 12 utility, even though slaves might be worse off in society than in a state of nature. This makes it hard to reconcile utilitarianism with any principle that social institutions ought to benefit all. The argument gives us no reason for thinking that feudal or caste societies were unstable, however. If people assume that those born into different castes have different moral importance, the merely decent behaviour of a superior may seem like a benefit that must be reciprocated, and social institutions may seem to benefit all in the proportion and manner they are due. The third and most important role reciprocity plays in Rawls's theory is that of a limit on duties of justice. The flip side of the commitment to returning benefits is an unwillingness to make unilateral sacrifices. "[E]ven with a sense of justice men's compliance with a cooperative venture is predicated on the belief that others will do their part" (296); "each person's willingness to contribute is contingent upon the contribution of the others" (238). People may be willing to comply with just rules even when they think they could get away with breaking them, but they will not be willing to comply with such rules unilaterally, if others are not also complying. For a conception of justice to be feasible, then, people will need assurance about the conduct of others. The argument for the stability of justice as fairness helps provide such assurance. So too does the little appreciated argument for the "congruence" of justice and the good. Rawls argued that in a just society, being a good person (i.e. having a sense of justice) would be a good thing for that person (496-505). Brian Barry ac42. Samuel Freeman, “Congruence and the Good of Justice,” in The Cambridge Companion to Rawls (Cambridge: Cambridge University Press, 2003), 283. 43. Rawls aimed to show only that justice converges with the individual's good in a society well-ordered by justice as fairness, and only for justice-compatible conceptions of the good; Freeman, “Congruence and the Good of Justice,” 285. The limited scope of Rawls' thesis makes the case for congruence more plausible, but does not resolve the question of why congruence should be necessary. Rawls's arguments for congruence were the following. First, being liable to moral feelings such as guilt and indignation is essential to participating in relationships of trust and friendship (425-28). Second, compliance with just institutions involves the exercise of valuable skills and capacities, because it is a rational and social activity. Rawls reaches this conclusion by applying his "Aristotelian principle" to social institutions. People enjoy exercising their capacities, and this enjoyment increases with the skill involved and the complexity of the activity in question (374-5). Complying with the rules of just institutions engages the cognitive ability to recognize rules as just and conducive to the 13 cused Rawls of accepting "the ancient doctrine that no act can be rational unless it is good for the agent to perform it." Rawls did try to show that that there are reasons of self-interest for wanting to have a sense of justice, in a just society, but his purpose was to assure each other that our predictable weaknesses of the will are not so extensive as to push levels of compliance below the level at which our duties of justice would no longer be binding. The same logic applies to Rawls's later notion of "overlapping consensus." Even if each citizen finds reasons within his or her own doctrine to affirm political liberalism and justice as fairness, people may still face difficult conflicts of values. Although every reasonable comprehensive doctrine supports justice as fairness given the fact of reasonable pluralism (suppose), doctrines that are not philosophically liberal would support quite different principles were reasonable pluralism not a fact. If, for many doctrines, the gap between political justice and their comprehensive ideal of justice is too large, then predictable failures of rationality on the part of otherwise reasonable citizens might drive levels of compliance below the reciprocity threshold. On this account, achieving overlapping consensus means achieving a sufficiently high level of congruence between political and comprehensive justice, on the part of reasonable doctrines, such that predictable failures of reasonableness on the part of otherwise reasonable citizens do not undermine citizens' assurance that others can generally be trusted to comply with just rules. The assumption that our sense of justice is conditional on a reasonable assurance of compliance by others helps explain Rawls's focus on "ideal theory", where we assume quasicommon good, and to comply with them for that reason. Thirdly, the Kantian interpretation of justice as fairness maintains that acting from principles that would be chosen in the original position expresses our nature as free and equal rational beings. Acting in accordance with justice thus allows us to affirm our dignity, as creatures who are not simply moved by causal forces, but can instead act as rational agents (503). 44. Barry, “John Rawls and the Search for Stability,” 884-85; cited by Freeman, “Congruence and the Good of Justice,” 281. 45. Rawls, Political Liberalism, 133-72. 14 universal compliance with the rules of a "well-ordered society" (8). This aspect of Rawls's work has been much criticized of late. Yet Rawls's focus on ideal theory was motivated in part by a realistic assumption about motivation. Even under the best conditions human beings will not generally be willing to comply with just rules if others are not likewise willing. Our theory of justice must therefore be such that most people would be motivated to act upon it, if a just society were to be successfully established. The real question is whether this assumption is too realistic, and whether true justice does not place greater demands on us than Rawls admitted. The worry that his view is too realistic can be brought out by considering the implicit limits Rawls placed on our natural duties: the duty not to injure people, harm the innocent, or act cruelly (94, 98), the duties of mutual aid or beneficence, and the duty of mutual respect (94-5, 98), and the duty of justice (99-100). These duties apply to all as equal moral persons "irrespective of their institutional relationships" (99). They are "principles for individuals" rather than "principles for institutions (93)" in that they apply outside or beyond institutions, where institutions do not exist, or within such institutions, to individual conduct permitted but not required by institutional rules. Principles that apply to individuals outside of or within 46. Charles W. Mills, “’Ideal Theory’ as Ideology,” Hypatia 20 (2005): 165-84; Amartya Sen, “What Do We Want From a Theory of Justice?,” The Journal of Philosophy 103 (2006): 215-38; Colin Farrelly, “Justice in Ideal Theory: A Refutation,” Political Studies 55 (2007): 844-64. For responses to this critique, see Zofia Stemplowska, “What’s Ideal About Ideal Theory?,” Social Theory and Practice 34, no. 3 (2008): 319-40; Adam Swift, “The Value of Philosophy in Nonideal Circumstances,” Social Theory and Practice 34, no. 3 (2008): 363-87; Ingrid Robeyns, “Ideal Theory in Theory and Practice,” Social Theory and Practice 34, no. 3 (2008): 341-62; A. John Simmons, “Ideal and Nonideal Theory,” Philosophy and Public Affairs 38, no. 1 (2010): 5-36. 47. I leave to one side what Rawls calls the principle of fairness, which gives rise to obligations when we voluntarily undertake roles or accept benefits within the context of a just institutional scheme (93-8, 301-8). This principle binds individuals to comply with the subset of institutional rules that apply only to those who engage in the voluntary conduct in question (e.g. running for office). Rawls concedes that the natural duty of justice is more fundamental, and that one might view the obligations that follow from the principle of fairness as a consequence of the natural duty of justice extended by voluntary actions. Rawls says that this one-principle view is "sound enough," but that "it seems appropriate to distinguish between those institutions or aspects thereof which must inevitably apply to us since we are born into them and they regulate the full scope of our activity, and those that apply to us because we have freely done certain things as a rational way of advancing our ends" (302). 48. Of course, some principles for individuals are also institutionally enforced; the natural duty not to injure 15 institutional rules will lack official public formulation and specification, as well as collective identification and sanction for those who fail to honour them. Our assurance of compliance on the part of others will be correspondingly lesser than in the case of institutional principles. Principles for individuals are nonetheless binding. In other words, principles that apply to individuals only by way of institutions are conditional on a reasonable expectation of reciprocity, whereas principles for individuals directly hold even where compliance on the part of others cannot be assured. Outside of an institutional scheme, the duty of justice includes no duty to share wealth or resources fairly, but only the duty "to further just arrangements not yet established, at least when this can be done without too much cost to ourselves" (99; see also 257-8). We only have a duty to help establish just institutions, not a duty to share fairly on an individual basis, and this duty to help establish just institutions cannot be too demanding, whereas no similar limit applies to the duty to comply with just institutions. The absence of personal, extra-institutional duties of distributive justice can seem surprising. Being born in a rich country is morally arbitrary in just the same way that being born talented is morally arbitrary. Why don't I have a duty of justice to help eliminate this inequality, even where no institutions exist? Why do we only have a limited duty to help establish just institutions, and not a similarly limited duty to help eliminate arbitrary inequality? The fact that we can't be morally required to devote all our energies to fighting inequality does not explain why we should not be required to do anything, apart from helping to establish just institutions. Recognizing the full importance of reciprocity to justice as fairness may simply reinforce the urgency of abandoning this aspect of the theory, but I want to explore an alternate possibility, which is to reinterpret reciprocity in a manner that is congenial to a subjectothers is made into laws against murder and assault. It remains in force beyond the confines of just institutions, however, and is thus not purely institutional even when institutionalized. 16 centred view, by emphasizing that such a conception can have a relational character. 3. Reciprocity and Relationships When and why are duties conditional on reciprocity? Hobbes famously argued that even the duty not to kill is conditional on the state's guarantee that others will not kill me. Without an effective sovereign, I have a right of nature to do whatever I think conducive to my own survival. Most people, I think, would accept that some of our most basic duties are not conditional on reciprocity. Although I am permitted to defend myself from attack, the fact that you are a known murderer who would murder me if given the chance does not give me the right to murder you if you don't have the chance, for the sake of some other personal benefit. Similarly, even if it one thinks that it is permissible to torture someone in order to avert a moral disaster, it is not permissible to torture a known torturer for just any minor benefit, simply because the torturer would torture us, were our situations reversed. We might also think the duty to aid is not conditional on reciprocity. If I can save you from starving or drowning at little cost to myself I must do so even if you would not do the same for me (although presumably not if you would threaten me, if you survived). Other duties are more plausibly construed as being conditional on reciprocity. For example, the duty to share fairly with you in the benefits of cooperation might apply only so long as I can reasonably expect you to share fairly with me. To determine what differentiates these cases, we need to distinguish different reasons it might matter whether others can be expected to comply with a given duty. The most straightforward reason is that it can affect the demandingness of the duty. Hobbes' point was that in the state of nature, any unilateral performance of duty threatens self-preservation. 49. Leviathan, Chapter 14, paragraph 1; Thomas Hobbes, The English Works of Thomas Hobbes, vol. III (London: John Bohn, 1839), 116. 17 Hume's example of a "society of ruffians" can be understood in the same way. In this case, however, the appropriate response is simply to concede that we only have extra-institutional duties of distributive justice up to some limit of demandingness or burdensomeness. The fact that others are not complying with a particular duty is merely one of many contingent factors that may affect its burdensomeness. I might have a demanding duty to share fairly, but only with those willing to share fairly with me, but I might also have an undemanding duty of aid even those who would not aid me. The real work, in this scenario, is being done by a criterion of demandingness, not reciprocity itself. A second reason that it might matter whether or not I can expect others to comply with a duty is that the effects of my action may depend on whether others will follow suit. In some cases, one-sided compliance may be morally futile, rather than too demanding. According to Hume, for example, the purpose of rules of property and contract is to make everyone better off, by facilitating investment, specialization, and so on. In the society of ruffians, my respect for rules of property in no way promotes this end, because it doesn't elicit compliance from others. Similarly, one might argue that unilateral action to reduce economic inequality would be futile. Saladin Meckled-Garcia raises this possibility in relation to global justice; individual action outside of a coherent institutional scheme may be pointless, because our various efforts may work at cross-purposes. Even if we both believe that we have a 50. "Suppose likewise, that it should be a virtuous man’s fate to fall into the society of ruffians, remote from the protection of laws and government; what conduct must he embrace in that melancholy situation? He sees such a desperate rapaciousness prevail; such a disregard to equity, such contempt of order, such stupid blindness to future consequences, as must immediately have the most tragical conclusion, and must terminate in destruction to the greater number, and in a total dissolution of society to the rest. He, meanwhile, can have no other expedient than to arm himself, to whomever the sword he seizes, or the buckler, may belong: To make provision of all means of defence and security: And his particular regard to justice being no longer of use to his own safety or that of others, he must consult the dictates of self-preservation alone, without concern for those who no longer merit his care and attention" Hume, The Philosophical Works of David Hume, 257-58. 51. Saladin Meckled-Garcia, “On the Very Idea of Cosmopolitan Justice: Constructivism and International Agency,” Journal of Political Philosophy 16, no. 3 (2008), 255-57. Similar arguments have been put forward by Thomas Nagel; Thomas Nagel, “The Problem of Global Justice,” Philosophy and Public Affairs 33, no. 2 (2005), 116. 18 duty to help eliminate morally arbitrary inequality, radical disagreement about what counts as arbitrary could make our efforts futile. If I am distributing to tall people and you to short people, the net effect of our actions could be nil. With the world as it is today, however, and assuming more plausible conceptions of distributive justice, this possibility seems remote. It hardly seems likely that for every dollar I contribute to Oxfam, say, someone in a poor country is sending a dollar to Canada, wrongly believing that it is Canadians who are suffering the effects of unmerited inequality. According to standard, luck egalitarian conceptions of justice, one-sided sharing is not pointless. Whether others are committed to sharing fairly or not, my doing so can reduce arbitrary inequality. So long as my duties of distributive justice are not too onerous, whether others are complying with theirs should not be relevant, on this account. In the example above, the outcomes of one action depended on whether others would follow suit. It is also possible, however, that actions can have a constitutive rather than a causal effect. This is the situation when the mutual performance of a duty constitutes a valuable relationship. Suppose, for example, that one accepts a Rawlsian account of distributive justice, according to which it is unjust is to exploit one's good luck to claim a greater share of the products of our cooperative labour (where this inequality of shares does not benefit others). I want to suggest that our duty not to exploit natural or social good luck to extract a greater share of the social product (unless such inequality benefits all) is grounded in the importance of having a relationship between persons who recognize each other as equals. The idea is that we can distinguish two ways that duties can flow from someone else's moral status. Some duties follow from someone's status as a person, a moral agent, and do not lapse just because the other person has behaved or would behave badly – this is the case of the du52. As A. J. Julius puts it, "surely isolated redistribution is not self-defeating in a world like our own;" A. J. Julius, “Basic Structure and the Value of Equality,” Philosophy and Public Affairs 31, no. 4 (2003), 345. 19 ties not to murder or torture. Other duties are constitutive of the relationship appropriate to equals, but are only binding conditional on mutual fulfillment, or a reasonable expectation of such. If the point of my satisfying my duties of distributive justice is to establish a relationship, then the existence of a reciprocity condition makes sense. One-sided fair-sharing may reduce the level of morally arbitrary inequalities, but it does not establish a relationship of mutual recognition as equals, not if it does not elicit the appropriate response, because by definition such a relationship must be two-sided. The explanation for some duties being conditional on reciprocity might be that in some cases, the point of the duty is to constitute a valuable relationship, whereas in other cases there is something at stake apart from the relationship. It is significant, in this respect, that the duties that seem not to be conditional on reciprocity are also not comparative. In each of the unconditional cases, there is some specific thing I must do or must not do, in virtue of your situation described in absolute terms: I must not kill you, I must not torture you, or I must save you. In each of these cases, there is something of intrinsic importance at stake apart from our relationship: a human life, or intense fear and suffering, and for this reason, it is better that one person comply with the duty than that none do. Arbitrary inequalities by themselves do not involve these kinds of concerns, since both parties may be well off, and since they may not be the result of anyone's actions. The concern with inequalities per se arises because of what they may say about our relationship. If we divide the benefits of social cooperation unequally in ways that don't raise the lower position, the inequality involves a failure to recognize others as equals engaged in a cooperative endeavour. The problem posed by the circumstances of justice is not simply to stabilize mutually 53. Christian Schemmel, “Why Relational Egalitarians Should Care About Distributions,” Social Theory and Practice forthcoming (2011). 20 beneficial cooperation, but to establish a relationship of mutual recognition as equals in a context of conflicts of interests and ideas. This interpretation of the circumstances and point of justice incorporates the Kantian or subject-centred commitment to respect for persons as equal moral agents, but it does so by focusing on the goal of establishing a relationship. Respect for persons is the fundamental ground for all our moral duties, presumably, but certain duties are conditional on reciprocity because their point is to constitute a relationship of mutual recognition. Because the view is relational, as well as subject-centred, it will naturally incorporate a reciprocity condition, not out of our separable individual interests in stabilizing cooperation, but out of our shared moral interest in standing in a particular kind of relation to one another. Here, the reciprocity condition is not seen as a feasibility limit imposed by human psychology, but as an intrinsic feature of at least some duties of justice, which are relational in the sense that the point of fulfilling the duties is to establish or constitute a relationship. Duties associated with distributive justice are not grounded in the duty to reciprocate benefits, but limited by a reciprocity condition because they are relational in this way. 4. Applications: Global Justice and Disability I have been arguing that the reciprocity condition is an intrinsic feature of certain duties of justice – specifically, egalitarian duties of distributive justice – because the point of these duties is to establish a relationship of mutual recognition as equals. By way of conclusion, I want to sketch the implications that this view has for the issues of global justice and disability. 4.1. Global Justice Why is our natural duty of justice only to contribute to the establishment of just institutions, and not to redistribute our wealth, up to whatever is considered a reasonable individual or national prerogative for the pursuit of self-interest? One answer is that inequality 21 per se at the global level simply doesn't matter, not intrinsically. This conclusion follows from the thesis that egalitarian principles of distributive justice apply only where there exist specific social practices or institutions (e.g. shared citizenship in a political society), and the empirical claim that the institutions in question do not exist at the global level. In contrast, I assume that global inequality does matter. Any time we interact with others, for example by trading with them, we ought to do so on fair terms, so that our relationship can be one of mutual recognition as equals. Distributive justice is relational in the sense that it is concerned with relationships between persons, but it is not only a concern within the boundaries of a state or where networks of trade exist. Whenever we are in sufficiently close causal contact, we are interacting with one another, even if only by refraining from venturing onto to each other's land or taking each other's possessions. Such minimal interaction is sufficient, I take it, for principles of justice to apply. 54. For example, Joseph Heath argues that "Rawls does not view inequality as problematic in the international context."Joseph Heath, “Rawls on Global Distributive Justice: A Defence,” in Global Justice, Global Institutions, ed. Daniel M. Weinstock, Canadian Journal of Philosophy, Supplementary Volume (Calgary: University of Calgary Press, 2005), 195. The lack of duties of distributive justice between states is "the intuition at the core of Rawls's view,"Heath, “Rawls on Global Distributive Justice: A Defence,” 199. Global economic inequality can of course matter for instrumental reasons, as for example if it generates large power imbalances, allowing one country to dominate others. In this case, however, our concern for inequality would be "derivative" rather than "foundational;" inequality would matter because of its effects, "not because it is intrinsically unjust" David Miller, “Against Global Egalitarianism,” The Journal of Ethics 9, no. 1/2 (2005), 56. 55. Different versions of this thesis are found in Michael Blake, “Distributive Justice, State Coercion, and Autonomy,” Philosophy and Public Affairs 30, no. 3 (2002): 257-96; Nagel, “The Problem of Global Justice.”; Sangiovanni, “Global Justice, Reciprocity, and the State.” Others accept the practice-dependence thesis but argue for a much broader account of the relevant practices; see for example Aaron James, “Constructing Justice for Existing Practice: Rawls and the Status Quo,” Philosophy and Public Affairs 33, no. 3 (2005): 281-316; A. J. Julius, “Nagel’s Atlas,” Philosophy and Public Affairs 34, no. 2 (2006): 176-92; Miriam Ronzoni, “The Global Order: A Case of Background Injustice? a Practice-Dependent Account,” Philosophy and Public Affairs 37, no. 3 (2009): 229-56. 56. For the distinction between these two senses of the term "relational," one focused on the scope and the other on the content of principles of justice, compare the terminology of Sangiovanni, “Global Justice, Reciprocity, and the State.” with that of Elizabeth Anderson, “What is the Point of Equality?,” Ethics 109, no. 2 (1999): 287-337; Samuel Scheffler, “What is Egalitarianism?,” Philosophy and Public Affairs 31, no. 1 (2003): 5-39; Christian Schemmel, “Distributive and Relational Equality,” Politics, Philosophy & Economics (forthcoming). 57. Jonathan Quong argues for a similarly broad account of the interaction that triggers application of principles of distributive justice; Jonathan Quong, “Left-Libertarianism: Rawlsian Not Luck Egalitarian,” Journal of Political Philosophy 19, no. 1 (2011): 64-89.. The essential point, I think, is that although the intensity or level of cooperation may vary across different contexts, the benefits and burdens of all forms of cooperation ought, in principle to be distributed fairly – subject, I would argue, to the reciprocity or assurance condition discussed 22 However, because the point of distributive justice is to constitute a relationship of mutual respect as equals, the duties associated with distributive justice are conditional on a reasonable assurance of reciprocity. There is no unilateral duty to share fairly in the benefits of cooperation, only a unilateral duty to try to establish institutions that will permit fair sharing with a reasonable assurance of reciprocity. Our duty with respect to global distributive justice is thus to help create (and in particular not to thwart) the development of the more robust global institutions that would afford such assurance. Institutional structures sufficient to ensure reciprocity are required for the full realization of justice; they are not a condition for principles of justice to apply. Duties of distributive justice are aspirationally global in scope, but, without a reasonable assurance of compliance on the part of others, the full duty to share fairly in the benefits of cooperation does not hold. The duty to help establish just institutions is not itself conditional on reciprocity – I have to try to create those institutions even if others are not trying – but it is limited in the extent of the demands it imposes on us. Assurance of reciprocity is not simply a feasibility constraint, as Arash Abizadeh suggests. On a relational account, the problem is not that a more equal distribution of wealth is impossible without adequate institutions, but that without a reasonable assurance of reciprocity duties of egalitarian distributive justice (as distinct from duties of aid and reparative justice) lack a point. 4.2. Disability If reciprocity matters because justice is in part a relational value, then there is an obvious difference between someone unwilling and someone unable to contribute to social production. There is no misrecognition or disrespect in a severely disabled person's failure to make a net contribution. Because reciprocity follows from the goal of securing a relationabove. 58. Arash Abizadeh, “Cooperation, Pervasive Impact, and Coercion: On the Scope (Not Site) of Distributive Justice,” Philosophy and Public Affairs 35, no. 4 (2008), 320-21. 23 ship of mutual recognition, our duties of justice are limited by the need for a reasonable assurance that others will reciprocate if they are able, not by the demand that everyone owed duties of justice in fact make a net contribution to social cooperation. This argument reconciles the reciprocity condition with the claim that duties of justice are owed to all those capable of a sense of justice. It leaves open the question of what specifically the disabled are owed, beyond what they are entitled to under Rawls's existing principles of justice. Amartya Sen, Will Kymlicka, and Martha Nussbaum have argued that even if they do come under the scope of justice, disabled persons are not fairly treated by Rawls's principles because his metric of advantage is given by an index of social primary goods, such as rights, income and wealth, excluding people's different shares of natural primary goods, such as the ability to see or walk. Two people count as equally well off, according to this standard, if they have the same bundle of social primary goods, even if one is disabled, and so faces a range of extra costs for medication or equipment. On the other hand, if we simply modify the difference principle so as to maximize the position of the worst off person, taking into account individual disability, the difference principle would become far too demanding, potentially absorbing all social resources. There are a number of possible responses to this dilemma. Perhaps, as Thomas Pogge argues, disability is often caused by injustice that can be diagnosed as such on Rawlsian principles, or caused to be disadvantageous by injustice. However, appeal to the social causa59. Amartya Sen, “Equality of What?,” Tanner Lectures on Human Values (1980): 195-220; Will Kymlicka, Contemporary Political Philosophy : An Introduction (Oxford: Clarendon Press, 1990), 70-71; Nussbaum, Frontiers of Justice, 164-65. 60. Kenneth J. Arrow, “Some Ordinalist-Utilitarian Notes on Rawls’s Theory of Justice,” Journal of Philosophy 70 (1973): 245-63; Brighouse, “Can Justice as Fairness Accommodate the Disabled?”, 546; Thomas Pogge, “Can the Capability Approach be Justified?,” Philosophical Topics 30, no. 2 (2002), 195-96; Stark, “How to Include the Severely Disabled in a Contractarian Theory of Justice,” 139, note 31. 61. Pogge, “Can the Capability Approach be Justified?”, 186. For other discussions of the social construction of disability, see Anita Silvers, “Reconciling Equality to Difference: Caring (F)Or Justice for People With Disabilities,” Hypatia 10): 30-55; cited byAnderson, “What is the Point of Equality?”, 334; Nussbaum, Frontiers of Justice, 165; Elizabeth Anderson, “Justifying the Capabilities Approach to Justice,” in Measuring 24 tion and construction of disability will not solve the problem of what we owe as a matter of justice to people with multiple and severe disabilities. A relational perspective suggests another possible response. The goal of constituting relationships of mutual recognition is not satisfied merely by dividing the social product equally across contributing social positions (unless an unequal distribution benefits all). For there will be individuals who due to their disabilities will be vulnerable to domination, if they are not provided with additional resources. Elizabeth Anderson has argued, along these lines, that "citizens have a claim to a capability set sufficient to enable them to participate as equals in society (assuming they have the potential to do so)." She characterizes this position as a "sufficientarian standard of justice in the metric of capabilities," where capabilities are defined in terms of the requirements of effective citizenship rather than human flourishing generally. This position seems intuitively plausible, and it goes a long way towards addressing the concern that principles of distributive justice when extended to cover disabilities might absorb all social resources. It is plausible that we have duties of justice to help those whom we respect as moral equals secure an adequate level of capability to function as citizens, so that our relations will not be founded on domination. Such assistance might also be required by a duty of aid, but it is required as a matter of justice because without the ability to participate in social life, people with disabilities will continually face situations of asymmetrical personal dependence. For the disabled to avoid being subject to domination, and for those who aren't disabled to avoid becoming parties to such domination, society will have to devote resources to securing the Justice: Primary Goods and Capabilities, ed. Harry Brighouse, and Ingrid Robeyns (Cambridge: Cambridge University Press, 2010), 96. 62. Nussbaum, Frontiers of Justice, 117, 128; Anderson, “Justifying the Capabilities Approach to Justice,” 92-93; Linda Barclay, “Disability, Respect and Justice,” Journal of Applied Philosophy 27, no. 2 (2010), 159. 63. Anderson, “Justifying the Capabilities Approach to Justice,” 83. 64. Anderson, “Justifying the Capabilities Approach to Justice,” 83. See p.88 for Anderson's discussion of the political nature of her capabilities approach. 25 necessary conditions for an adequate measure of independence. One question I have not addressed is whether making duties of justice conditional on the capability to have a sense of justice wrongly excludes animals and the severely mentally disabled from the scope of justice. Rawls said that possession of the two moral powers was merely sufficient for being owed duties of justice, not necessary, but his reasons for doing so were pragmatic rather than principled (442-3). Social contract theorists need not be too strict on what the threshold is for being said to possess the two moral powers. I think that being aware of oneself as a person with desires of one's own and recognizing others as the source of independent claims is sufficient. However, there is no way around the conclusion that on this relational account, beings utterly incapable entering into relationships of mutual recognition and respect will not be owed justice. Nor do I think that this is a flaw in the view. Duties of justice are not the only duties we have. Not causing pointless suffering is a duty of humanity, not of justice. The danger of broadening the category of duties of justice too much is that we will also water down these duties. While the relational account of justice may draw the circle of those owed justice too narrowly, in some people's eyes, it also supports demandingly egalitarian duties of justice, within the scope of those it covers – at least if a reasonable assurance of reciprocity is present. In contrast, Nussbaum explicitly recognizes that integrating disability into Rawls's theory might lead us to abandon the difference principle, in favour of an "ample social minimum." I think that would be a mistake. 65. Nussbaum describes a number of cases of persons with severe mental disabilities "who do not conform to the rather idealized picture of moral rationality" assumed in Rawls's Kantian conception of the person, and so are "not regarded as capable of reciprocity of the requisite sort." She points out that people who lack the capacity to form an explicit plan of life or recognize abstract principles of justice may nonetheless be capable of "complex forms of reciprocity;" Nussbaum, Frontiers of Justice, 134-35. 66. Nussbaum, Frontiers of Justice, 115, 146. 26 5. Conclusion Reciprocity does not ground but limits some duties of justice – specifically, our egalitarian duties of distributive justice. Some duties are not conditional on reciprocity, or only conditional on reciprocity to extent that the non-compliance of others affects the demandingness of the duty, whereas other duties are conditional on reciprocity itself, because their point is to constitute a relationship in which persons recognize each other as equals. Rawls's conception of distributive justice straddles this division. We have an unconditional duty to try to establish just institutions, which allow us to share fairly in the burdens and benefits of cooperation, but where such institutions are not in place we do not have a unilateral duty to share fairly (as distinct from our duty of aid, which is not conditional on reciprocity but also not egalitarian). Because the reciprocity condition is rooted in the relational character of distributive justice, non-contribution on the part of those unable to contribute does not mean that they fall outside the scope of distributive justice. 27