! "#$% %&$%# Citation: 19 Can. J. L. & Jurisprudence 319 2006 Content downloaded/printed from HeinOnline (http://heinonline.org) Fri Mar 16 12:16:56 2012 -- Your use of this HeinOnline PDF indicates your acceptance of HeinOnline's Terms and Conditions of the license agreement available at http://heinonline.org/HOL/License -- The search text of this PDF is generated from uncorrected OCR text. -- To obtain permission to use this article beyond the scope of your HeinOnline license, please use: https://www.copyright.com/ccc/basicSearch.do? &operation=go&searchType=0 &lastSearch=simple&all=on&titleOrStdNo=0841-8209 The Boundary of Justice and the Justice of Boundaries: Defending Global Egalitarianism Kok-Chor Tan I. Introduction One powerful and common strategy in defense of global egalitarianism is to show how and why the accepted arguments for distributive equality within a single state apply globally as well, that is, why the reasons for caring about inequalities between fellow citizens are reasons also for caring about inequality between persons in the world as a whole. For instance, taking (what they take to be) Rawls's central argu- ment for domestic distributive justice, namely, that individuals should not be dis- advantaged on account of arbitrary factors such as their social class, natural endowment, and misfortune, global egalitarians such as Charles Beitz and Thomas Pogge have argued that neither should individuals be at a disadvantage due to an accidental fact like their country of birth. So if distributive justice is motivated by the need to mitigate the effects of contingencies that are "arbitrary from a moral point of view"2 on people's life chances, this presents a consideration also for global distributive equality. The case for global egalitarianism, however, has been challenged by those whom we may call anti-global egalitarians. These critics (at least those I will be focusing on) are not anti-egalitarians through and through. They are egalitarians in that they accept that distributive equality matters domestically. What they reject is the claim that distributive equality has a global scope or universal applicability. Such critics include Walzer, D. Miller, R. Miller, Nagel, Freeman, and Rawls.3 And An earlier version of this paper was presented at The Department of Philosophy, The University of Pennsylvania and I am grateful to members of the audience for their good questions and criticisms. Thanks particularly to Karen Detlefsen and Susan Meyer for their helpful questions, and to Greg Hall, Matt Lister and David Reidy for their written critical comments. Most of all, thanks to Samuel Freeman for the numerous insightful and critical discussions on the subject, and for sharing his own forthcoming papers on this topic with me. 1. Charles R. Beitz, Political Theory and International Relations (Princeton, NJ: Princeton University Press, 1979) at Part Ill; Thomas W Pogge, Realizing Rawls (Ithaca, NY: Cornell University Press, 1989) at Part III. See also Allen Buchanan, Justice, Legitimacy, and Self- Determination (Oxford: Oxford University Press, 2004) at ch. 4; Simon Caney, Justice Beyond Borders (Oxford: Oxford University Press, 2005) and Darrel Moellendorf, Cosmopolitan Justice (Boulder, CO: Westview Press, 2002). In contrast, in his Global Justice: defending cosmopoli- tanism (Oxford: Oxford University Press, 1999), Charles Jones is concerned not so much with global egalitarianism as with "the specific problem of poverty" (5). Similarly Thomas W Pogge's more recent World Poverty and Human Rights (Cambridge: Polity, 2002) is primarily concerned with the problem of global poverty. 2. John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971) at 15. 3. Richard W. Miller, "Cosmopolitan Respect and Patriotic Concern," (1998) 27/3 Phil. & Pub. Affairs 202; David Miller, "Against Global Egalitarianism" (2005) 9/1-2 J. Ethics 55; Michael Walzer, "Response" in David Miller & Michael Walzer, eds., Pluralism, Justice, and Equality (Oxford University Press, 1995); Thomas Nagel, "The Problem of Global Justice" (2005) 33/2 Canadian Journal of Law and Jurisprudence Vol. XIX, No.2 (July 2006) it is important to note that anti-global egalitarians need not reject the duty of human- itarian assistance or the duty of rescue. In fact the anti-global egalitarians that I will refer to accept that there are duties of assistance towards persons who are unable to meet basic needs. They will denounce the current global economic arrangement as unacceptable because many people continue to be deprived of basic and urgent needs under this arrangement. Their position, however, is non-egalitarian in that while it holds that there is a moral duty to assist individuals or peoples to help them achieve minimally decent standards of living, economic inequalities between coun- tries as such are not matters of justice. To be sure, one could plausibly make the case that if inequalities between countries become excessive, those at the bottom end will be in effect deprived of basic needs given the operations of the marketplace. In this case, anti-global egalitarians will be concerned about the gross inequalities, not because these inequalities themselves are objectionable from the point of view ofjustice, but rather because they have the effect of impoverishing persons, abso- lutely as opposed to just making them relatively worse-off. So anti-global egali- tarians can in principle be concerned about global inequality, but this concern is derivative of the deeper commitment to protecting persons' access to basic needs rather than as a direct concern.4 Thus, it is still open to this position that substantial inequalities between countries be condoned when these inequalities do not in fact force anyone below some defined subsistence threshold. For example, we can imag- ine, say, a global arrangement in which the economic inequalities between countries exceed the limit permitted by (a globalized) Rawls's difference principle (and so would exercise Rawlsian global egalitarians like Beitz and Pogge) that is nonetheless unproblematic for the anti-global egalitarians because no one is forced below the basic minimum in spite of the inequality. In short, the difference between global egalitarians and anti-global egalitarians (who however remain committed to human- itarian assistance) is that the former supports a distributive principle for the purpose of regulating and limiting economic inequalities globally, whereas the latter is con- cerned only with ensuring that basic humanitarian needs of all persons are met. So a significant practical and philosophical difference remains to be addressed: must global justice include an egalitarian distributive commitment? I will consider two classes of arguments often deployed by the anti-global egal- itarians against attempts to universalize the demands of distributive equality. The first class of arguments tries to show that global egalitarians have misconstrued the reasons why equality matters domestically, and hence have illegitimately extended these reasons to the global arena. That is, the reasons why inequality mat- ters domestically, some anti-global egalitarians say, are reasons that gain traction Phil. & Pub. Affairs 113; John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999), esp. Part III; and Samuel Freeman, Justice and the Social Contract: Essays on Rawls (New York: Oxford University Press) [forthcoming] at chs. 8 and 9, and "The Law of Peoples, Social Cooperation, Human Rights, and Distributive Justice" (2006) 23/1 Soc. Phil. & Policy 29. See also Michael Blake, "Distributive Justice, State Coercion and Autonomy" (2001) 30/3 Phil. & Pub. Affairs 257; David Reidy, "A Just Global Economy: In Defense of Rawls" J. Ethics [forthcoming], and Joseph Heath, "Rawls on Global Distributive Justice: A Defense" Can. J. Phil. [forthcoming]. 4. See, e.g., David Miller, "Against Global Egalitarianism", supra note 3. The Boundary of Justice and the Justice of Boundaries only in the context of the political society. This class of arguments thus makes a conceptual point about the scope or boundary of egalitarian distributive justice, namely, that the boundary of distributive equality is inherently limited by the bound- aries of states. Call this the thesis of the limited boundary of distributive justice or the "limited thesis" for short. The second class of arguments does not directly question the conceptual bound- ary of egalitarian distributive justice. Rather it provides a limiting argument of a different kind. Instead of directly rejecting egalitarianism as principle with a global scope, it proposes that there are competing principles that in effect outweigh the demands of global equality.' In particular, this class of arguments tries to show that the claims ofjustice that citizens have against each other do and should limit the application of any global egalitarian principle. That is, the principle of special con- cern imposes demands on persons that limit what global equality can in fact demand of them. Because this line of argument holds that there is something significant about membership in a state, let me call this the thesis of the ethical significance of state boundaries, or the "significance of boundaries" thesis for short. I will reflect on these two sets of arguments in turn. But these two arguments are related in an important way. An important implication of this two-stage exam- ination of the boundary of distributive justice and the justice of political boundaries is that the boundary of distributive justice cannot be co-extensive with social or political boundaries (as claimed by the limited thesis) when the justice of boundaries is in question. II. The Boundary of Distributive Justice As mentioned, some anti-global egalitarians deny that the reasons in defense of distributive equality in the domestic context extend to the global context. Global egalitarians, they argue, fail to see that these egalitarian considerations gain a "foothold" only in the domestic context because of some special features of domes- tic political society that are absent in the global setting.6 These unique features of the political society are what generate the distributive commitment within the soci- ety. On this view, distributive egalitarian duties do not have universal scope; rather they are duties that arise only between members of a political society. What is distinctive about the political society such that distributive egalitarian considerations arise within its boundaries but not without? A recent line of argument in defense of the limited thesis focuses on the lawful coercive authority of the state (e.g., R. Miller; Blake; Nagel). On this view, it is the presence of lawful state coer- cion that motivates a distributive egalitarian commitment among citizens of the 5. As Dworkin notes, one need not deny the validity of a principle when one takes it to be out- weighed by other considerations or principles in particular instances. Taking Rights Seriously (Cambridge, MA: Harvard University Press, 1977) at ch. 2. So an anti-global egalitarian may oppose the universal application of egalitarian justice without denying the universal scope of a distributive principle, but (more modestly) by showing how other principled considerations can outweigh the principle of global equality and hence limit its application. 6. Nagel, "The Problem of Global Justice," supra note 3 at 128. state. And because there is no coercive lawful authority in the global arena, there is no grounding for any global distributive egalitarian commitment. It is important to clarify the real thrust of the argument so as to not misunderstand and underestimate its force. The argument is not claiming that there can be no dis- tributive commitments in the global arena because of the absence of an enforcement authority. If this were so, two responses would be immediately available to the global egalitarians. First, the argument so construed is highly controversial as a conceptual point for it takes the concept ofjustice to inherently include the notion of enforce- ability, such that a demand ofjustice that is unenforceable is not a demand ofjustice properly speaking. This understanding of the argument may find some grounding on Hobbes's remark that "covenants, without the sword, are but words". The Hobbesian may thus be tempted to think that justice has to be enforceable if it is to be a meaningful concept; and because there is no global authority or government to enforce the terms of distributive justice in the global arena, distributive justice or injus- tice "have there no place" (borrowing Hobbes' phrase). Thus the boundary of dis- tributive justice is necessarily limited to the boundaries of states for it is only within the boundaries of states that the demands ofjustice are effectively enforceable. For a label, we can call this the Hobbesian interpretation of the argument from coercion.7 Yet, as should be obvious, this coupling ofjustice with its enforceability is far from a widely held view of the concept of justice. After all, we can make sense of what justice demands even in situations where we know (regrettably) that we are unable to impose or enforce its demands. We can, for instance, embrace the United Nation's aspiration that developed countries contribute 0.7% of their GDP to development assistance (though of course this is not an egalitarian commitment but a commitment to end absolute deprivations) as a requirement ofjustice, while recognizing that most countries have failed to meet this requirement and so have failed in their duty ofjustice.8 The enforceability ofjustice and the demands ofjus- tice refer to two distinct stages in the fulfillment of justice. The first identifies its content and demands; the second speaks to its compliance and realization. To be sure, if global distributive demands cannot in fact be enforced without a global state, this would not bode well for the global egalitarian aspiration. So, if this is right, global egalitarians may have to rethink the reality of their goal given the untenability and undesirability of a world state. But this concern relies on another Hobbesian premise, an empirical one this time, that is also contestable. It relies on the Hobbesian view of international relations that there can be no enforcement of justice globally because of the absence of a global political sovereign. But, and this is the second response of the global egalitarian against the Hobbesian interpretation of the coercion argument, the claim that there are 7. Hobbes's own views on international justice are arguably more complex and so I intend this lit- erally as a Hobbesian view rather than Hobbes's own view on the matter. For one discussion on Hobbes and international law and justice that runs against the traditional view of Hobbes as an international amoralist, see Larry May, "Jus Cogens Norms" in his Crimes Against Humanity (Cambridge: Cambridge University Press, 2005). 8. Indeed, anti-global egalitarians will lament this actual failing on the part of rich countries, given that the goal here is that of poverty reduction. But they can do this only if they accept the dis- tinction between what justice demands and how justice can be affected. The Boundary of Justice and the Justice of Boundaries no effective means of enforcing distributive commitments globally absent a world state is empirically weak. International law is sometimes derided as toothless because there is no global sovereign to execute its orders. But this caricature of international law fails to see how the system of international law, along with a host of other international institutions (including organs like the World Court, etc.) can set expectations and motivate conformity with these expectations. To use the catchphrase, there is global governance even in the absence of global government. The enforcement of global justice need not be perfect, but to say that there is no enforcement at all without a world sovereign yielding the sword is an exaggeration.' Indeed, and closer to my purpose, none of the anti-global egalitarians I will be discussing endorse the Hobbesian view of a lawless international order. For instance, Rawls himself believes that global institutional means of enforcing the principles of his Law of Peoples are available. He thinks that global institutions "such as the United Nations ideally conceived"'" can have the authority and capacity to express and enforce the principles of the Law of Peoples, and to regulate the conduct of peoples. Thus, the feasibility of establishing new or improving global institutions for implementing and regulating conduct in accordance with principles of global justice is not ruled out by Rawls. Rawls's reason for denying a global distributive principle is not that there are no institutional means of lawful international enforce- ment, but because of his understanding of the basis and philosophical motivation of egalitarian justice. The Hobbesian interpretation of the argument from lawful coercion rests on premises about the concept of justice and basic facts about the international order that Rawls himself will find unacceptable." The argument from coercion, more powerfully then, is not to be seen as an argu- ment about the (lack of) enforceability of distributive justice (in the global context), but as an argument about the inconceivability of distributive justice (in the global setting). The idea here is that outside the context of a lawful coercive authority, considerations of distributive equality do not even arise. The problem is not that the international domain is one in which distributive commitments cannot be enforced (and hence inappropriate to talk about distributive justice here); the prob- lem, rather, is that distributive justice concerns do not even take hold in the inter- national domain. Considerations of distributive egalitarian justice are relevant only under conditions in which persons share a coercive legal order. The issue, in short, is how distributive justice is to be conceived not how it is to be enforced. What is it about lawful coercion that it is said to present a necessary condition for generating distributive commitments among agents? One line of argument invokes the ideal of reciprocity to ground the claim. Thus, to pave the way for the 9. For some discussion on international law as "public", and hence not entirely subject to the political interests of state agents, see Benedict Kingsbury, "The Problem of the Public in Public International Law" in NOMOS: XLIX. Moral Universalism and Pluralism (New York: New York University Press) [forthcoming]; and Anne Marie Slaughter, A New World Order (Princeton, NJ: Princeton University Press, 2004). For criticism of the Hobbesian world view, see Beitz's classic treatment in Part I, supra note 1. 10. The Law of People, supra note 3 at 36; also at 70. 11. See Rawls's rejection of political realism in Part I of The Law of Peoples, supra note 3. coercion argument, let me make some preliminary remarks about reciprocity and distributive egalitarianism, and its implications for global distributive justice. It is implicit in the idea of a society understood as a fair system of social coop- eration, in which citizens are considered as free and equal, that the basic social and political institutions of society must be those that citizens themselves can reasonably accept. That is, citizens must be able to justify to one another the social arrangement that they are helping to impose on each other. This requirement that citizens justify their common political and social order to each other describes the criterion of reciprocity. As Rawls puts it, the criterion of reciprocity holds that "when terms are proposed as the most reasonable terms of fair cooperation, those proposing them must think it at least reasonable for others to accept them, as free and equal citizens, not as dominated or manipulated or under pressure caused by an inferior political or social position."'" For many liberal egalitarians, including Rawls, it is this criterion of reciprocity that ultimately grounds a distributive commitment in a democratic society. As Rawls writes in Justice as Fairness, "taking equal division as the bench- mark [as is appropriate if we regard citizens as free and equal persons], those who gain more are to do so on terms acceptable to those who gain less, and in particular those who gain the least." 3 The difference principle, Rawls writes, is construed to meet the "reciprocity condition: those who are better off ... are not better off to the detriment of those who are worse off," but in fact are better of only when the least well off also benefits. 4 Accordingly, a society that allows great inequalities to exist between rich and poor would be one whose social and economic order could be reasonably rejected by some (for example by the worst-off). Those disadvantaged by the arrangement can reasonably feel that their interests are not being adequately taken account of if the institutions of society are structured such that those already more fortunate (because of, say, their natural endowment or social class) are permitted to benefit further to the detriment of the unfortunate when there are feasible alternative arrangements under which the disadvantaged can do better. For this reason Rawls rejects what he refers to as a capitalist welfare state in which citizens are guaranteed only a basic minimum but where inequalities between them above the threshold of basic minimum need not be regulated at all. In such a society, the criterion of reciprocity is violated because the least well-off class will see the better-off to be benefiting from an arrangement at their expense.5 Defenders of the limited thesis need not claim that the criterion of reciprocity is of no relevance in the global domain. It seems natural to think, after all, that we don't just owe fellow citizens justification for the institutions that we are imposing on them; we also owe foreigners a justification for any global order that we have a part in helping to impose on them. The idea of reciprocity seems to be a morally basic one such that it applies among persons who are regarded as moral equals and 12. Ibid. at 14. 13. John Rawls, Justice as Fairness: a Restatement, ed. by Erin Kelly (Cambridge, MA: Harvard University Press, 2001) at 123. 14. Ibid. Also at 49, 76-77. 15. Ibid. at 126-30. The Boundary of Justice and the Justice of Boundaries not just between democratic citizens. 6 It is entirely possible, then, that any global arrangement that allows for great inequalities between countries is one that the global worst-off persons could not reasonably accept. Yet for anti-global egalitarians, the ideal of reciprocity generates different eco- nomic justice commitments in the domestic and global arenas. Rawls, for instance, while not denying the criterion of international reciprocity, suggests that that ideal is met in the international domain "once the duty of assistance [between people] is satisfied and all peoples have a working liberal or democratic government". Reciprocity between peoples does not require peoples "to narrow the gap between the average wealth of different peoples."'7 Rawls does not fully explain in The Law of Peoples why the reciprocity criterion generates different requirements domes- tically and globally. He simply stipulates that fair cooperation between peoples does not require any egalitarian distributive commitment among them. To supplement Rawls's argument, one might be tempted to say that reciprocity generates a dis- tributive economic commitment only within a democratic society because of the type of mutuality felt between democratic citizens, and that there is no similar mutu- ality between persons as such. But this cannot, of course, be merely a descriptive claim (to wit, that democratic citizens in fact acknowledge and affirm such a mutual commitment for compatriots, whereas persons of the world as a whole do not). The argument has to be that it would be unreasonable to reject such mutual expectations among democratic citizens (for distributive equality) but that it would not be unrea- sonable to deny a similar expectation in the global setting. But this seems to beg the question for we can't determine whether an expectation is reasonable or unrea- sonable until we have a sense of what it is that persons are entitled to as a matter ofjustice. Thus, the anti-global egalitarian cannot simply dismiss any expectation for distributive equality on the part of the less advantaged of the world as unrea- sonable and use that as a basis for determining what global justice can require. So, it appears that an explanation is still owed by the anti-global egalitarians to show why the ideal of reciprocity generates distributive egalitarian commitments domes- tically but not globally. Some anti-global egalitarians thus emphasize a unique feature of the democratic political society, such that the expression of reciprocity in the domestic arena entails a distributive commitment, namely that citizenship in a political society is mem- bership in a non-voluntary and coercive legal association (e.g., R. Miller; Blake; Nagel). 8 On this argument, it is the expectation that democratic citizens justify to each other the coercive authority of the state, as required by the ideal of reciprocity, that grounds a distributive egalitarian commitment among them. The coercive authority of the state is an authority that each citizen has an equal share in. Democratic reciprocity means that this great power can be imposed on all only on terms that each can accept. This describes the idea of liberal democratic legitimacy-that political power is legitimate only when it is exercised on terms 16. Rawls is explicit that reciprocity is a relevant criterion not only among democratic citizens but also among peoples. See The Law of Peoples, supra note 3 at 35, 57. 17. Ibid. at 114. 18. See also Rawls, Justice as Fairness, supra note 13 at 93-94, 40. that all citizens can reasonably endorse. 9 The argument is that an institutional com- mitment to distributive equality is one necessary way (though, presumably, not suf- ficient) in which the coercive authority of the state can be justified to all and hence made legitimate. That is, if the price of membership in a political society is lawful coercion, the privilege of an institutionalized distributive egalitarian commitment between citizens (among other things) is what legitimizes this coercion in the eyes of members." As Richard Miller writes, it is the need "to provide adequate incen- tives for compatriots to conform to the shared [coercive] institutions that one helps to impose on them" that generates a concern for distributive equality in the domestic sphere." In a similar spirit, Michael Blake writes that "the real purpose of the dif- ference principle is to justify coercion to all those coerced [an inevitable fact of the state], including the least advantaged."22 As an aside, it is worth noting that the fact that membership is considered non- voluntary on this argument is not a rejection of the social contract idea. Social contract accounts need not rely on the voluntary acceptance of one s membership in a society but on the acceptance of the terms of membership in the given society. What indi- viduals are asked to consent to (hypothetically) is not whether they want to be mem- bers of a society; that is usually taken as a given. But precisely because membership is presumed to be fixed, they are entitled to social arrangements that they could not reasonably reject. Indeed, it is the presumption of non-voluntary membership that gives much normative weight to the contract idea that the terms of social cooperation as expressed by the basic institutions of society be acceptable to members. The argument from coercion thus offers one explanation for why the criterion of reciprocity generates distributive egalitarian commitments within democratic states but not beyond the borders of states. Although the reciprocity requirement remains in force in the global domain, there is no coercive global arrangement that needs to be similarly justified to individuals in the world as a whole. Only in the context of lawful coercion does reciprocity ground the commitment to distributive justice for only in this context is there a coercive social arrangement whose legit- imacy needs to be established. 3 19.Ibid. at 40-41. 20. As Paul Guyer lucidly reveals, this relationship between coercion, reciprocity and distributive justice can be found in Kant's political philosophy. See Guyer, "Life, Liberty and Property: Rawls and Kant" in his Kant on Freedom, Law and Happiness (Cambridge: Cambridge University Press: 2000) at 274ff, 285ff. Whether this Kantian account of distributive justice supports the limited thesis, I leave to one side for now. I suspect though, that if, as Guyer argues, the commitment to distributive justice is required if property rights enforcement (necessary for turning the right to property from a mere "provisional right" to a "conclusive" right) is to be reciprocally justifiable to all involved, one can argue that in as far as there are international norms and laws defining and protecting international property rights, then the Kantian account would deny the limited thesis and instead support some global distributive commitment. For Guyer's arguments in this direction, see his "The Possibility of Perpetual Peace" in J.C. Ciprut, ed., Ethics, Politics, and Democracy (Albany, NY: SUNY Press) [forthcoming]. 21. R. Miller, "Cosmopolitan Respect and Patriotic Concern", supra note 3 at 203-04. 22. Blake, "Distributive Justice, State Coercion, and Autonomy", supra note 3 at 283. 23. As a point of textual analysis, it is worth noting that R. Miller's reciprocity argument for dis- tributive justice emphasizes the need to give democratic citizens incentives to accept and comply with the coercive institutions of their society. Blake's argument focuses on personal autonomy, claiming that coercive arrangements can be justified to fellow autonomous agents only if these The Boundary of Justice and the Justice of Boundaries On this argument, then, there is no general direct concern for distributive equality among persons as such. The concern for distributive equality derives from the more fundamental requirement that coercive political arrangement be justified to each citizen on terms that each can accept (if it is to be legitimate), and it is this need to justify the coercive authority of the state to all citizens that grounds this distribu- tive commitment to them. 4 Nagel has accordingly said that distributive duties are "associative duties"; that is, they are duties that arise only among persons who share a certain kind of social or institutional relationship.25 Before examining this attempt to justify the limited boundary thesis, it is worth noting that the argument from coercion owes us some explanation as to why there is no direct concern for equality between persons as such (that is, why distributive equality is not an independent value but derives only from the need to legitimize the state's coercive authority). Why are distributive duties only associative duties, when there are other duties ofjustice among persons that obtain independently of associative ties? (Nagel, after all, accepts duties of'justice having to do with human rights that are independent of special ties.) That is, the argument from coercion attempts to make the case that coercion is a sufficient condition for distributive justice; but this leaves open the possibility that coercion is not a necessary condition, that distributive justice can simply be a natural duty of justice. My point here is not that such an argument (for taking distributive egalitarian duties to be a departure from the default position of inequality) could not be made; my point is that the default position of global equality, absent arguments to the contrary, is just as plau- sible a starting point.26 Be that as it may, granting that distributive egalitarian duties are derivative duties along the lines offered by the argument from coercion, the key question for my pur- pose is whether the argument successfully supports the limited thesis.27 It is not clear that it does. For one, it is arguable that the global order is indeed an involuntary and coercive one that affects profoundly and pervasively the life prospects of many disadvantaged persons in the world, and that this order is sus- tained and imposed on them through the complicity of others, in particular the global well-off. At the extreme end, the current global order deprives many of access arrangements do not permit excessive inequalities between them. But in either case, it is the idea of reciprocity that is fundamental-both are concerned with providing individuals with reasons to accept coercive institutions. Nagel's argument, as we will see in a moment, has again a different focus. Nagel stresses not the fact that citizens are subjects of coercive arrangements but that they are responsible authors of these arrangements and hence have a rightful say in how these arrange- ments ought to be. 24. The argument from coercion is unlike the Hobbesian argument contemplated above. That argu- ment was an argument about the concept of justice. The present argument has to do more with the idea of legitimacy. It is not that without enforcement there can be no justice; it is that to justify and legitimize the coercive enforcement authority of the state, the state (which is collectively supported by every citizen) owes each of its citizens a distributive egalitarian commitment (among other things). 25. Nagel, "The Problem of Global Justice", supra note 3 at 121. 26. Some might say that distributive duties are positive duties and positive duties can only be asso- ciative. But this (controversial) libertarian move is not one that the anti-global egalitarians I am discussing would themselves accept. 27. The next three paragraphs draw on my Justice Without Borders: Cosmopolitanism, Nationalism, and Patriotism (Cambridge: Cambridge University Press, 2004) at 172-77. to basic needs and goods." To be sure, these urgent deprivations are the sorts that the anti-global egalitarians I am discussing can deal with for these are what Blake calls instances of "absolute deprivations" rather than "relative deprivation". That is, the current global order is plainly coercive because many individuals are denied basic subsistence needs under its arrangement, and anti-global egalitarians do not deny that there is a duty of humanity to assist people thus deprived. But the global order is also coercive in ways that the domestic order is said to be coercive, that is, it is coercive in the sense that a certain conformity and com- pliance with a certain socio-political order is expected of persons. As in the domes- tic setting, there is in the global arena a legal order that is binding and enforceable on all, and from whose reach there is no meaningful escape, and which restricts individual life opportunities and personal autonomy. International trade laws, prin- ciples and laws governing territorial and political sovereignty, international norms regulating the conduct of multinational corporations, patent laws and other property laws concerning ownership of resources harvested from international waters, and so on are just some examples of a highly complex international legal system under whose regulation all persons are virtually subjects, and which limits the options and opportunities persons effectively have. And if the legitimacy of coercive social arrangements that limits personal autonomy is what motivates the need for a dis- tributive egalitarian commitment in domestic society, then so too, it seems, must the legitimacy of the global legal order rest on some distributive egalitarian com- mitment. It is important to stress that the international order is not simply coercive but lawfully coercive. The argument from coercion stresses the need to legitimize the legal coercive order of domestic society on those subject to its authority. That is, the relevant coercion that is in need ofjustification (to autonomous persons) is that which is legal and "ongoing. 29 So the analogy between the global and domestic settings is impaired if global coercion is not a lawful and ongoing coercion. But the examples of international coercion cited above pick out instances of lawful and ongoing coercion. The international order is not only coercive but also, as is attributed to the domestic political society, coercive in an institutional and legal way. International coercive actions are not confined to just those actions outside the bounds of international law (as in illegitimate interventions and illegal trading practices that violate, say, WTO rules), but include also actions that are sanctioned, and indeed often required, by international law. International laws and rules reg- ulating the recognition of sovereignty that allow and even require corrupt regimes to be tolerated by powerful countries with economic and political interests in sus- taining these regimes, WTO rules on trade and patent rights, and less obvious ones like norms and laws that recognize the sovereign right of states over any natural resource within its borders, are legal rules that impose themselves in an ongoing basis on individuals, and have the effect of coercing some by significantly limiting their opportunities. 28. See Pogge, World Poverty and Human Rights, supra note 1; also Onora O'Neill, Bounds of Justice (Cambridge: Cambridge University Press, 2000). 29. Blake, supra note 3 at 280. The Boundary of Justice and the Justice of Boundaries In addition to the global economic order that is being imposed on and enforced against all countries and their citizens, another example of a clear instance of legal- ized arrangement that can be seen as coercive of the global disadvantaged, and so in whose eyes is illegitimate if there are no corresponding attempts to mitigate the effects of that arrangement, are the immigration laws of individual (especially rich) countries. Immigration laws, enacted domestically (but sanctioned by international law and norms), are clearly legalized restrictions and impositions on non-members, and the effects of these restrictions are most vividly felt by the global disadvantaged. This is especially so in an existing global order in which resources and wealth are so unequally distributed between countries, for the global disadvantaged are legally and forcefully prevented from taking advantage of economic opportunities in better off countries." My purpose here is not to deny the legitimacy of restricting immi- gration as such, but only to identify another way in which our global arrangement is coercive of persons in an ongoing and legal way, and which consequently needs to be justified to those who are affected. Thus extending the domestic argument, the legitimacy of restrictive laws of immigration, laws that are enforced by the state but sanctioned internationally, must be conditional on there being some global dis- tributive commitments. The global disadvantaged who are being deprived of eco- nomic opportunities by laws restricting personal movement will have no reason to consent to and comply with these laws otherwise. These restrictions need not be seen by them as legitimate. In short, if a legal coercive authority must be justified to persons living under it in order for that authority to count as legitimate, either out of respect for their autonomy (Blake) or to give them incentives to comply with the arrangement (R. Miller), it does seem that the current global order upholds such a coercive legal order that must be thus justified. And if the justification of coercive authority requires an institutionalized commitment to distributive egalitarian justice, then it seems too that the justification of the international coercive order requires some global egalitarian distributive commitment. One development of the lawful coercion argument stresses not the general fact of coercive international law as such, but the notion of (coercive) property laws. Distributive principles are appropriate (and indeed necessary from the point of view ofjustice) for the institutions of the domestic society but not for international insti- tutions because the legal institutions of political society cover a complex series of norms and rules governing ownership. Distributive justice is, after all, not simply allocative or redistributive justice. Distributive principles are meant to regulate the background rules and norms of society that determine entitlements, just transac- tions, contract laws, and property rights." It is concerned not just with the expost 30. It is of course true that one's place of birth is a natural fact and not itself a matter of justice. As Rawls has noted, natural facts in themselves that are neither just nor unjust; what is just or unjust is "the way the basic structure of society makes use of these natural differences and permits them to affect the social fortune of citizens, their opportunities in life, and the actual terms of coop- eration between them" (Rawls, "Kantian Constructivism in Moral Theory" in Samuel Freeman, ed., John Rawls: Collected Papers (Cambridge, MA: Harvard University Press, 1999) at 337. The natural fact of a person's geographical place of birth becomes an issue ofjustice because global institutions turn that fact into an actual social disadvantage for the person. 31. Freeman, Justice and the Social Contract, supra note 3. facto allocation of a given bundle of goods and resources after they are produced among persons (that is, it is not concerned simply with the question of what should be taken from whom to give to whom), but is concerned more fundamentally with the conditions ex ante that determine rightful ownership, and this presupposes in place a whole series of laws regulating property rights, transactions, contracts and so on. Put another way, distributive justice is also backward looking in the sense that it looks to the background conditions that fix rightful ownership, whereas allocative justice is solely forward looking in that it aims to redistribute resources within accepted background norms of ownership. We can perhaps speak of a global allocative justice or redistributive justice (and perhaps the 0.7% of GNP contribution to ODA is an example of this), but not global distributive justice, so the argument goes. Thus Samuel Freeman observes that what "makes the incredibly complicated system of legal norms possible that underlie economic production, exchange, and consumption, is a unified political system that specifies these norms and revises them to meet changing conditions." Yet "[n]othing comparable to the basic structure of society exists on the global level."32 Freeman's argument offers an important reminder of the focus of distributive justice, that it is concerned with the background institutions of society that define rightful ownership in the first place and is not concerned with the mere allocation of resources and goods against a taken-for-granted theory of ownership. Yet it seems to me that the claim that there are no analogous background conditions that specify norms of rightful ownership of resources and wealth seems to underplay the exis- tence of a unified system of legal principles and rules that are operational in the global domain. For example, patent laws and other laws about intellectual property, international laws regulating how resources may be harvested from international waters, and more fundamentally, but less obviously for it is just so pervasive and taken for granted, the principle of state territorial sovereignty that assumes state ownership over resources that happen to lie within a country's borders, clearly do establish norms of ownership in a very profound and basic way. And these norms are, of course, not naturally given or inevitable, but are sanctioned and affirmed by international legal principles and practice. These international norms regulating global ownership seem to me to both generate the condition for distributive justice and provide the appropriate subject-matter of distributive principles. Now as a response to the above considerations, one might point out that inter- national property laws, in fact international law as a whole for that matter, are dif- ferent from domestic laws. Freeman writes that "[g]lobal cooperation and global institutions are supervenient upon social cooperation and basic social institutions."33 By this he means, as I understand him, that global norms and laws are fundamentally derived from domestic law in that these are norms and laws legislated by peoples as independent and self-determining cooperative units and put into international effect only through external arrangements (i.e., treaties) with each other. That is, there can be no international law without bounded independent political units capable 32. Ibid., emphasis mine. 33. Ibid. The Boundary of Justice and the Justice of Boundaries of making laws. For this reason, Freeman notes that global institutions are different "qualitatively" from the basic structure of a domestic society. Yet it seems to me that this line of argument conflates the source of international law, on the one side, and the scope and content of international law on the other. Although international laws can have their origin in domestic enactments, this does not mean that they cannot come to have a distinctive scope of application the pur- pose of which is to regulate the international rather than domestic domain. Moreover, that most (if not all) international laws have their origins in the form of treaties between independent nations, does not mean that these laws do not even- tually come to acquire a certain normative independence over time such that their authority comes no longer from the fact of the historical agreement between party countries to specific treaties, but from their general affirmation and acceptance by the international society over time, and can come to generate expectations not just from signatories to the particular treaties but from all lawful members of inter- national society. Current international laws about human rights, prevention of geno- cide and so on have their grounds of enforcement in agreements between signatory countries but nonetheless, through international practice, set lawful standards and expectations for all countries regardless of whether or not they were original sig- natories to these agreements. Indeed, it is hardly plausible that the norms grounding these laws derive their normative force solely from actual agreement between states, even if their legal enforceability and the means thereof are. In any case, there is certainly no opting out with respect to these laws, and hence it is a mischaracter- ization to say that they are only voluntary. The possibility of international laws that can come to have an autonomous authority and force (even though the original authority of these laws are domestic) is what paves the way for a lawful international order, as recognized by Kant. As he writes, "partly by an optimal internal arrange- ment of the civil constitution, and partly by common external agreement and leg- islation, a state of [international] affairs is created which, like a civil commonwealth, can maintain itself automatically."34 That is, even if international law is partly con- stituted by some original agreement between independent sovereign states, it can come, over time through international practice and the generation of mutual expec- tation, to enjoy an independent status and can retain its authority as law indepen- dently of the historical agreements between actors. Thus far, it seems then that if the need to justify a lawful coercive order is that which motivates a distributive egalitarian commitment, that motivation is also avail- able beyond the borders of states in as far as the global order comprises also a lawful and coercive involuntary order. More has to be said as to why the domestic legal order is distinct from the global legal order such that the former generates egalitarian distributive commitments but not the latter. Thomas Nagel has offered what seems to me to be a version of the argument from lawful coercion that side-steps my above response.35 For Nagel, it is not simply 34. Immanuel Kant, "Idea For a Universal History with a Cosmopolitan Purpose" in H.S. Reiss, trans. by and ed., Kant: Political Writings (Cambridge: Cambridge University Press, 1991) at 48, stress in original. 35. I am grateful for David Reidy's critical and helpful comments on the following paragraphs. the fact that citizens are affected by coercive institutions that is morally significant, and the reason for which an explanation for institutionally sanctioned inequalities is owed to them. What is crucial, on Nagel's account, is the fact that citizens are also mutually engaged in the making and maintaining of these coercive arrange- ments. Because laws sustaining our social institutions are also made in our name, and we are responsible for supporting these laws through our actions and decisions, we have a hand in any inequalities admitted by our lawful institutions. Hence we are consequently entitled to ask why we should accept the inequalities generated by these laws. It is this "engagement of the will" in the system of coercion that we jointly sustain that generates a distributive commitment among persons so engaged. He writes: Without being given a choice, we are assigned a role in the collective life of a par- ticular society. The society makes us responsible for its acts, which are taken in our name and on which, in a democracy, we may even have some influence; and it holds us responsible for obeying its laws and conforming to its norms, thereby supporting the institutions through which advantages and disadvantages are created and dis- tributed.Insofar as those institutions admit arbitrary inequalities, we are, even though the responsibility has been simply handed to us, responsible for them, and we therefore have standing to ask why we should accept them. This request for justification has moral weight even if we have in practice no choice but to live under the existing regime. The reason is that its requirements claim our active cooperation, and this can- not be legitimately done without justification-otherwise it is pure coercion.36 In short, some justification is owed to me (as a member of a political society) for any inequalities that the lawful order of my society permits because I am also a law-making member of society. I am entitled to this justification because I am not simply a subject of the law but a rightful author of the law. It is my capacity as one of the authors of these laws (an authorship that I express through my participation in the institutional life of the society) that gives me the moral "standing" to demand from my fellow members a justification for these laws. Thus any legal institutional arrangement that is disadvantageous to some members must be justifiable to them since they are entitled to ask, given that these are laws supposedly also made in their own name, why they should support and accept such a lawful arrangement. An institutionalized distributive egalitarian commitment among members of a soci- ety regulates and determines the bounds of acceptable inequalities between mem- bers, thus rendering the lawful arrangement agreeable to all. 7 36. Nagel, supra note 3 at 129, emphasis mine. 37. The reference to "law-maker" suggests that, for Nagel, a just society has to be a democratic society in which citizens are not merely participants of a system of social cooperation but that they have some say in the terms of cooperation. This contractarian reading ofjustice in Nagel, namely that a just society has to be democratic in some form, that is, a society whose members are in some sense authors of their own laws and not simply subject to them, is consistent with Nagel's overall position. Consider his opposition to Rawls's account of international toleration. Nagel writes that "there seems nothing wrong with being particularly supportive of transformation in a liberal direction" (ibid. at 135) even when we concede that there are "obvious practical reasons" for not imposing liberal democratic ideals on nonliberal societies. "We owe it to other people-con- sidered as individuals-to allow them, and to some degree enable them, to collectively govern themselves" (ibid., my stress). That is, we owe it to others, morally speaking, to allow them, if not even help them, establish domestic democratic institutions. The Boundary of Justice and the Justice of Boundaries Thus on Nagel's account, it seems that even if the laws made in our name may be unfavorable to others (under whose name these laws are not made), the latter are not entitled to an explanation. It is only when one is also disadvantaged by laws that one is also supposed to have a say in, or a role in shaping, that entitles one to some explanation and justification for a given arrangement. So even if I am right above that international law is coercive of many, if it is indeed true that international law is however not made in the name of persons as members of the global society as such but is made by agents in their capacities as citizens, then only fellow citizens are owed any justification for any social and legal order and the inequalities admit- ted by such order, for only fellow citizens are mutually engaged in the making of these arrangements. "The state makes unique demands on the will of its mem- bers-or the members make unique demands on one another through the institu- tions of the state-and those exceptional demands bring with them exceptional obligations, the positive obligations of justice."38 Because there are no similar "unique demands" on persons by the global order, no one has standing to demand justification for a global order that disadvantages some relative to others, and so the positive obligations of distributive justice in that context do not exist. Freeman's observation considered above, that international law supervenes on domestic law, is revived thus: what is qualitatively different between international and domestic laws is that domestic laws are made in the name of citizens whereas international laws, while enforced against foreigners, and are not laws made in their name. Nagel's thesis that it is the "engagement of one's will" in the lawful institutions of one's society that gives one the standing to demand a justification for any arbi- trary inequalities that institutions generate is complex and challenging. I don't pre- tend to be able to offer a decisive response to it. But some tentative remarks may be made. A first response is to insist that there are international laws and policies that are in fact made in the name of all persons governed by these laws. For instance, when the General Assembly of the United Nations adopts resolutions and policies that affect persons anywhere in the world, this is done supposedly in the name of the people (of the different countries) that the delegates to the Assembly officially represent. In as far as we accept that an ideal state is one in which the government represents the people in some appropriate way, any enactment of inter- national law by governments of countries will be in the name of the people they supposedly represent. So even if distributive justice considerations arise only among individuals who collectively make laws in their own name, it is contestable whether this rules out a consideration for global distributive justice. Of course it is true that as a descriptive matter, many governments do not effectively represent their com- patriots (especially in cases of non-democracies). But Nagel's thesis is a normative and not a descriptive one. It establishes the condition for state legitimacy. The point concerns how we ought to understand and conceive of individuals. We are to see them not merely as subjects of the laws of their country but also in some ways as the rightful authors of these laws. Under this condition, it seems that international laws are laws made in the name of a people that are represented by its delegates. 38. Ibid. at 130. Nagel acknowledges the formation of new networks of global governance that have transcended the traditional account of international law as simply treaties between states. Yet, he notes that these "these networks bring together represen- tatives not of individuals, but of state functions and institutions" and this indirect relationship to individuals is morally significant?9 But even if it is right that par- ticipants in global institutions are representatives of states rather than of individuals, which seems in fact correct for. the most part, should this not at least generate some distributive commitment between states? That is, if states are in fact making inter- national law in the name of each other, shouldn't the considerations applicable to individuals within domestic society analogously apply between states? If so, this would generate at the very least some distributive egalitarian duties between coun- tries. More to the point, however, it is not clear why the fact that individuals are represented through different institutions, some more directly than others, at dif- ferent political arenas, should be seen as morally significant in the way Nagel sug- gests. The different forms of representation can be seen as an institutional division of labor, that is, as different ways of representing the interests of individuals. Indeed, even in the domestic society, individuals are not always directly represented-their different interests are represented through different institutions, some more directly than others. Yet presumably Nagel does not require complete direct individual rep- resentation in the political order of society as a precondition of the "engagement of one's will" in the institutional life of their society. But leaving aside whether international laws are or should be seen as laws made in the name of all persons affected, there is a fundamental principle in Nagel's thesis that demands closer examination. This is the idea that only persons who are also authors of coercive lawful institutions have some standing to demand that these institutions not be to their disadvantage relative to other participants in the insti- tutions. The fact that one is simply affected or subject to some lawful arrangement, even disadvantageously, does not sufficiently entitle one to some justification. This point is well illustrated in Nagel's response to an anticipated objection, that immigration laws enacted by individual (rich) countries constitute a class of laws that affect outsiders disadvantageously. On the objection, aren't foreigners who are disadvantaged by the immigration laws of (rich) countries entitled to an explanation for this coercive arrangement, and would the justification of immigration restrictions not require some distributive commitments to them? Nagel points out, in response, that immigration policies do not need to be justified to outsiders because immi- gration "policies are simply enforced against the nationals of other states; the laws are not imposed in their name, nor are they asked to accept and uphold these laws. Since no acceptance is demanded of them, no justification is required that explains why they should accept such discriminatory policies, or why their interests have not been given equal consideration. It is sufficient justification to claim that the policies do not violate their prepolitical human rights."4 I think some questions may be raised concerning this focus on the rights of per- sons qua law-makers rather than simply qua subjects of the law. Why is it the case 39.Ibid. at 139. 40. Ibid. at 129-30, emphasis mine. The Boundary of Justice and the Justice of Boundaries that no explanation is owed to persons simply on the ground that they are affected by laws and institutions that others are imposing on them? Why is a justification (for unfavorable arrangements) owed only to those who have a hand in sustaining the arrangement and not simply to anyone who is affected, whether or not they have a role in the making of the arrangement? If laws made in our name have a disad- vantageous impact on others, can they not demand an explanation from us even though these laws were not laws made in their name? Indeed, if we accept as a gen- eral criterion of legal legitimacy that laws affecting persons must be laws that they could themselves accept, then Nagel's point that only those in whose name a law is enforced have "standing" to demand some explanation seems ungrounded. Legitimate and lawful coercion would seem to require that anyone affected by coer- cive arrangements is entitled to ask for some explanation for laws that may be imposed on them. One might say that it is not the fact that a law is imposed on some that is morally relevant. What is relevant is the presence of a lawful and coercive authority that has dominion over one. Hopeful immigrants, while they do find their movements restricted by immigration laws that can be enforced coercively against them, are not under the dominion of the authority enacting that restriction as long as they remain outside the jurisdiction of that society. Following Dworkin, one can say that a state because of its dominion over all within its jurisdiction owes each one of its citizens equal concern if it is to be legitimate in their eyes. But it owes no such equal concern, so one might say, to strangers at its gates even if it is the fact that its (immi- gration) laws are what is keeping them out, for it holds no dominion over them until they are admitted (and therefore until they are admitted, it doesn't owe them equal concern). Yet this reasoning perhaps begs the question by taking it for granted that the dominion of a powerful state coincides with its territorial boundaries. The example of immigration laws and practice shows, to the contrary, that states can exercise some kind of authority and dominion over persons who are not within their tra- ditionally recognizes jurisdiction. States do expect outsiders to obey and accept their immigration laws. More explicitly, the foreign policies of powerful countries do affect numerous people who are not citizens, and lawfully so if these are policies enacted according to accepted domestic legislative procedures. So how can it be said that a state can exercise dominion only over its citizens, when the laws it passes domestically can affect quite pervasively the lives of outsiders?4 ' Why is this per- vasive and profound and legal influence on persons' lives not a form of dominion? The real issue, in short, is what is meant by the exercising of dominion over another. I am suggesting that one need not be a member of a society before it can exercise dominion over one, if we understand by having dominion over another as having the capacity to lawfully affect significantly the life options of another. 41. Dworkin writes, "A political community that exercises dominion over its own citizens, and demands from them allegiance and obedience to its laws, must take up an impartial, objective attitude toward them all .... Equal concern ... is the special and indispensable virtue of sovereigns." Ronald Dworkin, Sovereign Virtue (Cambridge, MA: Harvard University Press, 2000) at 6. To be sure, Nagel does not say that no explanation is ever owed to persons affected by laws not made in their name. When externally imposed arrangements and laws affect persons in ways that violate their "'prepolitical' human rights" some accounting for the violation is demandable. "Prepolitical" rights are the rights per- sons have independently of any institutional arrangements, and these are, for Nagel, limited to basic rights that include subsistence rights, but do not cover the right to distributive equality of some sort. So when lawful arrangements merely allow for inequalities, or where they only disadvantage some (without offending against their basic rights), no explanation is owed to those so affected when they are only subjects of these laws. But why is an explanation owed only when people's prepolitical basic rights are violated but not when they are put at a disadvantage? Is this a morally significant distinction? Imagine a prepolitical state of nature in which a community A is settled upstream in relation to another community B that is placed further downstream. Imagine also that each of these communities is dependent on the resources from the river. Suppose now that A elects to extend a net across the river for the purpose of improving its yield of fish. Imagine also that this obstruction when constructed will not violate any prepolitical rights of people in B; some fish and other essential resources will continue to slip through, so the basic needs of the community is not at threat. Still members in B will now have a diminished quality of life; perhaps they now have to work extra hours to obtain the same amount of goods as they did before the construction of the net. Or perhaps even though they still are able to meet basic needs, they simply have fewer resources to work with now, thus compromising their material quality of life. On Nagel's model, it seems, members of B have "no standing to ask" why they should accept this arrangement. Their prepolitical rights are not violated for they still are able to meet their subsistence needs; and the arrangement (the net across the river) was not erected in their name but in the name of people in A. But why is no explanation or justification owed to the people down- stream by those upstream? The fact that this is an arrangement not made in their name or for their sake is immaterial-what is material is that this arrangement affects them disadvantageously. To be sure, if the construction of the net results in the community downstream losing its means of subsistence, then a basic pre- political right to subsistence has been violated and on Nagel's view the community upstream can be held accountable by the affected community. But why is the com- munity to be held accountable only when basic rights are infringed upon? Is the fact that it has rendered another worse-off not something that it has to be held accountable for, something for which it needs to justify? Immigration is just one example of laws domestically enacted that affect others disadvantageously and to whom, it seems from the above analogy, some justification and explanation is owed. Domestic environmental laws regulating industrial pol- lution levels, state subsidization of domestic sectors, laws regulating conduct of national corporations and so on are other examples. It is the recognition that these so called domestic policies affect outsiders and that some accounting is owed to all negatively affected that have motivated the call for greater global democratic The Boundary of Justice and the Justice of Boundaries decision making and institutions with respect to certain issues like the environment, conduct of corporations, state subsidy of agriculture and so on. To be sure, if domestically enacted laws or policies do not make foreigners worse-off, but only give them a worse deal compared to what fellow citizens get, it is hard to see what the grounds for complaint can be. But there are cases of laws domestically enacted that are not just relatively advantageous to members but that in effect make outsiders worse-off than they would be without these laws. If there indeed are such cases, do we not owe outsiders an explanation (for why they are being made worse-off) even if these are not laws made in their name? Don't we owe them an explanation even if we are not violating their prepolitical rights but simply because we are making them less well-off than they otherwise would be? Why should those negatively affected accept this coexistence peace- fully otherwise? Indeed, isn't one of the arguments for greater global cooperation and democracy prompted by the recognition that there is a need to justify to each other globally our decisions and policies that can affect others in ways that are disadvantageous?42 It seems to me that Nagel's attempt to drive a wedge between domestic and inter- national legal contexts, thus providing a unique grounding for egalitarian distributive justice in the domestic but not international context, does not fully succeed. It relies controversially on the point that only persons in whose name supposedly an arrange- ment is established and supported have the standing to ask why they should accept the arrangement when that arrangement admits of arbitrary inequalities. But the ideal of legitimacy seems to suggest that anyone affected by an arrangement that is disadvantageous of her or him in the sense that she or he is rendered worse-off by the arrangement is entitled to some justification and accounting, whether or not the arrangement was established in her or his name, especially when escaping from the effects of the arrangement is not a real option. That arrangement would be a case of "pure coercion" otherwise. 11. The Justice of State Boundaries Let me now turn to the second set of considerations that I have called arguments for the ethical significance of boundaries. As I said, this class of arguments does not directly challenge the global scope of distributive justice. Rather, it attempts to show that there are competing considerations or claims ofjustice that can limit what global egalitarian justice can demand of people. In particular, these arguments try to show that citizens owe to each other certain special obligations of distributive justice that may outweigh their duty of distributive justice towards foreigners. We can divide arguments for the special duties of citizens into two broad 42. This point is suggested in Kant: "The concept of Right, insofar as it is related to an obligation corresponding to it (i.e., the moral concept of Right), has to do first only with the external and indeed practical relation of one person to another, insofar as their actions, as facts, can have (direct or indirect) influence on each other." Kant, The Metaphysics of Morals, trans. by Mary Gregor (Cambridge: Cambridge University Press, 1993) at 56. See also sections 43-45 in "The Doctrine of Right". categories-those that treat these special duties as voluntary duties and those that treat them as non-voluntary duties. On the broadly voluntary approach, the special duties that citizens have to each other derive from their willing participation in a mutually beneficial association. It does not mean that citizens must have explicitly consented to these special duties, nor does it entail that membership in the political society is explicitly voluntary. The idea is that there are benefits to being a member of a social cooperative scheme which the well-ordered state is, and in accepting these benefits, one has also effectively consented to the special duties that come with being a participant and beneficiary of that cooperative scheme. The voluntarily acceptance condition is satisfied if there is willing and ongoing enjoyment of the benefits of being a member of the association. For short, we can call this the "mutual benefit argument" for special obligations. What some anti-global egalitarians go on to claim is that these special obligations are obligations ofjustice between citizens that can limit the demands of global egal- itarian justice. This argument, it is worth noting, is more modest than the "shared institutions" argument considered above. The present argument does not claim that only within an existing political society do considerations of distributive justice gain traction. It makes only the limited claim that special considerations ofjustice arise among individuals sharing a mutually beneficial scheme, and that what indi- viduals owe to each other more generally as a matter ofjustice will have to be bal- anced against such special duties that they can have. That persons have special obligations to their fellow citizens seems grantable. What is more controversial is the claim that these special duties ofjustice can limit the demands of global justice. It is this claim that we need to examine, momentarily. On the broadly non-voluntary account of special obligations, individuals do not acquire special obligations only when they have voluntarily assumed them in some way but can acquire special obligations simply by virtue of their occupying a role or social position to which these obligations attach. Moral philosophers sometimes refer to these as "associative duties".43 Associative duties are a species of special duties, but unlike other kinds of special duties, they pick out those special duties that attach to social roles, that occupants of a role have simply by virtue of being an occupant, even when they have not volunteered for the role. To say that a duty is an associative duty does not mean that the role to which the duty attaches cannot be voluntarily assumed. It means rather that the normative force of that duty comes not from the voluntary assumption of that role but directly from what it means to be an occupant of that role. We can call this the argument from associative duties. The different social roles that we occupy are characterized by different norms that give meaning and value to these roles, and the special duties that an occupant of roles has is a function of these norms." For example, we ordinarily accept that among the norms inherent in the ideal of friendship is some kind of special concern or loyalty towards friends. One who does not show her friends any special concern, 43. Michael Hardimon, "Role Obligation" (1994) XCI/7 J. Phil. 333; Samuel Scheffler, Boundaries andAllegiances (New York: Oxford University Press, 2001). 44. See Hardimon, supra note 43. The Boundary of Justice and the Justice of Boundaries or more fundamentally rejects the norm of special concern as inherent in the notion of friendship, does not know what it means to be a friend. It is part of being in a friendship that one shows a friend special concern. One who rejects this special concern is not occupying the role of friend as we ordinarily understand it. But associative duties are important in another way. Not only are they duties that inherently attach to certain roles or positions, but they are also duties that are not fully reducible to some more general moral principles. For example, one mis- understands and under-appreciates the value of friendship if one understands the special concern for friends to be ultimately, say, a strategy or a division of moral labor for maximizing overall social utility (that is, if everyone were to care specially for their friends, the good of society can be best realized). As Scheffler puts it, "it is pathological to attach nothing but instrumental value to any of one's personal relationships.... [And] if one values one's relationship to a particular person non- instrumentally, one will inevitably see it as a source of reasons for action .... for treating that person differently from others."45 Someone who takes the value of friendship to be entirely derivable from the principle of utility, or some even other general principle like the ideal of mutual benefit or the duty of gratitude, would be a false friend. She would in fact not be participating in a friendship relation and is liable to be a person to leave a "friend" high and dry when it is no longer mutually beneficial to stay in the relationship, or when the principle of gratitude does not require that the relationship be sustained. Some people might think that the argument from associative duties relies on a philosophically untenable account of the special duties of citizenship compared to the mutual benefit model. After all, citizenship is unlike say friendship or kinship on a number of dimensions. But still the position is not entirely implausible (and some philosophers have defended or at least entertained this possibility). For the sake of discussion, I will simply (to my own disadvantage) grant this associative duty account of special citizenship duties. I hope to argue that even if the special duties of citizenship are associative duties, they don't provide reasons for limiting the demands of global egalitarian justice. The case for limiting egalitarian demands here is not that distributive duties are only strictly associative duties and so have no global scope. That argument was discussed in the previous section (e.g., the discussion on Nagel). The present argu- ment is agnostic about the scope of distributive justice. The claim is that the (spe- cial) associative duties of citizenship present special demands among citizens that can limit what global distributive justice can ask of them. I mention the mutual benefit and associative duties arguments concurrently because my response to each will be structurally similar. Although the mutual ben- efit and associative duties arguments give different reasons for the special duties of citizenship, what is key in each of these arguments (in their defense of the sig- nificance of boundaries thesis) is the claim that special citizenship duties can limit the requirements of global egalitarian justice. My argument will be that, whether 45. Scheffler, supra note 43 at 121. we take patriotic duties to be mutually-beneficial or associative, that last move is unsustainable.46 My response will be two-pronged: (i) while membership in an association can have its special privileges and these privileges may be distributed by principles inter- nal to the association, it does not follow that what these privileges rightly are fall outside considerations ofjustice external to the association, and (ii) more funda- mentally, the boundaries that define spheres or units of special concern (and con- sequently membership in such units) are not themselves immune from evaluations from the point of view of justice. On the first prong of the response: Consider the argument from mutual benefit. To repeat, the first step of the argument is that citizens belong to a mutually ben- eficial arrangement and in accepting the benefits have also accepted the duties that come with it. The second step of the argument is that these special duties can limit the claims of non-members. Whether or not we endorse the first move as a correct justification for special citizenship duties, the second step of the argument cannot be permitted without further arguments. Individuals may voluntarily form and par- take in mutually beneficial arrangements, and accept special duties that come from participating in this arrangement. But it is another thing to say that simply by virtue of opting to join or form a mutually beneficial arrangement that one is straightaway absolved of one's duties to non-participants. If this were so, we might as well say that the demands of justice and morality may be unilaterally declined, for moral agents may simply absolve themselves of their outstanding responsibilities by vol- untarily taking on new and presumably less burdensome ones. But this, of course, is not how we understand the nature of moral duties-a person may not release herself from her outstanding duties simply through some voluntary act of hers. The main point to note, then, is that while persons may have special duties to one another because they share a mutually beneficial arrangement (or simply because of what it means to be an occupant of a certain role), the fact that they have these special duties alone does not absolve them of any outstanding moral com- mitments or duties ofjustice to non-participants. In fact, it is the other way around- what kinds of mutually benefit associations people may form and what kinds of special duties people may have on account of their participation in these associations are to be limited by what individuals owe to each other at large. A well-organized criminal gang may be described as a mutual benefit association. But we do not say that this means that members of the gang by virtue of their special duties to one another are no longer required to respect the right of nonmembers to property, not to be harmed and so on. The anti-global egalitarian argument from mutual benefit, if endorsed, must absurdly allow that a person's duty ofjustice to individuals may be limited if she is part of a criminal organization that requires her to pay special attention to her fellow members. Special duties do not limit what general duties ofjustice demand of persons; rather general duties ofjustice set limits to what kinds 46. Thus I don't attempt to argue that patriotic favoritism is indefensible per se or without any ratio- nal basis. For carefully laid out arguments in this direction, see Charles Jones, Global Justice at ch. 7. The Boundary of Justice and the Justice of Boundaries of special duties people can freely take on. One simply cannot take on a special duty that one is, as a matter of'justice, prohibited from carrying out. To put it plainly, that there are distinct spheres ofjustice does not mean that there need not be justice between these spheres.47 This general point about the relationship between special duties and general duties applies to the associative duties argument for the moral significance of boundaries. The fact, even if true, that it is part of what it means to be a citizen that we show fellow citizens special concern does not mean that the content of this special concern can be determined independently of what strangers can rightly claim of us. We can accept that friends can have special duties or concern for one another, and we can accept that these special duties are non-derivative but issue directly from what it means to be in that kind of relationship which we call friendship. But it is another thing to say that these special duties need not respect any bounds. On the contrary, we take the special duties of friendship to be limited by what justice per- mits-this is what allows us to say of a friendship that it has degenerated into crony- ism, or of kinship that is has degenerated into nepotism. Special concern for friends that violate norms ofjustice that apply to all persons falls outside what is permitted in the name of friendship.48 So even if it is correct that citizenship obligations are associative, that it is part of what it means to be a good citizen that we show each other special concern, it does not follow that that the claims of outsiders must be outweighed. On the con- trary, absent additional considerations, it is more the other way around. Before we can know what it is that citizens owe to each other by virtue of their status as com- patriots, we need first to know what it is that they may distribute among themselves, and this cannot be determined independently of what it is that they rightly own, which in turn cannot be determined without reference to what it is that they owe as a matter ofjustice to non-citizens. To put this point plainly, to know how much I must distribute among my fellows, I must first need to know what it is that I rightly have. This means that before we can realize our duties of justice to com- patriots, we must first have an account of global distributive justice and live up to its demands. Notice that my above considerations, consistent with the associative duties argu- ment, do not reduce the special duties of citizenship into some instrumental or strategic device to further the ends of global justice. It grants that these special duties are not reducible in that way. What it says is that even though these special duties have a certain normative independence, they are nonetheless to be limited by the demands ofjustice, and this means that they cannot be properly fulfilled without first determining what justice does demand of us. What we may call the "constrain- ability" of an associative duty by principles of justice on the one hand, and the reduction of that duty to principles ofjustice on the other, are two distinct features. Friendship duties may be constrained by, say, the general duty to warn others against 47. See Amy Gutmann, "Justice Across the Spheres" in Miller & Walzer, supra note 3. 48. Saying that duties of friendship are limited by considerations ofjustice does not mean that these duties are derivable from justice; rather it means only that the scope and content of these duties are constrained by the requirements of justice. foreseeable harm; but it does not mean that the values of friendship are reducible to or derivable from the principle of warning against harm. Now the anti-global egalitarian may at this point accuse me of question-begging. She may say that I have assumed some kind of global egalitarianism as the default requirement of justice and so naturally I can limit special duties between citizens to that egalitarian requirement. But I have not made any false assumptions. It is the anti-global egalitarians who have unwarrantedly assumed that they have a case against global egalitarianism when they say that I am begging the question. Remember that so far the case against the universal scope of egalitarian justice has not succeeded. So in defending the moral significance of boundaries, the anti-global egalitarian cannot assume that this case has been made. In sum, however citizens come to acquire special duties ofjustice to their com- patriots, it does not follow that these special duties can limit what global justice can demand of them. On the contrary, without further argument, the default position is to limit the kinds of special duties people can take on by what they owe to all others in general. The second prong of my response to the ethical significance of boundaries addresses a more fundamental point, a point that is already suggested, I believe, by the above considerations. Units or boundaries of special concern are themselves unjust if they confer special privileges on members at the unjustifiable expense of non-members. There are two ways that the justness of boundaries of special con- cern can themselves be called into doubt. One is when the terms of membership are in themselves indefensible (eg. we might think associations whose membership are inappropriately based on race are questionable). Another is if the existence of these special units of concern sustains or exacerbates existing injustices, as when those already unfairly advantaged join together to form associations that further their advantages, or when the formation of these associations put those already unfairly disadvantaged and left out at a further disadvantage. So, "males only" busi- ness associations can be criticized not just because of their gendered terms of mem- bership, but also because these associations allow already advantaged individuals to gain further advantages relative to non-members. Now perhaps the boundaries of spheres of special concern that characterize most rich countries need not be unjust with respect to their membership criteria. Few countries today adopt, say, a morally indefensible race based criterion of immi- gration (as say Australia did with its "White Australia" policy in earlier decades). But if only a few well-off countries violate this first condition ofjust boundaries and membership, it is less clear that many meet the second condition, which is that the boundaries and membership of units of special concern do not further existing background injustices. To return to a scenario discussed earlier, one reason why immigration restrictions in rich countries call out for justification (to outsiders who are potentially kept out) is not because these are based on arbitrary factors like race, but because these restrictions in the context of global inequality are odds with the freedom of persons to move and to better their economic condition. In a world marked by great economic imbalances, units of special concernfrom which many already deprived others are kept out and in which members, privileged to begin The Boundary of Justice and the Justice of Boundaries with, confer additional special privileges on one another, have the effect of aggra- vating the situation of the global disadvantaged. Not only do they remain deprived of access to more resources; they are prevented from such access by strictly policed boundaries from moving and bettering their opportunities. And, internally, partic- ipants in advantageous spheres of cooperation from which others are kept out often are able to further improve their situation, thus furthering the imbalance between members and non-members. Far from grounding special patriotic commitments that can limit the demands of global egalitarian justice, the justice of the boundaries of citizenship is impaired under conditions of global inequality. And when bound- aries are unjustifiable, they clearly cannot go on to justify limitations on what justice in fact demands. Whether persons voluntarily create mutual benefit associations or simply find themselves born into and participating in associative roles that come attached with special duties, the justice of the boundaries of these associations can- not be determined independently of what justice to outsiders requires, much less limit its demands. The boundaries of special concern are justifiable only if these boundaries and the special privileges of membership that they confer do not upset the duties of justice that people owe to each other at large. The above comments are not meant to establish the case for open borders.49 My point, simply, is that the justice of the boundaries of citizenship (and hence of the special concern membership can confer) cannot be taken for granted, that the boundaries of special considerations are justifiable only under certain conditions. This does not mean no well-off countries may regulate immigration under con- ditions of economic injustice. Justice is, after all, an ongoing quest, and to expect full realization of justice before special concern may be expressed in the human world is to deny any possibility of such expression. The real task for moral agents is to do their fair share in the promotion ofjustice. Among other things, what this means is that states may regulate immigration but only if they are also doing their share in correcting background global inequalities. So a rich country may restrict immigration only if it is prepared to live up to its global distributive commitments. To simplify this into a maxim, a country that wishes to keep strangers out must not insist on keeping more than its fair share of resources in. IV. Conclusion Let me recap the central claims of my discussion. I began by suggesting that global egalitarians can attempt to extend reasons for caring about inequalities in the domes- tic case to the global context. Anti-global egalitarians can try to block this extension by arguing that the reasons for caring about domestic inequalities are unique to the domestic political sphere. I have tried, in defense of the global scope of dis- tributive justice, to show that the case for limiting the boundary of distributive jus- tice in this way does not succeed. Until there are defensible reasons for limiting egalitarian considerations to the state, the default position, for egalitarians, is that of global egalitarianism. Here, the anti-global egalitarian may try to show that the 49. Joseph Carens, "Aliens and Citizens: the case for open borders" (1987) 49/3 Rev. Politics 251. global scope of distributive justice notwithstanding, the special duties of justice that citizens have to each other can limit the application of global distributive demands. But if global egalitarianism is the default position, any special concern compatriots owe to each other must be limited by the demands of global egalitarian justice, not the other way around, for we do not say that background justice can be ignored and outweighed simply when people take on special obligations. Fundamentally, the boundary of distributive justice cannot be defined by the bound- aries and membership of special concern when the justice of these boundaries and membership is open to question. Indeed a key issue of global justice, namely the question of the boundary of distributive justice, cannot proceed without consid- eration of the effects of existing political boundaries on persons' opportunities and access to resources. Again, my thesis is not that political boundaries have no sig- nificance and are to be imagined away in constructing an account of global justice. The point, rather, is that the justice of political boundaries cannot be simply taken for granted, and consequently they may be poor markers of the boundary of dis- tributive justice.