1 See Rawls, John A Theory of Justice, rev. ed. (Cambridge, MA: Harvard University Press 1999),Google Scholar §2; Political Liberalism (New York: Columbia University Press 1993), lecture VII; Justice as Fairness: A Restatement, Kelly, Erin ed. (Cambridge, MA: Harvard University Press 2001),Google Scholar §§4 and 15-16. For Rawls's two principles of justice, see Theory of Justice, 266-7, and Justice as Fairness, §13.

2 More explicitly, Rawls defines the basic structure as ‘the way in which the main political and social institutions of society fit together into one system of social cooperation, and the way they assign basic rights and duties and regulate the division of advantages that arises from social cooperation over time’ (Justice as Fairness, 10). On the idea that the various institutions that constitute the basic structure form a system together, see notes 35 and 61.

3 As Rawls puts it, the difference principle does not apply ‘to the decisions of individuals and associations, but rather to the institutional background against which these transactions and decisions take place.’ (Political Liberalism, 283; cf. Justice as Fairness, 54) For a particularly explicit statement of the general idea, see Justice as Fairness, 10-11.

4 See Rawls's brief comments on international justice in Theory of Justice (at 331-2), and the fuller discussion in The Law of Peoples (Cambridge, MA: Harvard University Press 1999).

5 See, e.g., Theory of Justice, 82, and Justice as Fairness, 10.

6 See Cohen, G.A. ‘Where the Action Is: On the Site of Distributive Justice,’ Philosophy & Public Affairs 26 (1997) 3–30CrossRefGoogle Scholar (reprinted with minor changes as chapter 9 of Cohen's If You're an Egalitarian, How Come You're So Rich? [Cambridge, MA: Harvard University Press 2002], and as chapter 3 of his Rescuing Justice and Equality [Cambridge, MA: Harvard University Press 2008]; references to this piece throughout will be to the original article, with corresponding page numbers in Rescuing Justice and Equality given in parentheses). Noteworthy discussions of Cohen's position include Thomas Pogge, ‘On the Site of Distributive Justice: Reflections on Cohen and Murphy,’ Philosophy & Public Affairs 29 (2000) 137-69, and Scheffler, Samuel ‘Is the Basic Structure Basic?’ The Egalitarian Conscience: Essays in Honour of G.A. Cohen, Sypnowich, Christine ed. (Oxford: Oxford University Press 2006), 102–29.CrossRefGoogle Scholar

7 See notably Beitz, Charles R. Political Theory and International Relations, 2nd ed. (Princeton, NJ: Princeton University Press 1999);CrossRefGoogle Scholar Pogge, Thomas Realizing Rawls (Ithaca: Cornell University Press 1989)Google Scholar and World Poverty and Human Rights: Cosmopolitan Responsibilities and Reforms (Cambridge: Polity Press 2002); Tan, Kok-Chor Justice without Borders: Cosmopolitanism, Nationalism, and Patriotism (Cambridge: Cambridge University Press 2004);CrossRefGoogle Scholar Caney, Simon Justice Beyond Borders: A Global Political Theory (Oxford: Oxford University Press 2006);Google Scholar Cohen, Joshua and Sabel, Charles ‘Extra Rempublicam Nulla Justicia,’ Philosophy & Public Affairs 34 (2006) 147–75;CrossRefGoogle Scholar Abizadeh, Arash ‘Cooperation, Pervasive Impact, and Coercion: On the Scope (not Site) of Distributive Justice,’ Philosophy & Public Affairs 35 (2007) 318–58.CrossRefGoogle Scholar

8 See notably Blake, Michael ‘Distributive Justice, State Coercion, and Autonomy,’ Philosophy & Public Affairs 30 (2002) 257–96;CrossRefGoogle Scholar Nagel, Thomas ‘The Problem of Global Justice,’ Philosophy & Public Affairs 33 (2005) 113–47.CrossRefGoogle Scholar

9 For Rawls's position on the family, see Political Liberalism, 258. On the difficulties that the case raises for the coercion view, see Cohen, ‘Where the Action Is,’ esp. 4 and 22-3 (117-18 and 137-8). Note that the coercion view is also difficult to square with the texts. Nowhere does Rawls explicitly say that the basic structure is the primary subject of justice because it is coercive. When he does speak of coercion, it is usually to stress that political power is necessarily coercive (see, e.g., Political Liberalism, 136 and 217, two passages that Blake takes to provide crucial support for his reading; see ‘Distributive Justice,’ 285-6).

10 I am grateful to an anonymous reviewer for pressing me to make this point explicit.

11 I am taking on board an assumption that the two standard views share: that what justifies focusing on the basic structure has to do exclusively with the kind of impact it has on individuals. The assumption could be questioned; indeed, Rawls himself does not accept it (see Justice as Fairness, §§15-16). I retain it here because there are fruitful possibilities within the logical space circumscribed by the assumption that have received too little attention, and because an explanation of the basic structure's importance in terms of its impact on people's lives would seem most philosophically satisfying. I am grateful to two anonymous reviewers for prompting me to clarify this aspect of my discussion.

12 See Cohen, ‘Where the Action Is.’

13 The intuitive reasoning for allowing such dealings is that the talented would be less productive without appropriate incentives, leading to worse prospects for all, including the worst off. Cohen discusses this argument at length in his ‘Incentives, Inequality, and Community,’ in The Tanner Lectures on Human Values, Grethe Peterson, ed., vol. 13 (Salt Lake City: University of Utah Press 1992), 263-329 (reprinted with minor changes as chapter 1 of Rescuing Justice and Equality). Rawls himself sometimes downplays the role of incentives in explaining why inequalities are necessary (see, e.g., Political Liberalism, 270).

14 ‘Where the Action Is,’ 5 (118).

15 Ibid., 20-2 (134-7).

16 Ibid., 12 (126).

17 I say ‘chiefly’ because it would be odd to speak of pervasive influence unless the impact was also wide-ranging in some way. But that idea plays a derivative role on the pervasive influence view: that the basic structure has an impact on various aspects of individuals' lives compounds the magnitude of its influence but doesn't make it a different kind of influence.

18 See Theory of Justice, 7 and 88.

19 Rawls's intuitive argument for his principle is very much in this spirit. If the basic structure conforms to the difference principle, then its impact is acceptable to those at the bottom of the expectations ladder, since things are set up to make their prospects as good as possible, and hence to work maximally to their benefit. More trivially, its impact is also acceptable to those at the top of the ladder, since they cannot reasonably complain about their lot to begin with. See Theory of Justice, 130-1.

20 Seminal discussions of this question include Sen, Amartya ‘Equality of What?’ in Choice, Welfare and Measurement (Cambridge, MA: MIT Press 1982), 353–69;Google Scholar Dworkin, Ronald Sovereign Virtue: The Theory and Practice of Equality (Cambridge, MA: Harvard University Press 2000), chap. 1–2;Google Scholar Arneson, Richard ‘Equality and Equal Opportunity for Welfare,’ Philosophical Studies 56 (1989) 77–93;CrossRefGoogle Scholar and Cohen, G.A. ‘On the Currency of Egalitarian Justice,’ Ethics 99 (1989) 906–44CrossRefGoogle Scholar (defending respectively capabilities, resources, opportunity for welfare, and access to advantage as metrics for egalitarian justice). I focus on the case of primary goods in the text because I am mainly concerned with the Rawlsian outlook, but the point I make holds across conceptions of the metric of justice. It rests on a general distinction between thinking that justice is primarily concerned with factors that determine how good individuals' prospects are according to the relevant metric, and thinking that it is primarily concerned with factors that exert controlling influence over individuals' ability to adopt and pursue a conception of the good (on which more below).

21 Theory of Justice, 79. For Rawls's reasons for focusing on primary goods rather than welfare, see notably ibid., 80-1, and Political Liberalism, 183-7. Note that the fundamental concern of justice could not be primary goods as such on a Rawls's view, given how he maps out his ‘social division of responsibility’ between society and the individual. According to that principle, society is responsible ‘for maintaining the equal basic liberties and fair equality of opportunity, and for providing a fair share of the other primary goods for everyone within this framework’; individuals, by contrast, are in charge of ‘revising and adjusting their ends and aspirations in view of the all-purpose means they can expect, given their present and foreseeable situation.’ (‘Social Unity and Primary Goods,’ in Collected Papers, Samuel Freeman, ed. [Cambridge, MA: Harvard University Press 1999], 359-87, at 371; cf. Political Liberalism, 189) On such a picture, how well different individuals end up doing in terms of primary goods will depend on matters that are within their sphere of responsibility — what specific ends they choose to pursue, how they choose to do it, and so on. I take the thought to be that, so long as society has done its share by providing fully adequate options (including, crucially, the most adequate set of options possible for the worst off), and so long as all remain above a certain threshold, how successful each person is at securing primary goods is no concern of justice.

22 To be precise, justice requires priority for the worst off for some primary goods — income and wealth, the social bases of self-respect, and what Rawls calls ‘powers and prerogatives of offices and positions of responsibility’ (‘Social Unity,’ 362). Basic liberties and what Rawls sometimes calls ‘fair opportunities,’ which are also primary goods, are to be secured equally for all citizens. Since our focus here is mostly on the application of the difference principle, we can leave these complications aside.

23 ‘Social Unity,’ 370; see also Political Liberalism, 187-8.

24 ‘Social Unity,’ 365. Rawls associates two capacities, or moral powers, with his conception of persons as free and equal: the capacity for a sense of justice and the capacity for a conception of the good (see ibid., and Justice as Fairness, 18-19). Primary goods are chosen in light of the fundamental interest individuals have in developing and exercising these two moral powers, as Rawls explicitly states when he writes (in a more precise variation of the passage I quoted above) that primary goods are ‘various social conditions and all-purpose means that are generally necessary to enable citizens adequately to develop and fully exercise their two moral powers, and to pursue their determinate conception of the good’ (Justice as Fairness, 57). The point I want to make rests specifically on a person's interest in developing and exercising her capacity for a conception of the good, so I leave aside the capacity for a sense of justice.

25 Political Liberalism, 180; emphasis added. See also ‘Social Unity,’ 365 ff.

26 I should acknowledge that I am pulling Rawls's theory in a broadly Kantian direction by stressing its reliance on an underlying idea of freedom. But I am not claiming that Rawlsians ought to embrace Kant's austere approach to political philosophy. Kant is committed to a number of radical theses that are difficult to square with the Rawlsian outlook. Most fundamentally perhaps, Kant holds that political philosophy properly construed starts from the idea that individuals have a right to freedom — a right entailing that restrictions on freedom can be justified only in terms of freedom itself. Nothing in the Rawlsian framework supports such a rigid view of the justification of coercion. On a Rawlsian view, coercion is justified when it is carried out by legitimate institutions making a credible attempt to live up to the demands of justice. The conception of persons as free and equal plays a crucial role in determining the content of the demands of justice, but the idea of freedom itself need not play a direct role in justifying each instance of coercion. On the right to freedom, see my ‘Kant on the Right to Freedom: A Defense,’ Ethics 120 (2010) 791-819; Ripstein, Arthur Force and Freedom: Kant's Legal and Political Philosophy (Cambridge, MA: Harvard University Press 2009),CrossRefGoogle Scholar chap. 2. Note that Ripstein also emphasizes the Kantian dimension of Rawls's theory in his ‘The Division of Responsibility and the Law of Tort,’ Fordham Law Review 72 (2004) 1811-44.

27 On the idea that the basic structure should be understood as a web of rules, see Theory of Justice, 47, where Rawls defines the basic structure in terms of institutions, and institutions in terms of systems of rules.

28 This way of putting the point brings out how significant it is that individuals are born subject to a given basic structure. Being born to a basic structure is not simply being born to a certain set of expected payoffs. It is being born to a set of rules and constraints that will impact at every turn how one can lead one's life. In that sense, it is very much like being born in the middle of a game that one has no choice about playing — a game whose rules are remarkably far-reaching in their application, moreover, so that they regulate much of one's life. Obviously, no ordinary game could justifiably be imposed in this way. Indeed, one reason to think that the choice of principles of justice in the original position can yield a definite result is that, by any reckoning, the basic structure must meet extraordinarily demanding conditions if imposing so extensive a set of rules on those born to a given society is to be acceptable.

29 I say ‘normally’ because, on the view I am putting forward, it is conceivable (if unlikely) that a society could have a basic structure without having a legal system properly speaking. I consider the possibility in section V.

30 For a parallel claim about the basic structure in general, see Theory of Justice, 7.

31 Theory of Justice, 207.

32 ‘Social Unity,’ 365; emphasis added. Note that, in this passage, Rawls draws a distinction between persons ‘highest-order interests’ in realizing and exercising the two moral powers and their ‘higher-order (as opposed to … highest-order) interest in advancing [their] determinate conceptions of the good’ (ibid.), a distinction that is very much in the spirit of the point I am making here.

33 As Rawls puts it, ‘in justice as fairness individuals are not identified with their actual or possible plans but are viewed rather as beings that have a capacity for forming, adopting, and changing these plans, should they be so moved; and who give priority to preserving their liberty in these matters’ (‘Fairness to Goodness,’ in Collected Papers, 267-85, at 283; cf. Theory of Justice, 131-2).

34 This way of putting the point echoes an important aspect of Rawls's argument for adopting primary goods as a metric of justice. Rawls doesn't argue that primary goods are essential for the pursuit of any conception of the good whatsoever but rather that pursuing a conception of the good will normally require primary goods. He adds, ‘That the primary goods are necessary conditions for realizing the powers of moral personality and are all-purpose means for a sufficiently wide range of final ends presupposes various general facts about human wants and abilities, their characteristic phases and requirements of nurture, relations of social interdependence, and much else’ (‘Social Unity,’ 367). We can take the thought on board for present purposes and say that, given various general facts about human nature, and given what the basic structure is like, a person has no reasonable choice but to operate through the basic structure to adopt and pursue a conception of the good.

35 Justice as Fairness, 10. Bear in mind that Rawls considers the basic structure primarily as a system of institutions working together (see the passage quoted in note 2 above). Accordingly, we need not establish that each part of the basic structure exerts controlling influence on its own. From the standpoint of justice, given that we are considering institutions that work together, the chief concern is the joint influence that the institutions exert on the lives of individuals. Matters are different when one considers structures that are not institutionally integrated in the basic structure. In such cases, one has to ask whether the structure in question exerts controlling influence either on its own (as I suggest below the family structure does) or by affecting the basic structure's workings at a sufficiently deep level (as I suggest below a sexist ethos does). Of course, we can still say, as Rawls does for the family, that a structure that is not institutionally integrated should count as part of the basic structure because it meets the relevant criterion. Nothing hangs on this, but it is important to be clear about how the criterion operates in the different cases.

36 The idea that a fundamental concern for moral arbitrariness should ground our theorizing about justice is found in Dworkin, Sovereign Virtue; Arneson, ‘Equality and Equal Opportunity for Welfare’; G.A. Cohen, ‘On the Currency of Egalitarian Justice.’ On the place of moral arbitrariness in Rawls's approach, see notably Blake, Michael and Risse, Matthias ‘Two Models of Equality and Responsibility,’ Canadian Journal of Philosophy 38 (2008) 165–99;CrossRefGoogle Scholar James, Aaron ‘Constructing Justice for Existing Practice: Rawls and the Status Quo,’ Philosophy & Public Affairs 33 (2005) 281–316,CrossRefGoogle Scholar esp. 288 ff.

37 Political Liberalism, 283.

38 Theory of Justice, 87.

39 Justice as Fairness, 55; emphasis added; cf. 124, and Theory of Justice, 156.

40 Let me emphasize that I am concerned here specifically with the kind of ethos that Cohen himself considers: an ethos that allows the talented to hold out for higher wages. I am not claiming that any ethos that capitalism might inspire would fail to exert controlling influence over individuals (what I say about a sexist ethos in the next section should make this clear). Also, as I stress in the text, I follow Cohen in considering the impact of the relevant ethos only in a society whose basic structure conforms to the difference principle. Undoubtedly, this kind of ethos is more likely to exert controlling influence if the basic structure is sufficiently unjust, since those at the bottom of the ladder will then be highly vulnerable.

41 I am thinking here of the familiar claim that the moral arbitrariness of a person's being born on one side of a border rather than the other is sufficient to justify applying something like the difference principle at the international level. See, e.g., Beitz, Charles R. ‘Cosmopolitan Ideals and National Sentiment,’ The Journal of Philosophy 80 (1983) 591–600;CrossRefGoogle Scholar Tan, Justice without Borders.

42 See Cohen and Sabel, ‘Extra Rempublicam,’ 164 ff. (responding to Nagel's ‘The Problem of Global Justice’).

43 Compare a point made by Andrea Sangiovanni, who argues against the idea that ‘only states exercise authority over property entitlements.’ He mentions the following counter-examples: ‘The TRIPS agreement sets justiciable standards for intellectual property rights … international law confers legal rights to the appropriation and sale of natural resources as well as the right to borrow internationally; international antitrust law regulates cross-border mergers and acquisitions; IMF conditionality agreements fundamentally shape domestic social and economic policy’ (‘Global Justice, Reciprocity, and the State,’ Philosophy & Public Affairs 35 (2007) 3-39, at 14n20). The last case is a particularly clear example of the kind of impact on the rule-setting workings of the basic structure that I have in mind here.

44 The controlling influence view implies, plausibly enough, that the principles of justice that apply to international actors become more stringent when they interact with individuals who are particularly vulnerable. This allows us to say that the impact of a multinational corporation on the world's poorest individuals is a primary concern of justice without endorsing a similar conclusion concerning its impact on (say) French citizens. This is in line with Rawls's claim that the international duty of assistance toward the world's poor has a ‘cutoff point’ (see Law of Peoples, 119).

45 The view put forward by Julius, A.J. in his remarkable discussion ‘Basic Structure and the Value of Equality,’ Philosophy & Public Affairs 31 (2003) 321–55,CrossRefGoogle Scholar could also have served as a foil here. I focus on Blake's view instead because it is more likely to be familiar, and because its comparative simplicity allows for a sharper contrast with the idea of controlling influence. Still, for those who are acquainted with Julius's view, it may be useful to point out some respects in which our views differ. What primarily calls for justification on Julius's view is jus framing, a quite general moral relation which obtains whenever A acts with the intention of leading B to act in A's interest, and which is problematic for the quite general reason that, when not justified, it involves treating another as a mere means (see 328). The basic structure is normatively significant, the thought goes, because how we set it up will lead people to do some things rather than others — that is, because framing can take place through the basic structure. The key question, on Julius's view, is how such framing can be justified. For complex reasons (including that individuals don't normally have a choice about being subject to the basic structure's impact), he thinks that framing that is mediated by the basic structure calls for a special kind of justification, one that is distributive and egalitarian in character; hence the role of the difference principle (see 333-5).

Two contrasts are particularly worth emphasizing here. First, on Julius's view, the basic structure's special importance is not essentially tied to the systematic impact it has on a person's ability to set and pursue ends. What matters is that how it is set up will lead some people to do one thing rather than another, and that those who are subject to this impact are given no choice about the matter. A structure need not exert the kind of systematic impact that I take to be characteristic of the basic structure to meet these two conditions (for instance, large corporations often have the kind of impact that Julius worries about without exerting controlling influence). Second, the notion of framing makes essential reference to individual intentions, an idea that has no counterpart in the view I advocate. As I see it, what calls for justification is a certain kind of collective structuring of the social world. Whether I support the basic structure because I think that it will lead others to serve my interests, or out of a sense of duty, or for whatever other reason, makes all the difference to how virtuous I am but no difference to the principles of justice that apply to the basic structure. So long as the basic structure has the relevant kind of impact, all that matters is whether we together have control over the structure in a way that makes us collectively accountable (see Theory of Justice, 88 and 479).

46 Blake, ‘Distributive Justice,’ 272. Most of the examples that Blake considers involve the threat of force, but he does not see this as necessary for coercion, as evidenced by his suggestion (at 265) that exploitative trade relationships can be coercive. Since Blake professes general agreement with Alan Wertheimer's views on the issue, however, we can probably take him to be committed to the view that some form of threat of making the patient less well-off than she ought to be is required for coercion, even if such threats need not involve the use of force. See Alan Wertheimer, Coercion (Princeton: Princeton University Press 1987).

47 See ‘Distributive Justice,’ 266-73. I take Blake's focus on autonomy to aim, among other things, to counter G.A. Cohen's contention that the coercion view imposes an arbitrarily narrow conception of justice. Cohen asks, ‘why should we care so disproportionately about the coercive basic structure, when the major reason for caring about it, its impact on people's lives, is also a reason for caring about informal structure and patterns of personal choice? To the extent that we care about coercive structure because it is fateful with regard to benefits and burdens, we must care equally about the ethi that sustain gender inequality, and inegalitarian incentives.’ (‘Where the Action Is,’ 23 [138]; see also 21 [136]) The charge is unfounded if liberals are ultimately concerned with autonomy, as Blake contends. Scheffler makes a similar point in more strictly Rawlsian terms when he writes that, ‘given the status of individuals as free and equal, the establishment of coercive institutions poses a special justificatory problem’ (‘Is the Basic Structure Basic?’ 124).

48 ‘Distributive Justice,’ 274. For Scanlon's, T.M. view, see What We Owe to Each Other (Cambridge, MA: Harvard University Press 1998).Google Scholar

49 This cannot be a strict identity, since the law of taxation is also relevant to the allocation of goods (as Blake recognizes; see ‘Distributive Justice,’ 276). 50 It is of course the basic structure as a whole that is to be arranged to benefit maximally the worst off, not the institutions of property and contract law considered in isolation. On this point, see Ripstein, ‘The Division of Responsibility and the Law of Tort.’

51 See ‘Distributive Justice,’ 270, where he quotes approvingly Rawls's conception of rational autonomy as the capacity to ‘form, to revise, and to pursue a conception of the good, and to deliberate in accordance with it’ (Political Liberalism, 72).

52 Blake, ‘Distributive Justice,’ 278. Blake adds, ‘All the forms of legal rules we use are ultimately backed up with coercive measures that implicate the liberal principle of autonomy’ (ibid.). He goes on to quote approvingly Robert Cover's view that ‘every judicial act is an act of implicit violence’ (cited at 279; see Cover, Robert M. ‘Violence and the Word,’ Yale Law journal 95 [1986] 1601–30CrossRefGoogle Scholar).

53 See Hart, H.L.A. The Concept of Law (Oxford: Clarendon Press 1961), 27–9.Google Scholar Interestingly, Blake takes this passage to support his position: he insists that Hart sees contract as ‘a limited grant of (coercive) legislative power’ (‘Distributive Justice,’ 277). This contradicts the spirit of the passage, which aims to stress the difference between criminal and tort law on the one hand, and contract law on the other. Thus Hart insists that ‘[l]egal rules defining the ways in which valid contracts or wills or marriages are made do not require persons to act in certain ways whether they wish to or not. Such laws do not impose duties or obligations. Instead, they provide individuals with facilities for realizing their wishes, by conferring legal powers upon them to create, by certain specified procedures and subject to certain conditions, structures of rights and duties within the coercive framework of the law’ (The Concept of Law, 27-8; see also 35-42). Blake would need to say more to explain exactly how such a system is in tension with individual autonomy.

54 See Abizadeh, ‘Cooperation, Pervasive Impact, and Coercion,’ 348-52, as well as his ‘Democratic Theory and Border Coercion: No Right to Unilaterally Control Your Own Borders,’ Political Theory 36 (2008) 37-65. For an insightful critical discussion of Abizadeh's position, see Miller, David ‘Why Immigration Controls Are Not Coercive: A Reply to Arash Abizadeh,’ Political Theory 38 (2010) 111–20.CrossRefGoogle Scholar I should stress that, when I speak of borders in what follows, I mean to refer specifically to their function in keeping foreigners out (call this external borders). Borders also determine when citizens can leave their own country (call this internal borders). On the view I defend, there is a fundamental difference between the two aspects. My country's external borders lie outside the basic structure to which foreigners are otherwise subject. By contrast, its internal borders are an integral part of the domestic legal system and hence of the basic structure. Internal borders are therefore subject to the same standards of justification as any other restriction that the state may place on its citizens’ freedom of movement, namely, the standards of justification that apply to the basic structure as a whole (which, on Rawls's view, include an explicit protection for freedom of movement; see, e.g., Justice as Fairness, 58). The point is in keeping with the common view that a country's locking in its citizens is much harder to justify than its keeping out foreigners.

55 For Blake's discussion of borders, see ‘Distributive Justice,’ 280n. Borders obviously have ‘different aims and purposes’ than does the private law, so any principles applying to them will have to take into account their ‘peculiar nature and special requirements’ (Justice as Fairness, 11). The question remains, however, whether borders raise problems of justification that are of the same nature as those raised by the basic structure. If they do, then they are part of the primary subject of justice; we just need to figure out what principles are appropriate for them.

56 Going back to an idea introduced above (see section III), you could say that being born to a given basic structure is like being born to a particular game whose rules you have no choice but to follow, while being born subject to the borders of other states is like not being allowed to take part in other games, some of which may promise higher yields than the one you are currently playing. My claim is that the former situation is subject to more stringent standards of justification than the latter, especially if the rules of the game determine how you can lead your life more or less at every turn (as the rules of the basic structure do).

57 The thought that borders raise different concerns with respect to different individuals should not be controversial. It seems overwhelmingly plausible that the restrictions France places on immigration raise fundamental concerns of justice with respect to citizens of poor African countries, but not with respect to American citizens, even if the restrictions deprive the latter of options that they greatly value. A conception of the good that requires living specifically in France for its pursuit has to be, from the point of view of justice, on a par with one that requires plovers’ eggs. Such ‘expensive preferences’ cannot ground claims of justice on a Rawlsian view, since it is the individual's responsibility to revise such ends in light of the means she can expect to have, not others’ responsibility to supply her with the needed means (see again the principle of social division of responsibility in ‘Social Unity,’ 371). On expensive preferences, see Dworkin, Sovereign Virtue, chap. 1.

58 See Law of Peoples, 8-9 and 39n.

59 Sangiovanni, ‘Global Justice, Reciprocity, and the State,’ 10-11.

60 Note that the question here is not whether morality, or perhaps a sincere acceptance of the justification for the difference principle, may require one to promote an egalitarian ethos in one's society (for the latter suggestion, see Shiffrin, Seana Valentine ‘Incentives, Motives, and Talents’ Philosophy & Public Affairs 38 [2010] 111–42CrossRefGoogle Scholar). The issue is whether the presence of an inegalitarian ethos is a primary concern of justice, where this is taken to warrant certain types of institutional interventions that infringements of morality more broadly construed do not.

61 This means that, unlike the different branches of the legal system, the family is not institutionally integrated into the basic structure (see note 35 on this). Hence, we must consider the family's impact in isolation, as we did with foreign borders.

62 ‘Where the Action Is,’ 22 (137).

63 See notably Okin, Susan Moller Justice, Gender, and the Family (New York: Basic Books 1989).Google Scholar I do not mean to minimize the importance of Okin's concerns, but it is worth noting the subsidiary role that children play in her explanation of the family's relevance to justice. She emphasizes that a just family is necessary for the healthy moral development of children (see 99-100). Otherwise, she mostly focuses on child care (including women's role as primary care givers) and on questions of gender identification. These are obviously important issues, but I am not sure that they capture the main difference between the family and other private associations. Rawls himself was largely convinced by Okin's arguments; see ‘The Idea of Public Reason Revisited,’ in Collected Papers, 573-615, at 595-601.

64 If that were the case, the treatment of the family would fall under nonideal theory, which Rawls never suggests. On the distinction between ideal and nonideal theory, see Theory of Justice, 216-18.

65 Note by contrast that, however demanding they might be, religions normally count as voluntary associations for political purposes. See Political Liberalism, 221-2, and ‘The Idea of Public Reason Revisited,’ 599n68.

66 See ‘The Idea of Public Reason Revisited,’ 597.

67 Rawls tends to be vague on this issue, saying both that principles of political justice set constraints on the family's internal life but also that, on this general matter, ‘society has to rely on the natural affection and goodwill of the mature family members’ (‘The Idea of Public Reason Revisited,’ 598). I suspect that his theory has more definite implications than this suggests for the principles that should apply within the family (which would presumably include some suitable equivalent of the principle of fair equality of opportunity), but I cannot pursue the matter here.

68 For extremely helpful comments and suggestions, I am grateful to Brian Hutler, A.J. Julius, Michael Kessler, Doug MacKay, Jonathan Peterson, Arthur Ripstein, Patrick Turmel, Stephen J. White, and two anonymous reviewers for this journal. I also wish to acknowledge my debt to UCLA's Law and Philosophy Program, which provided generous financial support and an exceptionally stimulating environment for a crucial portion of the work on this project.