¹ Rawls, John, The Law of Peoples, with ‘The Idea of Public Reason Revisited’ (Cambridge, MA, 1999)Google Scholar, hereinafter ‘LoP’. Page references in the text are all to this work.

² See Rawls, Political Liberalism, expanded edn. (New York, 2005)Google Scholar.

³ The conditions of reasonableness for a people are distinct, then, from Rawls's conditions of reasonableness for a liberal citizen (for which see Political Liberalism, pp. 48–66). The term ‘reasonable’ always refers herein to Rawls's notion of global reasonableness, unless otherwise noted. I analyse this notion more closely in section II.

⁴ The ideal of GPR also calls on liberal citizens to ‘repudiate government officials and candidates for public office who violate the public reason of free and equal peoples’ (LoP, p. 57). My focus, however, is on GPR as employed at the global level.

⁵ I use the term ‘indeterminacy’ advisedly, in keeping with Gerald Gaus's influential distinction between ‘indeterminacy’ and ‘inconclusiveness’ in public reason. For explanation, see the text around n. 27, below.

⁶ See e.g. the overview of that objection in Jonathan Quong, ‘Public Reason’, Stanford Encyclopedia of Philosophy, <http://plato.stanford.edu/entries/public-reason/> (2013).

⁷ See my ‘Public Reason and Prenatal Moral Status’, Journal of Ethics 19 (2015), pp. 23–52, at 49.

⁸ For such an alternative, see e.g. Gerald Gaus's remarks on the worldwide application of his ‘convergence’ conception of public reason, in The Order of Public Reason (Cambridge, 2011), pp. 470–9. As it happens, in a new work, Christopher Eberle argues that convergence liberalism also carries unwelcome implications regarding war; see his ‘War and Respect’, in Rawls and Religion, ed. Tom Bailey and Valentina Gentile (New York, 2015), pp. 29–51, at 43–7. For doubts about whether Eberle's war-based critique of convergence succeeds, see my ‘Review of Rawls and Religion’, Notre Dame Philosophical Reviews, <https://ndpr.nd.edu/news/59524-rawls-and-religion/> (2015).

⁹ Thus, Rawls writes (LoP, p. 85): ‘I have argued that both reasonably just liberal and decent hierarchical peoples would accept the same Law of Peoples. For this reason, political debate among peoples concerning their mutual relations should be expressed in terms of the content and principles of that law.’

¹⁰ Political Liberalism, pp. 11–15.

¹¹ The caveat that the relevant features of the global public culture be shared is vital. Some parts of the current global public culture are excluded from GPR, because they would not be endorsed by all reasonable peoples. Thus, for instance, while the Universal Declaration of Human Rights is part of this culture, not all of its provisions provide public reasons, according to Rawls, since some decent peoples reject them. On this, see LoP, p. 80 n. 23.

¹² For close analysis of the basis of Rawls's understandings of international toleration and reasonableness, see Porter, Thomas, ‘Rawls, Reasonableness, and International Toleration’, Politics, Philosophy & Economics 11 (2012), pp. 382–414 CrossRefGoogle Scholar.

¹³ This conclusion might be queried, on grounds that Rawls claims at one point (LoP, p. 65) that there is a human right to ‘formal equality as expressed by the rules of natural justice (that is, that similar cases be treated similarly)’. On whose authority, however, are two cases to be judged relevantly similar? I believe that Rawls can only coherently claim that formal equality is a human right if that authority is understood to rest with the society in question. For if it were not so interpreted, the right would immediately lead to equal liberal citizenship, ruling out the conceptual possibility of a non-liberal, decent people that respects human rights. Yet, if formal equality is interpreted as I suggest, it rules out merely arbitrariness and corruption in the enforcement of human rights, not the sort of systematic inequality described in the text.

¹⁴ Samuel Freeman, Rawls (Abingdon, 2007), p. 430.

¹⁵ See LoP, p. 69 for the claim that decent and liberal peoples alike endorse the global original position as fair. Technically, Rawls describes two global original positions, in which liberal and decent peoples separately appraise the Law of Peoples. This detail becomes relevant in section IV.

¹⁶ For the view that a group's failure to fulfil its duties of global distributive justice presents its victims with just cause for war, see Fabre, Cécile, Cosmopolitan War (Oxford, 2012), ch. 3CrossRefGoogle Scholar.

¹⁷ See e.g. Martin, Rex, ‘Just Wars and Humanitarian Interventions’, Journal of Social Philosophy 36 (2005), pp. 439–56, at 440CrossRefGoogle Scholar.

¹⁸ As I argue in section V, this second question is not only one that Rawls omits to discuss, but one that GPR is incapable of addressing.

¹⁹ See e.g. his ‘A Defense of the Traditional War Convention’, Ethics 118 (2008), pp. 464–95. Note that the agreement of individuals as well as of peoples plays a role in Benbaji's case for a symmetrical war convention. In this respect, among others, he departs from the Rawlsian framework.

²⁰ An anonymous reviewer proposes that equality between peoples might instead be achieved by ‘levelling down’ in setting the terms of the exemption – i.e. by stipulating a uniformly low ceiling on the number of civilians which a people, irrespective of size, is permitted to target as a means of saving itself. Given how widely the populations of peoples will differ, this ceiling would have to be very low to deny any people permission to cause a greater evil in exercising its exemption. And it is difficult to see why, in the original position, rational parties would endorse an exemption that is so tightly constrained. If they would, however, GPR would be subject to an objection that is the converse of the one advanced in the text: namely, that the exemption carved out is too strict. For the exemption would deny more populous peoples the ability to save themselves by targeting civilians in excess of the ceiling, even if doing so would clearly be the lesser evil. The fundamental point here is that contractualist reasoning militates against an exemption that is appropriately sensitive to the numbers saved and killed. Added to this, the envisaged move is subject to the general problem – discussed in the next section – that liberal and decent peoples disagree over the extent to which the killing of different groups constitutes an evil, and would therefore seem incapable of agreeing in principle how many civilian casualties the survival of a people should be set as worth.

²¹ Walzer, Michael, Just and Unjust Wars, 4th edn. (New York, 2006), p. 254 Google Scholar.

²² This definition describes only half of what, according to some philosophers, proportionality is about – namely what McMahan calls ‘wide proportionality’ (as contrasted with ‘narrow proportionality’, which concerns whether the harm inflicted upon an individual exceeds that to which (s)he is liable). I focus on wide proportionality because, unlike narrow proportionality, it is a shared concern of orthodox and so-called ‘revisionist’ just war theories. I believe, however, that the argument in the text would also apply to narrow proportionality, mutatis mutandis. For McMahan's distinction between narrow and wide proportionality, see e.g. his Killing in War (Oxford, 2009), pp. 21ff.

²³ Thus, I abstract from e.g. the problem of whether the ‘goods’ that count in (wide) in bello proportionality can be understood in a morally neutral way, such that proportionality can be satisfied by belligerents without just(ified) war aims. An anonymous reviewer invites consideration of whether the indeterminacy described in this section remains when the harms of war are weighted – as in the traditional doctrine's understanding of in bello proportionality – against the neutral currency of military advantage, or contribution to military success. I believe so, since the question of when belligerents are entitled to count their victory as a good is separate from the question of how heavily that good weighs in the scales vis-à-vis harms to the enemy, given their relative moral standing.

²⁴ I am grateful to Jeff McMahan for that point.

²⁵ I am grateful to Jonathan Quong for suggesting this line of reply, and to several participants at a seminar of the Centre for Ethics, Law and Public Affairs at Warwick for pressing me further on it.

²⁶ Schwartzman, Micah, ‘The Completeness of Public Reason’, Politics, Philosophy & Economics 3 (2004), pp. 191–220, at 209–14CrossRefGoogle Scholar.

²⁷ See Gaus, Gerald, Justificatory Liberalism (New York, 1996), pp. 151–8Google Scholar.

²⁸ See Political Liberalism, e.g. at p. xlv.

²⁹ See Wenar, Leif, ‘Why Rawls is not a Cosmopolitan Egalitarian’, in Rawls's Law of Peoples, ed. Martin, Rex and Reidy, David A. (Oxford, 2006), pp. 95–113, at 104CrossRefGoogle Scholar.

³⁰ Wenar, ‘Why Rawls’, p. 104.

³¹ Wenar, ‘Why Rawls’, p. 104, emphasis in original.

³² Wenar, ‘Why Rawls’, p. 106.

³³ Wenar, ‘Why Rawls’, p. 103, emphasis in original.

³⁴ Wenar, ‘Why Rawls’, p. 103.

³⁵ Wenar, ‘Why Rawls’, p. 110.

³⁶ Wenar, ‘Why Rawls’, p. 111.

³⁷ Wenar contends (at pp. 108–9) that, since any viable theory of global justice must assign states a right to defend their territories, principles of right conduct in war must accordingly identify who may permissibly be killed partly on the basis of political and territorial affiliation. He refers in this context to ‘principles for individuals’, implying that he may think the Law of Peoples after all includes – or could be extended to include – such principles. Yet this is not so, for the reasons given in section II.

³⁸ See Walzer, Just and Unjust Wars, pp. 58ff.

³⁹ I am grateful for feedback received on earlier versions of this article from audiences in Braga, Warwick, Birmingham, Kent and Boulder, Colorado, and for written comments from Matthew Clayton, Collin O'Neil, Jonathan Quong, Scott Sturgeon and an anonymous reviewer for this journal. I am especially thankful for extensive feedback and encouragement from Jeff McMahan. Finally, I record my thanks to the Leverhulme Trust for the Early Career Fellowship that enabled this research.