The Justice of Peace Treaties

The Journal of Political Philosophy: Volume 20, Number 3, 2012, pp. 273–292 The Justice of Peace Treaties* Daniel Schwartz I. INTRODUCTION U NDER international law a treaty between two or more states or international organizations is valid if it meets the requirements stipulated by treaty law. This law, emerging mainly from international conventions, establishes among other things who can conclude treaties and under which circumstances. Article 52 of the Vienna Convention of the Law of Treaties of 1969 reads as follows: ‘a treaty is void if its conclusion has been procured by the threat or use of force in violation of the principles of international law embodied in the Charter of the United Nations.’1 By withholding validity from extorted treaties the convention creates a disincentive against the use of this particular form of wrongful coercion with the aim of keeping the level of such coercion below the level that would obtain under a more liberal regime. The pursuit of this apparent benefit comes at a cost. States forfeit at least partially their access to treaties that could help them palliate the effects of being caught in a situation in which the alternative to a coerced treaty is even worse. As with other conventions that produce benefits and impose burdens it seems pertinent to examine whether treaty law distributes benefits and burdens fairly. I argue that the isolated introduction of stricter international treaty law increases the risk of weaker countries becoming victims of predatory violence while at the same time reducing such risk for middle-weight countries. The distribution of the benefits and burdens resulting from restrictions in access to coerced treaties would not be agreed to by countries deciding as free and equals on the terms of their cooperation. Therefore it is unfair. A fair distribution of the burdens imposed by treaty law can only be achieved against a background of transfers in deterrent capacities. So long as this does not happen on the required scale it is unfair to withdraw validity from coerced peace treaties. Further I argue that recognizing coerced peace treaties if done for the right reasons need not involve complicity with the coercer. *In writing this paper I have benefitted from comments and thoughts by Itzhak Benbaji, Nir Eyal, Esteban Klor, Shlomi Segall, Yonatan Shemmer and participants of the IR Colloquium at Hebrew University and the Political Science Colloquium at the University of Tel Aviv, all of whom I thank. I owe also gratitude to two anonymous reviewers of JPP and to its editor, Robert Goodin, for very useful comments, corrections and suggestions. 1 United Nations, Treaty Series (New York: UN, 1969), vol. 1155, p. 331. © 2011 Blackwell Publishing Ltd., 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. doi: 10.1111/j.1467-9760.2010.00390.x 274 DANIEL SCHWARTZ The article is organized as follows. Section I locates my project in relation to recent developments in relevant scholarly fields. Section II surveys precedents of extorted treaties and extricates a principle that seems to explain the non conferral of validity to such treaties. Section III describes the costs potentially paid by countries which have been denied access to extorted treaties and explains how treaty validation controls access to treaties. Section IV argues that under present non-ideal conditions restrictive treaty conventions impose a disproportionately heavier burden on weaker countries compared to stronger ones (stronger countries may in fact benefit from such conventions). In Section V, I argue that this makes such conventions distributively unjust. Finally, Section VI scrutinizes and defuses a powerful objection to the conferral of validity to extorted treaties: that it involves complicity with international extorters. One would hope to find explicit discussion of the justice of extorted treaties in the jus post-bellum literature, which examines the justice of the termination of violence and its aftermath.2 Yet, its focus is on the moral duties of a just victor who is unconstrained by contractual duties towards the defeated. Brian Orend, for example, provides a set of recommendations called ‘settlement principles’ addressed to ‘a just state, seeking to terminate a just war successfully’.3 There is, however a significant ambiguity in Orend’s use of the phrase ‘peace settlement’: it seems to cover both situations in which the victor imposes peace terms unilaterally and those that result from a negotiated treaty between the parties.4 Most interstate wars are fought by countries that do not have the power or the will to occupy the entire territory of the enemy nor to dissolve its military and political structures. The aggressor aims at improving his position at the negotiating table. Each side knows that some form of peace treaty, armistice, or agreed suspension of hostilities (tacit or explicit) will put an end to the violence.5 Hence there is ample reason to study the justice of treaties and related contractual instruments. 2 I have in mind the following articles: Brian Orend, ‘Justice after war’, Ethics & International Affairs, 16 (2002), 43–56 and his ‘Jus post bellum: the perspective of a just-war theorist’, Leiden Journal of International Law, 20 (2007), 571–91; Robert E. Williams, Jr. and Dan Caldwell, ‘Jus post bellum: just war theory and the principles of just peace’, International Studies Perspectives, 7 (2006), 309–20; Gary J. Bass, ‘Jus post bellum’, Philosophy & Public Affairs, 32 (2004), 384–412; Carsten Sahn, ‘“Jus ad bellum”, “jus in bellum” . . . “jus post bellum”? Rethinking the conception of the law of the armed force’, European Journal of International Law, 17 (2007), 921–43; Richard P. DiMeglio, ‘The evolution of the just war tradition: defining jus post bellum’, Military Law Review, 186 (2005), 116–63; Louis V. Isaiello, ‘Jus post bellum: the responsibilities of victors in war’, Naval War College Review, 57 (2004), 34–52; and the recent book by Carsten Stahn, Jann K. Kleffner, eds, Jus Post Bellum: Towards a Law of Transition from Conflict to Peace (The Hague: Asser, 2008). 3 Orend, ‘Justice after war’, p. 55. 4 Orend, ‘Jus post bellum’, pp. 575–6 says, ‘[T]here is no guarantee that if you fought justly, for the sake of a just cause, then you will automatically impose a just set of peace terms upon your vanquished enemy.’ From this I gather that, for Orend, dictated peace terms constitute ‘settlements’ as do the ‘peace treaties’ discussed under the same heading in p. 77 of his ‘Justice after war’. 5 For the differences between these various instruments see Yoram Dinstein, War, Aggression, Self-Defense, 2nd edn (Cambridge: Cambridge University Press, 1994), pp. 35–58. THE JUSTICE OF PEACE TREATIES 275 The recent work of Allen Buchanan seems more directly relevant to the question pursued here. Buchanan holds that distributive justice offers a standard from which to judge international institutions. Treaty-making conventions such as the Vienna Convention of the Law of Treaties are international institutions. Indeed, as Paul Reuter has observed, ‘the law of treaties could . . . be seen as part of the “constitution” of international society’.6 For John Rawls, contractual instruments certainly belong to the domestic basic structure, which is the primary subject of justice: ‘the principles of justice apply to the practice of promising in the same way that they apply to other institutions’.7 If there is such a thing as ‘the global basic structure’, as Charles Beitz, Thomas Pogge, and Allen Buchanan have it, or a ‘basic structure of the Society of Peoples’ as Rawls has it,8 surely treaty law belongs there. While Buchanan does not examine the justice of conferral of validity to treaties, he does examine the justice of conferrals of recognition of statehood.9 The two questions are related, however, for being recognized as a state is often a precondition of being able to enter international treaties. The connection cuts both ways: it is a common concern that allowing a community to enter international transactions indirectly implies recognition of statehood. Partly on these grounds, the U.S., Australia, New Zealand, and Canada voted against the UN Declaration on the Rights of Indigenous Peoples, adopted by the General Assembly in 2007, which states that ‘treaties, agreements and other arrangements between States and indigenous peoples are properly matters of international concern and responsibility’.10 Like Buchanan’s work, mine is intended as a contribution to non-ideal theory.11 In the ideal scenario, one of full compliance with principles of justice, there is no such thing as unjust coercion. In a second best scenario, some countries engage in unjust coercion, but other countries are willing to provide military assistance to victims, thereby diminishing the frequency of such threats. In the third scenario, coerced countries cannot for the most part count on 6 Paul Reuter, Introduction to the Law of Treaties, trans. J. Mico and P. Haggenmacher (London: Kegan Paul, 1995), p. 29. 7 John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), p. 345. 8 Thomas Pogge, Realizing Rawls (Ithaca, NY: Cornell University Press, 1989), pp. 238–9. John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999), p. 113. Charles Beitz, Political Theory and International Relations (Princeton, NJ: Princeton University Press, 1979), p. 151. 9 Allen Buchanan, Justice, Legitimacy and Self-Determination: Moral Foundations of International Law (Oxford: Oxford University Press, 2004). Buchanan discusses duress, consent and treaties at pp. 303–4 but only as prelude to his discussion on whether state consent validates international norms. 10 UN Declaration for the Rights of Indigineous Peoples, 13 September, 2007, <www.un.org/ esa/socdev/unpfii/en/drip.html> (last accessed 20/10/2010). The explanation given for the votes against can be found in UN Department of Public Information, ‘General Assembly Adopts Declaration on Rights of Indigineous Peoples’, 13 September, 2007, <www.un.org/News/ Press/docs/2007/ga10612.doc.htm> (last accessed 20/10/2010). 11 Buchanan, Justice, Legitimacy and Self-Determination, p. 61. 276 DANIEL SCHWARTZ external military assistance and must decide whether to enter coerced treaties or resist. Here I appraise treaty conventions in the context of the third, non-ideal scenario. It is beyond the purview of this article to offer an original theory on whether promises, contracts, and treaties signed under duress generate moral duties that should be honored. Detailed discussions of this question can be found in the work of Alan Wertheimer, Gregory Kavka, John Deigh, and David Owens.12 I refer to these discussions, as well as the insights and conclusions of others, throughout the paper. The question on whether positive municipal law should enforce coerced contracts has been discussed recently, for example in papers by Oren Bar-Gill and Omri Ben Shahar and by Claire Finkelstein.13 A central concern of these works is the reduction of harm to victims. In their view in some circumstances this concern recommends conferring validity to such contracts. While my argument about coerced contracts in international law is sympathetic to this view, it brings on board an additional consideration, that of the fairness of the distribution of such harms. Possibly my argument applies in some revised form to the domestic level (after all this is the prime site of distributive justice) but I shall not pursue the way it does within the confines of this paper. II. REASONS FOR WITHHOLDING VALIDITY FROM EXTORTED TREATIES The two traditionally posited causes of contract invalidity are coercion and fraud. Arguably, treaties are a form of contract.14 The innovation of Article 52 of the Vienna Convention consists in invalidating treaties resulting from threats to states, rather than only to the representatives of states in charge of negotiating the treaty. That threats to state representatives void treaties was a classical doctrine of international law preceding the Vienna Convention and its sources of inspiration, such as the Harvard Draft of 1935.15 At first glance the Vienna Convention requires too much and too little. Too much, because virtually all post-war peace treaties are entered into against the background of an implicit threat of resuming war (either reciprocal or 12 Alan Wertheimer, Coercion (Princeton, NJ: Princeton University Press, 1987). Gregory S. Kavka, Hobbesian Moral and Political Theory (Princeton, NJ: Princeton University Press, 1986), p. 391–8. John Deigh, ‘Promises under fire’, Ethics, 112 (2002), 483–506. David Owens, ‘Duress, deception and the validity of a promise’, Mind, 116 (2007), 293–315. 13 Oren Bar-Gill and Omri Ben-Shahar, ‘Credible coercion’, Texas Law Review, 83 (2007), 718–80, esp. pp. 753–4, 779–80. Claire Finkelstein, ‘Contracts under coercion: should one keep an agreement with a robber?’ Unpublished manuscript. 14 Perhaps not all treaties as some international lawyers distinguish between contractual treaties and law-making treaties. Even if the distinction is accepted, it is clear that peace treaties belong to the former category. See Reuter, Law of Treaties, pp. 26–8. 15 See Ibid., p. 179. James W. Garner et al., ‘Codification of international law. Part III: law of treaties’, American Journal of International Law, 29 (1935), Suppl., pp. 1152–4. THE JUSTICE OF PEACE TREATIES 277 unilateral). Are we then to treat these treaties as invalid? Too little, because the Vienna Convention only invalidates treaties signed under threats of force, it does not invalidate treaties signed under unarmed threats, such as economic and political sanctions, nor does it void treaties in which the different bargaining powers of countries came to be acquired through unjust means.16 In addition, it is seldom noted that the convention does not void treaties signed as a consequence of a coercive offer—assuming these indeed exist—such as Opulenta lending money to bankrupt Povera on the condition that it sign a treaty giving Opulenta far reaching commercial concessions.17 Whether the Vienna Convention requires too much or too little is a tempting question, but, as I hope will emerge from this article, it is the wrong question to ask. As will be discussed later, we should not focus on the demandingness of the conditions for conferral of validity but rather on integrating them into overall arrangements which determine the distribution of deterrent capabilities across the board. Possible examples of extorted treaties (whether the coercion involved is lawful or not) are the U.S.-Japan peace treaty of 1951, the 1938 Munich agreement by which the Sudetenland was transferred from Czechoslovakia to Germany and, paradigmatically, the post WWI treaties of Versailles and of St. Germain. If we look further back in time we have the ‘unequal treaties’ typically (but not exclusively) imposed by Western powers on China, Japan, Siam, and Korea, and the ‘removal treaties’ between the US and the various Native-American nations. It could be said that, in all these cases, might became right. More recent cases of treaties denounced by one of its parties as invalidated by threats of force at the time of conclusion include those between the USSR and Turkey (1946), the USSR and Czechoslovakia (1968), Iran and Iraq (1969–1975), Morocco/Spain and Algeria (1975), Nicaragua and Colombia (1979–80), and NATO and Yugoslavia (1999).18 The reason coercion invalidates contracts need not be—as is often supposed—that it renders our entering contracts involuntary, nor that it destroys choice (these effects of coercion are frequent but not necessary).19 Some forms of 16 As Judge Padilla Nervo argued in the Fisheries Jurisdiction Case at the International Court, ‘A big power can use force and pressure against a small nation in many ways. . . . [C]ertain “Notes” delivered by the government of a strong power to the government of a small nation, may have the same purpose and the same effect as the threat of force. . . . There are moral and political pressures which cannot be proved by so called documentary evidence, but which are in fact indisputably real and which have, in history, given rise to treaties and conventions claimed to be freely concluded and subjected to the principle: pacta sunt servanda’. ‘Dissenting opinion’, at §§ 46–7, 12 Fisheries Jurisdiction Case (United Kingdom v. Iceland), Merits, 1974 ICJ Rep, 3 (25 July, 1974). Note that according to prevalent interpretation ‘threats of force’ are intended to cover also demonstrations of force. See Nikolas Stürchler, The Threat of Force in International Law (Cambridge: Cambridge University Press, 2007), p. 113. 17 On coercive offers see Wertheimer, Coercion, pp. 202–26. 18 These cases are taken from Stürchler, Threat of Force in International Law, p. 285. 19 On this, see Wertheimer, Coercion, pp. 30–3. 278 DANIEL SCHWARTZ coercion leave room for choice. Peter is blocking John’s path, forcing John to turn either left or right. That choice is genuine, even if Peter has coercively worsened John’s ‘choice situation’. The fathers of international law such as Grotius, Pufendorf, Burlamaqui, Barbeyrac, Wolff and Vattel paid extensive attention to the question of coerced peace treaties, armistices and truces (as did Hobbes, outside the context of international war).20 Pufendorf and some contemporary theorists make clear that what morally voids the treaty is not coercion per se. What matters, they say, is unjust coercion.21 We might take ‘unjust’ coercion to mean a worsening of a person’s choice situation brought about by an act that violates the moral law and/or morally binding positive law. If A unjustly coerces B into signing a contract, even if it is only mild unjust coercion (say the threat of doing something unlawful but only moderately harmful), then this contract is invalid under municipal law. Pufendorf observes that if we regard promises extracted under unjust coercion as binding ‘we may as well say that a Law, which forbids thieves may, at the same time, reckon their trade and practice among honest ways’.22 His view is not that coerced treaties are naturally void, not even that treaties resulting from unjust coercion are void, but that they should not be allowed to generate rights in the coercer (he allows them to cause rights in third parties). Pufendorf puts it thus: ‘[S]uch Fear alone invalidates, and destroys an obligation, as is caused by some vice or imperfection in the other Party, rendering him incapable of acquiring a Right.’23 The thought is that if the law validates the intended effect of the coercion, namely the promise that is so extracted, it validates also the means through which it has been obtained. However, Pufendorf says, an act cannot be both morally wrong and morally permissible. 20 Hugo Grotius, The Rights of War and Peace, ed. Richard Tuck and Jean Barbeyrac, trans. John Morrice (Indianapolis: Liberty Fund, [1625] 2005), bk II., ch. 11, sect. 7, §2, p. 712 (and criticism of Barbeyrac in note 4, p. 713). Samuel Pufendorf, The Law of Nature and Nations: Or, A General System of the Most Important Principles of Morality, Jurisprudence, and Politics, trans. Basil Kennet (London: J. & J. Bonwicke, 1729), bk III, ch. 6, §9–13, pp. 274–82. Jean-Jaques Burlamaqui, The Principles of Natural and Politic Law (Indianapolis: Liberty Fund, [1747] 2006), vol. 2, pt IV, ch. 10, pp. 325–34. Thomas Hobbes, Leviathan (Oxford: Oxford University Press, [1651] 1996), bk I, ch. 14, §27, p. 92. For Emer de Vattel and Christian Wolff, see Randall C. H. Lesaffer, ‘A Schoolmaster abolishing home-work? Vattel on peacemaking and peace treaties’, <http://ssrn.com/abstract= 1091170> (last accessed October 20, 2010). Pufendorf criticizes Grotius: ‘It must be said, that the Extortions and Ravages of unjust War can never be retained or defended by Conscience. And it is beyond Dispute, than an unjust Conqueror is obliged to make Restitution, so I cannot think it appears by any Law of Nations, that it is unlawful to plead an Exception of Fear against him. And though there were such a Law I confess I do not apprehend what injury it would be to the common Quiet and Security of Mankind, if no Notice is taken of it’ (The Law of Nature and Nations, VIII.8.1, p. 858). 21 See Owens, ‘Duress’, p. 298. Deigh, ‘Promises’, p. 489, subjects to scrutiny the view that the problem is not coercion as such but the offensiveness of some forms of coercion. But here ‘offensiveness’ is something other than injustice; it is lack of respect to the self-ruling capacity of the coerced, and the feeling offended by the coerced caused by this disrespect. 22 Pufendorf, The Law of Nature and Nations, III.6.10, pp. 277–8. 23 Ibid. THE JUSTICE OF PEACE TREATIES 279 I will not discuss here the contradiction denounced by Pufendorf. Instead I want to invoke a different principle available to him. It says that we should prevent unjust coercers from reaping the benefits of their injustice. In some judicial systems this is known as the ‘forfeiture rule’. As Lord Justice Fry wrote in the well known case of a son who stood to inherit the estate of his father, whom he murdered, ‘[I]t appears that no system of jurisprudence can with reason include amongst those rights which it enforces, rights directly resulting to the person asserting them from the crime of that person.’24 The principle behind the forfeiture rule generates duties on third parties to neutralize the benefits derived from unjust coercion. A similar thought is expressed by Paul Reuter: ‘[I]nvalidity can . . . hardly be regarded as a result of vitiated consent; it is rather a sanction of an international offence deemed sufficiently important by some to fall within the category of “international crimes” entailing special penalties for the individuals and States that have committed them.’25 In the following I use ‘coercive threats’ rather than the term ‘duress’. One predominant understanding of ‘duress’ is that which is frequent in criminal defenses: as excusing or justifying the carrying out of an unlawful action. This duress consists in overwhelming coercion such that someone with ‘reasonable firmness of will would not be able to resist.’26 ‘Coercive threat’ here covers also weaker forms of coercion. Given some tendency to associate ‘duress’ with overwhelming coercion, I will avoid the term in what follows. III. REASONS FOR CONFERRING VALIDITY TO EXTORTED CONTRACTS Theorists supporting the validation of coerced peace treaties do so in order to limit the extent of the destruction of war and in general to foster stability. We can neutralize part of the benefits derived from unjust coercion by reducing the capacity of the coercer to extort pacts. Yet we should ask ourselves, shouldn’t we sometimes, in the interest of the coerced, abstain from reducing such capacity? Consider Burlamaqui’s poignant observation that there are only two ways of obtaining peace. The first is, the total and entire destruction of our enemy; and the second is, the entering into articles of treaty with him. If therefore treaties and compacts, made between enemies, were not in themselves sacred and inviolable, there would be no other means of procuring a solid peace, than carrying on the war to the utmost extremity, and to the total ruin of our enemies.27 24 Cleaver vs. Mutual Reserve Fund Life Association (1892) 1.Q.B. 147 (Canada). Reuter, Law of Treaties, pp. 180–1. 26 Wertheimer, Coercion, p. 154. 27 Burlamaqui, The Principles of Natural and Politic Law, IV.10.3, p. 328. Or, as Deigh, ‘Promises’, p. 486 puts it, ‘Enemies in a war, whose very relations consist in attempting to change each other’s behaviour through force and violence, may find it useful, as a way of reducing the destructiveness of their hostilities, to be able to make one another promises that they can regard as binding even and 25 280 DANIEL SCHWARTZ The point is crisply made by Kavka: We cannot hope to keep people out of the bad circumstances (e.g., anarchy, illness, poverty) that might lead them to make forced promises simply by declaring such promises void; we would just deprive them of one tool for making the best of a bad situation.28 Often by neutralizing the expected benefit to the transgressor, by denying a right in him, we also damage the prospective victim of the wrong. As Deigh puts it, we ‘deprive people caught in a war a tool for limiting the destructiveness of their situation’.29 Utopia conquers part of the territory of Arcadia, and then offers peace terms. If Arcadia refuses, war will continue and Arcadia will have to deploy a considerable number of troops to the border—troops that, for example, may be needed to alleviate the effects of a natural disaster in a different part of the country. Given the circumstances, it is preferable for Arcadia to sign the peace treaty. Suppose treaty conventions invalidate the treaty because Utopia has unjustly worsened the choice situation of Arcadia. Knowing this, Utopia decides to withdraw the offer of a peace treaty, and Arcadia is condemned to the worse of the two available options. By preventing Utopia from benefiting from its unjust threat, the international society increases harm to Arcadia relative to the alternative. Should not Arcadia’s interest be taken into account? Should we not offer victims a way out when they are trapped in a situation of unjust coercion? There is an additional, and I think overlooked, reason to validate coerced treaties. It concerns not the treaties between strong unjust aggressors and weak innocent victims, but rather the spillover effect on treaties between innocent strong countries and weak ones. Suppose a morally innocent power is weighing whether to offer a border treaty to a weak neighbor. The strong country may fear that the weak neighbor may claim later, falsely but credibly, that it signed under unjust coercion, and thus have the treaty invalidated. Hence an effort to neutralize the benefits of unjust coercion may neutralize also the benefits of innocent strong countries that do not act unjustly and impair small countries’ capacity to enter into profitable treaties with them. The example of Arcadia and Utopia is premised on a causal link between the validation of treaties and their availability to interested parties. It is necessary to say something about the mechanism involved. Consider a peace treaty proposed by the coercer to the unjustly coerced. The probability of the unjustly coerced being offered the peace treaty depends on how attractive it is for the coercer. The attractiveness of a treaty from the point of view of the coercer is a function of the level of concessions and the expected compliance by the coerced. especially if the promises are coerced.’ A similar point is made by Emer de Vattel, The Law of Nations of the Principles of Natural Law, trans. C. G. Fenwick (Washington DC: Carnegie, 1916), bk IV, ch. IV, sec. 37, p. 356. 28 Kavka, Hobbesian Moral and Political Theory, p. 396. 29 Deigh, ‘Promises’, p. 487. THE JUSTICE OF PEACE TREATIES 281 Participants in treaty conventions expect signatories to comply. In the event of the treaty being breached they will not interfere with justified retaliatory steps by the defrauded party and may cooperate with these steps to various degrees, including the levelling of sanctions. For this and other reasons it seems empirically correct to say that validation increases the proclivity of the signatories to comply, and lack thereof diminishes it.30 Expected compliance and therefore the attractiveness of the peace treaty to the coercer is increased by validation. Convention participants therefore control the access to peace treaties of the unjustly coerced. Introducing a treaty convention that is less tolerant of unjust coercion amounts to a confiscation (the degree of which depends on the demandingness of the convention) of a palliative instrument that states could use to deal with situations of unjust coercion. It may be asked, if the convention participants are incapable of influencing the behavior of the coercer, how can they be capable of influencing the behavior of the coerced? The answer is that since the coercer is stronger than the coerced, participants will have a greater influence on the former than on the latter. In other words, it will be much easier to augment the compliance of the coerced to the extorted treaty than to make coercers comply with international rules that forbid extorting treaties. States pay a price for the invalidation of coerced treaties: they renounce an instrument capable of reducing the harm of a situation. It is not a novel point that restricting access to extorted treaties can be harmful for those who could be potentially coerced. As we have seen, it had already been adverted to by early theorists of international ethics. Yet a less often noticed fact, and one that I wish to emphasize, is that the impact of treaty restrictions is not uniform across countries; unequal burdens are imposed. Therefore there are initial grounds to condemn such restrictions not merely as conflicting with humanitarian sentiment but as distributively unjust. These suspicions are examined in detail in the next section. IV. STRICTER TREATY CONVENTIONS: DIFFERENTIAL EFFECTS It is natural to think that restricting general access to extorted treaties contributes to a collective good: the reduction in the level of risk of each finding himself in a situation in which coercion is used to extort a pact from him. The thought is that as the attractiveness of extorted treaties decreases (given a decrease of expected compliance caused by invalidation) coercers will be less inclined to use their unjust coercion to procure such treaties. I shall argue that whether or not stricter treaty conventions reduce the average levels of extortion, their isolated introduction disproportionately worsens the 30 For an explanation of the sources of treaty compliance in the absence of enforcement see Andrew T. Guzman, How International Law Works (Oxford: Oxford University Press, 2008), pp. 33–49. 282 DANIEL SCHWARTZ situation of the weakest countries while it benefits middle-weight countries. Restricting access to extorted contracts resembles the sealing of a small leak in a dam only to deflect the pressure to a different more dangerous one. I deploy these points in the next few paragraphs. Let us imagine a stylized non-ideal world made of three countries in conditions of complete information: Aggresso, an international bully bent on acquiring some of its neighbors’ territory; Competento, a country of competent deterrence capacity; and Debile, a very weak country. Aggresso’s options are abstention, attempting to obtain the territory as concessions given under a treaty obtained through threats of force (extortion), or direct seizure by war. Hence, Aggresso has two military options: limited sanction (LS), a backward-looking punitive measure intended merely to make good on an unsuccessful threat, and direct seizure (DS), a purposive attack with the goal of taking possession of that which was unsuccessfully sought by threats. In general LS is less destructive and therefore less costly to both sides than DS, which tends to involve full-fledged war. There are, in addition, two possible treaty conventions: a liberal one and a stricter one, they differ in the degree of tolerance to coercion as exhibited by their conferrals of validity. Scenario I: Liberal Convention Competento and Debile are both vulnerable to extortion by Aggresso (extortion pays off as resulting treaties are conferred increasing compliance validity). Competento is not vulnerable to DS, while Debile is. Scenario II: Stricter Convention Let us assume that extortion is no longer an option and the level of requested concessions is kept fixed. Given that war against Debile is relatively affordable and extortion is not on the table, for Aggresso the only way of obtaining Debile’s desired territory is to launch a DS. However, insofar as Competento is concerned, DS is not option. Even if Aggresso could win the war, victory would come at a prohibitive cost. Hence Aggresso will choose neither to attack nor to threaten Competento; it will abstain. Competento is better off and Debile worse off in II. The isolated introduction of the stricter convention benefits Competento’s position but worsens Debile’s. The rationality of deterrence and war is a notably sophisticated and complex field. This very basic analysis has no pretensions to contribute to this particular field, but aims to disclose the logic supporting the prediction that the isolated introduction of stricter treaty conventions harms the weaker countries. The view that middle-weight countries actually benefit from such conventions is perhaps less evident and must confront a number of objections even within the boundaries of this rudimentary analysis. It may first be objected that Competento was not susceptible to extortion in I in the first place, since given its defensive capabilities no credible threats against it were possible. Therefore, the objection goes, Competento’s position remains unchanged: it does not actually benefit from the stricter convention. THE JUSTICE OF PEACE TREATIES 283 The response relies on the distinction above between the costs of limited sanction (LS) and direct seizure (DS). Although DS by Aggresso against Competento would be too destructive for it to be able to back Aggresso’s extortive threats, LS is affordable, and is thereby capable of credibly backing Aggresso’s extortive threats. But, the objector may retort, Competento will counter Aggresso’s threat of LS with the threat to respond with the sort of massive retaliation normally reserved as a response to DS. It is irrational for Aggresso to launch a DS against Competento. Aggresso’s threats to inflict an LS would now cease to be credible, having been transformed by Competento in incredible DS threats. Against the objector it may be noted that for Competento’s counter threat of massive retaliation to be credible it would have to be reasonable for it to engage in large scale war against a superior enemy solely to avert a limited sanction.31 Competento’s counter threat might not be credible, and it will not be able to avert LS. Hence Competento is vulnerable to Aggresso’s threats in I. A game theorist might present additional objections beyond the scope of this paper. I shall only add that the analysis above only requires commitment to the actual possibility of limited sanctions (or ‘limited retaliation’, the term of art in nuclear deterrence literature). This is not an implausible assumption. Regardless of the debatable impact of stricter conventions on middle-weight countries, it is beyond doubt that weak countries are harmed by the isolated introduction of stricter treaty conventions. For them the increase in expected harm will be dramatic. The reason is that harms caused by exitless wars are on a different scale than those of coerced treaties. Once one takes into consideration the disparity in the scale of the harms involved, it becomes clear that for some countries a small increase in the probability of attack is enough to dramatically worsen their situation, almost irrespective of the fluctuations of the probabilities of extortion and abstention. Against what baseline should we appraise stricter treaty conventions? The argument above does not need to take any particular convention as a baseline. The view is that, relative to any treaty convention, moving to a stricter set of validity requirements without correlative transfers of deterrent capacity will make things worse from the point of view of distributive justice. This explains the assertion above that asking whether the Vienna Convention is too liberal or to strict is asking the wrong question. What matters is whether it is complemented by arrangements that will at the very least mitigate the inequality of its effects. The way of reducing the probability both of extortion and of attack for the worst off is not only to make extortion less attractive by raising the bar of validity, but also by making unjust attack less attractive by transferring deterrent 31 On the incredibility of massive retaliation see Robert Powell, Nuclear Deterrence Theory (Cambridge: Cambridge University Press, 1990), pp. 29–32 and William Kaufmann, ‘The requirements of deterrence’, Military Policy and National Security, ed. William W. Kaufmann (Princeton, NJ: Princeton University Press, 1956). 284 DANIEL SCHWARTZ power to the worst off. In the example above, the introduction of a stricter convention would need to be complemented with a transfer of deterrent capability from Competento to Debile, such that both would improve their position relative to the more liberal convention. Confiscating palliative instruments without transferring deterrence benefits countries of intermediate capabilities while damaging weaker ones. Of course, one needs to consider the possibility that in fact the lower levels of compliance caused by stricter treaty conventions will be compensated for by Aggresso through increasing the amount of requested concessions. This would preserve the expected utility of its threats (assuming countries to be risk-neutral). If such compensation is fully pursued the conclusion would necessarily be that stricter treaty conventions are ineffective and inane. Nevertheless, no amount of promised concessions would be sufficient to compensate for a virtually zero level of compliance with extorted treaties. In addition, it is a plausible conjecture that many or most countries are to some degree risk averse, and so would not prefer a treaty offering substantial concessions with a small probability of compliance over a treaty offering more concessions with a higher probability of compliance, even if both promise the same expected utility. It will be hard for risk averse countries to compensate for the decrease in compliance introduced by stricter treaty conventions. V. STRICTER TREATY CONVENTIONS: FAIRNESS AND JUSTICE Assuming the description of the effects of the isolated introduction of a stricter convention in the stylized case is correct, what does it tell us about its fairness? A convention that places some states at a disadvantage would not be chosen in a fair procedure, that is, in one that expresses the freedom and equality of those called to decide on their basic terms of cooperation. In other words a representative of state who, placed under the veil of ignorance, does not know whether she represents Competento or Debile would not support introducing such a stricter convention. For Hume and Rawls,32 a promise’s moral validity depends on its conformity with a justified social convention. The convention is justified if the practices that it makes possible, to quote Deigh, ‘contribute positively and importantly to the collective good of human beings.’33 For Rawls, contracts, as promises, further the common good in that they ‘stabilize cooperative agreements for mutual advantage.’34 Yet conventions must also conform with principles of justice. In the present international context a convention that gives unequal access to treaties fails this test: it prevents those who most need it from benefiting from the 32 David Hume, A Treatise on Human Nature (Oxford: Oxford University Press, [1739–1740] 2000), bk III, pt ii, ch. V, pp. 331–7. Rawls, A Theory of Justice, p. 345. 33 Deigh, ‘Promises’, p. 492. 34 Rawls, A Theory of Justice, p. 345. THE JUSTICE OF PEACE TREATIES 285 resulting good. In other words, it violates Rawls’s difference principle if—as I argue—it not only fails to improve the situation of the weaker countries but in fact worsens it (whether or not it also violates equal liberty).35 If this is so, why do we normally regard it as morally permissible for the state to withhold validity from domestic extorted contracts? One of the main functions of states is to protect citizens from being caught in situations in which they would benefit from having the option of an extorted contract. In some states citizens enjoy a roughly equal level of publicly provided protection (the police take the same interest in the protection of poor and rich, white and black, etc.), and some among them enjoy roughly the same absolute level of security (taking in account both publicly provided and privately purchased protection). If levels of risk of criminal violence are low and the risk is evenly distributed, the price that citizens pay for having their access to extorted contracts restricted is both low and fairly distributed. In badly policed states in which inequality in security is extreme, those living virtually at the mercy of crime lords (say dwellers in a shanty town that the police are too afraid to enter) would be unfairly treated if their access to palliative treaties with criminals, such as the payment of ‘protection money’, were reduced or eliminated. I have been suggesting that treaty law should be appraised in terms of its conformity with principles of distributive justice between states. But two popular views seem to deny that such principles in fact exist. There is first the cosmopolitan belief that distributive justice does not operate between states but only between individuals. Second is the anti-cosmopolitan belief that because there are no egalitarian duties to redistribute wealth between states, it follows that distributive justice does not operate between states. I discuss the import of these beliefs below. According to the cosmopolitan view defended by Brian Barry, Charles Beitz and Thomas Pogge, we should discuss distributive justice between individuals, not between states, countries, societies, or Rawlsian ‘peoples’.36 Without having to dismiss this view one can, in line with Buchanan, hold that ‘in addition to these individual principles [of distributive justice] principles are needed for determining relations among “peoples” organized in states’.37 The present analysis assumes only that, wherever else principles of distributive justice apply, they apply also to the society of states. The idea of distributive justice between states seems to face opposition not only from cosmopolitans, but also from anti-cosmopolitans like David Miller and the later Rawls, who believe it to be incompatible with states’ autonomy to 35 Freedom of contract is not one of the basic liberties protected by the principle of equal liberty. Rawls, A Theory of Justice, p. 61. 36 Brian Barry, The Liberal Theory of Justice (Oxford: Clarendon, 1973), pp. 128–33. Beitz, Political Theory and International Relations, pp. 71–82. Thomas Pogge, Realizing Rawls, p. 248. 37 Allen Buchanan, ‘Rawls’s law of peoples: rules for a vanished Westphalian world’, Ethics, 110 (2000), pp. 697–721 at p. 700. 286 DANIEL SCHWARTZ determine their own internal arrangements. However, it is necessary to examine closely the exact nature of the thesis they oppose. These theorists react to global proposals of redistribution, analogous to domestic taxation, such as Thomas Pogge’s ‘Global Resources Dividend’ and Hillel Steiner’s ‘Global Fund’.38 Nevertheless both agree that cooperation between states, nations or peoples for mutual advantage—such as on the curbing of greenhouse gas emissions—should be conducted, to quote Miller, on ‘fair terms’39 without undue advantage to the strongest nations. Let us consider this in more detail in Rawls’ The Law of Peoples. His ‘second original position’ is inhabited by representatives of liberal peoples charged with defining the basic terms of cooperation between peoples who see themselves as free and equal.40 His rejection of ‘equality among peoples’ means only that, in this second original position, representatives would not set up institutions equalizing the wealth of individual citizens or the citizenries taken as a group.41 His ‘distributive justice’ refers only to wealth equalization. It is just this narrowly conceived distributive justice that is denied a role in interstate (or ‘inter-people’) affairs. Rawls’ restriction of the goods allocated by distributive justice solely to wealth seems arbitrary. There is no reason not to allow distributive justice to govern the distribution of the full range of international primary goods (which surely includes freedom from some forms of coercion).42 Regardless of whether representatives of peoples would in fact reject redistribution of wealth so as to preserve autonomy, it is clearly the case that they would reject institutions that place some peoples at a disadvantage when dealing with external enemies. It follows that an arrangement aimed at reducing the use of coercion in international affairs that places some of the participants (be they ‘peoples’ or states)43 at a disadvantage violates distributive justice in its original Rawlsian meaning, as the virtue in charge of all primary social goods, not just wealth. It may be noted that my conclusion frontally contradicts Rawls’s view in A Theory of Justice that ‘it would be wildly irrational in the original position to agree to be bound by words . . . extorted by force.’44 In reply we should remind ourselves that, while Rawls in this work is engaged in ideal theory, here we are engaged in non-ideal theory. In particular, we accept that some features of the behavior of relevant actors is beyond control, or in other words (i) that 38 See David Miller, National Responsibility and Global Justice (Oxford: Oxford University Press, 2007), p. 57. 39 Ibid., pp. 6–7, also pp. 70, 73. 40 Rawls, Law of Peoples, p. 33. 41 Ibid., p. 114. 42 Ibid., p. 29. ‘Liberal peoples . . . [s]eek to protect their territory, to ensure the security and safety of their citizens, and to preserve their free political institutions and the liberties and free culture of their civil society.’ 43 On the puzzling Rawlsian ‘peoples’ and its fundamental equivalence with states see Onora O’Neill, ‘Agents of justice’, Global Justice, ed. Thomas Pogge (Oxford: Blackwell, 2001), pp. 188–203 at pp. 194–5; Buchanan, ‘Rawls’s law of peoples’, pp. 698–9; Simon Caney, Justice beyond Borders (Oxford: Oxford University Press, 2006), p. 11. 44 Rawls, A Theory of Justice, p. 345. THE JUSTICE OF PEACE TREATIES 287 international bullying is a phenomenon which will not go away for the time being, and (ii) that third parties are often unwilling to sufficiently defend those who are victims of international bullies. Our question concerns the institutional arrangements which should, taking these features as fixed, be established by third parties. Since there is a variety of ways in which international institutions that take the facts of unjust coercion and insufficient assistance into account can be arranged, we should prefer arrangements that seem fair. It may nonetheless be retorted that those taking part in a convention that allows bullies to extract extorted treaties become their accomplices. There are two possible grounds for this claim: (i) that third parties could and should do more to prevent international bullying (for example by transferring deterrent power); and (ii) that even if third parties cannot do more, or are morally justified in not doing more, their making extorted treaties available (or more available than they would otherwise be) through their conferral of validity in itself makes them accomplices of the extortion. The next section inspects this second reason. VI. CONFERRAL OF VALIDITY AND COMPLICITY Some moral duties are prior to duties of distributive justice. If there is a negative moral duty not to validate coerced treaties (because it involves complicity with unjust coercers) then we would be justified in withholding validity. Is the conferral of validity to an extorted treaty a form of complicity with the unjust coercer? It would seem so. The participant of the validating convention increases the possibility of extracted treaties by making them more profitable. Hence the charge of complicity warrants serious consideration. As has been noted by Chiara Lepora and Robert E. Goodin, ‘complicity’ is often used as, a ‘catch all-term’ covering various forms of being implicated in the wrongdoing of someone else, each with its own distinctive features.45 These include ‘full joint wrong-doing’, co-operation, collaboration, conspiracy, collusion, connivance, condonation, and complicity proper. In the discussion below I will use the term ‘complicity’ in its non-technical sense, as covering these various wrongful ways of associating oneself with other’s wrongdoing. Let us stipulate that the convention participant is impotent to impede the use of force. The influence of the convention participant on the unjust coercer restricts itself to making available for him a form of unjust coercion (extorted treaty) that diminishes harm to the victim relative to the alternative.46 The 45 Chiara Lepora and Robert E. Goodin, ‘Complicity and its conceptual cousins’, unpublished manuscript, p. 9. 46 As a parallel one may consider the one-time provision by the Law of Nations by which captured soldiers could be reduced to slavery was justified as way of getting captors to willingly refrain from killing them. On this, see Rawls, A Theory of Justice, p. 218 and Hugo Grotius, The Rights of War and Peace, III.7.5, p. 1364. 288 DANIEL SCHWARTZ question is, therefore, whether increasing the prospects of this extorted treaty over direct attack constitutes complicity. The case at hand can be modeled as a special case of deflection of harm. What is peculiar here is that we do not deflect harm from one person to another, but from one type of harm to another for the same person. Hence the puzzles that arise in the ‘trolley cases’ made famous by Judith Jarvis Thomson are not relevant here.47 Consider these examples of intra-personal harm deflection: Switch: An interrogator is about to torture a prisoner with an electric prod. A person living in the house next door knows this and has access to the fuse box containing the switch that reduces the voltage of the electric current, thus making the electric shocks less harmful. The interrogator will not associate him with the drop in voltage, and will spend the same time torturing independent of the level of harm caused. It is unquestionable that the neighbor would be justified and morally required to turn the switch. Bribe: The neighbor has no access to the switch, but is capable of persuading the torturer to turn the switch himself by offering a bribe. In both cases we have a deflection to the lesser of two harms for the same person, but the methods are different. In the first case the method does not require coordination with the wrongdoer. Moreover the deflection runs contrary to the wrongdoer’s aims. In the second case there is communication and eventually persuasion, such that the wrongdoer favors the deflection. Whatever we think of Bribe (I think most people would feel that the neighbor is either required or justified to give the bribe), there may be, to use Thomson’s phrase ‘a lingering moral discomfort’. The source of moral discomfort may be skepticism about the stipulation that the neighbor cannot do more—for example give a larger bribe to prevent any sort of torture. Yet if we remain within the stipulated bounds, the cause of the discomfort could be (a) the moral dubiousness of associating with someone known to be bent on committing a serious wrong, or (b) the fact that the deflection is agreeable to the wrongdoer rather than imposed on him counter his wishes. I shall not discuss the currency of either (a) or (b) but rather simply inspect whether they are relevant to the case of a participant in an international convention that validates at least some treaties procured through unjust coercion. First, is there some form of association between participants in a convention and unjust coercers? A convention the participants of which validate some extorted treaties need not involve any sort of coordination, communication, or contact with the unjust coercers. One can construct the concept of association more or less expansively, but it seems that the central case of association would include at least some of these. Of course, participants in the convention 47 ‘The Trolley Problem’ in Judith Jarvis Thomson, Rights, Restitution & Risk, ed. William Parent (Cambridge, MA: Harvard University Press, 1986), pp. 94–116. THE JUSTICE OF PEACE TREATIES 289 communicate something in the sense that they make public their readiness to validate extorted treaties. But to say that this is a criminal association is as implausible as saying that legislators who refrain from imposing sanctions on adultery associate themselves with the adulterers.48 Second, clearly the unjust coercer who opts for extorted treaty over attack does so agreeably. Validation deflects harms not through mounting opposition to the coercer’s will but by allowing him to give expression to some of his antecedent preferences. There seem to be two different moral impulses operating here. First, we like to see the plans of wrongdoers thwarted, not only because we want wrongdoing to fail, but because we want life to go badly for wrongdoers (or at least to go less well than for non wrongdoers). At the same time we find actions that promote the good through opposition to evil more heroic and appealing than those which involve no such opposition. None of these reasons make non-heroic actions that diminish wrongdoing without worsening the life of wrongdoers morally impermissible. Hence the imputation of association is irrelevant to this case. It is true that validation deflects harms without antagonizing the coercer, and that antagonistic deflection may be preferable, but this is not a good reason to regard non-antagonistic deflection as impermissible. Does this conclusively clear validation from the charge of complicity? One reason why it does not is that the analogy overlooks an essential feature of the case of treaty validation. It is true that the convention participant deflects harm by creating an incentive for the unjust coercer to switch from the more harmful to the less harmful use of unjust coercion (to choose extorted treaty over attack). Nevertheless, unlike the case of bribing the torturer, the source of incentive created by the convention participant consists of the profits obtained through wrongdoing. The convention augments the expected benefit that the unjust coercer can expect to derive directly from his unjust threats. Deflection is achieved, not by supplementing wrongful gains (as in Bribe) but rather by increasing them. This seems to be an incentive that is impermissible to offer. By increasing his wrongful gains you involve yourself in the wrongful activity itself. Hence we need to move to arguments which accept the participation of convention makers in wrongful actions but nevertheless excuse or justify such participation. I cannot survey here all the possible arguments, but the next case provides one of them in outline: Collaborative Friend: A robber walks into my house determined to maximize his unlawful profit. He has the option to take either my gold ring or my fridge. The fridge is more expensive, but he will not be able to carry it away on his own. After 48 This, seems to leave open the possibility of collusion, which occurs ‘when two or more agents act in unspoken concert with each other to wrong some third person or party’ (Lepora and Goodin, ‘Complicity and its conceptual cousins’, p. 9). I am not denying that such unspoken concert or coordination may obtain in the case under consideration, but that it needs not to. 290 DANIEL SCHWARTZ trying to remove the ring he comes to the conclusion that only severing the finger will do. Obviously worried by this prospect I ask a friend to help the robber to carry the fridge. He accepts. My friend increases the profit of the criminal in order to deflect harm to the less harmful of the two options. It seems clear that the friend is not an accomplice in the robbery (if, that is, blameworthiness is necessary for complicity). The friend is involved in a wrong without himself acting wrongfully. He has been authorized by me, who has consented to waive the relevant right (a right against my friend, not against the robber). What I authorize is making more profitable the perpetration of one particular kind of wrong against myself. Equally, the participants in a convention which validates an extorted treaty are increasing the profit margin of this wrongful act by being disposed to put pressure on the victim to comply with the coerced treaty. This seems fine if authorized by the victim.49 It may be objected that the victim has no choice but to consent, and thus this consent authorizes nothing. Note, however, the alleged absence of choice does not invalidate the consent given. Otherwise a patient may claim that the consent given to remove a tumor when fearing he might otherwise die is invalid and could sue the hospital even after the successful operation. Consent is not invalidated by lack of choice because the surgeon is in no way related to the source of the counterfactual peril.50 In the example above the consent was given when virtually no choice was available, but the source of the danger was not my friend. If the consent holds in the case of the operation so it should here. Similarly, if Arcadia authorizes a group of states who are not the source of the danger to pressure it (i.e., Arcadia) to comply with the unjust expectations of Numidia (by validating the treaty), then this consent would be valid. An additional objection questions the value of the analogy for the case under study.51 For it may be argued that we look more benignly towards the friend who engages in a form of private connivance with evil than towards legislators granting a permission to indulge in unjust coercion, even if they act with a similar 49 This may seem to be allowed by Frances Kamm’s ‘Principle of Secondary Permissibility’ (PSP) according to which it is permissible to intentionally cause harm to a person whenever it would otherwise be permissible to cause a greater unintentional harm to her (in the presence of some additional conditions) or when, unless one inflicts intentional harm, the person will be harmed to a greater extent (intentionally or unintentionally) by a different agent. Hence it would have been justified to bomb the trains en route to Nazi extermination camps (which would have killed many Jews awaiting certain murder) in order to obstruct the genocide. Yet, note that (i) Kamm prescinds from the need to be authorized by the victim, and (ii) that what PSP is said to permit is the creation by A of a new harm to avert the harm that would otherwise be caused by B. Here we have the element of cooperation between A and B, which adds some complicating features. See Frances Kamm, Intricate Ethics (New York: Oxford University Press, 2007), pp. 170–3. 50 This usually parallels the difference between ‘forced promises’ and ‘coerced promises’ as discussed by Deigh, ‘Promises’, p. 488. 51 I thank an anonymous reviewer for this observation. THE JUSTICE OF PEACE TREATIES 291 motivation in mind—diminishment of harm to those under their care.52 Hence, according to the objection it is misleading to transpose our moral reactions to the friend’s actions to those of convention-makers. It is certainly the case that public connivance with evil is rather more problematic than a private act (particularly if the former is an exceptional act). Unlike private connivance, public connivance provides an incentive or removes a disincentive to unjust coercers. It may be also said to pervert the law, making it not just silent about certain abuses but actually protective of their performance. It is doubtful, however, that the contrast between the effects of these two forms of connivance is so stark (imagine the effect of a private connivance so extensive as to constitute a social practice). In any case, whatever the reasons to judge more benignly the collaborative friend than collaborative convention makers, these are not due to the moral properties belonging to the act of collaboration itself (in both cases we permit the performance of an evil act, for the same motivation), but rather to the effects of such collaboration. In other words if the friend is not an accomplice neither are convention makers (even if there may be reasons to question the behavior of the latter that do not apply to the former). The case of the collaborative friend is intended as an illustration of the view that harm deflection caused by a third party by increasing the unjust profit directly produced by unjust coercion may be morally permissible and even morally required when the victim’s consent has been given or can be safely presumed. This very brief discussion cannot produce a conclusive proof. All the example provides is some prima facie plausibility to the claim that harm deflections by a third party which involve collaboration with the person who inflicts the wrongful harm need not involve complicity. In order to assess the morality of validation, we have first examined the impotent participant in a validating convention. It is clear however that many—perhaps most—third parties are not impotent, and could do more to reduce harm to victims of unjust coercion. But if this is so, what is morally censured is not the validation of extorted treaties, but rather their refraining from sufficiently assisting victims of unjust coercion. Consider a bystander who is capable of defending a passerby from aggression but fails to do so, and yet after the attack gives some medical care to the victim. If the bystander’s conduct is reprehensible, then surely this is because of the failure to protect and not because of the subsequent provision of care. Similarly, the need to reduce harm by validating extorted treaties would decrease if third parties were willing to come to the rescue of victims of unjust coercion. It is their failure to do so when they can, or should, and not the provision of palliatives that seems morally wrong. 52 Connivance ‘at a minimum, involves assenting in the wrongdoing, and contributing to it in that way’ (Lepora and Goodin, ‘Complicity and its conceptual cousins’, p. 11). 292 DANIEL SCHWARTZ VII. CONCLUSION Let us return to the initial question about the Vienna Convention now in force: does it ask for too much or too little? The answer that emerges is that ‘too little’ and ‘too much’ should refer not to the demandingness of the requirements of treaty validity but to the extension of the general set of arrangements needed to control and reduce unjust coercion. The Vienna Convention only controls one lever. In order to be able to achieve the desired result it should be operated together with other levers (especially deterrence transfers). Using just this part of the machinery available to us in order to reduce the harms of unjust coercion not only may actually increase them, but will do so in an unfair manner. It may be noted that, assuming the effects of the isolated introduction of stricter treaty conventions on weak countries are indeed of the sort described above, such conventions may be rejected on grounds other than those of fairness. This observation does not contradict the argument above, the goal of which is simply to identify one reason to reject the introduction of such conventions. Yet some may regard unfairness as a relatively unimportant reason to reject the introduction of conventions that condemn weak countries to the possibility of total war. They would say that it is the humanitarian impulse to prevent destruction and death that provides the most forceful reason for rejection (as noted in Section III, this is after all what impelled the Fathers of International Law to accept coerced treaties). It could be observed that the unfairness of a stricter convention is a badly chosen, perhaps even unfortunate, line of criticism. In assessing this observation it is useful to consider this next analogy. Given health budgetary constraints, a country fails to vaccinate a tenth of the population against a rampant epidemic. The unvaccinated are chosen randomly. Clearly the main source of moral outrage is humanitarian: it demands the reduction of other budget components so as to enable comprehensive vaccination. Compare this scenario with one in which all the non-vaccinated are made up of the poorest tenth of the population. Here the sense of unfairness will be a central contributor to the moral outrage: citizens are being discriminated against on the basis of wealth. Moving back to the international arena, stricter treaty conventions introduced in isolation of complementary security arrangements will not increase risks for a random selection of countries, but for those already disadvantaged, namely those more likely to find themselves threatened (while probably benefiting middle-weight countries). The unfairness of the isolated introduction of stricter treaty conventions is an important component of what makes them worthy of moral condemnation.