This article seeks to clarify the current legal status of a particular aspect of the ‘responsibility to protect’ principle—the idea that bystander states have a collective responsibility to protect populations beyond borders from mass atrocities when host states fail to do so. It outlines the development of this idea and argues that, while the legal force of key international statements on the ‘responsibility to protect’ principle may be weak at best, the International Court of Justice and the International Law Commission have offered bold declarations in recent years which do point towards the gradual development of legal duties for the extraterritorial protection of populations.
The ‘responsibility to protect’ concept has emerged over the last decade at a remarkable pace. In less than four years, from 2001 to 2005, it progressed from an idea advanced by an independent commission of experts to being unanimously endorsed by the United Nations (UN) General Assembly.1 It has subsequently been referred to in numerous resolutions by the Security Council and the General Assembly,2 and has taken a prominent place in international debate about the protection of populations from mass atrocities. The concept has been invoked by states, non-governmental organisations (NGOs) and the international media, both to justify and to condemn behaviour, and both to advocate and to deter international action in response to crises in various places, including Darfur, Kenya, Georgia, Myanmar, Gaza, Sri Lanka, the Congo, North Korea, Côte d'Ivoire and, most recently, Libya, Syria, Yemen and other states caught up in the so-called ‘Arab Spring’.3 However, its international legal implications are not immediately clear.4
The ‘responsibility to protect’ encompasses two broad propositions. One is more novel and contested and represents a far more significant challenge to accepted principles of international law than the other. However, commentators have too often failed to clearly distinguish between the two and have instead offered some rather obscure pronouncements about the extent to which the concept conceived in its entirety can be understood as an emerging norm of international law.5 Rather than make such pronouncements, it is more fruitful to determine the degree of legal support for each of the concept's distinct propositions.
The first proposition is that states have a responsibility to protect their own populations from mass atrocities—specifically from genocide, war crimes, ethnic cleansing and crimes against humanity.6 This duty is deeply embedded in existing international law. It is well established in a range of universal and regional human rights conventions,7 and is clearly endorsed in the General Assembly's 2005 World Summit Agreement. No state today denies this duty. Indeed, even those states most wary of the ‘responsibility to protect’ concept, such as Venezuela, Cuba, Myanmar, Nicaragua and Sudan, readily accept this duty.8 Moreover, the notion that the society of states may rightfully hold states to account for the performance of this duty is also well established. Since the end of the Cold War, the Security Council has repeatedly condemned states and authorised sanctions and interventions in response to the occurrence of mass atrocities and humanitarian crises under Chapter VII of the UN Charter.9 To be sure, two of the permanent members, China and Russia, have tended to oppose attempts within the Security Council to authorise military intervention in the affairs of functioning sovereign states that do not grant their consent, and of course the legality of unauthorised intervention remains at best highly contested.10 However, in March 2011, both of these states were willing to allow the adoption of UNSC Resolution 1973 authorising the use of ‘all necessary measures’ to protect civilians from the threat of mass atrocities in Libya against the wishes of Moammar Gaddafi's regime. There remains little doubt today that the Council has the legal right to authorise both non-forcible and forcible interventions to enforce the protection of populations in those instances where it agrees to do so. In short, the notion that states have an obligation to protect their own populations from mass atrocities is firmly established in international law and need not detain us here.
The second proposition of the ‘responsibility to protect’ concept is that bystander states or the ‘international community’ have not simply a right but a collective responsibility to assist host states in protecting their populations and to act to protect these populations in situations where the host state is manifestly failing to do so. In contrast to the way in which debates about ‘humanitarian intervention’ tended to be framed in the 1990s, the extraterritorial protection of populations is posited not merely as a discretionary right but as a positive duty borne by all states. This suggested duty would seem to have extraordinary implications for interstate relations. It could be taken to imply, for example, that bystander states have not an option but an obligation to act in response to all situations, like Libya, in which civilians are understood to be at risk of mass atrocities. The widespread embrace of the ‘responsibility to protect’ principle, which includes such a proposition, is a remarkable development. However, this proposition is much less well established in international law.
The purpose of this article is to clarify the legal status of this proposed collective ‘responsibility to protect’ beyond borders. While the legal force of key international documents on the ‘responsibility to protect’ principle may be weak at best, the International Court of Justice (ICJ) and the International Law Commission (ILC) have offered bold declarations in recent years which complement these documents and which point towards a significant shift in the international legal obligations of bystander states. The notion that there may be gradually developing a legal duty to protect beyond borders rewards close examination.
The article proceeds with three further sections. Section 2 briefly outlines the historical development of legal thinking about duties to protect beyond borders. It emphasises not just that this idea has deep historical roots which can be traced over several centuries, but also that legal theorists were increasingly reluctant to assert these duties as the independence and liberties of sovereign states were progressively established from the seventeenth century onwards, at least until the end of the Cold War. Section 3 examines the emergence of the ‘responsibility to protect’ concept in recent years—focusing particularly on the notion of a collective responsibility to protect beyond borders—and observes that there is little ground for thinking that the leading international statements and UN resolutions on this principle have, by themselves, established new legal obligations with respect to the extraterritorial protection of populations. The fourth and longest section, however, examines some important legal developments, particularly the ICJ's judgment in Bosnia v Serbia in 200711 and also the work of the ILC, which do provide an albeit limited legal basis for the notion that there is a responsibility to protect beyond borders; a responsibility borne by states and perhaps even by international organisations. It concludes by acknowledging the limitations and ambiguities of existing international consensus, state practice and case law, and particularly the lack of clarity or even coherence around questions of enforcement and reparation, but nevertheless insisting that important developments have been made in recent years towards a legal obligation to protect populations beyond borders.
2. Historical Background
It is well known that there is a rich history of legal thinking about the rights of states to protect those beyond their borders from tyranny and persecution. The origins of these ideas can be traced at least to early modern Europe, if not much earlier.12 Less recognised is the fact that the protection of foreigners was commonly framed by jurists not merely as a right but as a duty. In the first century BC, the Roman philosopher Cicero argued that there were two types of injustice: ‘men may inflict injury; or else, when it is being inflicted upon others, they may fail to deflect it, even though they could’. Justice was said to entail not merely negative duties to refrain from perpetrating harm but also positive duties to protect others from harm: ‘the man who does not defend someone, or obstruct the injustice when he can, is at fault just as if he had abandoned his parents or his friends or his country’.13 He insisted that duties were not only owed to those close to us but also to strangers and foreigners. Those who suggested otherwise were said to ‘tear apart the common fellowship of the human race’.14 Cicero's account of duties would greatly influence subsequent legal theorising about the responsibilities of assistance and intervention on behalf of strangers. Early modern jurists who developed arguments for intervention to rescue innocents and to punish tyranny often framed such intervention as a duty. Alberico Gentili, for example, asserted that ‘It is the duty of a man to protect men's interests and safety’,15 and, on these grounds, he defended war waged on behalf of foreigners, since ‘the subjects of others do not seem to me to be outside of that kinship of nature and the society formed by the whole world’.16 However, this duty was commonly said to be limited in particular ways. Some jurists, such as Hugo Grotius, acknowledged that the defence of strangers was a duty but, emphasising the natural right of self-preservation, insisted that princes and states were bound to discharge this duty only to the extent that it could be carried out with ‘convenience’ and without excessive cost to themselves.17 Other legal theorists, such as Samuel Pufendorf, insisted that while states did bear a positive duty to assist and protect those beyond their borders, it was an ‘imperfect’ duty in the sense that it was a matter of conscience rather than covenant, and failure to discharge the duty ought not to attract punishment.18 The idea that states have a responsibility to protect populations beyond their borders, then, is not new. Moreover, we will find in subsequent sections that the kinds of caveats and limitations that jurists applied to this duty in the early modern period continue to restrain the development of the duty in international law today.
The consolidation of the independence and liberties of states and the gradual replacement of natural law with positive law from the seventeenth to the nineteenth century saw an increasing reluctance among legal scholars to impose obligations on states to protect those outside of their territories. The nineteenth century was witness to a number of so-called humanitarian interventions, conducted particularly by European states in Ottoman affairs and by the United States in the American hemisphere.19 These interventions were accepted as legal by many, though by no means all, legal scholars.20 However, these scholars tended not to advance the claim that intervention was obligatory.21 Such claims were restricted to the domain of political rhetoric.22 It was generally accepted by jurists that sovereign states had the right to determine for themselves when, where, and how to uphold international law and the rights of humanity. States could not be bound to protect beyond borders where it was not in their interests to do so. Moreover, it was becoming increasingly accepted that sovereign states also had the right to freedom from external interference and intervention. The principle of non-intervention, which had first been articulated by Emmeric de Vattel in the mid-eighteenth century, would become more firmly established in international law through the course of the twentieth century,23 first in the Covenant of the League of Nations and the Kellogg-Briand Pact, and then, even more plainly, in the UN Charter and numerous subsequent General Assembly declarations.24 This development had the effect that, to the extent that the humanitarian intervention was embraced by legal theorists during these decades, it tended to be cautiously framed as a permissible exception to the rule of non-intervention rather than as a demanding legal obligation.25
On 9 December 1948, the General Assembly adopted the Convention on the Prevention and Punishment of the Crime of Genocide.26 In Article I, contracting parties declared genocide to be ‘a crime under international law which they undertake to prevent and to punish’. I will return to the meaning of this provision later in the article.27 For now, it suffices to observe that neither a plain reading of the text nor an analysis of the travaux préparatoires provided clear grounds for thinking that it imposed upon states a duty to take action to prevent the occurrence of genocide beyond their borders, other than perhaps an obligation under Article VIII to ‘call upon the competent organs of the United Nations’ to take ‘appropriate’ action. Nevertheless, with the consolidation of an international human rights regime over the next few decades and particularly with the increased activism of the Security Council after the Cold War, there did emerge a belief in some quarters of international society that the Genocide Convention imposed a duty to intervene beyond borders in response to genocide. This belief was displayed most clearly in the context of the Rwandan genocide. In 1994, the United States, in particular, consciously avoided characterising the violence in Rwanda as ‘genocide’ out of fear that such a determination would generate an obligation to ‘do something’. While some American officials were primarily concerned about the political implications of a genocide finding, others were motivated by reluctance to accept the legal obligations which they feared would ‘arise in connection with the use of the term’.28 However, the Clinton administration later clarified that it believed that the Convention did not impose any particular legal obligations on a state to respond to genocide outside of its territory.29 In 2000, William Schabas was able to conclude: ‘The obligation to prevent genocide is a blank sheet awaiting the inscriptions of State practice and case law.’30
3. The Rapid Emergence of the ‘Responsibility to Protect’
The story of the rapid emergence of the ‘responsibility to protect’ concept, from its articulation by an independent commission in 2001, through its endorsement by member states at the UN World Summit less than four years later, to its widespread acceptance today has been well told elsewhere.31 This section briefly outlines the emergence of the particular proposition which is the focus of this article—that states have not merely a responsibility to protect their own population but a collective responsibility to protect populations beyond their borders—from 2001 through to the present, and to consider the legal force of leading international statements and resolutions on this idea.
In 1991, UN Secretary-General Javier Perez de Cuellar called for an agreement not on ‘the right of intervention but the collective obligation of States to bring relief and redress in human rights emergencies’.32 Through the course of that decade, the Special Representative on Internally Displaced Persons, Francis Deng, developed a principle which he termed ‘sovereignty as responsibility’, which entailed that the enjoyment of the sovereign right of non-interference was conditional upon the performance of sovereign responsibilities for the protection of populations.33 Deng suggested that in those instances in which states failed to carry out their responsibilities, the ‘international community’ had not merely a right but a responsibility to step in: ‘To the extent that the international community is the ultimate guarantor of the universal standards that safeguard the rights of all human beings’, he insisted, ‘it has a corresponding responsibility to provide innocent victims of internal conflicts and gross violations of human rights with essential protection and assistance.’34 This theme was further developed by the International Commission on Intervention and State Sovereignty (ICISS), an independent commission established by the Canadian government, whose report, The Responsibility to Protect, released in December 2001, quickly became a touchstone for debate.35
The Commission developed the idea of the ‘responsibility to protect’ with the intention of overcoming intractable international debate around the concept of humanitarian intervention which had developed in the 1990s and which had become particularly heated in the wake of NATO's intervention in Kosovo in 1999. Echoing the call of Perez de Cuellar, Deng, and others, its objective was to shift the focus of discussion from the rights of interveners to the rights of those in need of protection from mass atrocities, and the corollary duties to ensure that these rights were protected. The Commission wrote of a ‘primary’ or ‘default’ responsibility which was borne by individual states for the protection of their own populations, and a ‘residual’ responsibility which was borne by the wider society of states to act to protect populations, through military intervention if necessary, in instances where states were unwilling or unable to carry out their own responsibilities.36 ‘Where a population is suffering serious harm, as a result of internal war, insurgency, repression or state failure, and the state in question is unwilling or unable to halt or avert it,’ the Commission declared, ‘the principle of non-intervention yields to the international responsibility to protect’.37 It was suggested that the responsibility to protect embraced three specific responsibilities: the responsibility to prevent atrocities; the responsibility to react to atrocities; and the responsibility to rebuild in the aftermath of atrocities. And it was emphasised that prevention was ‘the single most important dimension of the responsibility to protect’.38 However, the bulk of the report was concerned with the more controversial question of reaction: that is, military intervention in instances where prevention had failed.39 Here, the Commission fell back on earlier arguments about the rightfulness of intervention and grounded its claims in traditional, just war criteria, including just cause, right intention, last resort, proportionality, reasonable prospects for success, and right authority.40 The novelty of the Commission's argument was that, whereas scholars during the previous decade had suggested that the satisfaction of just war criteria could generate a right of intervention,41 the ICISS claimed that it actually generated an obligation to intervene. However, little justification was provided for this claim. No legal argument was offered, and even a moral argument was hard to find. It was simply implied that such intervention was demanded by our ‘common humanity’ and by the need to deliver ‘practical protection for ordinary people, at risk of their lives, because their states are unwilling or unable to protect them’.42
In December 2004, the ‘responsibility to protect’ was adopted by a High-Level Panel of experts commissioned by UN Secretary-General Kofi Annan. Appealing to the commitment by states to prevent genocide under the Genocide Convention and an emerging understanding that ‘genocide anywhere is a threat to the security of all’,43 the Panel stated that ‘there is a growing acceptance that while sovereign Governments have the primary responsibility to protect their own citizens from such catastrophes, when they are unable or unwilling to do so that responsibility should be taken up by the wider international community.’44 The Panel wrote of ‘a growing recognition that the issue is not the “right to intervene” of any State, but the “responsibility to protect” of every State when it comes to people suffering from avoidable catastrophe’.45 And it tied this notion of shared responsibility to the Security Council: ‘We endorse the emerging norm that there is a collective international responsibility to protect, exercisable by the Security Council authorizing military intervention as a last resort, in the event of genocide and other large-scale killing, ethnic cleansing or serious violations of international humanitarian law which sovereign Governments have proved powerless or unwilling to prevent.’46 Three months later, the Secretary-General endorsed the principle in his own agenda for UN reform, urging states to recommit themselves to supporting the rule of law and human rights and to ‘Embrace the “responsibility to protect” as a basis for collective action against genocide, ethnic cleansing and crimes against humanity, and agree to act on this responsibility’ by using ‘diplomatic, humanitarian and other methods’ and, if necessary, ‘enforcement action’ authorised by the Security Council.47
Member states unanimously endorsed the ‘responsibility to protect’ at the UN World Summit in September 2005.48 However, the agreement incorporated into the Summit Outcome Document adopted by the General Assembly was more restrained than the visions of the international enforcement of sovereign responsibilities earlier offered by the ICISS, the High-Level Panel, and the Secretary-General. The language of these earlier documents was necessarily diluted in order to build international consensus.49 Several states expressed concern that endorsement of the concept would facilitate self-interested interference by powerful states in the domestic affairs of the weak and some suggested that its interventionist provisions were incompatible with international law.50 Consequently, the agreement expressly tied intervention to the authority of the Security Council, and the responsibility to protect was expressly limited to the specific crimes of genocide, war crimes, ethnic cleansing and crimes against humanity. Most crucially for the present discussion, not only was the scope of intervention restrained, but also the notion that such intervention was a collective international responsibility rather than a discretionary right was watered down. The United States, in particular, was determined to retain its sovereign freedom to decide when and where to respond to mass atrocities and humanitarian crises. Consequently, it sought to reject any suggestion that it might bear a legal obligation to act in a particular way when other states failed to protect their populations.
Echoing the three earlier statements on the ‘responsibility to protect’ discussed above, an early draft of the Summit Outcome Document had stated, ‘we recognize our shared responsibility to take collective action’.51 Upon reading this draft, US Ambassador to the UN, John Bolton, issued a communiqué insisting that it should be made clear that ‘the responsibility of the other countries in the international community is not of the same character as the responsibility of the host’. He acknowledged that states might have a legal obligation to protect their own populations, but insisted that the responsibility of the ‘international community’ to protect was merely a ‘moral responsibility’. He firmly declared: ‘We do not accept that either the United Nations as a whole, or the Security Council, or individual states, have an obligation to intervene under international law.’ Bolton insisted that member states should therefore ‘avoid language that focuses on the obligation or responsibility of the international community and instead assert that we are prepared to take action’. The determination of what particular measures to adopt in specific cases, he insisted, ‘cannot be predetermined in the abstract but should remain a decision within the purview of the Security Council’.52 Bolton was successful in having the language watered down. In the final negotiated document, member states declared that ‘The international community should, as appropriate, encourage and help states’ to protect their populations and that it bears ‘the responsibility to use appropriate diplomatic, humanitarian and other peaceful means, in accordance with Chapters VI and VIII of the Charter, to help to protect populations from genocide, war crimes, ethnic cleansing and crimes against humanity’.53 With respect to Chapter VII enforcement measures which could be undertaken when states had ‘manifestly failed’ to protect their populations, member states merely declared that they were ‘prepared to take collective action … on a case-by-case basis’. They did not explicitly accept that they had an obligation to do so.
Since the World Summit, the ‘responsibility to protect’ has assumed a prominent place in discussions around the prevention and protection of populations from mass atrocities and humanitarian crises. It has been invoked by various actors to advocate international action in response to numerous crises around the world.54 Officials from many states, including the United States, have repeatedly acknowledged that the ‘international community’ does have a responsibility to protect populations when states fail to do so.55 However, they have tended not to articulate this in terms of a legally binding obligation which is borne by particular states or institutions. The Security Council has made express reference to the ‘responsibility to protect’ in resolutions on crises in Darfur (2006), Libya (2011), Côte d'Ivoire (2011) and Yemen (2011), on the need to ensure stability in the recently established South Sudan (2011), and on the theme of protection of civilians in armed conflict (2006 and 2009).56 However, while these resolutions may be read as confirming the acceptance of the broad principle by the society of states, they have done nothing to clarify the legal nature of the particular responsibility to protect populations beyond borders.
In January 2009, UN Secretary-General Ban Ki-moon released a report, ‘Implementing the Responsibility to Protect’, which outlined a three-pillar strategy for advancing the agenda for the ‘responsibility to protect’ that was said to be mandated by the World Summit agreement.57 The first pillar was the responsibilities of states for the protection of their own populations. The second was the responsibility of the ‘international community’ to assist and encourage states to undertake their protection obligations. Both of these pillars reflected the language of the Summit Outcome Document. The third pillar was ‘the responsibility of Member States to respond collectively in a timely and decisive manner when a State is manifestly failing to provide such protection’.58 This pillar included peaceful measures under Chapter VI, enforcement measures under Chapter VII, and regional arrangements under Chapter VIII of the UN Charter. Here, the Secretary-General snuck back in the language of ‘responsibility’ that had been removed from the provisions for international collective action in the Summit Outcome Document at the insistence of Ambassador Bolton. Neither the United States nor any other state has objected to this language in Secretary-General Ban Ki-Moon's report. This is probably because the term ‘responsibility’ seems harmless enough and does not imply a binding legal obligation. In July 2009, the General Assembly debated the report. An overwhelming majority of states spoke in favour of the ‘responsibility to protect’ concept. However, while it was widely accepted that the responsibility of states to protect their own citizens was a principle grounded in existing international law, little was said about the nature of the international collective responsibility to protect beyond borders.59 Two months later, the General Assembly adopted a resolution declaring simply that it ‘[t]akes note of the report of the Secretary-General’ and ‘Decides to continue its consideration of the responsibility to protect’.60
Those are the key international documents and agreements on the ‘responsibility to protect’, as it has emerged from 2001 to 2011, that speak of the collective responsibility of states to protect populations beyond their borders. What are the legal implications of these various statements and resolutions? As Carsten Stahn has observed, none of them can be treated as generating binding legal obligations under the classic sources of international law delineated in Article 38 of the Statute of the International Court of Justice; that is, international conventions, international custom or general principles of law.61 Some legal scholars today accept that General Assembly and Security Council resolutions and, to a lesser extent, reports issued by the Secretary-General and expert panels can contribute to the determination and interpretation of international law. Others continue to insist, however, that such sources at best merely play a contributory role in the gradual development of customary international law.62 Moreover, the most authoritative of the documents analysed above, the Summit Outcome Document adopted by the General Assembly, is also the document that outlines the principle of international collective responsibility most weakly. Analysis of the text of that document, and the negotiations which preceded it, provides little ground for concluding that states intended it to establish new legal obligations for the protection of populations beyond borders.
Nevertheless, there are reasons for thinking that there may be gradually developing legal obligations for the extraterritorial protection of populations. While the key documents on the ‘responsibility to protect’ concept themselves may not be legally binding in any strict positivist sense, complementary customary law developments in recent years do provide grounds upon which the responsibility to protect beyond borders can be rightly understood to rest, at least to some degree. The remainder of this article examines these developments.
4. Legal Developments Toward a Responsibility to Protect Beyond Borders
This section takes the ICJ's judgment in Bosnia v Serbia in 2007 as its starting point and subsequently draws on other case law and the work of the ILC in order to develop a picture of legal developments regarding extraterritorial duties to protect populations. It seeks to understand the scope of the obligation to protect—specifically, which actors bear an obligation to protect, in what kinds of situations, and what this obligation demands of them—and also to consider the limited mechanisms of enforcement and prospects for reparation that accompany this obligation.
A. The Scope of the Obligation to Protect
On 27 February 2007, the ICJ handed down its judgment in the Case Concerning the Application of the Convention on the Prevention and Punishment of the Crime of Genocide (Bosnia and Herzegovina v Serbia and Montenegro).63 The Court made clear that its jurisdiction in the case was confined solely to the crime of genocide, declaring that it had ‘no power to rule on alleged breaches of other obligations under international law, not amounting to genocide, particularly those protecting human rights in armed conflict. That is so even if the alleged breaches are of obligations under peremptory norms, or of obligations which protect essential humanitarian values, and which may be owed erga omnes’.64 It will be suggested later in this article that an argument can be made that obligations to protect populations from the other crimes which fall within the scope of the ‘responsibility to protect’ concept—namely war crimes, crimes against humanity, and ethnic cleansing—do constitute peremptory norms which may be owed erga omnes.65 Nevertheless, the Court's ruling was necessarily restricted to consideration of the crime of genocide.
While Bosnia argued that the totality of crimes committed by the Bosnian Serbs between 1992 and 1995 amounted to genocide, the Court followed the ICTY in finding the 1995 Srebrenica massacres alone to be genocide. The Court then found that the state of Serbia was neither directly responsible for the Srebrenica genocide, nor was it complicit in the genocide.66 However, it did find Serbia responsible for failing to prevent the genocide and to punish the perpetrators. In order to justify its finding, the Court needed to address questions about the scope of the obligation to prevent genocide found in the Genocide Convention: questions such as who bears the obligation, what does the obligation demand, and under what conditions does it arise.
The Court first emphasised that, while prevention and punishment are closely linked in Article 1 of the Genocide Convention, the obligation to prevent is distinct from the obligation to punish: ‘The obligation on each contracting State to prevent genocide is both normative and compelling. It is not merged in the duty to punish, nor can it be regarded as simply a component of that duty.’67 Significantly, the Court determined that the scope of the obligation to prevent extends beyond referral to the competent organs of the United Nations in accordance with Article VIII. ‘Even if and when these organs have been called upon,’ the Court declared, ‘this does not mean that the States parties to the Convention are relieved of the obligation to take such action as they can to prevent genocide from occurring, while respecting the United Nations Charter and any decisions that may have been taken by its competent organs’.68
In an extraordinary paragraph, the Court then indicated that the Convention imposes an obligation upon every state to do all it reasonably can to prevent genocide beyond its borders. It began by establishing that the obligation to prevent genocide is one of conduct and not one of result. A state cannot be under an obligation to succeed in preventing genocide regardless of the circumstances. Rather, the obligation of states is ‘to employ all means reasonably available to them, so as to prevent genocide so far as possible. A State does not incur responsibility simply because the desired result is not achieved; responsibility is however incurred if the State manifestly failed to take all measures to prevent genocide which were within its power, and which might have contributed to preventing the genocide’.69 The Court suggested that this is a duty of ‘due diligence’ which requires ‘an assessment in concreto’. It then proceeded to establish parameters for assessing whether a state has discharged its obligation:
The first, which varies greatly from one State to another, is clearly the capacity to influence effectively the action of persons likely to commit, or already committing, genocide. This capacity itself depends, among other things, on the geographical distance of the State concerned from the scene of the events, and on the strength of the political links, as well as links of all other kinds, between the authorities of that State and the main actors in the events. The State’s capacity to influence must also be assessed by legal criteria, since it is clear that every State may only act within the limits permitted by international law; seen thus, a State’s capacity to influence may vary depending on its particular legal position vis-à-vis the situations and persons facing the danger, or the reality, of genocide.70
The Court then returned to the idea with which it began the paragraph, declaring that ‘it is irrelevant whether the State whose responsibility is in issue claims, or even proves, that even if it had employed all means reasonably at its disposal, they would not have sufficed to prevent the commission of genocide’. It suggested that this was irrelevant ‘since the possibility remains that the combined efforts of several States, each complying with its obligation to prevent, might have achieved the result – averting the commission of genocide – which the efforts of only one State were insufficient to produce’.71
(i) Who Bears the Obligation?
This remarkable paragraph demands close consideration. To begin with, the Court makes clear that the obligation to prevent genocide is not territorially limited. As Marko Milanovic has observed, while the positive obligation of due diligence under most human rights treaties is dependent upon a state having jurisdiction over a certain person or territory, the Court did not find any such threshold criteria in the Genocide Convention.72 Rather, the scope of a state's obligation is simply determined by its ‘capacity to influence effectively’ the genocidal actors, as far as is permissible under international law.73 The obligation, then, would appear to be borne by every state to a greater or lesser degree.74 A state with intimate ties to the genocidal actors, as Serbia was shown to be in this case, will have a particularly burdensome obligation to act. So too, presumably, will a great power that possesses the ability to persuade or compel persons to refrain from committing the crime. A less influential and weaker state may appropriately be less active, and it may merely cooperate where possible to facilitate the prevention of genocide; yet, even it will not be entirely free from the obligation to prevent under the Genocide Convention.75 After all, the Court declared that states have an obligation ‘to employ all means reasonably available to them, so as to prevent genocide so far as possible’.76 It also indicated that diligent performance of the obligation to prevent may necessarily see multiple states combining their efforts to successfully avert the commission of genocide. This suggestion finds support in Article 41(1) of the ILC's Articles on State Responsibility which provides that ‘States shall cooperate to bring to an end through lawful means any serious breach’ of the peremptory norms of international law.77
In applying its reasoning to the facts of the case, the Court emphasised the unique position of influence that Serbia wielded over the Bosnian Serbs, owing to the strength of their political, military and financial ties. It also noted that Serbia was bound by very specific obligations, by the Court's orders on provisional measures delivered in 1993, to ensure that genocide was not committed by those over whom it exercised influence.78 One observer has taken this to mean that ‘the bar remains very high for establishing the sufficient level of influence necessary for the legal duty to prevent to arise’.79 Another has concluded that Serbia was assigned the duty to prevent genocide only because it had substantially enabled the genocide: ‘Having substantially enabled that conduct, Serbia could not lawfully stand by, even though other states with the capacity to restrain the Bosnian Serbs probably could … [Serbia's] causal connection to the abuse provides the normative justification for assigning it an obligation to protect.’80 Both commentators misread the Court's judgment. In paragraph 430 quoted above, the Court boldly and clearly outlined a much wider scope for the duty to prevent genocide than they suggest. In this paragraph, the Court did not indicate that the degree of influence needs to be as high as it was in the case of Serbia, and it certainly did not indicate that the responsible state must have enabled the genocide. Rather, it indicated that all states bear the duty to prevent, a duty which places greater or lesser demands on states, according to their capacity to influence the genocidal actors.
(ii) What Measures does the Obligation Demand?
The Court provided no indication of what particular measures are demanded by the obligation to prevent genocide. Rather, it found Serbia responsible for failing to take ‘any initiative to prevent what happened, or any action on its part to avert the atrocities which were committed’.81 As quoted above, the Court declared that a state has a rather demanding duty to take ‘all measures to prevent genocide which were within its power’, but it also offered the more restrained suggestion that a state must ‘employ all means reasonably available to them’.82 This notion of ‘all means reasonably available’ is instructive. It is at once expansive and yet bounded in scope. As noted earlier, legal theorists and moral philosophers have, for centuries, insisted that a state cannot be bound to discharge duties towards strangers to an extent that is excessively costly to itself.83 As James Nickel explains, ‘When we ask whether a certain party can bear a burden, we really want to know whether that party can bear that burden without abandoning other responsibilities that ought not be abandoned.’84 The European Court of Human Rights concurs, suggesting that an obligation ‘must be interpreted in a way which does not impose an impossible or disproportionate burden on the authorities’.85
By refraining from dictating particular measures, the Court accords states some discretion in determining how to discharge their obligation to prevent. This is entirely appropriate, given that different situations will demand different measures, and states will have varying abilities to carry out such measures. Nevertheless, in invoking the notion of due diligence, the Court makes clear that states do not have complete discretion to choose how they will act to prevent genocide. They must use all means reasonably available to them. Indeed, while one commentator has suggested that ‘the obligation never requires a state to use military force’,86 the Court does not provide any grounds for reaching this conclusion. Depending on the situation, a capable state may be rightly expected to apply diplomatic pressure, to impose sanctions, and even to take coercive measures to prevent genocide. To be sure, the Court makes clear that the scope of the obligation cannot exceed the limits of what is permitted by international law. Yet, in those instances where the Security Council authorises the use of force to prevent genocide, a state would seem to be under an obligation to contribute troops to a military intervention if such measures are ‘reasonably available’ to it. The requirement for Security Council authorisation, in turn, suggests that members of the Council, and perhaps even the institution itself, may bear a particular obligation to facilitate the prevention of genocide. This interesting notion is worth briefly considering.
(iii) What about the Security Council and its Members?
It will be recalled that the leading international statements on the ‘responsibility to protect’ that followed the ICISS report expressly wed the collective responsibility of states to the authority of the Security Council, though the language of the 2005 World Summit agreement indicated an unwillingness amongst states to recognise that the Council and its members bear legal obligations for the protection of populations. Over the past few years, taking as a model its earlier articles on state responsibility, the ILC has developed and codified legal rules concerning the responsibility of international organizations. The Draft Articles on the Responsibility of International Organizations were adopted on second reading in June 2011. In developing these Articles, the Commission has offered some bold statements which speak of the collective obligation to prevent genocide. It has indicated that international organisations can be attributed responsibility for internationally wrongful acts—acts of commission and also omission—and that they are bound to obligations arising from peremptory norms of international law, such as the prohibition of genocide, just as individual states are.87 The Special Rapporteur offered the failure of the UN to prevent genocide in Rwanda as an example of an international organisation failing to discharge its legal obligations:
Assuming that general international law requires States and other entities to prevent genocide in the same way as the Convention on the Prevention and Punishment of the Crime of Genocide, and that the United Nations had been in a position to prevent genocide, failure to act would have represented a breach of an international obligation. Difficulties relating to the decision-making process could not exonerate the United Nations.88
However, these suggested legal obligations of the Security Council remain controversial and subject to crucial limitations. Jose Alvarez charges the ILC with making ‘huge leaps of judgment’ which are not supported by the behaviour of states or international organisations,89 and claims that the notion that an international organisation may be legally bound to prevent genocide just as states are is ‘absurdly premature and not likely to be affirmed by state practice’.90 He suggests that there is a lack of clarity around the idea that international organisations such as the United Nations are subject to the rule of law since ‘the institutional practices of these organizations, unlike the internal rules of states, are often a constitutive part of international law’.91 Alvarez observes that, while the ILC's Draft Articles on Responsibility of International Organizations appear to assume that an organisation can commit an internationally wrongful act even when it acts in conformity with its own internal rules, ‘this ostensible rule – drawn from the established rule that a state's internal law provides no excuse from a violation of international law – is extremely difficult to apply to organizational rules that may be both “internal law” and international law’.92 The voting procedures of the Security Council, for example, constitute international law. If member states fail to vote to adopt a resolution facilitating action in response to the commission of genocide, it is difficult to comprehend how the Council itself can be said to have committed an internationally wrongful act.
Further, it remains unsettled whether Security Council decisions can be subject to judicial review. While some commentators argue that the ICJ has the authority to review Council resolutions in certain situations, the Court has yet to do so in practice.93 And while judicial review of Council resolutions may be conceivable, it is more difficult to conceive how the failure of the Council to adopt a resolution may itself be subject to review. Moreover, as Ralph Wilde has observed, the unaccountability of international organisations is ‘not something which has come about by accident’.94 International organisations tend to be developed and supported by states at least in part because this allows them to transfer responsibility for certain activities away from themselves and to entities which are not subject to effective accountability mechanisms.95 Suggestions that the actions of the Security Council ought to be subject to legal scrutiny appear unlikely to be affirmed in the foreseeable future by its member states and certainly not by its veto-wielding members.
Given the difficulties in constructing a theory of liability of the Security Council, a number of advocates of the ‘responsibility to protect’ principle have placed significant weight on the potential international legal responsibilities of the Council's individual member states and particularly the veto-wielding permanent members. Louise Arbour, for example, has wondered ‘why the exercise of a veto blocking an initiative designed to reduce the risk of, or put an end to, genocide would not constitute a violation of the vetoing States' obligations under the Genocide Convention’.96 Anne Peters has similarly suggested that the exercise of a veto could ‘under special circumstances constitute an abus de droit by a permanent member’.97
In response to such claims, some commentators have replied that the voting behaviours of states represent political decisions about the implementation of law rather than decisions which are themselves bound by law. There is no requirement for a permanent member to justify the exercise of a veto, they observe, and there are no established guidelines, much less mechanisms, to objectively evaluate the legitimacy of their exercise.98 To suggest that the exercise of a veto may in certain circumstances be illegal, they argue, ‘misunderstands the political nature of the Security Council’.99 Such arguments are not entirely persuasive. All decisions made by states about how to respond to the threat or commission of genocide are political, regardless of whether they are made within or outside of the Security Council, but this does not mean that such decisions cannot also be bound by law. The fact that permanent members have discretion under the UN Charter as to how they exercise their veto does not mean that they cannot also have assumed obligations under other treaties, such as the Genocide Convention, that bounds that discretion.100 To the extent that states can be liable for seeking to circumvent their international obligations when acting through an international organisation, and the Draft Articles on Responsibility of International Organizations insists that they can be,101 there is no obvious reason why the voting behaviour of a member state in the Security Council, when confronted with genocide, is any less bound by law than are the actions of non-member states who might similarly possess some capacity to effectively influence the genocidal actors.
Nevertheless, while its Bosnia v Serbia judgment may have indicated that the scope of the international obligation to prevent genocide is very broad, it must be acknowledged that the ICJ has to date made no clear ruling on the specific obligations of states as members of the Security Council.102 This law remains to be written. Moreover, by their rejection of suggested guidelines for Council responses to mass atrocities and their rejection of proposed agreements to restrict the exercise of the veto in cases of mass atrocities during the lead up to the 2005 World Summit, the permanent members have made clear their opposition to moves to subject their actions within the Council to legal scrutiny.103
(iv) Under What Conditions does the Obligation to Protect Arise?
In its Bosnia v Serbia judgment, the ICJ referred to a general rule of the law of state responsibility codified in Article 14(3) of the ILC's Articles on State Responsibility,104 and declared that ‘a State can be held responsible for breaching the obligation to prevent genocide only if genocide was actually committed.’105 However, it insisted, ‘a State's obligation to prevent, and the corresponding duty to act, arise at the instant that the State learns of, or should normally have learned of, the existence of a serious risk that genocide will be committed. From that moment onwards, if the State has available to it means likely to have a deterrent effect on those suspected of preparing genocide, or reasonably suspected of harbouring specific intent (dolus specialis), it is under a duty to make such use of these means as the circumstances permit.’106 So while the obligation to prevent genocide can be breached only when genocide actually occurs, the obligation arises as soon as there is a known serious risk that the crime will be committed.
While this paragraph imposes demanding obligations on states, its temporal scope falls well short of the entirety of responsibilities said to be borne by the ‘international community’ according to the ‘responsibility to protect’ concept. The temporal scope suggested by the Court does appear to encompass the ‘responsibility to protect’ concept's notion of collective action taken in instances where a state is already ‘manifestly failing’ to protect its population from genocide. However, the ‘responsibility to use appropriate diplomatic, humanitarian and other peaceful means, in accordance with Chapters VI and VIII of the Charter, to help protect populations', as posited in the World Summit agreement,107 would only find legal grounds in the Bosnia v Serbia judgment in situations in which there is a known serious risk that genocide will be committed. Moreover, the Court's judgment provides no legal support for the broader notion that the ‘international community’ has an obligation to ‘encourage and help’ states to protect their populations,108 or to assist them in developing domestic institutions and building their capacity to prevent genocide,109 before the threat of genocide has fully emerged. Such expansive and vague responsibilities do not find legal support in other case law or in state practice either, and it is difficult to imagine how legal support for them could emerge in the foreseeable future.
(v) What About ‘Responsibility to Protect’ Against Crimes other than Genocide?
In the Bosnia v Serbia case, the ICJ made clear that it was confining itself to determining the specific scope of the duty to prevent in the Genocide Convention. It did not ‘purport to establish a general jurisprudence applicable to all cases where a treaty instrument, or other binding legal norm, includes an obligation for States to prevent certain acts’. And it certainly did not ‘purport to find whether, apart from the texts applicable to specific fields, there is a general obligation on States to prevent the commission by other persons or entities of acts contrary to certain norms of general international law’.110 The judgment of the Court, therefore, does not provide legal grounds for an obligation to protect populations against the other crimes encompassed within the ‘responsibility to protect’ concept—war crimes, crimes against humanity, and ethnic cleansing.111 Nevertheless, as noted earlier, despite refusing to rule on such crimes, the Court did acknowledge that states may bear obligations other than the prevention of genocide which ‘protect essential humanitarian values, and which may be owed erga omnes’.112 An argument can be tentatively made that the other ‘responsibility to protect’ crimes do constitute breaches of peremptory norms, and their prevention may be considered obligations owed erga omnes.
Article 53 of the Vienna Convention on the Law of Treaties113 defines a peremptory norm of international law as one which is ‘accepted and recognized by the international community of States as a whole as a norm from which no derogation is permitted and which can be modified only by a subsequent norm of general international law having the same character’. The ILC suggests that peremptory norms ‘arise from those substantive rules of conduct that prohibit what has come to be seen as intolerable because of the threat it presents to the survival of States and their peoples and the most basic human values’.114 While the rules prohibiting war crimes and crimes against humanity may not be as firmly established as those prohibiting genocide, Christian Tams argues that there is a ‘huge amount of evidence’ indicating their peremptory nature.115 The ILC includes the prohibition of crimes against humanity alongside the prohibition of genocide in a list of peremptory norms and also suggests that the ICJ's claim in the Legality of the Threat or Use of Nuclear Weapons case that the basic rules of international humanitarian law applicable in armed conflict are ‘intransgressible’ would seem to justify their being treated as peremptory.116
An argument has emerged in recent years that the obligation to end breaches of peremptory rules of international law is one which falls on all states, regardless of where the breaches occur.117 The ILC seeks to establish this in its Draft Articles on State Responsibility. Article 41(1) provides that ‘States shall cooperate to bring to an end through lawful means any serious breaches’ of the peremptory norms of international law. Stahn observes that this positive duty ‘comes close to the idea of collective responsibility under the concept of responsibility to protect’.118 Article 41(1) does not prescribe what form cooperation should take. In its commentary, the ILC envisages the possibility of both cooperation ‘organized in the framework of a competent international organization, in particular the United Nations’, and also non-institutionalised cooperation.119 Similarly, Article 41(1) does not prescribe what measures states should take to bring to an end a serious breach of a peremptory rule. The commentary suggests that ‘[w]hat is called for in the face of serious breaches is a joint and coordinated effort by all States to counteract the effects of these breaches.’120 However, the Commission acknowledged that this obligation may not yet be well established in international law: ‘It may be open to question whether general international law at present prescribes a positive duty of cooperation, and [Article 41(1)] in that respect may reflect the progressive development of international law.’121 Moreover, even more restrictive than the Bosnia v Serbia judgment, the temporal scope of the obligation posited by the ILC is restricted to instances in which serious breaches of peremptory norms have already been committed. The ILC provides no legal grounds for an obligation to take measures to assist or pressure states to protect their populations from ‘responsibility to protect’ crimes prior to the commission of those crimes.
In sum, recent years have seen a significant shift in the international legal obligations of bystander states. Judgments of the ICJ and the work of the ILC indicate that the obligation to prevent genocide falls on every state that possesses the capacity to effectively influence genocidal actors. This potentially includes states in their capacity as members of the Security Council and perhaps the Council itself, though the law in this regard remains particularly underdeveloped. States are charged with employing ‘all means reasonably available to them’ to prevent the commission of genocide, an obligation which arises as soon as there is a known serious risk that genocide will be committed. An argument can be tentatively made that the obligation to prevent also extends to the other ‘responsibility to protect’ crimes such as war crimes, crimes against humanity, and ethnic cleansing, though the temporal scope of obligation falls far short of the expansive responsibilities to encourage and assist states to protect their populations prior to the commission of mass atrocities as outlined by the ‘responsibility to protect’ concept.
Nevertheless, despite the significant developments that have been made in recent years with respect to obligations to protect populations beyond borders, a serious lack of clarity or even coherence remains around questions of enforcement and reparation. It is to these questions that this article finally turns.
B. Issues of Enforcement and Reparation
It has been suggested by some commentators that the responsibility to protect beyond borders does not constitute a legal obligation because it is unenforceable. Alvarez, for example, observes: ‘If there is such a thing as a responsibility to protect, the legal mind naturally assumes that a failure to exercise such responsibility is an internationally wrongful act entailing the usual panoply of potential remedies, including the legal liability of the wrongful actor and the potential for countermeasures against that actor by others.’ He suggests that claims that there is a legal obligation to protect in this sense are ‘absurdly premature’.122 Monica Hakimi similarly suggests that the ‘responsibility to protect’ concept ‘does not reflect a legally operative obligation to protect’ since it is ‘essentially unenforceable and, in practice, unenforced against particular bystander states’.123 These critiques are not unpersuasive. Nevertheless, there are again some recent legal developments which in some small measure begin to respond to these shortcomings and which are worth briefly considering.
The Bosnia v Serbia judgment of the ICJ is again instructive. It will be recalled that the Court found that it did not need to determine whether Serbia could have prevented the genocide from taking place in Srebrenica in order to find Serbia in breach of its obligation to prevent.124 However, it suggested that the same rule did not apply with respect to the question of reparations. ‘Since it now has to rule on the claim for reparation,’ the Court stated, ‘it must ascertain whether, and to what extent, the injury asserted by the Applicant is the consequence of wrongful conduct by the Respondent … The question is whether there is a sufficiently direct and certain causal nexus between the wrongful act, the Respondent's breach of the obligation to prevent genocide, and the injury suffered by the Applicant, consisting of all damage of any type, material or moral, caused by the acts of genocide’. The Court found that it could not conclude ‘with a sufficient degree of certainty that the genocide at Srebrenica would in fact have been averted if the Respondent had acted in compliance with its legal obligations’. Since it could not establish a causal nexus between Serbia's violation of its obligation to prevent and the damage resulting from the genocide at Srebrenica, the Court found that ‘financial compensation is not the appropriate form of reparation for the breach of the obligation to prevent genocide’.125 Rather, the Court held that the form of reparation to which Bosnia was entitled was satisfaction, and the appropriate form of satisfaction was the Court's declaration that Serbia had failed to comply with its obligation to prevent genocide under the Genocide Convention.126
This decision has been criticised by a number of commentators.127 Milanovic suggests that the requirement for causality with respect to reparations was ‘excessive’.128 He observes that the Court cited no authority for its position and insists there is actually ample jurisprudential evidence to the contrary. Both the European Court of Human Rights and the Inter-American Court have awarded compensation against states that failed to prevent human rights violations within their jurisdiction without requiring establishment that the violations would definitely have been prevented if the states had discharged their obligations.129 Moreover, Milanovic notes, the ILC Draft Articles on State Responsibility do not distinguish between the different forms of reparation on the basis of causality.130 The ILC makes clear that causality is a necessary condition for any form of reparation.131 The three forms of reparation—restitution, compensation and satisfaction—are said to differ according to their suitability to alleviate the injury caused, not according to the causal relationship between the wrongful act and the injury.132 It makes little sense, therefore, for the Court to argue that there is enough causality for satisfaction, but not enough for compensation.133 Milanovic suggests that, just as the scope of the obligation to prevent genocide is proportionate with the state's capacity to effectively influence the genocidal actors, so too should be the scope of reparation for failing to discharge this obligation.134 He laments that, having interpreted the obligation to prevent genocide in a very expansive way, as an obligation of every state to do what it reasonably can, the Court reduced the obligation to ‘mere symbolism, by setting aside any truly meaningful form of reparation’.135 The Court's finding, then, poses a significant limitation on the extent to which states will feel compelled to discharge their emerging legal obligations for the extraterritorial protection of populations.
Not surprisingly, the mechanisms for enforcement and reparation available with respect to the responsibility of international organisations are even more tenuous. The ILC's Draft Articles on Responsibility of International Organizations includes a chapter on ‘reparation for injury’ which is modelled on the chapter by the same name found in its Draft Articles on State Responsibility and includes provisions for the same three forms of reparation—restitution, compensation and satisfaction—for injury caused by internationally wrongful acts. Articles 37(2) and 37(3) declare that ‘[s]atisfaction may consist in an acknowledgment of the breach, an expression of regret, a formal apology or another appropriate modality,’ and that it should be proportionate to the injury. In its commentary, the Commission offers the apologies and expressions of regret offered by UN Secretary-General Kofi Annan for the failure to prevent the genocides in Srebrenica and Rwanda as examples of ‘the appropriate legal consequences’ of a breach of an obligation under international law.136 However, it remains far from clear whether the ICJ or any other body bears the authority and the willingness to demand such reparations from an international organisation for injury caused by internationally wrongful acts such as these, and how such a body would determine which particular actor or actors—including the United Nations, the Security Council, member states or the Secretary-General him/herself—are at fault.137 Moreover, it is unclear what, if anything, is really accomplished by demanding an apology from the Secretary-General in the wake of the slaughter of hundreds of thousands of civilians.138
In short, while the key international statements and resolutions on the ‘responsibility to protect’ may not themselves establish legal obligations for the extraterritorial protection of populations, customary law developments in recent years do provide grounds upon which at least some aspects of the ‘responsibility to protect’ beyond borders can be rightly understood to rest. It has been established that every state with the capacity to effectively influence genocidal actors has an obligation to take all means reasonably available to prevent genocide as soon as they become aware of a serious risk that the crime will be committed. Moreover, it can be cautiously argued that this obligation extends to the other ‘responsibility to protect’ crimes including war crimes, crimes against humanity and ethnic cleansing, though no legal grounds are to be found for an obligation to assist states to develop the capacity to protect their populations prior to the commission of these crimes, as suggested in the ‘responsibility to protect’ concept, and it is difficult to conceive how such grounds might be established in the foreseeable future.
However, issues of enforcement and reparation with respect to the obligation to protect populations beyond borders remain seriously under-developed. This, no doubt, limits the extent to which states feel compelled to carry out their extraterritorial obligations. Consequently, it is unlikely that the law as it stands will, by itself, be able to consistently stimulate timely, coordinated and effective international responses to mass atrocities. While, the Security Council may have acted reasonably quickly and authorised the use of ‘all necessary measures’ to protect civilians from the threat of mass atrocities in Libya, there are little grounds for thinking that member states will feel legally compelled to respond with the same determination to a similar crisis in the near future if they do not wish to do so.139 The legal developments examined in this article are significant, and they may turn out to represent small steps towards binding and enforceable customary law. Nevertheless, it would seem that, until measures of enforcement and reparation are inscribed in law, the extraterritorial protection of populations from mass atrocities will continue to be dependent, as it has long been, on the vagaries of the political will of states.
I am grateful to Alex Bellamy, Sara Davies and Stewart Webster for advice that strengthened this article.