Adam Roberts, "NATO's 'Humanitarian War' Over Kosovo," Survival, Vol. 41, no. 3 (1 October 1999)Adam Roberts, "NATO's 'Humanitarian War' Over Kosovo," Survival, Vol. 41, no. 3 (1 October 1999) The 11-week bombing campaign conducted by NATO in spring 1999 against the Federal Republic of Yugoslavia (FRY) has many claims to uniqueness. It was the first sustained use of armed force by the NATO alliance in its so-year existence; the first time a major use of destructive armed force had been undertaken with the stated purpose of implementing UN Security Council resolutions but without Security Council authorisation; the first major bombing campaign intended to bring a halt to crimes against humanity being committed by a state within its own borders; and the first bombing campaign of which it could be claimed that it had on its own, and without sustained land operations, brought about a major change of policy by the target government. NATO leaders were reluctant to call their action 'war'. However, it was war - albeit war of a peculiarly asymmetric kind. It indisputably involved large scale and opposed use of force against a foreign state and its armed forces. Because it was justified principally in terms of stopping actual and anticipated Serb killings and expulsions in the Serbian province of Kosovo, the campaign was sometimes colloquially called a 'humanitarian war'. Whatever the nomenclature, Operation Allied Force marked a high point in the increasing emphasis on human rights and humanitarian issues which has been a striking feature of international relations in the post-1945 era. For theoreticians of international relations it represented a further remarkable twist in the strange and long-running association between the supposedly hard-nosed and 'realist' factor of force, and the supposedly soft and 'idealist' factor of international humanitarian and human-rights norms. The date of 24 March 1999 was doubly significant for human rights in international relations. It was the day when the Appeal Chamber of the UK House of Lords, following a second hearing of the matter, announced its decision that, in principle, Chilean ex-President Augusto Pinochet could be extradited to Spain. This ruling was a landmark in the evolution of the idea that there are some crimes so extreme that a leader responsible for them, despite the principle of sovereign immunity, can be extradited and tried in foreign courts. NATO's Operation Allied Force was also launched on 24 March. The operation was announced at the start as based on the idea (closely related to the one advanced in the Pinochet decision) that there are some crimes so extreme that a state responsible for them, despite the principle of sovereignty, may properly be the subject of military intervention. The international human-rights movement - a huge array of individuals, non-governmental organisations (NGOs), inter-governmental bodies and more - was deeply divided over Operation Allied Force. This reaction was not surprising: the human-rights movement was naturally unhappy to see human rights and international humanitarian law become a basis for initiating war. In particular it was doubtful about the air campaign, because in the short term it failed to stop, and probably even exacerbated, extreme violence against Kosovars. Throughout the air campaign, NATO leaders repeatedly emphasised five objectives which Yugoslav President Slobodan Milosevic was required to accept: a verifiable cessation of all combat activities and killings; withdrawal of Serb military, police and paramilitary forces from Kosovo; the deployment of an international military force; the return of all refugees and unimpeded access for humanitarian aid; and a political framework for Kosovo building on the Rambouillet Accords.1 The set of agreements concluded on 3-10 June under which Yugoslav forces left Kosovo reflected the main NATO demands. It could easily be interpreted as a triumph for bombing as a means of opposing extreme human-rights violations. However, such a judgement may be premature. There is intense debate about what constituted the effective elements of the military campaign. Further, its final outcome as regards the political future of Kosovo and the FRY is not clear. What can now be done is to identify, and begin to explore, some of the difficult questions arising from the campaign. Six are considered here. Why did NATO embark on the use of force, and was it legitimate under international law? Why was such reliance placed on air-power? Did the NATO air campaign lead to an intensification of Serb atrocities in Kosovo? What problems vis-a-vis the laws of war were posed by the air campaign? What factors led to the settlement of lo June? What can be learned about international decision-making on the use of force, and on responding to massive human-rights violations? NATO's resort to force The bombing campaign, which had many causes, marked a significant break from NATO's previous policy and practice. The world's most effective military alliance, with a successful record of helping to maintain peace in Europe for half a century through deterrence and defence, committed its forces and prestige to a major exercise in strategic coercion, seeking to influence the outcome of a largely civil war within one Balkan state. Even more remarkable, it did so without the explicit approval of the UN Security Council. The main underlying explanation for the willingness of NATO's 19 member states to take action over Kosovo is not their interpretations of particular events, such as the failure of the negotiations over the province at Rambouillet and Paris in February and March 1999. Nor was it a shared vision as to what the future of the province should be. Rather, the NATO states were united by a sense of shame that, in the first four years of atrocious wars in the former Yugoslavia (1991-95), they had failed, individually and collectively, to devise coherent policies and to engage in decisive actions. In the last months of 1998 and the first months of 1999 it became evident that the bitter war between the Kosovo Liberation Army (KLA) and the Yugoslav army was at risk of developing into wholesale 'ethnic cleansing' of the Kosovar Albanians, who constituted over 8o% of the province's population.2 Further, it became increasingly clear that the recommendations, resolutions and roles of outside institutions - the European Union (EU), NATO, the Organisation for Security and Cooperation in Europe (OSCE) and UN - were being ignored or violated, especially by the Yugoslav authorities. In these circumstances, whatever doubts existed in individual NATO member states about the wisdom of the diplomatic and military policies pursued by NATO, there was no obvious alternative course of action. Furthermore, because of the alliance's chequered previous record in Yugoslavia, and also because of an appreciation of the inherent value of sticking together, no NATO state wanted to be the first to step out of line. The absence of UN Security Council authorisation for the use of force against Yugoslavia was always going to be a difficult problem for NATO. From the early stages of the deterioration of the situation in Kosovo in early 1998, the Security Council had been willing to impose an arms embargo on Yugoslavia in respect of Kosovo, and also to exert other pressure on Belgrade to moderate its policies in the province.3 However, Russia and China had consistently made it clear that they would veto any proposal for military action against Yugoslavia regarding its conduct in its own territory. Equally consistently they stressed the importance of the non-intervention norm as the essential basis of the UN and of the present system of international security. Was NATO right to launch Operation Allied Force without at least making an attempt to get authorisation from the Security Council? The argument for having at least tried is that the effort would have shown respect for the UN, and would have enabled people around the world to see exactly which states were refusing to authorise action to stop atrocities. However, the, argument against seeking authorisation weighed more heavily with NATO governments: it could have been more difficult to get public support for a military action which had actually been vetoed in the UN, and the whole process might expose divisions in the alliance. Thus NATO's first major military campaign took place in circumstances where there was significant scope for disagreement about the legality of the operation. Lawyers tend to like a world of clarity, where an action can be distinctly categorised as legal or illegal. Politicians and members of the public around the world look to law to provide clear guidance, or at least a verbal bludgeon with which to assault their opponents. In reality, because contradictory principles were inescapably at the heart of this crisis, there was no definitive legal answer that could satisfy a convincing majority of the world's peoples, governments or even international lawyers. Law can provide principles, guidelines, procedures, but not always absolute answers. UN Secretary-- General Kofi Annan showed awareness of this when, at the beginning of the bombing campaign, he issued a statement which recognised that there were occasions when force might be necessary, but also referred to the importance of Security Council authorisation.4 Although NATO's decision to use armed force in the form of air-power did not have as clear a legal endorsement as its governments might have wished, it was far from being in unambiguous violation of international law. Two main legal arguments were used in support, the first based on UN Security Council resolutions, the second on general international law. UN Security Council resolutions Resolution 1199 of 23 September 1998, in particular, had demanded that Yugoslavia inter alia 'cease all action by the security forces affecting the civilian population', and had referred to possible 'further action' if measures demanded in the resolution were not taken. In addition, Resolution 1203 of 24 October 1998, by demanding Serb compliance with a number of key provisions of accords concluded in Belgrade on 15-16 October (including with the NATO Air Verification Mission over Kosovo), accepted that the Alliance had a direct standing and interest in the Kosovo issue. An argument can be made that, even if the Security Council was not able to follow these resolutions on Kosovo with a specific authority to use force, they provided some legal basis for military action. On 26 March 1999, two days after the bombing began, the Security Council did, in a curious way, give at least a crumb of legal comfort to the NATO cause. A draft resolution sponsored by Russia (and supported by two non-Council members, India and Belarus) called for 'an immediate cessation of the use of force against the Federal Republic of Yugoslavia'. Only three states (Russia, China and Namibia) voted in favour, and 12 against. In the debate, the speeches in support of the resolution did not address in any detail the question of what to do about Kosovo. The representative of Slovenia, which was among the states opposing the resolution, made the key point that the Security Council does not have a monopoly on decision-making regarding the use of force. It has 'the primary, but not exclusive, responsibility for maintaining international peace and security'.5 While this debate confirmed that the NATO action was not considered manifestly illegal, a failed draft resolution is not a strong basis for arguing the legality of a military action, and this episode was rarely mentioned in statements by NATO leaders. General international law Several NATO governments put forward an argument that military intervention against another state could be justified in cases of overwhelming humanitarian necessity. The main basis for such an argument is general international law, but there may also be some element of reliance on the UN Charter or on Security Council resolutions. A UK Foreign and Commonwealth Office note of October 1998, circulated to NATO allies, suggests elements of both these approaches: Security Council authorisation to use force for humanitarian purposes is now widely accepted (Bosnia and Somalia provided firm legal precedents). A UNSCR would give a clear legal base for NATO action, as well as being politically desirable. But force can also be justified on the grounds of overwhelming humanitarian necessity without a UNSCR. The following criteria would need to be applied. (a) that there is convincing evidence, generally accepted by the international community as a whole, of extreme humanitarian distress on a large scale, requiring immediate and urgent relief; (b) that it is objectively clear that there is no practicable alternative to the use of force if lives are to be saved; (c) that the proposed use of force is necessary and proportionate to the aim (the relief of humanitarian need) and is strictly limited in time and scope to this aim - i.e. it is the minimum necessary to achieve that end. It would also be necessary at the appropriate stage to assess the targets against this criterion. There is convincing evidence of an impending humanitarian catastrophe (SCR 1199 and the UNSG's and UNHCR's reports). We judge on the evidence of FRY handling of Kosovo throughout this year that a humanitarian catastrophe cannot be averted unless Milosevic is dissuaded from further repressive acts, and that only the proposed threat of force will achieve this objective. The UK's view is therefore that, as matters now stand and if action through the Security Council is not possible, military intervention by NATO is lawful on grounds of overwhelming humanitarian necessity.6 The argument that general international law provides a basis for military intervention can be reinforced by reference to bodies of law which have developed considerably since the UN Charter was drawn up in 1945. In particular, crimes against humanity, violations of the 1948 Genocide Convention, and violations of the 1949 Geneva Conventions may all constitute grounds for intervention, even though these and related agreements do not provide explicitly for military preventive measures against states violating their provisions. In this perspective, it cannot be right to tolerate acts which violate widely supported legal norms just because the Charter does not explicitly provide for military action in such circumstances, or because a veto on the Security Council makes UN-authorised action impossible. The NATO governments, although not all justifying the military action in quite the same terms, generally concentrated on these two main arguments. They eschewed detail in their statements, and made little reference to the long tradition of legal writing about humanitarian intervention.' They also said little or nothing about arguably relevant state practice, such as India's war against Pakistan in 1971, which had enabled refugees to return to what became Bangladesh, or the US-led and UN-authorised intervention in Haiti in 1994 which had led to the capitulation of the military regime there.8 In April and May 1999, after Yugoslavia brought a case in the International Court of Justice against certain NATO states, accusing them of illegal use of force, the NATO governments involved generally eschewed the opportunity to make a ringing legal defence of their actions, and largely confined themselves to technical and procedural issues.9 The simple and general statements made by NATO governments in 1998-99, such as that by the UK, were for the most part based on the proposition that the situation faced in Kosovo was exceptional. Additional arguments, overlapping with the two main arguments indicated above, were occasionally used in support of the legitimacy of military action. The most important was that the situation in Kosovo was indeed a threat to international peace and security. Both President Bill Clinton and Prime Minister Tony Blair, in their major speeches on the war, put emphasis on the proposition that a large new wave of refugees from Kosovo could destabilise neighbouring countries and lead to an expansion of the war.lo The fact that there was massive multilateral support within NATO (an organisation in which all 19 member states have in theory the power of veto) confirms that this military action did represent an international-community interest, and not just the interests of one single state. A further element was sometimes woven into the argument, namely the claim that democratic states have a greater right to engage in military interventions than do autocracies; or at least have a greater claim to international support when they do so. The fact that 19 states with multi-party democratic systems did act collectively is impressive, and the democratic nature of their systems may have helped to place certain restraints on the means used and on the goals of the military operation. However, existing international law relating to the legitimacy of resort to force does not depend to any significant degree on the fundamental distinction between democratic and autocratic states. In UN-based as well as European institutions, democracy may be emerging as an important criterion whereby a state's claims to be a legitimate member of international society are judged, but this has yet to be reflected in the body of international law relating to intervention. In summary, there was an international legal basis for the action taken by NATO over Kosovo. The two main planks of the legal basis (one consisting of requirements in Security Council resolutions, the other drawing on general international law), both placed central emphasis on the protection of the inhabitants of Kosovo. However, any justification of 'humanitarian intervention' along these lines is subject to four important caveats. Since no existing international legal instrument provides explicitly for forcible military intervention within a state on humanitarian grounds, neither of the main arguments indicated above gives an incontestable basis for the NATO action. It is thus in the nature of things that different individuals and states see the matter differently. The question of the military means pursued by NATO to secure the proclaimed political and humanitarian ends was bound to affect judgements about the legality of the operation. NATO's reliance on bombing did give rise to questions (discussed further below) about its appropriateness so far as protecting the inhabitants of Kosovo was concerned, and about its conformity with the laws of war. The argument that a regional alliance has a general right and even a duty to act as vigilante for UN Security Council resolutions, while it may have the considerable merit of ensuring that such resolutions are taken seriously, could also create a risk of undermining international inhibitions against the use of force. Questions were inevitably raised about the selectivity of the action taken by NATO. The obvious question raised by Serbs was why NATO had acted over Kosovo when nothing had been done to stop the Croatian government's ethnic cleansing of Serbs from the Krajina in 1995: that episode has been conveniently expunged from Western collective memories, but it is not forgotten in Belgrade, where the refugees from Croatia are still a conspicuous presence. There were many other equally pertinent questions, not least why NATO had not acted with equal resolve against the FRY when Yugoslav forces had attacked Dubrovnik and Vukovar in Croatia in 1991-92. The motives for the NATO military action included many elements which were not purely humanitarian, and not exclusively concerned with Kosovo. Apart from elements already mentioned (guilt over past inaction regarding Bosnia, and concern over peace and security in the region generally), factors influencing the decisions of NATO states included their reluctance to accept large numbers of refugees on a permanent basis. A further key element was NATO's credibility: having become deeply involved in 1998 in international diplomacy regarding Kosovo, particularly in making military threats to Belgrade and in underwriting agreements, NATO would indeed have lost credibility had it not acted after it became apparent that agreements were not being observed. Needless to say, other more sinister motives were attributed to NATO. One of the more outlandish theories purporting to explain Operation Allied Force was that the Western states had failed to solve the 'Millennium Bug' problem in the computer programmes of their cruise missiles and other 'smart weapons': thus, in a new version of 'use it or lose it', the weapons had to be used in 1999. The available evidence suggests that the critical considerations impelling NATO to take action were those of humanity and credibility. An amalgam of these factors was apparent in the justification for the use of force made by UK Foreign Secretary Robin Cook in a House of Commons debate on 25 March 1999: Since March last year, well over 400,000 people in Kosovo have at some point been driven from their homes. This is about a fifth of the total population. In Britain the equivalent would be over ten million people. I defy any Hon. Member to meet the Kosovar Albanians, to whom I have talked repeatedly over the past three months, and tell them that we know what is being done to their families, that we see it every night on the television in our own homes, that in the region we have a powerful fleet of allied planes; and yet that, although we know what is happening and have the power to intervene, we have chosen not to do so. Not to have acted, when we knew the atrocities that were being committed, would have made us complicit in their repression ... The first reason why we took action was that we were aware of the atrocities that had been carried out and we had the capacity to intervene, but that is not the only reason. Our confidence in our peace and security depends on the credibility of NATO. Last October, NATO guaranteed the cease-fire that President Milosevic signed. He has comprehensively shattered that cease-fire. What possible credibility would NATO have next time that our security was challenged if we did not honour that guarantee? The consequences of NATO inaction would be far worse than the result of NATO action.11 The decision to take action was a step into the unknown for an organisation which had spent its first 50 years carefully crafting military threats which did not in the event have to be executed. At least until the failure at Rambouillet in February, it had been quite commonly assumed in NATO capitals that the threats against Yugoslavia would not actually have to be implemented. However, the NATO decision to use force was facilitated by the belief, widely but not universally held at NATO headquarters and among member governments, that bombing would achieve results in a short time. The reliance on air-power The NATO campaign was overwhelmingly in the air. Allied pilots flew 37,465 sorties, of which over 14,oo6 were strike missions. As the campaign progressed, it grew in intensity. By the time the air campaign was suspended on lo June, Operation Allied Force had 912 aircraft and 35 ships - almost triple the forces that the campaign started with.2 How did it happen that the ancient and ever-contested idea of 'humanitarian intervention' came to be associated with bombing? Why did Robin Cook refer to that 'powerful fleet of allied planes'? In the long history of legal debates about humanitarian intervention, there has been a consistent failure to address directly the question of the methods used in such interventions. It is almost as if the labelling of an intervention as 'humanitarian' provides sufficient justification in itself, and there is no need to think further about the aims of the operation or the means employed - or indeed to understand the society in which the intervention occurs. In the 199os, a high price has been paid for the failure to address seriously the questions of means and purposes in several interventions, including in Somalia. The idea that air-power can be the means of implementing the decisions of the international community is not new. In 1944-45, when the UN Charter was being drawn up, and before sobering truths had emerged about the limited effectiveness of strategic bombing in the Second World War, both the Western Allies and the Soviet Union had a vision of the will of the international community being imposed by air-power. One result was the little-known Article 45 of the UN Charter, which states that 'Members shall hold immediately available national air-force contingents for combined international enforcement action'. Belatedly, and without UN blessing, that vision was implemented over Kosovo. There were two principal explanations for this high degree of reliance on air-power. First, the NATO member states were not willing to risk lives in this operation. A problem which has stalked all interventions with a basically humanitarian purpose in the 199os is that the Western powers that are willing to intervene militarily are reluctant to accept the risk of casualties. This leads to particular modes of operation, such as hesitant and temporary military involvements, and reliance on air-power, that may conflict with the supposed humanitarian aims of the operation. Air-power, such as that used over Iraq in 1991 and subsequently, can be relatively risk-free. Nonetheless, in Kosovo it was an astonishing achievement to engage in acts of war against a well-armed sovereign state for 11 weeks and not incur a single combat casualty. The second reason for the degree of reliance on air-power was a questionable reading of the history of the Bosnian war. It is perfectly possible that the NATO bombing campaign, Operation Deliberate Force, which began at the end of August 1995 and attacked Serb targets in Bosnia, contributed something to Serb weakness and eventual acceptance of a cease-fire; and it may also have contributed to Serb willingness to agree to a less-than-ideal solution in the subsequent negotiations at Dayton.13 However, the mythologising about that campaign ignored one inconvenient fact: that it followed a period of sharp Serb military reverses on the ground, including the mass expulsion of the Serbs from the Croatian Krajina. Also the 1995 bombing was not against Serbia proper, and thus did not arouse the same nationalist response as would the bombing in 1999. The real lesson of those 1995 events might be a very different one: that if NATO wants to have some effect, including through air-power, it needs to have allies among the local belligerents, and a credible land-force component to its strategy. That conclusion takes one off the high moral pedestal associated with the idea of humanitarian intervention, and involves messy bargaining and compromise, but does point to mechanisms for achieving results. The false analogy with the bombing in Bosnia in 1995 appears to have played a significant part in leading to the most extraordinary miscalculation of the whole Kosovo campaign: namely that Belgrade would be likely to give in after a short period, perhaps only a few days, of bombing. This illusion appears to have been widely held in NATO headquarters and national capitals.14 Western statements about this military action showed remarkably little understanding of the way Serbs think about their country and its defence. It is true that Serbia has for some years been deeply demoralised and divided, that its citizens were not all equally attached to Kosovo, and that its capacity to withstand the opposition of 19 NATO member states was limited. Yet many Serbs, steeped in a martial tradition, have held a heroic, xenophobic and dated view of their place in the world, according to which Serbia faced off the Ottomans in the early nineteenth century and the Austrians in the First World War, and Yugoslavia stood alone against Hitler in 1941 and Stalin in 1948-53. A people with the image of themselves as suffering courageously in a deeply hostile world, and as having a personal obligation to defence, was never likely to make a simple cost-benefit analysis of bombing, or to crumple quickly in face of a bombing campaign alone. The problem was not simply President Milosevic, but the mentality of many Serbs. The bombing campaign had twin but distinct aims, which can be roughly summarised as reducing Serb military capacity (including capacity for repression) in Kosovo; and putting pressure on the Yugoslav regime to modify or abandon its policies there. In the October 1998 crisis over Kosovo, the threat of air-power was explicitly made to Milosevic as a means of inducing him to comply with the demands of the Contact Group and the UN Security Council. Yet when Operation Allied Force began, it was widely presented as having the purpose of reducing Serb military capacity there. Certain UK official statements illustrate the emphasis initially placed on reducing Serb repressive capacity in Kosovo, and doing so through bombing alone. On 24 March, hours before the first attacks took place, Defence Minister George Robertson reminded the House of Commons Select Committee on Defence that in October, when the OSCE Verification Mission had been established, Serb forces in Kosovo were to be reduced to roughly zo,ooo internal security and 12,000 Yugoslav army troops; and that, by now, there were 16,000 internal security and 20,000 Yugoslav army troops, plus 8,ooo reinforcements just over the border. He then stated: 'Our military objective our clear, simple military objective - will be to reduce the Serbs' capacity to repress the Albanian population and thus to avert a humanitarian disaster'.15 Robertson did state that the refugee problem would grow. But then he was asked specifically by a member of the Select Committee on Defence: 'With 50,000 Serbian soldiers either in or around Kosovo, once we attack the opportunity for them to give instant payback to the Kosovars is obviously a very great incentive on their part. They will be able to dish out an awful lot of punishment very quickly. What is the plan to safeguard the interest of those Kosovars?' There was no answer, and when the question was repeated George Robertson stated: 'We would clearly take that into account if that was the situation'.16 Another member of the Select Committee who warned that the NATO action might make the clearance more bloody got equally short shrift.17 On targeting, Robertson said: 'Our targets are military and do not involve civilian or urban targets. That is a message that will get through despite the fact that the media is state owned and controlled ... If military action has to be taken ... it will be taken with precision-guided weapons, and it will be taken against only military targets with a very clear objective, not to bomb common sense or even self-interest into the mind of President Milosevic, but to reduce the military capability that is being used against a civilian population'." Shortly afterwards, the chairman said: 'Having clarified their legal status, I presume there will be no formal declaration of war'. Robertson replied, 'It is not a war'. He indicated, as did Clinton and others that day, that there was no plan for a land/air campaign over Kosovo: `NATO has ruled that out'.19 Much that Robertson said on that day was sensible, and his recognition that `we cannot have a casualty-free war' was an implicit acceptance that NATO was getting into something very like war. He correctly recognised that the laws of armed conflict would apply. Nothing that he said was egregious by the standards of the NATO countries at the time. Yet it is hard to avoid the judgement that the campaign began in an atmosphere of unwarranted official optimism about both the capacity of bombing to reduce the Serb military threat to the Kosovars and the probability that the bombing would stay limited. The initial exclusion of the option of a land invasion was the most extraordinary aspect of NATO's resort to force. It resulted from the inherent difficulties of such an action, nervousness in many capitals about public support for a land war, and from a failure of imagination and strategic thinking in NATO and in national capitals. The initial exclusion of even the threat of a land option had adverse effects: in Kosovo, the FRY forces could concentrate on killing and concealment rather than defence, while in Belgrade the Yugoslav government could hope simply to sit out the bombing. Within the Alliance, creating at least a credible threat of a land option proved to be one of the most important and difficult tasks of the war. The Serb atrocities in Kosovo From February 1998 onwards, the conflict between the KLA and the Yugoslav forces in Kosovo had degenerated into a war of atrocities and ethnic cleansing. The fierce Serb offensive of summer 1998 had left an estimated 1,500 Kosovar Albanians dead, and 300,000 had fled their homes to hide in the mountains and forests. These events led to the adoption of UN Security Council Resolution lgg of 23 September 1998, and also to the threat of NATO air strikes in October to force Belgrade to retreat from its extreme actions. The result was the agreements of 15-18 October 1998, brokered by US Balkan envoy Richard Holbrooke with Milosevic over the heads of the ethnic Albanians and the KLA. Those agreements had brought a partial withdrawal of Serbian security forces, and had provided both for the deployment of up to 2,000 unarmed OSCE monitors in Kosovo and for NATO-led aerial verification. There was widespread scepticism as to whether they would bring a lasting end to the mass murder and expulsions. The killing on 15 January 1999 of at least 45 ethnic Albanians in the village of Recak, 18 miles south-west of the regional capital of Pristina, became the symbol of the breakdown of the October agreement. The Yugoslav authorities blocked numerous requests to allow investigators from the International Criminal Tribunal for the Former Yugoslavia to look into these and other killings. All this led to a hardening of the NATO member states' view that no political settlement for Kosovo would work unless it allowed for deployment of a substantial NATO-led force. By the time the NATO offensive began on 24 March, further Serb killings of Kosovars had occurred, as well as new displacements of population. The UN High Commissioner for Refugees (UNHCR) Special Envoy for the region stated a few weeks later: 'At the time UNHCR had to suspend its operations in Kosovo on 23 March 1999, there were thought to be over 260,000 internally displaced persons (IDPs) within Kosovo, over 100,000 IDPs or refugees in the region, and over 100,000 refugees and asylum seekers outside the region'.20 Such figures for enforced population displacements, though very high, were not on the scale of what happened after 24 March. This raises the question of whether the bombing made things worse for the Albanian majority in Kosovo. It is not disputed that, in the words of a White House spokesman on 26 March, the situation in Kosovo took 'a dramatic and serious turn for the worse' in the days after the bombing commenced.21 Many refugees fleeing from Kosovo saw the Serb onslaught against them as a direct consequence of the NATO action. As one put it: 'The Serbs can't fight NATO, so now they are after us'.22 Within one month of the start of the bombing campaign, over half a million people had fled from Kosovo into neighbouring countries, and many thousands more were displaced within Kosovo itself.23 During the whole period of the bombing, according to NATO figures, almost one million inhabitants left Kosovo, and half-a-million were internally displaced.24 Thousands of Kosovar Albanians were killed. Although the degree of involvement was far from uniform, Serb police, military and paramilitary forces all took part in committing these atrocities. NATO governments sometimes contended that such killings and expulsions had been imminent anyway, and that Belgrade had set in motion Operation Horseshoe, the plan for the systematic ethnic cleansing of Kosovo, even before the start of the NATO bombing. Whatever the strength of these contentions, which may well be vindicated as more information becomes available, there are grounds for doubting whether, in the absence of the NATO bombing, the ethnic cleansing would have proceeded with such speed and viciousness. All major cases of genocide and ethnic cleansing in the twentieth century have occurred during or immediately after major wars: the chaos and hatred unleashed in war, and the secrecy that wartime conditions engender, can provide the necessary conditions for such mass cruelty. Any conclusion that NATO's military operations hastened the killings and expulsions of Kosovar Albanians would not mean that the NATO operation should be judged a failure. It may have been better to bring the crisis to a head than to let it fester on, albeit in a less intense form, for year after year; and there was evidently some diminution of the intensity of Serb repression around the end of April. As Jonathan Steele reported from Kosovo in July: If there was a plan to remove every last Albanian from Kosovo in a Nazi-style 'final solution', it was abandoned or at least relaxed about a month into the bombing campaign ... Whatever motive best explains the atrocities committed by the Serbs after Nato started its bombing, no Albanians say Nato was wrong. Those Western critics who condemn the bombing for turning a humanitarian crisis into a catastrophe get short shrift in Kosovo. Albanians were the primary victims and there is an almost universal feeling that, although the price was far bloodier than expected, it was worth paying for the sake of liberation from Serb rule.25 Even if NATO's bombing had unwittingly exacerbated it, the reign of terror against the Kosovar Albanians had the effect of shoring up NATO's unity and resolve. The huge refugee crisis meant that NATO governments and publics were reinforced in their determination not to allow the ethnic cleansing of Kosovo to stand. The indictment of Slobodan Milosevic and four of his senior colleagues by the Yugoslav war crimes tribunal, announced on 27 May, merely confirmed the strong sense in many countries that there was a fundamental moral difference between the two sides. The laws of war While most of the NATO bombing campaign was accurate and was directed at legitimate targets, certain actions did raise questions about whether NATO, in pursuing its humanitarian war, was observing all the requirements of the laws of war (international humanitarian law). These requirements overlap with, and are not necessarily antithetical to, those of military efficiency. During the bombing campaign, questions relating to the laws of war were raised most publicly by Mary Robinson, UN High Commissioner for Human Rights, but her speeches did not go into detail and did not have major impact. In her report of 30 April, for example, she said simply: In the NATO bombing of the Federal Republic of Yugoslavia, large numbers of civilians have incontestably been killed, civilian installations targeted on the basis that they are or could be of military application, and NATO remains the sole judge of what is or is not acceptable to bomb. In this situation, the principle of proportionality must be adhered to by those carrying out the bombing campaign." A strong defence can be made of the NATO air campaign. As General Wesley Clark has written: it was not a campaign against the Serbian people. It focused specifically on the forces of repression from top to bottom to coerce a change in their behaviour or, failing that, to degrade and ultimately destroy their means of repression. Allied planners, targeters and pilots worked diligently to prevent injuries and loss of life among the civilian population and to prevent collateral damage.27 The emphasis on air-power in this campaign, coupled with the reluctance to risk the lives of servicemen, exposed certain problems about the extent to which NATO was able to perform its military tasks effectively and to minimise damage to civilians. In particular, the use of smart weapons, and the practice of bombing from IS,ooo feet, were associated with certain problems so far as the safety of civilians and of neutral states were concerned. These included: Collateral damage, for example in the cases in which passenger trains and buses were crossing bridges at the moment when bombs hit. Errors in identifying and attacking targets, including misidentification of the functions of particular buildings (for example, the Chinese embassy), and weapons going astray. Pressure to attack fixed targets such as buildings, bridges and electricity installations, because they are easier to identify and destroy by such means than are moving targets. Since most military assets are either mobile or capable of concealment and hardening, the pressure to attack fixed targets meant, in practice, pressure to attack targets whose destruction had a significant effect on the civilian population. The damage to civilians and to neutral states which resulted from such problems do not begin to compare, in any grim comparison of losses, with the effects of the ethnic cleansing in Kosovo. Such damage may indeed be inevitable in war. Yet it is a salutary reminder that there are moral problems with the whole idea of the low-risk waging of war. A further difficulty arose from the possible environmental effects of certain NATO actions, including the release of chemicals resulting from certain air attacks, and the use of toxic materials (especially depleted uranium) in weapons and quantities of unexploded ordnance which was a serious hazard after the war.28 The underlying problem goes deeper than the particular requirements and incidents of the Kosovo campaign. The US, and with it NATO, have developed over recent decades a conception of how force can be applied, which involves putting military pressure not just on the armed forces of the adversary state but also on its government. Such an approach was evident in some official thinking about nuclear deterrence, and also in the conduct of certain operations in which NATO members have been involved, including aspects of the bombing campaign against Iraq in 1991. The approach is in tension with one underlying principle of the laws of war, as famously expressed in the i868 St Petersburg Declaration, 'that the only legitimate object which States should endeavour to accomplish during war is to weaken the military forces of the enemy'. After this campaign, NATO members will, sooner or later, have to address the question of how their concept of war relates to the laws of war, and whether any modifications of either are suggested by this experience. The most detailed international agreement bearing on military targeting, and placing limits on attacks on civilians and civilian installations, is the 1977 Geneva Protocol I. The FRY is a party to this agreement, as are all members of the NATO alliance except France, Turkey and the US. America, although unwilling to ratify it, has stated that it accepts and implements many of the Protocol's provisions. On becoming parties, several NATO members (Belgium, Canada, Germany, Italy, Netherlands, Spain and the UK) made declarations which implied recognition that, despite their best efforts, there could be many ways in which military activities would impinge seriously on the civilian population. The NATO campaign is as much subject to consideration by the International Criminal Tribunal for the Former Yugoslavia as are any acts of the local parties to the conflicts. Indeed, it is ironic that the US - having devoted considerable diplomatic effort in 1gg8-gg to opposing certain provisions of the 1998 Rome Statute of the International Criminal Court (not yet in force) on the grounds that the prosecutor of such a court might have unwelcome powers to examine critically US military actions - then proceeded to go to war in 1ggg in the former Yugoslavia, the only region of the world in respect of which there is an independent prosecutor for war crimes. Indeed, the Yugoslav Tribunal has considerably greater powers vis-a-vis national legal institutions than would the planned International Criminal Court. The Kosovo campaign may yet teach NATO member states that they can live with the existence of an international criminal tribunal capable of considering their actions as well as those of their adversaries. Factors leading to the settlement Two months into the bombing campaign, the prospect was looming that it might have to continue over the summer, with serious risks that NATO's unity could not endure so long and inevitably controversial an operation. Then, to the relief (and, in some cases, scarcely concealed surprise) of NATO governments, on 3 June Milosevic formally accepted joint EU-Russian peace terms presented to him the previous day. This led, albeit with numerous difficulties on the way, to the military agreement signed at Kumanovo air base in Macedonia on g June, and to UN Security Council Resolution 1244 of the following day.29 The settlement of 3-10 June, the result of sustained diplomatic efforts which had continued throughout the air campaign, involved elements of compromise on the NATO side. Some were cosmetic, such as the avoidance of any specific mention of NATO's role in the Kosovo Force (KFOR) in Security Council Resolution 1244 and certain associated documents (though the NATO role had been mentioned in the paper agreed by the Yugoslav government on 3 June). At least three concessions by NATO in the June settlement were of more substance. First, the UN was given a central role in the administration of Kosovo (a concession which had certain advantages for NATO in helping to bring its operations back within a clear mandate of the Security Council). Second, although it had always been envisaged, even at Rambouillet, that KFOR would be composed of forces from NATO and non-NATO countries, there was now a more definite prospect of Russian participation. Third, there was no longer any mention of the status-of-forces provisions in Appendix B of the Rambouillet Accord which would have accorded NATO personnel unimpeded access, including for training and for operations, throughout FRY territory.30 These provisions, the subject of intense controversy during the war, went further than the equivalent provisions in the status-of-forces agreement between NATO and the FRY which had constituted part of the 1995 Dayton Peace Accords.31 On the key issue of the political future of Kosovo, the June settlement terms remained as much of a fudge as the abortive Rambouillet terms of 23 February. Rambouillet had included repeated reference to 'the sovereignty and territorial integrity of the Federal Republic of Yugoslavia', and had envisaged that Kosovo would have a status in some respects akin to that accorded to Republics (Montenegro and Serbia) in the FRY Constitution.32 In the June settlement, there was repeated reference to the Rambouillet Accords, and in particular to the principle that the people of Kosovo can enjoy 'substantial autonomy within the Federal Republic of Yugoslavia'. These words will be the subject of much debate and interpretation. However, the changed facts of power and demography following the war must mean that the prospects of substantial independence for Kosovo have increased. Overall, the terms of the settlement represented a considerable concession by Yugoslavia from previous positions enunciated by its government and parliament. What led the Yugoslav authorities to make the critical concessions of 3 and 9 June? Air-power clearly played a significant part, and advocates of air-power were not slow to claim a victory. However, as the Yugoslav forces withdrew from Kosovo in June and the NATO-led KFOR became established there, it became evident that the Yugoslav army in Kosovo had been much less seriously damaged than NATO had previously believed. Some 47,ooo soldiers were reported as having left the province, several thousand more than intelligence reports had indicated were there at the height of the military campaign. At the same time, Yugoslavia as a whole, and in particular Serbia proper, 'clearly suffered enormous damage, particularly to its roads, bridges and industry after ii weeks of increasingly intense bombing'.33 If this view is correct, then the disturbing lesson of the air campaign may be that its most effective aspect involved hurting Serbia proper (including its population and government) rather than directly attacking Serb forces in Kosovo and protecting the Kosovars. While the pressure arising from air operations clearly influenced the Yugoslav government's decision to accept the settlement, the actuality and threat of land operations also played a part. In the last fortnight of the war, KLA operations near the Kosovo-Albanian border forced Yugoslav soldiers out into the open. This enabled NATO aircraft to attack them, causing what were probably the most substantial military casualties of the whole campaign. A NATO air attack on 7 June, in which US B-52s used cluster bombs against Yugoslav forces near Mount Pastrik, killing several hundred, appears to have put effective pressure on the Serb negotiators in the stalled talks at Kumanovo.34 Also in the last two weeks before the settlement of early June, NATO began to signal the prospect of a ground operation. During the whole campaign, the problem of getting agreement on land operations, with all their immense difficulties and risks, threatened to undermine NATO's hard-won unity, and stalked NATO's fiftieth-anniversary summit in Washington like a ghost on 2324 April. At times it appeared that there might have to be a 'coalition of the willing' within NATO if any effective threat of a land intervention were to materialise. Only in late May was there any coherent action on the matter. On 25 May, NATO ambassadors approved a plan, KFOR-Plus, increasing the projected size of KFOR to 50,ooo troops; and on 31 May, the US government finally gave Wesley Clark permission to strengthen and widen the road in Albania leading from the port of Durres to Kukes on the Kosovo border. These were ways of conveying to Milosevic that the invasion option was getting serious.35 The developments on the ground in Kosovo, and the evidence of NATO preparations for ground operations, influenced ongoing negotiations, including those held outside Moscow on 27 May 1999.36 They also played a part in bringing Russia to recognise the need for a settlement along the lines which NATO had been demanding; and Russia's change of direction was bound to have a serious impact in Belgrade. After the war, indeed, some Serbs started to attribute their defeat to an alleged Russian betrayal.37 Can a doctrine be developed from this case? At the beginning of the air campaign, if NATO governments had known that it would have to last ii weeks, would involve so many difficult issues and incidents, and would require a serious prospect of land war, it is far from certain that they would have embarked on it. Like a revolution, it marked a significant turning-point, but one that is in danger of being too much glorified. The lessons of the revolution in warfare of which Kosovo is a symbol may bear resemblance to the lessons of the Yugoslav revolution as recollected by a principal participant, Milovan Djilas: Revolutions begin new epochs, whose direction no one can foresee, let alone determine. Would life be life if it had to conform to hypothesis? Revolutions Must take place when the political forms are unable to develop reasonable and just solutions. Revolutions are justified as acts of life, acts of living. Their idealisation is a cover-up for the egotism and love of power of the new revolutionary masters. But efforts to restore pre-revolutionary forms are even more meaningless and unrealistic. I sensed all of this even then. But choice does not depend only on one's own outlook but also on reality. With my present outlook, I would not have been able to do what I had done then38? Many lessons will be drawn from the Kosovo action, including some hard ones about the virtues, and limits, of operating in a large and disparate alliance. At times, NATO showed the classic problem of a large international organisation in its inability to agree on more than a lowest common denominator. NATO also experienced tensions due to the fact that the US supplied about 85% of the effective power in the bombing campaign, a figure which demands reflection about European readiness for independent security policies. Only with the entry of KFOR into Kosovo in June was the imbalance in military burden-sharing visibly redressed. During the war, the question was often raised as to whether a general doctrine justifying humanitarian intervention could be developed. As Blair said in his Chicago speech on 22 April: The most pressing foreign policy problem we face is to identify the circumstances in which we should get involved in other people's conflicts. Non-interference has long been considered an important principle of international order. And it is not one we would want to jettison too readily ... But the principle of non-interference must be qualified in important respects. Acts of genocide can never be a purely internal matter. When oppression produces massive flows of refugees which unsettle neighbouring countries they can properly be described as 'threats to international peace and security'.39 Blair went on to list five major considerations which might help in decisions on 'when and whether to intervene': First, are we sure of our case? War is an imperfect instrument of righting humanitarian distress; but armed force is sometimes the only means of dealing with dictators. Second, have we exhausted all diplomatic options? We should always give peace every chance, as we have in the case of Kosovo. Third, on the basis of a practical assessment of the situation, are there military operations we can sensibly and prudently undertake? Fourth, are we prepared for the long term? In the past we talked too much of exit strategies. But having made a commitment we cannot simply walk away once the fight is over; better to stay with moderate numbers of troops than return for repeat performances with large numbers. And finally, do we have national interests involved? The mass expulsion of ethnic Albanians from Kosovo demanded the notice of the rest of the world. But it does make a difference that this is taking place in such a combustible part of Europe. Subsequent attempts to develop any general doctrine regarding the circumstances in which humanitarian intervention may be justified have run into predictable difficulties. Two enduring and inescapable problems are: first, that most states in the international community are nervous about justifying in advance a type of operation which might further increase the power of major powers, and might be used against them; and second, NATO members and other states are uneasy about creating a doctrine which might oblige them to intervene in a situation where they were not keen to do so. Operation Applied Force will contribute to a trend towards seeing certain humanitarian and legal norms inescapably bound up with conceptions of national interest.4 It may occupy a modest place as one halting step in a developing but still contested practice of using force in defence of international norms. However, the unique circumstances in which Operation Allied Force took place, and the problems which the campaign exposed, militate against drawing simple conclusions about humanitarian intervention or about the capacity of bombing alone to induce compliance. In the international community, the NATO campaign was the subject of deep differences of opinion, based on diverging perceptions and interests which are not going to change suddenly. The fact that the campaign failed in the intended manner to avert a humanitarian disaster in the short term, even though it did eventually stop it, makes it a questionable model of humanitarian intervention. The uncomfortable paradox involved - that a military campaign against ethnic cleansing culminated in a settlement in which the majority of Serbs resident in Kosovo departed - must reinforce the sense that humanitarian operations cannot suddenly transform a political landscape full of moral complexity. The advanced-weapons-systems bombing, although extraordinarily accurate, gave rise to serious questions about its effectiveness against armed forces and its impact on civilians. The reluctance of NATO governments to risk the lives of their forces, the difficulty in developing a credible threat of land operations and, above all, the narrowness of the line between success and failure, suggest that the many lessons to be drawn from these events should be on a more modest scale than any grand general doctrines of humanitarian intervention. Notes 1 The five objectives appear in the Statement on Kosovo issued by the NATO summit in Washington DC on 23-24 April 1999. This slightly abbreviated version from Tony Blair's speech in Chicago on 22 April 1999 ('Doctrine of the International Community'), is available at: http://www.number-10.gov.uk/public/ info/ index.html 2 According to incomplete Serbian statistics based on the census taken in the Federal Republic of Yugoslavia (FRY) in 1991 (boycotted by Kosovo Albanians), of 1,954,747 Kosovo inhabitants, 1,607,690 (82.2%) were Albanians, and 195,301 (10%) were Serbs. Between 1991 and 1997 up 500,000 Kosovo Albanians left Kosovo for Turkey, Macedonia, Switzerland and EU countries: Stefan Troebst, Conflict in Kosovo: Failure of Prevention? An Analytical Documentation, 1992-1998, ECMI Working Paper no. 1 (Flensburg: European Centre for Minority Issues, 1998), p. ln. 3 A comprehensive arms embargo was imposed on the FRY in respect of Kosovo by Security Council Resolution 1160 of 31 March 1998. This resolution also expressed support for OSCE efforts for a peaceful resolution of the crisis in Kosovo. Further pressure was reflected in resolutions on 23 September and 24 October 1998 (mentioned below). 4 See also Kofi Annan's earlier exposition of the possible necessity for intervention, with particular reference to Kosovo, in his Ditchley Foundation Lecture, 'Intervention', Ditchley Park, England, 26 June 1998 (unpublished). 5 Account of the Security Council debate, UN Press Release SC/6659, 26 March 1999. 6 One-page FCO note of 7 October 1998, 'FRY/Kosovo: The Way Ahead; UK View on Legal Base for Use of Force'. This note states that it was being circulated 'to all our NATO allies'. See also Baroness Symons of Vernham Dean, written answer to Lord Kennet, Hansard, 16 November 1998, col. WA 140 (see: http://www.parliament.the-stationery-- office.co.uk/pa/cm/cmhansrd.htm). The same basic line of UK government thinking on legal authority for military action over Kosovo can also be found in an FCO memorandum of 22 January 1999 to House of Commons Select Committee on Foreign Affairs, which made brief additional reference to the possibility that circumstances could arise in which a use of force over Kosovo would be justified in terms of individual or collective self-defence. See also the Committee's examination of Tony Lloyd on 26 January 1999. 7 For a useful recent study see Fernando R. Teson, Humanitarian Intervention: An Inquiry into Law and Morality, 2nd edn (Irvington-on-Hudson, NY: Transnational Publishers, 1997). There was reference to some modern exponents of the legitimacy of humanitarian intervention in Catherine Guicherd, 'International Law and the War in Kosovo', Survival, vol. 41, no. 2, Summer 1999, p. 24. 8 The authoritative account is David Malone, Decision-Taking in the UN Security Council: The Case of Haiti, 199097 (Oxford: Oxford University Press, 1998). 9 The FRY's case in the ICJ is proceeding slowly. On 29 April 1999 the FRY instituted proceedings against ten NATO member states, accusing them of bombing Yugoslav territory in violation of their obligation not to use force against another state. After hearings on provisional measures were held on 1012 May, the Court handed down its decision in each of the ten cases on 2 June. It threw out two cases (those against Spain and the US) on the grounds that it manifestly lacked jurisdiction. In the eight other cases (those against Belgium, Canada, France, Germany, Italy, Netherlands, Portugal and UK) the Court found that it lacked prima facie jurisdiction and could therefore not order provisional measures; but it remained seised of those cases. On 30 June, the Court decided that FRY should submit a Memorial in each of the eight cases by 5 January 2000, and the eight should each submit a Counter-Memorial by 5 July 2000. 10 President Clinton, television address, 24 March 1999, published in The Washington Post, 25 March 1999, p. A34; Prime Minister Blair, 'Doctrine of the International Community'. 11 Robin Cook, House of Commons debate on Kosovo, 25 March 1999, Hansard, cols. 537-539 (see: http://www.parliament.the-stationery office.co.uk/pa/cm/cmhansrd.htm). 12 General Wesley K. Clark, Supreme Allied Commander Europe, `When Force is Necessary: NATO's Military Response to the Kosovo Crisis', NATO Review, Summer 1999, pp. 16,18. The word 'war' does not appear in this article. 13 For a powerful if not always judicious argument that the 1995 bombing had such an effect on the peace negotiations, see Richard Holbrooke, To End a War (New York: Random House, 1998), chapter 7, pp. 94-111. 14 Personal interviews in the US, Netherlands and the UK, March-June 1999. 15 George Robertson, House of Commons Select Committee on Defence, Hearings, 24 March 1999, at Q. 356. 16 Ibid., Q. 366 and 367, in response to Mike Hancock. 17 Crispin Blunt, ibid., Q. 371-374. 18 Ibid., Q. 376, in response to Laura Moffatt. 19 Ibid., Q. 390. In its context, this remark may have been compatible with a view that the land/air operation was ruled out only for the time being. 20 Nicholas Morris, 'Coping with the Kosovo Crisis', informal presentation at UNHCR headquarters, Geneva, 7 May 1999. 21 Daniel Williams, `Brutal Conditions Enveloping Kosovo', Washington Post, 27 March 1999, p. Al. 22 Ibid., at p. A16. 23 Mary Robinson, UN High Commissioner for Human Rights, 'Report on the Human Rights Situation Involving Kosovo', UN Commission on Human Rights, Geneva, 55th session, 30 April 1999 (see: http://www.unhchr.ch/data.htm). 24 Sergio Balanzino, NATO Deputy Secretary General, 'NATO's Humanitarian Support to the Victims of the Kosovo Crisis', NATO Review, Summer 1999, p. 9. 25 Jonathan Steele, 'Confused and still in denial, Serbs have a long way to go', The Guardian, London, 9 July 1999, p. 9. 26 Mary Robinson, report of 30 April 1998. Subsequently, at a news conference in Montenegro on 9 May 1999, she made further statements on the subject, and notably rejected use of the term 'collateral damage'. 27 Wesley Clark, 'When Force is Necessary', p. 16. See also his more detailed discussion of targets on p. 17. 28 See especially Mark Fineman, 'Serbia's Nightmare: The Toxic Aftermath of War', International Herald Tribune, 7 July 1999, pp. 1, 5; and Dan Eggen, 'Bombs that Still Kill: NATO "Duds" in Kosovo', International Herald Tribune, 20 July 1999, p. 4. The environmental effects of warfare have been of considerable concern in recent years, due partly to the salience of the matter in the 1990-91 Gulf War. See, for example, the study in the US Naval War College 'Blue Book' series, Richard J. Grunawalt, John E. King and Ronald S. McClain (eds), Protection of the Environment During Armed Conflict, US Naval War College, International Law Studies, vol. 69 (Newport, RI: Naval War College, 1996). 29 Military Technical Agreement between the International Kosovo Force (KFOR) and the government of the Federal Republic of Yugoslavia and the Republic of Serbia, concluded at Kumanovo, Macedonia, 9 June 1999 (see: http://www.nato.int); UN Security Council Resolution 1244 of 10 June 1999, which contains as annexes the statement of general principles on the political solution to the Kosovo crisis which had been issued by G-8 Foreign Ministers at their meeting at St Petersberg on 6 May 1999; and the paper presented in Belgrade on 2 June 1999 which formed the basis of the peace agreement. 30 Article 8 of 'Appendix B: Status of Multi-National Military Implementation Force', part of the final draft of the Rambouillet Accord, 'Interim Agreement for Peace and Self-Government in Kosovo', 23 February 1999; (see: http://www.balkanaction.org). 31 Article 2 of 'Agreement Between the Federal Republic of Yugoslavia and the North Atlantic Treaty Organisation Concerning Transit Arrangements for Peace Plan Operations', part of Annex 1-A of 'General Framework Agreement for Peace in Bosnia and Herzegovina', initialled at Dayton, Ohio, 20 November 1995 (Washington DC: Office of Public Communication, US Department of State, 1996). 32 Rambouillet Accord of 23 February 1999, Framework, Article 1, and Chapter 1, preamble and Article 6. 33 Steven Lee Myers, 'NATO Air War May Have Done Less Damage Than Alliance Thought', International Herald Tribune, 29 June 1999, p. 4. 34 See especially William Drozdiak, 'Yugoslav Troops Devastated by Attack: B-52 Raid Leaves Forces "Pulverized"', Washington Post, 9 June 1999, p. 19; and Jonathan Steele's report from the destroyed village of Planeja, 'Ghost Village Marks the Battle that Ended the War', The Guardian, 17 July 1999, p. 17. 35 See especially Patrick Wintour and Peter Beaumont, 'How NATO planned to Invade', and `Kosovo: The Untold Story', The Observer, 18 July 1999, pp. 1, 2 and 13-20. 36 These factors are emphasised in accounts of the 27 May negotiations between Martti Ahtisaari, the EU's special envoy on Kosovo, Viktor Chernomyrdin, Russia's negotiator, and Strobe Talbott, US Deputy Secretary of State. See, for example, Martin Walker, 'Revealed: How Deal was Done in Stalin's Hideaway', The Guardian, London, 5 June 1999, pp. 1-2. On the role of Peter Castenfeld, a Swedish financier, in acting as a secret envoy to Belgrade a few days before the 3 June agreement, see John Lloyd, 'Secret Envoy Laid Ground for Deal', Financial Times, 14 June 1999, p. 3. 37 Chris Hedges, 'Officials Redirect Serb Rage: Russia is Now the Villain', International Herald Tribune, 17-18 July 1999, p. 5. 38 Milovan Djilas, Wartime (London: Secker & Warburg, 1977), p. 450. 39 Tony Blair, 'Doctrine of the International Community'. 40 A point made with particular force by Joseph Nye, 'Redefining the National Interest', Foreign Affairs, vol. 78, no. 4, July/August 1999, pp. 22-35. Adam Roberts is Montague Burton Professor of International Relations at Oxford University and a Fellow of Balliol College. His publications include United Nations, Divided World, 2nd edn (Oxford: Oxford University Press 1993), and Nations in Arms: The Theory and Practice of Territorial Defence, 2nd edn (London: Macmillan for the IISS, 1986). Copyright Oxford University Press (England) Autumn 1999 ©1999 UMI Company; All Rights Reserved. Only fair use, as provided by the United States copyright law, is permitted. UMI Company makes no warranty regarding the accuracy, completeness or timelines of the Publications or the records they contain, or any warranty, express or implied, including any warranty of merchantability or fitness for a particular purpose, and shall not be liable for damages of any kind or lost profits or other claims related to them or their use.