International Law and the United Nations: The Use of Military Force
Keynote Address by Gareth Evans to Heinrich Boll Foundation, 5th Annual Foreign Policy Conference, The Role of International Law and the United Nations in a Globalizing World, Berlin, 24 June 2004
There are innumerable areas in which the roles of international law and the United Nations intersect in ways that are fundamentally important for our ever more interdependent world, including international trade and communications, environmental protection and the conduct of diplomatic relations. But no issue involving this intersection is of more immediate importance, or more controversial and difficult, than the current status of the international rules governing the use of military force, and it is on this that I want to concentrate in this keynote address.
As we look out on the world around us, those rules seem to be in serious disarray, their application erratic and their interpretation contested. No universally accepted practice currently governs their use. States are seen as making up the rules as they go along, going to war when they should not be, and not taking military action when they should.
Secretary General Kofi Annan put part of the problem in a nutshell when he said to the General Assembly on 23 September last year, six months after the Coalition invasion of Iraq, if ‘States reserve the right to act unilaterally, or in ad hoc coalitions’ without waiting for agreement in the Security Council, ‘this logic represents a fundamental challenge to the principles on which, however imperfectly, world peace and stability have rested for the last fifty-eight years’.
He put the other part of the problem – states not going to war when they should – equally starkly in his Millennium Report to the General Assembly three years earlier: “…if humanitarian intervention is, indeed, an unacceptable assault on sovereignty, how should we respond to a Rwanda, to a Srebrenica – to gross and systematic violations of human rights that affect every precept of our common humanity?”
In the United Nations Charter of 1945 the use of force was subjected to the rule of law in a much clearer way than had ever previously been attempted, including in the ill-fated League of Nations Covenant, and the law was backed with a system of collective security much more potentially effective than anything that had gone before. The Charter made absolutely clear in Article 2(4) that all UN member states “shall refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any state, or in any other manner inconsistent with the Purposes of the United Nations”. It allowed only two exceptions to the prohibition on the use of force in international law: self-defence under Article 51, and military measures authorised by the Security Council in response to “any threat to the peace, breach of the peace or act of aggression” (under Chapter VII, and by extension for regional organisations in Chapter VIII).
But it was one thing to create a new legal order, quite another to make it work as intended. The founders of this new order were not naïve, recognising as Adlai Stevenson said at the time, that “(e)verything depends on the active participation, pacific intentions and good faith of the Big Five.” And the mutual hostility among the five permanent members themselves which very soon emerged, to be sustained throughout the Cold War years, certainly did threaten to bring the whole edifice down. For forty-four years, until 1989, states repeatedly used military force against others (by one count 200 times and by another 680 times) - a great many of which occasions could not begin to be explained or justified as self-defence under Article 51, although that fig-leaf was regularly relied upon.
But a paralysed Security Council passed very few Chapter VII resolutions, and there was only one large scale collective military action responding to a breach of the peace (Korea in 1950, during the Soviet Union’s misjudged period of absence from the Council). The order which prevailed was essentially a new variation on an old balance of power theme.
Through all this the UN security system somehow managed to stay afloat, playing a reasonably significant role at least in defusing and managing conflict. ‘Good offices’ diplomacy contributed to the peaceful abatement of some 25 per cent of the post 1945 conflicts. And peacekeeping, a role invented by the Secretariat – involving the verification, monitoring, and supervision of ceasefire and broader peace agreements – proved crucial in reducing the risk of a number of further wars, especially in South Asia and the Middle East. But it is hard to argue that the new international legal order promised by the Charter was, for the duration of the Cold War, under anything but stress.
All that changed, dramatically, after the Berlin Wall came down. The removal of a major source of ideological and great power conflict liberated the UN to play the global security role its founders intended, as became immediately apparent with the response to Iraq’s invasion of Kuwait in 1991. Expectations were reinforced by UN-organised or authorised military deployments in the former Yugoslavia, Somalia, and Haiti. The average number of resolutions passed in a year went from 15 to 60, or from a resolution a month to a resolution a week. Ninety-three percent (247 of 267) of all Chapter VII resolutions of the Council were passed between 1990 and 2002. Before 1989 the Council applied sanctions twice; since then it has imposed sanctions fifteen times.
Not all this activity was either effective or uncontroversial: too often what happened – not least in Somalia in 1993, Rwanda in 1994 and Bosnia in 1995 – was too little too late, misconceived, poorly resourced, poorly executed, or all of the above. And of course in Kosovo in 1999, in what was arguably a very strong case for intervention, a divided Security Council was bypassed completely. But there was reason to believe that step by step, painful and disappointing as so much of the process was, a new and much more responsive international legal order was falling into place.
But now the wheel has turned again, and maybe even more seriously than in the Cold War years. The decision of the U.S., and its coalition allies, to go to war in Iraq in March 2003 in defiance not only of one or more threatened vetoes, but of the clear and overwhelming majority of the members of the Security Council, was a watershed of a wholly different, and more disconcerting kind. Three different kinds of challenges to the international legal order are involved, all of them serious.
First, to the extent that the invasion of Iraq was based on a claim of anticipatory self-defence more far-reaching than any previously asserted, Article 51 has been stretched beyond endurance. Secondly, to the extent that the invasion was based on claims of threats to international peace and security sufficient under Chapter VII to justify Security Council authorization, but with those claims poorly argued and the Council eventually bypassed, the credibility of the whole Charter system has again been frontally challenged. And thirdly, to the extent that the invasion was based on Saddam Hussein’s record of tyranny over his own people – but again poorly and inconsistently argued, and with the Council bypassed – we have seen close to being choked at birth what many were hoping was an emerging new norm justifying intervention on the basis of the principle of “responsibility to protect”.
Article 51 and Anticipatory Self Defence. It has long been accepted, both as a matter of customary international law predating Article 51 and international practice since, that notwithstanding the language of the article referring only to the right arising “if an armed attack occurs”, the right of self-defence extends beyond an actual attack to an imminently threatened one. Provided there is credible evidence of such an imminent threat, and the threatened state has no obvious alternative recourse available, there is no problem – and never has been – with that state, without first seeking Security Council approval, using military force ‘preemptively’. If an army is mobilising, its capability to cause damage clear, and its hostile intentions unequivocal, nobody has ever seriously suggested that you have to wait to be fired upon. In this sense, what has been described generically as ‘anticipatory self-defence’ has always been legal.
The problem arises with another kind of anticipatory self-defence: when the threat of attack is claimed to be real, but there is no credible reason to believe it is imminent. It should be said immediately that the issue here is not with the principle of military action against non-imminent threats as such. It is perfectly possible to imagine real threats which are not imminent – including the nightmare scenario combining rogue states, WMD and terrorists. The problem boils down to whether or not there is credible evidence of the reality of the threat in question (taking into account, as always, both capability and specific intent); whether the military attack response is the only reasonable one in all the circumstances; and – crucially – who makes the decision.
The question is not whether preventive military action can ever be taken: it is entirely within the scope of the Security Council’s powers under Chapter VII to authorise force if it is satisfied a case has been made (and the Council and others these days are quite properly giving increased attention, in relation to both WMD proliferation and terrorism, to circumstances in which such cases might be made). The question is whether military action in response to non-imminent threats can ever be taken unilaterally.
This is the question on which the U.S. has led with its chin, and generated a storm of continuing international controversy in the lead up to the 2003 Iraq war, in particular with the release in September 2002 of the National Security Strategy, arguing for a broadening of “concept of imminent threat to the capabilities and objectives of today’s adversaries” and stating that “there is a compelling case for taking anticipatory action to defend ourselves, even if uncertainty remains as to the time and place of the enemy’s attack”.
The biggest problem with extending the scope of unilateral self-defence action under Article 51 in the way contemplated by the Bush administration is that it utterly fails to acknowledge that what is sauce for the goose is sauce for the gander, legitimising the prospect of preventive strikes in any number of volatile regions, starting with the Middle East, South and East Asia. To undermine so comprehensively the norm of non-intervention on which any system of global order must be painstakingly built is to invite a slide into anarchy. We would be living in a world where the unilateral use of force would be the rule, not the exception.
Chapter VII and External Threats. When a state poses a threat to other states, or individuals outside its borders, Chapter VII of the UN Charter fully empowers the Security Council to take any coercive action at all, including ultimately military action, that it deems “necessary to maintain or restore international peace and security”. It doesn’t matter whether the threat involves a “threat to the peace, breach of the peace or act of aggression”; whether it is occurring now, or in the imminent future or in the distant future; or whether it is constituted by an act or an omission; or whether it is constituted by the state’s own actions or by non-state actors it harbours or supports.
Nor does it matter whether the threat is constituted by an actual or potential act of violence, or simply a challenge to the Council’s authority. If the Security Council decides that a state is not acting in compliance with previous resolutions and for that reason alone, by putting the credibility of the Council’s decisions at risk it is posing a ‘threat to the peace’ – whatever other kind of threat that state may or may not constitute – the Council is fully entitled to make that judgement and act accordingly.
Usually the only question – a huge one politically, but not legally - is how the unquestioned authority of the Council should be exercised, when the Charter offers no detailed criteria, when States see their interests so differently and when some States exercise so much more influence than others. There is growing acceptance that, as the Secretary-General said last September, that there may need to be ‘early authorisation of coercive measures to address certain types of threats – for instance, terrorist groups armed with weapons of mass destruction’. He suggested, very gently, that Security Council members ‘may need to begin a discussion on the criteria’ for such early authorisation, and this is a key question to which I will return.
Chapter VII and Internal Threats. While the UN Charter is as clear as it possibly could be on the question of external threats, it is much less clear about the right of collective action against a state when the only threat involved is to those within it: the issue of so-called humanitarian intervention. Article 2.7 expressly prohibits intervention “in matters which are essentially within the jurisdiction of any state”, but this is in tension with language elsewhere acknowledging individual human rights, and with a mass of law and practice over the last few decades which have set real conceptual limits claims of untrammeled state sovereignty, not least the Genocide Convention.
What is clear is that the Security Council can always authorise Chapter VII military action against a state if it is prepared to declare that the situation, however apparently internal in character, does in fact amount to a ‘threat to international peace and security’ – as it did for example in Somalia, and eventually Bosnia, in the early 1990s. For all practical purposes, threats to international peace and security are what the Security Council says they are. But more often than not, even in conscience shocking situations like Rwanda in 1994, it has declined to initiate or authorise any enforcement action at all. And that has led to a willingness, in cases where necessity seemed to demand, to bypass the Council, as in Kosovo in 1999, in ways that – if repeated – can’t help but diminish its authority.
The need to actually mean it when we say ‘never again’ after each Cambodia, or Rwanda, or Srebrenica is in constant tension with desire of sovereign states not to see a diminution of their sovereign authority. The tenacity of that attachment can be seen to this day in the reluctance that has been evident in the Security Council to take on and hold to account Sudan for its contribution to the horrors continuing to unfold – as I speak - in Darfur.
It was this to try to resolve this tension between the claims of state sovereignty and individual rights that the Canadian Government sponsored the International Commission on Intervention and State Sovereignty (ICISS), which I co-chaired with the Secretary-General’s Special Adviser on Africa, Mohamed Sahnoun, and which produced its report, The Responsibility to Protect in December 2001.
The most groundbreaking contribution of the report was the its characterization of the central issue not as the ‘right’ of anybody to intervene, but rather the responsibility of the international community to act to protect a State’s peoples when the State itself abdicated that responsibility, either through ill-will or incapacity. The ICISS Commission spelt it out in the form of two critical principles. The primary responsibility for the protection of a state’s own people must lie with the state itself. But 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 principle of non-intervention should lead to a larger principle, that of the international responsibility to protect.
We had no doubt that the proper authority to authorise action in these circumstances was the Security Council, but had to face the problem that the Council has not always acted as it should in the face of conscience shocking violations of humanity. Our bottom line was that the task was not to find alternatives to the Security Council, but to make the Security Council work better. And that led us specifically to formulate a set of criteria for military action, which we argued strongly should be accepted at least informally as guidelines by the Security Council - and hopefully could ultimately become so embedded in practice that they assumed the status of customary international law.
The ‘responsibility to protect’ principles were quietly gaining momentum in the academic and political debate, albeit rather submerged by the preoccupations with terrorism and WMD that emerged after 911. But they have been rather thrown off course by the emergence of the argument in relation to last year’s invasion of Iraq – as other rationales in terms of bombs and terrorists dropped away – that it was Saddam’s murderous tyrannising of his own people that made him a suitable case for humanitarian intervention treatment. The argument was a very strong one a decade and more ago, when Saddam was massacring Kurds in the late 80s and southern Shiites in the early 90s – as the world in both cases looked steadfastly the other way – but it had much less application in recent years, when no such catastrophe was occurring or imminent and, awful as it was, the Iraqi regime’s behaviour was not much worse than a score or two of others.
It had all the appearance of a convenient rationalisation rather than genuine motivation, and certainly looked that way to those concerned, for good reasons or bad, about the potential overreach of the humanitarian intervention principle. Was Robert Mugabe to be the next case for military treatment, the ostensible rationale for the enterprise being not his current onslaught on democracy and human rights, but the slaughters in Matabeleland in the 1980s? The net result of all this was that an emerging international norm of real potential utility was once again struggling for acceptance.
Five Criteria of Legitimacy
So what is the task ahead for those of us who do want to raise our sights, not lower them; who want to reverse the trend toward unilateralism rather than collective action in the use of force; who want to see a new consensus emerge, in the context of internal threats to human security, not only as to when states should not go to war, but when they should; and who, above all, want to recreate the kind of confidence in the role and judgment of the Security Council that will lead to a dramatically reduced inclination to bypass it on the part of those capable of doing so?
The key issue, I would suggest, is not the available international law tools. We don’t need new rules, any modification of the existing rules or any new rule-making or rule-applying institutions. Article 51 and Chapter VII, properly applied, between them enable the full range of necessary responses to all likely future security threats. And the Security Council, with its role supplemented by the regional organizations under Chapter VIII, remains – for all its problems of composition – the only body we have, or are ever likely to, fully empowered to act in security matters on behalf of the whole international community
What we do need is better process in determining the proper application of the existing rules governing the use of force - particularly in the Security Council, but wherever else such decisions are made as well. Much more attention should be paid – not only in academic discourse but in actual operational decision-making – to the group of principles which have to be satisfied if any decision to use military force is to be not just legal, but legitimate.
The distinction – if it can be operationalised, with criteria of legitimacy simplified, standardised and commonly accepted – is an important one. The effectiveness of the global collective security system, as with any other legal order, depends ultimately not only on the legality of decisions, but the common perception of their legitimacy – their being made on solid evidentiary grounds, for the right reasons, morally as well as legally. While it is obviously optimal for any military action to be both unquestionably legal under international law and universally accepted as legitimate (as was the case, for example, with the 1991 Gulf War), it is fair to suggest that military action which is technically illegal but widely perceived to be legitimate (as with Kosovo in 1999) does far less damage than action which is generally perceived to be neither legal nor legitimate (Iraq in 2003).
So what are the appropriate criteria of legitimacy which should govern these decisions? I would argue for five:
(1) Seriousness of Threat: is the threatened harm to state or human security of a kind, and sufficiently clear and serious, to prima facie justify the use of military force?
(2) Proper Purpose: is it clear that the primary purpose of the proposed military action is to halt or avert the threat in question, whatever other purposes or motives may be involved?
(3) Last Resort: has every non-military option for meeting the threat in question been considered, with reasonable grounds for believing lesser measures will not succeed?
(4) Proportional Means: is the scale, duration and intensity of the proposed military action the minimum necessary to meet the threat in question?
(5) Balance of Consequences: is there a reasonable chance of the military action being successful in meeting the threat in question, with the consequences of action not likely to be worse than the consequences of inaction?
The immediate intellectual origins of these criteria lie in a recommendation of the ICISS Commission in its The Responsibility to Protect report: the context there was the specific one of ‘humanitarian intervention’ in response to internal threats, but what we described there as ‘threshold’ and ‘prudential’ criteria can easily be expressed in the more generalised form in which I have formulated them here. The ICISS recommendations owed something, in turn, to the general approach of the Kosovo Commission before it, chaired by Richard Goldstone and Carl Tham in 2000. Even more obviously, their ultimate intellectual origins lie in the whole tradition, and vast literature, of ‘just war’ theory. But that said, it is very important to emphasise that the criteria here argued for are intended to reflect universal values and not in any way be culture or religion specific. I am not aware of anything in the literature to contradict this: certainly many conversations over recent years with colleagues from different cultural and religious traditions, particularly Islam, have persuaded me that there is absolutely nothing incompatible there with the principles here espoused, and much to support them.
I would hope that these criteria are simple enough, relevant enough and compelling enough to be employed by any policy makers anywhere – at a national as well as intergovernmental level – in any context when they are deciding whether it’s right to fight. But my immediate proposal is that these criteria, or guidelines, however else they are used, be systematically discussed and applied whenever the Security Council is making any determination under Chapter VII as to whether it is appropriate to use military force in response to any ‘threat to the peace, breach of the peace or act of aggression’ (Art 39). They would apply whether the threat is external or internal; whether the issue is the ‘right to react’ or the ‘responsibility to protect’; or whether the threat is constituted by armies marching, WMD acquisition, terrorism or tribal machetes.
The criteria should also be regarded as relevant when the Security Council is considering its response to a claimed act of self-defence under Art 51, or the question of use of force by a regional organisation under Chapter VIII. They also have potential application, should the occasion ever arise, to any ‘Uniting for Peace’ or similar resolution that might be considered by the General Assembly.
I am not proposing that the five criteria be embodied in the Charter or in any formal treaty – that would be much too tall an order to contemplate - only that they be adopted, informally if necessary, and actually applied by the Security Council in its deliberations. If they did win general support they might appropriately at some stage be embodied in a General Assembly declaratory resolution: this, and their regular application in practice, would over time enhance their prospect of being considered rules of customary international law.
It is of course possible to go into much more detail in drafting these criteria, particularly the crucial first one, asking “is the threatened harm to state or human security of a kind, and sufficiently clear and serious, to prima facie justify the use of military force?” For example, in the context of internal threats, the ICISS argued that the ‘just cause’ threshold should be very narrowly confined, with the bar set high and tight, limited to situations of large scale killing or ethnic cleansing, actual or apprehended.
There have been quite a number of references by the Secretary General and others over recent years to the possible utility of a set of guidelines or criteria of this kind. Prompted initially by the UK in the late 1990s, and later by the ICISS Commission in 2001, there has already been some discussion within the Security Council about the pros and cons of going down this path, at least in the context of humanitarian intervention, in response to internal threats Beyond that, the debate on criteria has not moved much beyond the stage of occasional footnote suggestions and throwaway single lines in speeches stage. It is to be hoped that that debate can now seriously start, perhaps in the context of the work of the Secretary General’s High Level Panel on Threats, Challenges and Change, due to report in December this year.
However they might ultimately be drafted, and whatever the form in which they might be adopted, the point of adopting these five criteria of legitimacy (and having them being seen to be applied) will I hope be clear. Nobody suggests that there is any coin-in-the-slot push button predictability about the kind of answers one will get when one puts the five questions I am suggesting: opinions will always differ, just as national interests will differ, and it is hardly unknown for interests to influence opinions.
But, as someone who has spent a fair number of years in and around these kinds of deliberations, I am a firm believer that good process produces, if not always optimal, at least better outcomes. If the process demands that criteria have to be systematically discussed, it is much more difficult to duck, weave, fudge, dissimulate and just ignore critical issues: colleagues will ask harder questions, and even the press will sometimes start asking the right questions. At the end of the day strong arguments will look stronger and weak arguments weaker, and these appearances do matter.
Putting it succinctly, the point of adopting and applying the five criteria of legitimacy is not to guarantee that the objectively best outcome will always prevail: it is to maximise the possibility of achieving Security Council consensus around when it is appropriate or not to go to war; to maximise international support for whatever the Security Council decides; and to minimise the possibility of individual member states bypassing or ignoring the Security Council.
And if we can achieve any of that, we really will be well on the way to building an international legal order that, despite all the challenges which continue to buffet it, will give us some grounds for optimism that we are not forever condemned to repeat the mistakes of the past.
Hon Gareth Evans AO QC is President of the International Crisis Group, a Member of the UN Secretary General’s High Level Panel on Threats, Challenges and Change, and was Co-Chair of the International Commission on Intervention and State Sovereignty 2001. This address draws on the author’s 2004 Oxford University Cyril Foster Lecture, to be published in the next quarterly issue of Survival.