home       biography       publications       speeches       organisations       images       @contact

When is it right to fight? Legality, legitimacy and the use of military force

2004 Cyril Foster Lecture, Oxford University. Monday, 10 May 2004.

Some wars will always have to be fought. National interest will demand it, or our common humanity will compel it. But every generation of political leaders seems to have to relearn that war, whatever the justice of the cause for which it is fought, is always ugly, always destructive, always the source of immense human pain and misery, and almost always produces unintended results. As professional soldiers often seem to know better than civilians, the decision to go to war should never be made lightly or cavalierly, with disregard for the evidence which might justify it, or inattention to its possible consequences. And it should never be made with indifference to the formal rules of international law, such as they are, that ban or allow the use of military force.

The trouble is that, as we look out on the world around us, those rules seem to be in serious disarray, their application erratic and their interpretation contested. Not for the first time in history, we have something approaching a crisis of confidence in the reality and relevance of the global legal order. There is growing cynicism and scepticism about the bindingness, and indeed even existence, of the international rules authorising military action. Too many states are seen as making up the rules as they go along, going to war when they should not be, and not going to war when they should.

There is concern, above all, about the way in which one state, the United States – and the power in the world with by far the most capacity to do so – has seemed to want to single- handedly rewrite the rule book. What is challenged above all is the notion, asserted in the lead up to the attack on Iraq in 2003, that the right to act in self-defence, without need for prior UN Security Council approval, extends without check to situations where the threatened attack is neither actual nor imminent – and where the reacting country remains, in effect, the sole judge of whether there is a real threat at all. This led Secretary-General Kofi Annan to sound the alarm in the strongest possible terms in his address to the General Assembly on 23 September last year: if ‘States reserve the right to act unilaterally, or in ad hoc coalitions’, he said, 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’.

The Secretary-General’s concern prompted him at the same time to establish a High Level Panel on Threats, Challenges and Change, consisting of some very distinguished and well-known statesmen and women – and me– to advise him by the end of the year on how the international system might respond to the new security environment.[1] Our task is a daunting one: to evaluate the threats, old and new, to both state security and human security that the world will confront in the decades ahead, to identify appropriate policy responses, to recommend necessary and desirable institutional changes. But it will also be to address what UK Foreign Secretary Jack Straw has been calling “the jurisprudence issue”: the status and application of the most fundamental international rules of all, those determining when it is right to fight. And it is just that issue that I want to tackle in this Cyril Foster Lecture, which – given the very long and eminent list of predecessors in this role - I feel deeply honoured and privileged to have been invited to deliver.


The United Nations was created in 1945 above all else “to save succeeding generations from the scourge of war, which twice in our lifetime has brought untold sorrow to mankind”. And the detailed terms of the Charter appeared to deliver on that Preamble rhetoric. 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, with the centerpiece a Security Council empowered to deal forcefully with acts of aggression and threats to the peace. The sense was very real, and for good reason, that international relations had at last emerged from a centuries old jungle, and that a new age of international law had dawned.

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).

While the Security Council’s power to approve force in acting “to maintain or restore international peace and security” was left effectively open ended, the traditional right of self-defence was spelt out in Article 51 in terms that are very far from open-ended, and in fact linked back to the Security Council:

Nothing in the present Charter shall impair the inherent right of individual or collective self-defence if an armed attack occurs against a Member of the United Nations, until the Security Council has taken measures necessary to maintain international peace and security. Measures taken by members in the exercise of this right of self-defence shall be immediately reported to the Security Council…

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.”[2] 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) [3] - 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: one in which, “each superpower would refrain from attacking the essential interests of the other, but would be freed to use force at will in its own sphere of influence”,[4] and that force was used repeatedly, particularly in Central and Eastern Europe and in Latin America and the Caribbean.

The irony is that some of the very few cases of military action to generate almost universal condemnation as indefensible intrusions on state sovereignty were ones that today might be regarded as permissible examples of ‘humanitarian intervention’, or intervention for human protection purposes, although they were not justified this way by the intervening countries at the time: India’s invasion of East Pakistan in 1971 (in the context of West Pakistan’s brutal suppression of Bengalis) and Vietnam’s invasion of Cambodia in 1978 (in the context of the Khmer Rouge’s genocidal brutality toward its own people) were both claimed, implausibly, to be cases of national self-defence, but that did nothing to dilute the criticism.[5]

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.[6] 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.[7] 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.[8]

Whether all this activity was effective is another question: certainly in relation to interventions for human protection purposes, 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.

The high-water mark, in terms of both commitment to collective security institutions and a willingness to adapt them to deal with new kinds of threats, probably came with the unanimous Security Council resolution on 12 September 2001, immediately after the attacks on the twin towers and the Pentagon, accepting that Article 51 self-defence extended to using force against non-state terrorist organisations as well as "those responsible for aiding, supporting or harbouring the perpetrators, organizers and sponsors of . . . acts" of terrorism, with this followed two weeks later by the adoption of mandatory global controls to prevent the financing of terrorism and the recruiting of terrorists.[9] Confidence in the flexibility of the flexibility of the UN system, and its capacity to deliver, was strong.

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”.

I want in the remainder to this lecture to explore these issues in more detail, and then to make some suggestions as to how, in this very difficult environment, we might nonetheless go about consolidating or rebuilding an international legal order that will be capable of meeting the challenges of the 21st century. But before I embark on this course it is necessary to do a little smiting of those who would say at the outset that this whole enterprise is misconceived.

There are certainly some such critics. Faced with all the stops and starts and tumult I have described, and the way in which the Security Council has been used, misused and bypassed not only during the desolate Cold War years but over the more hopeful years since, some voices – the most strident of them Professor Michael Glennon, writing in Foreign Affairs[10] last year – are going so far as to claim that there are just no rules any more, that the whole UN Charter ‘edifice [has come] crashing down’. Glennon argues that the reality of US power and the failure of the Security Council structure to reflect it should be frankly recognised and that, in any attempt in the future to recreate a body of international law governing the use of force in all its manifestations, ‘what the design should look like must be a function of what it can look like’.

There is a lofty analytical response to this line of argument, for example that offered by a commentator in a recent issue of the American Journal of International Law when he says “Glennon raises, but hardly resolves, law’s most profound epistemological issue, namely the effect of deviance on the authority of norms”. [11] I have to confess that my own response is a little more visceral. I find Michael Glennon’s position to be deeply alarming, might-is-right realism dressed up as international law analysis, and incapable of acceptance by anyone who values decency in the conduct of international relations.

Thomas Franck puts his own response not much less gently, in words I find powerfully resonant:

What, then, is the proper role for the lawyer? Surely, it is to stand tall for the rule of law. What this entails is self-evident. ..When the politicians seek to bend the law, the lawyers must insist that they have broken it... When the powerful are tempted to discard the law, the lawyer must ask whether someday, if our omnipotence wanes, we may not need the law. Lawyers who do that may even be called traitors. But those who do not are traitors to their calling.

The search for an orderly, principled system of international law and practice on conflict is as old as conflict itself. There is always a choice, when confronted with the unhappy reality that governments don’t always behave as we hope they might, of raising your sights or lowering them: the tragedy of intellectualising failure in the way that those like Glennon do, is that it encourages so many to lower sights when the acute need is to raise them.

But that’s only the beginning of the necessary answer. We have to satisfy ourselves that the rules we have are up to meeting the threats we now have to face, and that they are in fact capable of being applied. So what is to be said about the present status and workability of the Article 51 rule on self-defence, and the general competence of the Security Council to authorise the use of force under Chapter VII?


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, and where – as linguistic purists insist - the issue accordingly is not preemption but prevention. (The English language seems to be unique in having two different words here – ‘preemption’ to describe responses to imminent threats, and ‘prevention’ for non-imminent ones: that luxury, however, cherished though it may be by policy aficionados who happen to be native English speakers, seems to have done far more to confuse than clarify the debate for everyone else, who tend to use the words, if at all, interchangeably) The classic non-imminent threat situation is early stage acquisition of weapons of mass destruction by a state presumed to be hostile – the case that was made against Iraq by Israel in justifying its strike on the half-built Osirak reactor in 1981, and the case that the U.S. tried to make against Iraq before March 2003.

The problem 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, even though, in the event, it did not formally rely on Article 51. First there was President Bush’s ‘axis of evil’ State of the Union address in January 2002, in which he declared that the U.S. “will not permit the world’s most dangerous regimes to threaten us with the world’s most destructive weapons”; then, five months later at West Point, his assertion that “we must take the battle to the enemy, disrupt his plans, and confront the worst threats before they emerge”; then in September 2002 the release 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”.

This was all too much even for Henry Kissinger, not normally a dedicated booster of international law or reluctant user of U.S. power, who wrote in a Washington Post column on 16 September 2002:

As the most powerful nation in the world, the United States has a special unilateral capacity to implement its convictions. But it also has a special obligation to justify its actions by principles that transcend the assertions of preponderant power. It cannot be in either the American national interest or the world's interest to develop principles that grant every nation an unfettered right of preemption against its own definition of threats to its security.

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.

But even short of doomsday scenarios of this kind, the wrong-headedness of carrying out, threatening or even just hinting at, unilateral preventive strikes is readily apparent. To the extent that the attack on Iraq can be characterised as such a strike, it does not seem to have done much so far – even before the current disastrous revelations about the humiliation and abuse of Iraqi prisoners – to advance the cause of a democratic Greater Middle East. To the extent that North Korea was meant to be cowed by the prospect of such action, the effort seems to have been spectacularly counterproductive, generating a rush to acquire as fast as possible, as the best deterrent to such a strike, the very nuclear weapons whose potential acquisition was what prompted the unilateral sword-waving in the first place. Similar reasoning may be prompting the foot-dragging by Iran in responding to the international demand to terminate its suspect nuclear programs. Of course any serious challenge to U.S. dominance by anyone will be a long time coming, but history is replete with examples of over-exuberant military power eventually being countered by defensive alliances with aspirations of their own

More prosaically, if a threat is real but not imminent, and there is, by definition, time to address it, there is every reason to believe that with an appropriate policy strategy, mobilizing other international players as appropriate, inspiring rather than alienating those whose support is needed, that threat can be deterred or contained, resolved by persuasion or negotiation, or reduced by the socialisation of the actors concerned. Recognising the inherent horror and hugely uncertain consequences of going to war, statecraft has always been about avoiding war if security objectives can possibly be guaranteed in other ways. But as Arthur Schlesinger Jr has put it, President Bush has repudiated the strategy that won the Cold War, replacing “a policy that aimed at peace through the prevention of war by a policy aimed at peace through preventive war”.[12]

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.

And the options available to the Council extend, at its complete discretion, to authorising or endorsing the use of force by blue helmets, Multinational Forces, ‘coalitions of the willing’ or individual states, as well as endorsing (sometimes after the event) military action by regional organisations operating under Chapter VIII – e.g. ECOWAS in Liberia and Sierra Leone.

While there is always the practical problem of ensuring that action follows decision, and what might be described as the Hotspur question in relation to the actual supply of authorised military resources (“I can call spirits from the vasty deep”… “But will they come?”), [13] there is, in short, no doubt about the legal capacity of the Security Council to declare effectively anything it wants to a threat to international peace and security, and to authorise military action 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 question to which I will return.

There may of course be a legal issue in a given case – as, notoriously, with Iraq last year - as to whether the Security Council has actually authorised the use of force at all, or whether some further resolution is necessary to accomplish that. As a former Attorney General, and longstanding senior Cabinet lawyer, I am not entirely unfamiliar with the delicate task of commissioning or delivering opinions which somehow have to accommodate legal conscience with perceived political necessity. And no doubt there is much to be said – and has been said, by Adam Roberts[14] among others – for the ‘presumption of continuity’ of those earlier Security Council resolutions requiring Iraq’s full and complete disarmament and implying dire consequences if it didn’t.

But even in my finest full-wigged days I don’t think I could have managed to persuade myself, or anyone else, that either Resolution 687 back in April 1991, or Resolution 1441 as late as November 2002, or any resolution in between, actually in themselves authorised the use of force (in the way that Resolution 678 clearly did for the 1990 Gulf War); or that they required anything other than further resort to the Council before force could be used; or that they could possibly be construed as explicit enough in their terms to override the obvious reality that the current votes for force, vetoes quite apart, fell conspicuously short of a majority.

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. With no higher authority to gainsay it, 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 – to respond quickly and forcefully to “gross and systematic violations of human rights that affect every precept of our common humanity”, as Kofi Annan described the issue to the General Assembly in his 2000 Millennium Report – 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 2001. [15]

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.

In this respect, in order to maximise the chances of reaching consensus when the next Rwanda, or Kosovo – or Darfur – came along, the ICISS Commission deliberately made the threshold criterion coercive intervention very tough. The argument was that unless the bar is set very high and tight, excluding less than catastrophic forms of human rights abuse, prima facie cases for the use of military force could be made across half the world. So we proposed that the trigger condition for intervention in these internal cases should be narrowly confined to two kinds of situation involving serious and irreparable harm to human beings in progress or imminent: either large-scale loss of life due to deliberate state action, inaction or inability to act, or large-scale "ethnic cleansing" carried out not only by killing, but forced expulsion, acts of terror or rape.

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 it has to be said that 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. [16]

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. Those of us fighting to get the responsibility to protect principle recognised had to face all over again the argument, after Iraq, that any formalisation of the principle would, in the long tradition of missions civilisatrices, only encourage those who were all too inclined to misuse it.


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 are appalled at the indifference being shown to, as Kofi Annan described them, “the principles on which, however imperfectly, world peace and stability have rested” for the last six decades; 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?

Four steps, I think, are involved. The first is simply to recognise the full range of security threats with which any credible contemporary international legal order has to deal. They include the most traditional of all concerns, attacks by states employing conventional weapons, whether actual, imminently threatened or possible. They certainly include attacks by states employing weapons of mass destruction, whether actual, imminently threatened or just possible. They must include the so-called new threats, in particular attacks – employing any class of weapons – by non-state actors supported by a state, whether actual, imminently threatened or possible. They must also include attacks perpetrated or allowed by a state on its own people, again whether that violence be actual, imminently threatened or merely possible. And security threats should also be seen as constituted by persistent breach of UNSC resolutions putting at risk the credibility of the enforcement system.

The second step is to acknowledge that we do already have to hand all the legal tools and rules we need to properly manage every one of these security threats. When it comes to authorising the use of force, we don’t need new rules, any modification of the existing rules or any new rule-making or rule- applying institutions. In the case of rules, Article 51 and Chapter VII, properly applied, between them enable us to fully respond – reactively, preemptively or preventively – to protect the national interest, the collective international interest and the intrastate human interest. And to do so when it comes to actual attacks of any kind, imminently threatened attacks of any kind, possible future attacks of any kind, and necessary measures to enforce the implementation of Security Council resolutions.

When it comes to rule-making and applying institutions the Security Council’s role can certainly be supplemented by that of the regional or sub-regional organisations recognised under Chapter VIII of the Charter, subject to their seeking at least subsequent authorisation from the Security Council. And there may be a role still for the General Assembly acting in Emergency Special Session under the “Uniting for Peace” procedure: it is quite possible, for example, had the Kosovo issue been tested there in 1999, that majority support would have been forthcoming. But, apart from the case of legitimate self-defence in response to actual or imminent attack, there simply is no better or more appropriate body than the UN Security Council to authorise military action.

Whatever its failings in terms of the contemporary representativeness and legitimacy of its composition, particularly so far as the five veto wielding permanent members are concerned, the Security Council is the only body we have, or are ever likely to have, fully empowered to act in security matters on behalf of the only organization we have, or are ever likely to have, with effectively universal membership. As the ICISS Commission put it, with succinct and I hope impressive clarity, “The task is not to find alternatives to the Security Council, but to make the Security Council work better than it has”.

The third step is to recognise that as a global community responding to security threats we have to do much better at prevention than we have, to avoid so far as possible the issue of the use of military force ever having to arise. What this means in all the different contexts we have been addressing is a vast subject in itself, which I have no time here even to begin to survey. But for the purposes of illustration, let me take just the critically important case of nuclear weapons, and what might more usefully be done than simply threatening certain unpalatable regimes with annihilation if they don’t immediately give up whatever weapons capability they might have. [17]

The basic need here is to develop new and tougher universal rules that, on both the supply and demand side, prevent proliferation the old-fashioned way, through the hard diplomatic work of negotiating legal regimes and then putting appropriate resources into enforcing them: something that has not been happening, and (on the evidence of the breakup in disarray last Friday – 7 May – of the latest Preparatory Committee meeting for next year’s planned NPT Review Conference) is depressingly unlikely to happen any time soon. On the supply side the kind of measures that need to be worked on include tighter regulation of internal fuel cycle activities for Nuclear Non-Proliferation Treaty (NPT) members; more stringent export control systems for NPT members; prohibitions on painless withdrawal from the NPT; effective interdiction measures able to be lawfully applied against non-NPT members and non-state actors; negotiation of a fissile material cut-off treaty; improved protective security and destruction regimes for fissile and other dangerous materials; and improved multilateral intelligence capability.

On the demand side, we need to fully understand the range of reasons states seek to acquire WMD, in particular nuclear weapons – to acknowledge that more often than not the motivation is not inherently aggressive, but related to considerations like perceived defensive security needs, national prestige and domestic political pressures – and for the ‘haves’ to respond in ways that reduce the incentives of the ‘have-nots’ to acquire. Those ways, in turn, include giving security assurances in appropriate cases; not testing or further developing new classes of nuclear weapons; not applying systematic double standards in accepting acquisition of nuclear weapons by some countries but not others; not applying double standards in condemning or reacting to proliferation activity; getting serious about arms reduction; and making a serious commitment to the elimination objective in Article VI of the NPT (something that all the existing nuclear powers, including Britain and France, have steadfastly refused to do).

The fourth and final step we need to take if we are to rebuild an effective international legal order for the 21st century, and I believe the most important of all, is to recognize the acute need for better process in determining the proper application of the existing rules governing the use of force. By process here I don’t mean the kind of institutional innovations proposed from time to time which may be conceptually attractive but are wildly unfeasible in practice – e.g. the suggestion of Allen Buchanan and Robert Keohane that the Security Council appoint an impartial independent commission which would evaluate preventive uses of force after the event, applying sanctions if the empirical claims made before the event proved to be untrue.[18]

What is necessary, and I believe rather more realistically achievable, is for there to be much more attention 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 between legality and legitimacy first came into prominence in this context with the argument of the Sweden-sponsored Kosovo Commission in 1999 that the NATO invasion may not have been legal, in the absence of Security Council approval, but it was – taking into account and balancing out some fourteen ‘threshold’ and ‘contextual’ principles - legitimate.[19]

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).

A corollary of this argument is that if there is a widespread perception that action is illegitimate, making it formally legal won’t do anything to change that - and may in fact do even more damage to the reputation of the legalizing body. This was very much France’s argument in the context of the Iraq debate, and – as some say about Wagner’s music, it wasn’t as bad as it sounds: that the credibility of the Security Council would be put even more at risk by endorsing action widely seen around the world as unjustified than it would by being ignored and bypassed altogether.

A further corollary of the argument may be that legitimacy helps breed legality. If there is general confidence that Security Council decisions on the use of force will be made for the right reasons, giving attention to the full range of threats with which major countries are currently preoccupied, having regard to the weight of the evidence, applying appropriate threshold and prudential criteria, and with the power of veto not being exercised capriciously, then those who are currently tempted to bypass the system will be much less tempted to do so – or, perhaps equally importantly, will be at much greater risk of embarrassment if they do so act.

So what is the better decision-making process I am arguing should apply? How can the issue of legitimacy, as distinct from legality, be operationalised in practice? What are the appropriate criteria of legitimacy which should govern these decisions?


My proposal is straightforward: that the Security Council, whenever considering whether to authorise the use of military force in the exercise of its powers under Chapter VII of the UN Charter, should always address – whatever other considerations it may take into account - at least the following five basic criteria of legitimacy:

(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 explored, 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.

Obviously the ICISS recommendations owed something, in turn, to the general approach of the Kosovo Commission before it. 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. To the extent that I have scanned the literature on this subject [20] I am not aware of anything 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?”. I have already noted how, when the context is 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.

External threat and self-defence situations are much more variable, and it may be that there is nothing much to be gained by attempts at further refinement. What matters is that the question be asked, and given a rational and credible answer. In all these cases – as we are now acutely aware in the aftermath of the Iraq war – a great deal ultimately depends on the quality of the evidence. Actual behaviour is one thing, merely threatened behaviour is something else: to establish a threat, plausible evidence of both capability and intent to cause harm is required. (In this respect, an actual cross-border invasion, as with Iraq in 1991, is the easiest case to establish. By contrast, making a plausible case based on the possibility of non-imminent supply of possibly possessed weapons of mass destruction to international terrorists with whom a regime has possible links, as with Iraq in 2003, is about as hard as it gets…)

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, perhaps in the context of the work of the High Level Panel, that debate can now seriously start.

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) would I hope be clear. Nobody, least of all me, suggests that there is any coin-in-the-slot push button predictability about the kind of answers one will get when one puts the 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. In betting on Security Council deliberations – or for that matter almost any governmental or intergovernmental process – Objectivity, as we say in the Antipodes, [21] is not normally the first horse you would back.

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, going to war when we should not be, and not going to war – to protect our fellow human beings from catastrophe – when we should be. Judging when it is right, and not right, to fight is just about the hardest call that anyone in high public office is ever called upon to make. But the stakes are huge, and one has to nurture the hope that good leaders, if not always born, can – with good process added to basically sound law and institutions – at least be made.

[1] The High Level Panel, announced by the Secretary-General on 4 November 2003 and due to report to him by 1 December 2004, comprises Anand Panyarachun (Chair), Robert Badinter, Joao Clemente Baena Soares, Gro Harlem Bruntdland, Gareth Evans, Mary Chinery-Hesse, David Hannay, Enrique Iglesias, Amre Moussa, Sadako Ogata, Satish Nambiar, Yevgeny Primakov, Qian Qichen, Nafis Sadiq, Salim Salim and Brent Scowcroft.

[2] Address to the Chicago Bar Association, June 1945, quoted in Stephen C. Schlesinger, Act of Creation: The Founding of the United Nations, Westview, 2003, p.261.

[3] The figure of 200 is cited by Thomas M. Franck (“Some Observations on the I.C.J.’s Procedural and Substantive Innovations,” 81 American Journal of International Law 116, 1987.), noting that customary norms of nonintervention are “adhered to, at best, only by some states, in some instances, and have been ignored…with impunity in at least two hundred cases.” Herbert K. Tillema (“Risks of Battle and the Deadliness of War: International Armed Conflicts,” paper presented to the International Studies Association, San Diego, Apr. 16-29, 1996, and quoted elsewhere) counts 690 “overt military interventions” during roughly the same period. The disparity between the two is at least in part accounted for by the inclusion in the latter of interventions sanctioned by the state in which intervention occurred, interventions into non-fully-autonomous states, interventions justified under the terms of Article 51, as well as cases of UN-sanctioned interventions. Neither figure was derived with thresholds for casualties, as most counts of interstate war are. (Research note prepared by High Level Panel Secretariat)

[4] Thomas M. Franck, ‘What Happens Now? The United Nations after Iraq’, American Journal of International Law, Vol 97, Issue 3, 2003.

[5] International Commission on Intervention and State Sovereignty (ICISS), The Responsibility to Protect: Supplementary Volume, Essay 4, ‘Interventions Before 1990’, pp. 49-68: see www.iciss-ciise.gc.ca for the full text of both Report and Supplementary Volume.

[6] Adapted from “International Crisis Behavior” (ICB) dataset (updated thru 2001: http://www.cidcm.umd.edu/icb/) from Michael Brecher and Jonathan Wilkenfeld, A Study of Crisis, Ann Arbor: University of Michigan Press, 2000. (Research note prepared by High Level Panel Secretariat)

[7] The most sophisticated quantitative research on peacekeeping suggests that, all things being equal, the deployment of peacekeepers reduces the risk of another war by 70 to 85 per cent: Virginia Fortna,, “Peace Operations: Futile or Vital?” Paper commissioned by the United Nations Foundation for the work of the High-Level Panel on Threats, Challenges and Change, January 2004. (Research note prepared by High Level Panel Secretariat)

[8] Figures in Peter Wallensteen and Patrik Johansson, “Security Council Decisions in Perspective,” in David Malone, ed., The UN Security Council: From the Cold War to the 21st Century, Boulder, CO: Lynne Rienner, 2004, and David Cortright and George A. Lopez, The Sanctions Decade: Assessing UN Strategies in the 1990’s, Boulder, CO: Lynne Rienner, 2000. (Research note prepared by High Level Panel Secretariat)

[9] cf. Franck, op.cit.

[10] Michael J Glennon, ‘Why the Security Council Failed’, Foreign Affairs, May-June 2003, 16-35.

[11] Tom J Farer, “The Prospect for International Law and Order in the Wake of Iraq”, AJIL Vol 97, Issue 3.

[12] “Eyeless in Gaza”, New York Review of Books, 23 October 2003

[13] Shakespeare, Henry IV Pt I, Glendower and Hotspur (3.1)

[14] “Law and the Use of Force After Iraq”, Survival, vol 45 no 2,2003

[15] The Commission’s members were Gareth Evans and Mohamed Sahnoun (Co-Chairs), Gisele Cote-Harper, Lee Hamilton, Michael Ignatieff, Vladimir Lukin, Klaus Naumann, Cyril Ramaphosa, Fidel Ramos, Cornelio Sommaruga, Eduardo Stein and Ramesh Thakur. It consulted comprehensively over a full year, meeting in Asia and Africa as well as North America and Europe, and holding roundtables and other consultations in Latin America, the Middle East, Russia and China. The ICISS`report, with its large supplementary research volume, is available on www.iciss-ciise.gc.ca.

[16] See Gareth Evans, “Humanity did not Justify this War”, Financial Times, 15 May 2003; Kenneth Roth , “War in Iraq: Not a Humanitarian Intervention”, Human Rights Watch World Report 2004.

[17] Compare the argument for differential treatment – culminating in the application of non-Security Council endorsed military force if necessary – against certain ‘closed societ[ies] with no effective internal opposition’ in Lee Feinsten and Anne-Marie Slaughter, “The Duty to Prevent” (Foreign Affairs, Jan/Feb 2004), and my commentary at the American Society of International Law Conference, Washington DC, 1 April 2004, “Uneasy Bedfellows: ‘The Responsibility to Protect’ and Feinstein-Slaughter’s ‘Duty to Prevent’ ”, accessible at www.crisisweb.org.

[18] Allen Buchanan & Robert O. Keohane, “The Preventive Use of Force: A Cosmopolitan Institutional Proposal”, Ethics in International Affairs, Jan 2004.

[19] Kosovo Report, Independent International Commission on Kosovo, Co-chairs Richard Goldstone and Carl Tham, Oxford, 2000.

[20] Eg. Sohail Hashmi, “Interpreting the Islamic Ethics of War and Peace” in Hashmi ed., Islamic Political Ethics; Oliver P. Ramsbotham, ‘Islam, Christianity and Forcible Humanitarian Intervention’, Ethics and International Affairs, Vol 12 1998. I am indebted for these and other references and input to ICG Research Analyst Dan Vexler.

[21] A former populist Premier of NSW, Jack Lang, gave future Prime Minister Paul Keating what he once told me was a formative piece of advice: “In any horse race, son, always back the one called Self Interest. He’ll be the only one trying.”