(a)  Authorization by the Security Council under Chapter VII UN Charter

15  What has been called ‘Security Council-authorized collective humanitarian intervention’ (Franck [2002] 136–7) or simply ‘collective humanitarian intervention’ is nothing but the use of military force authorized by the Security Council under Chapter VII UN Charter for the maintenance or restoration of international peace and security, in circumstances where there is a humanitarian aspect to the Council’s aims. Security Council practice since 1990 has extended the interpretation of ‘threat to the peace’, to the point that it is now accepted that egregious and widespread human rights violations within a single State, along with purely internal armed conflicts, can constitute such a threat. This was a ‘settled practice’ already in 1995, according to the International Criminal Tribunal for the Former Yugoslavia (ICTY) in Prosecutor v Tadić (Decision on the Defence Motion for Interlocutory Appeal on Jurisdiction IT-94-1-AR72 [2 October 1995] para. 30). Such violations or internal armed conflicts may result in a humanitarian crisis, and may create refugee flows, destabilizing the region and risking armed responses by neighbouring States. But even beyond these transboundary effects, the rise of human rights during the last century, and the concomitant shrinking of the domaine réservé of domestic jurisdiction, has led the Security Council to acknowledge that purely internal situations may qualify as threats to the peace, even if the risk of serious transboundary consequences is marginal. In such circumstances, the Security Council may authorize States to take forcible measures to stop the human rights violations and avert or put an end to the humanitarian crisis. The use of force then can be seen as having a ‘humanitarian purpose’; and the intervention of the international community as represented by the States willing to heed the Security Council’s call to arms becomes a ‘humanitarian intervention’, but a collective one.

16  There are numerous examples of UN-authorized armed interventions that have been characterized as ‘humanitarian’ by jurists. They include the UN-authorized interventions in Somalia (UNSC Res 794 [1992] [3 December 1992] SCOR 47^th Year 63 Haiti (UNSC Res 940 [1994] [31 July 1994] SCOR 49^th Year 51); Rwanda (UNSC Res 929 [1994] [22 June 1994] SCOR 49^th Year 10); Bosnia and Herzegovina (UNSC Res 836 [1993] [4 June 1993] SCOR 48^th Year 13; Res 1031 [1995] [15 December 1995] SCOR 50^th Year 18; Res 1088 [1996] [12 December 1996] SCOR 51^st Year 42); Albania (Res 1101 [1997] [28 March 1997] 52^nd Year 58); and East Timor (UNSC Res 1264 [1999] [15 September 1999] SCOR 54^th Year 128). In each of these cases, the Security Council authorized the use of ‘all necessary means’ to ensure delivery of humanitarian assistance (but not—initially at least—to assist any warring parties in an internal conflict), or to monitor the implementation of a ceasefire or peace agreement.

17  In April 2011, the Secretary-General of the UN instructed the UN Operation in Côte d’Ivoire (‘UNOCI’), and the French forces supporting it, to use force in order to prevent the use of heavy weapons against the civilian population in Abidjan by one of the parties to the conflict there. This was pursuant to authorizations by the Security Council to UNOCI and the French forces to use ‘all necessary means’ to protect civilians ‘under imminent threat of physical violence’ in UN Security Council Resolution 1933 of 30 June 2010 (UN Doc S/RES/1933 at paras 16–17), Resolution 1975 of 30 March 2011 (UN Doc S/RES/1975 at para. 6), Resolution 1739 of 10 January 2007 (SCOR [1 August 2006–31 July 2007] 208 at para. 8), and Resolution 1962 of 20 December 2010 (UN Doc S/RES/1962 at para. 17) respectively. The Secretary-General stated that the military operation undertaken by UNOCI was not in support of any party to the conflict, but rather was action in self-defence to protect civilians, in accordance with UNOCI’s mandate (UN Doc SG/SM/13494 of 4 April 2011), even though the attacks were directed primarily against one party to the internal conflict. The authorization to use force in Libya by virtue of UN Security Council Resolution 1973 of 17 March 2011 (UN Doc S/RES/1973) was meant to ensure the protection of ‘civilians and civilian populated areas under threat of attack’ (at para. 4), but again force was employed against one of the sides to the conflict. While questions may emerge as to the scope of the authorization and the specific measures and targets covered, this is a matter of interpretation of the objective set by the Security Council. The important point for present purposes is that the use of force for arguably humanitarian reasons has been explicitly authorized by the Security Council in all instances described above.

18  There are also a number of cases in which force was argued to have been authorized by the Council, either implicitly or retrospectively, in humanitarian contexts. The US, the UK, and France intervened in Iraq in 1991 to alleviate the suffering of Kurdish (and subsequently also Shia) populations that were being oppressed by Baghdad. They intervened first by the establishment of safe havens and then by the introduction of no-fly zones over the north (and subsequently also the south) of the country (Iraq, No-Fly Zones). These no-fly zones were kept in place until the 2003 invasion of Iraq (Iraq, Invasion of [2003]), though France withdrew from their enforcement in 1998. The intervening States relied primarily on UN Security Council Resolution 688 of 5 April 1991, ‘in support of’ which action was taken. The authorization of UN Member States to use force in order to implement Security Council Resolutions 660 (1990) et seq and to restore peace and security in the area in UN Security Council Resolution 678 of 29 November 1990 (SCOR 45^th Year 27), following Iraq’s invasion of Kuwait, had already ceased in accordance with UN Security Council Resolution 687 of 3 April 1991 (at paras 1 and 33). The argument in support of the establishment of safe havens and no-fly zones was that Security Council Resolution 688, though it was not adopted under Chapter VII UN Charter and did not include the shibboleth ‘all necessary means’, had implicitly authorized the use of force for the limited purpose of protecting the Kurds and Shi’ites in Iraq.

19  Along similar lines, it was argued with respect to the North Atlantic Treaty Organization (NATO) intervention in Kosovo that the NATO bombardments could be justified on the basis of UN Security Council Resolution 1199 of 23 September 1998 (SCOR 53^rd Year 13), para. 16 of which provided that the Council would consider additional measures if the ones provided for in UN Security Council Resolution 1160 of 31 March 1998 (SCOR 53^rd Year 10) did not lead to the desired results such as the defeat of violence and terrorism. France, among other intervening States, considered that the use of force had been implicitly authorized by Security Council Resolution 1199 and Resolution 1203 of 24 October 1998 (SCOR 53^rd Year 15) when further breaches of measures provided for in Security Council Resolutions did occur. Another argument considered that the Security Council implicitly and retrospectively authorized the use of force against the (then) Federal Republic of Yugoslavia (‘FRY’) because it did not condemn the threat of force by NATO that led to the conclusion of agreements between the FRY and NATO and the Organization for Security and Co-operation in Europe (OSCE) on verification in 1998, but rather endorsed these agreements (UNSC Res 1203). These arguments have caused much controversy in the literature.

20  With respect to potential ex post facto Security Council authorizations (or ‘ratifications’), the Economic Community of West African States (ECOWAS) interventions in Liberia and Sierra Leone through the Economic Community of West African States Monitoring Group (‘ECOMOG’) between 1990 and 1999, which were said to have essentially humanitarian aims, were ‘commended’ by the Council, while limited authorization to undertake visit and search of ships (Ships, Visit and Search) to enforce embargoes was granted when armed operations were already under way. This, it was argued by some (but disputed by others), amounted to ex post facto authorization of ECOWAS action. Similarly, French action in the Central African Republic in 1997 was approved and further authorized by the Security Council. These instances are perhaps enough to establish the possibility that the Security Council may retroactively authorize, validate, or ratify forcible action, even though significant reservations remain, for example with respect to the right of self-defence of the target State against the use of force that is, at the time, illegal, but is then retroactively authorized by the Security Council.

21  Conversely, no ex post facto authorization can be claimed to have been given through UN Security Council Resolution 1244 of 10 June 1999 (SCOR 54^th Year 32), which established an international civilian and military presence in Kosovo following the NATO bombing campaign. It is one thing to accept the status quo based on a pragmatic attitude towards the situation on the ground, which is what 1244 actually did, and another to endorse an action explicitly, as the Council arguably did with respect to the ECOWAS interventions in Liberia and Sierra Leone (by ‘commending’ ECOWAS and ECOMOG for their ‘efforts’). Some have argued that the mere absence of condemnation by the Council serves as some form of retrospective authorization. This pushes an already fragile argument on ex post facto authorization to breaking point. It is very difficult to establish the meaning of and motives for an omission with any semblance of certainty. The failure of the draft resolution that would have condemned the NATO action in Kosovo serves as a pertinent example: while some commentators rushed to consider this as an implicit ex post facto authorization of NATO, others pointed to the reasons put forward by States for rejecting the draft resolution, namely their uneasiness at the resolution being unbalanced, favouring the Serbs and seemingly lending support to the admittedly oppressive regime of Milošević.

22  In any event, forcible action authorized (even if ex post facto) by the Security Council under Chapter VII UN Charter, would constitute collective enforcement action that is legal under the UN Charter and international law as a recognized exception to the prohibition of the use of force. It would not constitute justified unilateral humanitarian intervention, or support the right to engage in unilateral humanitarian intervention.

(b)  Use of Force in Self-Defence

23  Humanitarian intervention in order to alleviate the suffering of a local population cannot, without more, be justified as self-defence. Self-defence under Art. 51 UN Charter requires that an armed attack occur against a State. In most cases, widespread violations of human rights will not reach the gravity threshold of an armed attack. Even if the oppression does reach the threshold of an armed attack, however, there will be no armed attack against a State, but at most an armed attack against the population of the State by or with the support or inaction of State authorities. The right to self-defence under international law vests in States and not in sub-State entities such as the local population. Moreover, the oppression will, ex hypothesi, not emanate from another State, but will be by the government upon its own people.

24  There have been proposals within the North Atlantic Assembly to extend the right of self-defence to cover the ‘defence of common interests and values, including when the latter are threatened by humanitarian catastrophes, crimes against humanity, and war crimes’ (North Atlantic Assembly Resolution 283 para. 15 (e)), but international law certainly does not yet extend so far, nor does it seem to be moving in this direction. It has also been argued that the right of (collective) self-defence should perhaps cover not just attacks on States, but attacks on populations as well, since defence of a population is as much warranted as the defence of a political structure (Henkin 833). But these suggestions stretch the interpretation of Art. 51 UN Charter far beyond breaking point and lack any foundation in the practice and opinio iuris of States. Defending its intervention in East Pakistan (later Bangladesh) in 1971, India argued before the UN General Assembly that the influx of millions of East Bengali refugees fleeing Pakistani repression amounted to ‘civil aggression’ comparable to an armed attack, but this—along with all other justifications put forward by India—was overwhelmingly rejected by the General Assembly, which called for the immediate cessation of the hostilities and for the withdrawal of armed forces.

25 ‘Humanitarian’ interventions have been justified on the basis of the right of self-defence under Art. 51 UN Charter. But this has been in cases where the State resorting to force could claim to have suffered an armed attack in the traditional sense by action of the target State. This was the justification of Tanzania for its intervention in Uganda in 1979. That intervention toppled the regime of Idi Amin, which had an atrocious human rights record; but Tanzania’s legal justification of self-defence did not refer to that atrocious record or to the dire humanitarian situation in Uganda, but rather remained within the traditional paradigm. The same is true of the justification put forward by Vietnam for its 1978 intervention in Democratic Kampuchea (Cambodia), which brought an end to the violent and abusive Khmer Rouge regime (Cambodia Conflicts [Kampuchea]). In both of these cases, humanitarian concerns were commingled with the official justification of the use of armed force as self-defence. Humanitarian considerations, in the absence of an armed attack against the intervening State or (in the case of collective self-defence) its allies, cannot in itself justify a use of force as an exercise of the right of self-defence. Further, the use of force in self-defence must in any event be limited to countering the armed attack and may not extend to regime change in the target State.

(c)  A New Exception Allowing Forcible Unilateral Humanitarian Intervention?

26  In view of the difficulty of fitting humanitarian intervention undertaken in the absence of Security Council authorization or of an armed attack against a State within the traditional exceptions of the prohibition of the use of force, arguments have been put forward that an additional, new exception may have emerged to the prohibition of the use of force. This would have to be an exception carved out through subsequent practice of UN Member States resulting in a new interpretation of the relevant Charter provisions (cf Art. 31 (3) (b) Vienna Convention on the Law of Treaties [1969]; ‘VCLT’), or possibly through the emergence of a new customary rule (supervening custom). The ‘reinterpretation’ might, for example, require the reference to territorial integrity and political independence in Art. 2 (4) UN Charter to be read as qualifying the prohibition on the threat or use of force, in the manner described above (where it was rejected as a correct reading of the Charter as originally drafted). Such a ‘reinterpretation’, it might be said, is not an ‘amendment’ of the Charter, which would have to follow Arts 108 and 109 UN Charter. The alleged new rule would bring about a qualification of the clear provisions of Art. 2 (4) UN Charter as currently interpreted, and thus would require that the membership of the UN accept this reinterpretation of the Charter, evidenced through the widespread (if not complete) support of the UN Member States.

27  If the argument were not based on a ‘reinterpretation’ of the Charter, it would have to emerge as a new rule of customary international law through the practice and opinio iuris of States. But when cast as an exception to the prohibition of the use of force, which has achieved the status of ius cogens (at least in its core, ie the prohibition of aggression), the new customary law exception to the prohibition would have to achieve the same status, making the requirements for its emergence (assuming it to be logically and legally possible) even more exacting than those for an ordinary rule of customary international law (cf Art. 53 VCLT).

28  A number of instances of practice since 1945 have been invoked by authors—and to a lesser extent by States—as evidence of a general practice of the assertion of a unilateral right to intervene to avert or put an end to a humanitarian crisis or to widespread violations of human rights. The instances commonly cited include, among others, the Indian intervention in East Pakistan (Bangladesh) in 1971; the Tanzanian intervention in Uganda in 1978; the Vietnamese intervention in Democratic Kampuchea in 1978; the French intervention in the Central African Empire (later the Central African Republic) in 1979; the US interventions in Grenada (1983) and Panama (1989); the ECOWAS/ECOMOG interventions in Liberia (1990) and Sierra Leone (1997); the US, UK, and French intervention in Iraq to protect Kurdish and Shia populations from 1991 to 2003 (France intervening until 1998); the interventions in Somalia (1992); Rwanda (1994); and East Timor (1999); and of course the NATO intervention in Kosovo in 1999.

29  In order to support the argument that a new rule of customary international law has emerged, it would be necessary that States made claims that these instances of the use of force were lawful because of the doctrine of humanitarian intervention. The ICJ asserted in the Military and Paramilitary Activities in and against Nicaragua Case that no one has the ‘authority to ascribe to States legal views which they do not themselves advance’ (at para. 207). States may, nonetheless, act in the belief that they are entitled to do so, and only later articulate the precise justification for their action. But such practice is limited. No opinio iuris in favour of a new customary law exception to the Charter prohibition can be deduced from State action authorized by the Security Council, such as the interventions in Somalia, Rwanda, East Timor, and arguably the ECOWAS/ECOMOG interventions in Liberia and Sierra Leone (which were explicitly condoned, and at least arguably in part explicitly authorized, by the Security Council). Only instances of unauthorized intervention could provide evidence of opinio iuris in favour of a new exception allowing unilateral humanitarian intervention. But again, in none of the instances above did the intervening States argue that their actions were justified by a rule of customary international law that allows intervention on humanitarian grounds. In the interventions of India in East Pakistan, Tanzania in Uganda, and Vietnam in Cambodia, the intervening States claimed to have been acting in self-defence in response to border incursions and other acts or threats of force; and even then the response of the international community was far from unequivocal acceptance of the interventions. It rather ranged from strong condemnation (Vietnam) to mere silence (Tanzania).

30  The use of force in Iraq to establish safe havens and no-fly zones in 1991–92 was argued to have been ‘in support of’ (and thus implicitly authorized by) UN Security Council Resolution 688; while the use of force to enforce the no-fly zones in 1993 was claimed by the US to be based on the right of self-defence against threats of attacking coalition aircraft patrolling the zones. These justifications should be seen as distinct: an argument based on the right of self-defence of Allied aircraft patrolling the no-fly zones cannot itself justify the establishment of the zones. The argument is circular: no-fly zones themselves imply a threat to use force, which must be justified. France combined the argument of implicit authorization with a claim to be responding to violations of UN Security Council Resolution 687; and the UK initially enunciated a free-standing right of humanitarian intervention, but then modified this claim by combining it with the argument that aircraft patrolling the no-fly zones had the right to use force to defend themselves against any attack (which is a distinct matter, as explained earlier).

31  In other cases of use of force, such as those of the US in Grenada and Panama, the justifications ranged from action to rescue nationals abroad, to invitation by the legitimate government, to action for the restoration of democracy. In all these instances there were condemnatory resolutions passed in the UN General Assembly. Even when humanitarian intervention was explicitly claimed as a justification, it was never the sole justification but was always combined with a universally accepted exception to the prohibition on the use of force, such as self-defence or authorization by the Security Council.

32  The question arises whether, irrespective of the legal position prior to 1999, the NATO intervention in the (then) FRY may have provided at least part of the requisite State practice and opinio iuris for the emergence of a customary exception from the prohibition of the use of force along the lines of ‘humanitarian intervention’. This seems unlikely. Some of the States participating in the intervention expressly denied that they considered the Kosovo campaign to be an instance where they had the right to act as they did under international law. Importantly, the German Foreign Minister stated before the Federal Parliament on 16 October 1998 that the NATO decision on air strikes against the FRY ‘must not become a precedent’, while the denial of precedential value to the Kosovo intervention was the major theme in the German parliamentary debate (Simma 13 and 20). Belgium stated in the UN General Assembly on 26 September 1999 that UN Security Council Resolution 1244 had achieved ‘a return to legality’ (Brownlie [2000] 908) and that it hoped that resort to force without Security Council authorization would not become a precedent (White 37). US arguments on the legality of the Kosovo intervention did not espouse any clear doctrine of humanitarian intervention, but rather relied—when referring to such humanitarian intervention—on a mixture of circumstances and principle in order to ‘qualify any universalist theory or wide-ranging rule that might prove less attractive in other hands’ (Henkin 829). This connects well with the German and other denials of the precedential value of the Kosovo intervention, and highlights the lack of any opinio iuris with regard to a right to unilateral humanitarian intervention.

33  The response of other, non-NATO, States to arguments that there was a legal basis for the Kosovo bombing campaign and for a right of humanitarian intervention was overwhelmingly negative. The Non-Aligned Movement (NAM), numbering well over half of the Member States of the UN, unequivocally condemned the use of force against the (then) FRY, as did many other States, some of which are nuclear powers. In these circumstances, no right of unilateral forcible humanitarian intervention can be said to have emerged as a rule of customary international law.

34  Some authors have argued that in all of the instances mentioned above, humanitarian objectives were the motives for action, and thus irrespective of the legal justifications articulated by the acting States, the instances constitute clear State practice in favour of a right of humanitarian intervention. Such a view, however, runs counter to explicit and clear statements of the ICJ on the formation of custom, which requires both practice and opinio iuris. Motives are not the same as reasons; and the requirement for opinio iuris looks to the latter. And while action by a State coupled with silence on its part as to its legal justification may be presumed to be accompanied by opinio iuris to the effect that its action is lawful on some basis or other, the same cannot be said when the State rebuts the presumption of opinio iuris in a specific case or elects to base its action on some legal bases to the exclusion of others. The latter, excluded, bases cannot carry the opinio iuris of the State. The fact that intervening States have been so reluctant to rely explicitly on a right of humanitarian intervention means that there is great difficulty in finding any opinio iuris that can properly be counted towards the establishment of a right of humanitarian intervention.

35  Post-Kosovo practice does not reveal any reliance on an alleged right of humanitarian intervention. With respect to the 2011 crisis in Libya, it is telling that force was only used to protect civilians after the UN Security Council adopted Resolution 1973, in which it authorized the use of ‘all necessary means’ by UN Member States to protect Libyan ‘civilians and civilian populated areas under threat of attack’ (at para. 4) and to enforce a no-fly zone (at para. 8). In the run-up to the Resolution’s adoption, numerous States, including the US, the UK, and NATO Member States collectively, underlined the need for Security Council authorization before any armed force could be used in Libya.