Question 6: What does the criterion of ‘necessity’ mean?
See my response to question 4 above.
In the Wall Advisory Opinion, the Court referred to Article 25 of the State Responsibility Articles and, in particular, to the requirement therein that “necessity” requires that the conduct in question “is the only means for the state to safeguard an essential interest against grave and imminent peril”. Leaving aside whether Article 25 of the State Responsibility Articles was appropriately invoked by the Court in this context, it is undisputed that the concept of necessity contemplates circumstances in which a state is faced with an immediate requirement to act against a grave peril. The concept can be traced back directly to the language of the formula used following the Caroline incident, ‘the necessity of self-defence, instant, over-whelming, leaving no choice of means, and no moment for deliberation’.
The interpretation of “necessity” in a self-defence context is, in my view, closely linked to the scale of the attack or threatened attack and the idea of prospective proportionality, i.e., what is necessary to effectively address the attack or threatened attack. It is also bound up with the question of whether a State has, or ought to have, a margin of appreciation when it comes to assessing a ‘grave and imminent peril’. The ICJ, in the Oil Platforms Case and the Wall Advisory Opinion, rejected both elements, in my view wrongly, as well as ignoring its assessment, in the Gabcikovo Case109, as well as in various examples of State practice (cited in the ILC State Responsibility commentaries), which suggest that necessity does admit of some margin of appreciation.
In my view, the use of force by way of self-defence may be justified as necessary in circumstances in which:
there is a well-founded appreciation of grave peril;
the use of force is in the circumstances the only available means or, if other means are available, is likely, on a considered assessment of all the circumstances, to be the only effective means available to address the threatened peril; and,
any delay in the use of force by way of self-defence would result in a significant increase in the risk of peril.
The use of force on grounds of necessity is also limited by the principle of proportionality.
It means that those making the decisions have no reasonable alternative but to take action. In this context, the much-mooted notion of ‘wars of choice’ seems to me redundant, at least in the context of liberal democracies. There might be choices over how much to contribute, in which ways, to a use of destructive force, but I cannot imagine any democratic leader(s) who would take a decision to engage armed forces, or other means of applying restrained coercive violence, without judging that they had no other, or better, option. In this context, the interaction with ‘imminence’ must be weighed. Necessity will determine imminence – it will be necessary to act before it is too late. Here, another question informing reflection must be, once again, what must an actor – state or non-state (or virtual state?) – do to demonstrate that it is not a threat, in a broad sense, and so that action is not justified. Certainly, there will be an increasing onus on those charged with being a threat to show that they are not.
That the force used is no more than is reasonably necessary to achieve the goal permitted – for example, the expulsion of an invader, the prevention of future terrorist threats from a particular source.
Applying that test is very complicated since it has to take account of such factors as the need to keep your own casualties as low as possible. The Belgrano sinking is an instance of how difficult the test can be to apply.
The criterion of necessity means that there should be no reasonable alternative to the proposed course of action that is likely to be reasonably effective in averting the threat. That principle has several strands.
Firstly, it answers the question, “what force may be used?” It means that less forceful, or less extensive means would be insufficient to remove the threat or reduce it to an acceptable level. That point is important as it is generally not necessary to kill every last soldier (or insurgent) in order to remove the threat for all practical purposes, or at least to reduce it to a level at which, say, normal policing action can reasonably be expected to avert a renewed attack. That suggests that the analysis is dynamic, in the sense that throughout a use of force in self-defence the question should be continuously asked, is there a real need for further force to be used. In other words, defensive force should be controlled, and not consist in the triggered release of a pre-ordained response.
Threats constitute a particular problem. The making of a threat to kill may be said to be a lesser degree of force and to inflict a lesser degree of harm that an actual moderate physical assault. There have, I think, been explicit threats in recent history to use ‘disproportionate force’ in response to any attack. I do not think that the law is either very clear or very important in this context. It is plain that such threats are unhelpful in any society which seeks to maintain the principle of proportionality.
Secondly, it answers the question, “when may force be used?” Necessity is an aspect of imminence, in as much as it requires that there be no time to pursue non-forcible measures with any reasonable chance of averting or stopping the attack.
The third strand is another aspect of the above question of when force may be used, which might be thought of as the “who” question. It is the question of whether it is necessary for the target state to take action, or whether another state is able and willing to act, and intends to do so.
Finally, it raises the question of, “against whom” a necessary response may be directed. The purpose of defensive force is to disarm the (imminent) attacker and stop, or avert, the attack. Action against third parties, such as the central government of the state in which the attacker is found, may be justified if it is controlling the attack, but may not be justified if the attackers are operating independently of that government. Defensive force cannot be used to punish states for a failure to repress terrorist actions emanating from their territory.
Sir Adam Roberts
The criterion of “necessity”, in relation to a planned use of force in another state, has to refer first and foremost to the lack of reasonable alternatives to the projected military action. Other military or non-military means of achieving an objective need to be carefully considered and, if at all possible, pursued.
This meaning of necessity is especially important in the kinds of crises that have given rise to discussion of pre-emptive and preventive uses of force. For example, the fact that a terrorist insurgency is getting military support from another state does not in itself prove the necessity for attacking the territory of that state. The UK did not conduct its counter-insurgency operations in Malaya from 1948 onwards, or in Northern Ireland more recently, on that basis – and it would have run into trouble if it had argued in these cases that it was entitled to attack China and Ireland respectively. I do not deal here with the jus in bello principle of ‘military necessity’, which is somewhat distinct, though with certain points of overlap.
See my response to question 5 above.
Necessity is linked to proportionality in that the response has to be limited to measures appropriate in order to deal with the threat (see my response to question 5, above). Indeed, necessity is a gloss on proportionality and restricts the response to the elimination of the threat.
Necessity will also relate to the means available so that the kinds of forces and the level of armament to hand will be relevant to the type and intensity of response that it would be reasonable to expect, as well as the realistic possibilities of resorting to non-military means in the circumstances. What this means in practice is, however, uncertain. Tanzania in responding legitimately to a Ugandan attack upon the Kagera salient in 1978 continued on to Kampala and overthrew the regime. There was little meaningful criticism. Again, the UN authorised coalition forces in 1991 stopped at the Iraqi border after expelling invading troops from Kuwait.
Necessity refers to the action required to terminate the attack and/or subdue the threat. Thus, necessity and proportionality could come into serious conflict. The standard formulation, of course, is conjunctive: self-defence has to be both proportionate and necessary.
The difficult case is the act of self-defence that seeks to extinguish a serious threat or ongoing use of force through measures that appear disproportionate to the original armed attack. As Myjer and White put it: “Does an attack on a small part of the United States justify an armed response against a whole country?”.110 One way round this problem is to say that proportionality is to be judged against the threat as well as the armed attack itself. This seems more pragmatic but risks collapsing proportionality and necessity.
With anticipatory self-defence, the position is murkier still. Necessary to counter-act this attack? This and future attacks? Proportionate to the expected attack?
It might be useful to refer to the case-law of the European Court of Human Rights as a guide to the idea of necessity.111 The Court takes as its first standard action being "absolutely necessary" (coincidentally, for lawfully self-defence, inter alia,in Article 2(2) of the Convention). What this requires of the state is an investigation of the circumstances to see that there is no alternative, careful planning about the deployment and use of the force determined to be absolutely necessary, and a strict limitation of the actual use of force to the purposes identified in the Convention. This is a more demanding test than the international law standard of "necessity". Whether action is merely "necessary" for some purpose or other requires a proportionality assessment, taking into account the importance of the individual right which would be interfered with by the proposed action. The state is expected to produce evidence to support its claim of necessity. Considerable weight is attached to the existence of alternative means for securing the state's ends. Futility of action to achieve the claimed aim is the clearest demonstration of absence of necessity. The Court distinguishes the test of necessity from "reasonableness" or good faith. Reasonableness does not take sufficient notice of what is at stake for the individual; good faith does not protect against (patently) inadequate decision-making.
The interest which is at stake when a state claims the right to respond by self-defence may be the lives and property of a state and its nationals when the latter is not directly implicated or not implicated at all in a wrongful act against the defending state. Necessity means that the force used was the only way in which the armed attack could have been terminated and that the force used was directed solely to terminating the attack and, if the defending state is faced with a campaign against, of preventing or minimising the effects of future elements of the campaign against it. The force must be used for self-defence purposes and no other but hard questions arise about the necessity of removing a regime implacably opposed to the self-defending state and willing to resume hostilities when it is able, if left in place. Considerations of self-determination have an impact on programmes for regime change. That is to say, even necessity is limited by other rules of law.112 As the defeats of Germany and Japan show, overthrow of particularly recalcitrant governments might find legal justification in some circumstances, even as exercises of self-defence.
The requirement of necessity incorporates two related aspects: first, that all peaceful means of defending the state have been exhausted, and second, that the scale of the threat is of sufficient gravity to warrant the use of force.
Sir Michael Wood
The use of force in self-defence must be necessary, meaning that other (non-forcible) means to reverse/avert the attack must be unavailable.
Question 7: Is it permissible to use force in self-defence against a terrorist grouping within another state although that state may not be unwilling, but simply unable, to deal with a terrorist organisation itself?
Sir Franklin Berman
The answer in principle is ‘yes’, but lies at a level of generality that renders it of singularly little practical usefulness. The trivial case is that of a state that is willing (though unable); such a state must be presumed to be ready to act jointly with other states specially affected, or at least to consent to their acting, in such a way as to remove any question mark over legality.
It is implicit that the first step has to be to call on the target state to meet its obligations, and that squares with the practice of the UN Security Council, e.g. in respect of Libya and Afghanistan. A state that merely claims to be willing, but declines to act alone or in combination with others, puts itself by that fact into the ‘unwilling’ category. It also makes itself a wrongdoer (in respect of other states directly affected), so diminishing its rights in relation to corrective measures they are entitled to take within the limits laid down by international law.
That said, the nature and degree of the force that may legitimately be employed will not be identical to the case where the territorial state’s own actions give rise to the threat to the other state or states. Specifically, the fact that in international armed conflict a belligerent state’s infrastructure and civilian population are exposed to the risk of collateral damage can’t readily be assumed to apply to the case of limited punitive operations aimed at neutralizing a terrorist organization.
The closest analogy (though a fruitful one) is intervention by a belligerent on neutral territory to put a stop to hostile activities carried out by an opposing belligerent there.
A state is required to ensure that its territory is not used for the commission of unlawful acts. Where a state is unable to assert control over a terrorist organisation located in its territory, the state which is a victim of the terrorist attacks would, as a last resort, and on the basis of the principles addressed above, be permitted to act in self-defence against the terrorist organisation in the state in which it is located.
This is a particularly difficult issue – though the simple answer is ‘yes’, as it has already happened. However, there is clearly a case theoretically at least for saying that the US operations against Afghanistan and Sudan, while self-defensive, were also acts of aggression against those states, because the states themselves had not been involved per se in the attacks on the Nairobi and Dar es Salaam US embassies. This appears to be the case acutely regarding Sudan, where the information, as I understand it, later proved to be out of date and wrong. Therefore, regarding each of the elements of the self-defence equation in the changing context, there has to be some indication of what the conditions are for any actor (state or otherwise) reasonably taking action it believes to be justified at the time and for which it does not believe that there is an alternative to demonstrate responsibility afterwards. Clearly any lives lost cannot be restored. But it seems to me that, while action might be necessary, there must be responsibility after the fact (whatever the case, but especially if it is shown clearly that action, in this uncertain world, was appropriate, in terms of that which was known (while more could not reasonably be known) but the legitimacy of which is compromised by the emergence of information which was not known and could not be known
In an extreme case, I think it is, but only as a last resort. There is an analogy with the right of a belligerent to destroy enemy forces which are using neutral territory or waters in a case where the neutral is unable to enforce its neutral rights.
Yes. The state may not be responsible for the acts of the terrorists, but it is responsible for any failure to take reasonable steps to prevent the use of its territory as a base for attacks on other states. Its inability to discharge the duty does not relieve it of the duty. If it refuses an ‘offer’ to send troops into the state to nullify a threat of an imminent attack, it must find some other way of discharging its responsibility. If it does not, it should be regarded as ‘unable or unwilling’ to discharge its responsibilities.
But in any event, I repeat an earlier point.113 The right to use force in self-defence is an inherent right. It is not limited to ‘forcible counter-measures’ and thus not dependent upon any prior breach of international law by the state in which defensive force is used.
Sir Adam Roberts
It is extremely hard to answer this question in the abstract. In principle it would be wrong to give a negative answer. However, a positive answer should not be seen as a green light. Terrorist organisations are not easily rooted out by foreign armed forces, especially when the latter are ignorant of the geography, culture and language of the society concerned. Intervention may make most sense when there are significant local allies (whether the government, or regional forces) who will collaborate in addressing the problem posed by the terrorist organisation.
If it is accepted that force may be used against non-statal actors where the host state is unwilling to act, then there is no reason in principle why force cannot be used when a host state is willing but unable to act. However, the circumstances in which force can be used may differ, particularly in regard to the nature and extent of prior engagement with the host state, and the question of whether (and under what conditions) its formal consent may be required (see Grenada).
See my response to question 2 above. Inability to act is not a defence since the target state is still under attack from the first state. However, where the state, although unable to deal with the terrorist organisations, is indeed willing to take action, it must take steps to seek assistance in order to mitigate its powerlessness and thus render the proposed response from the target state unnecessary.
There are two general problems worth raising here. The first concerns the problem of nomenclature. The terms “pre-emptive self-defence”, “preventative war”, “anticipatory self-defence” and “precautionary self-defence” are not terms of art and their, often interchangeable, use has created great confusion e.g. a group of politicians I spoke with this summer reversed the usual international law categorisations of anticipatory (immediate) and pre-emptive (precautionary) self-defence.
Secondly, though, I wonder if the search for universalisable rules is misconceived. Perhaps in approaching the issue of force there ought to be more emphasis on the apparent prerogatives of Great Powers and the vulnerability of outlaw states. Might it not be the case that principles of international law applying to self-defence outside the Charter are likely to operate in ways rather similar to the operation of collective security under the terms of the Charter? The putative legitimacy of a use of force in each case may become conditional on the status of the actors employing such force and the status of those who are subject to such attacks. Tentatively in Kosovo and, more confidently, in Afghanistan, the Great Powers, may have attempted a redefinition of sovereign equality itself.
What distinguished the response against Afghanistan were the particular facts. In the case of the 11 September 2001 attack, the magnitude of the damage inflicted by the non-state actor was exceptional but its legal relevance was as evidence of what the continuation of the campaign against the USA might mean. Responsibility for the attack fell to Al-Qaeda, established by the group's own statements, and more were plausibly threatened. That the group benefited from facilities in Afghanistan was also made out; the non-co-operation of the government of Afghanistan to engage in an operation against Al-Qaeda was established to a much lesser degree. The Taliban government was in breach of obligations under The Security Council Resolution to hand over Bin Laden to a state willing to prosecute him (though those resolutions did not give, nor were claimed by any state to give, an independent right to use force to enforce them). Non-co-operation might have been in breach of an obligation owed to the UN (though not to the US); further, it might have been evidence of the necessity of taking action to prevent or limit the next episodes in Al-Qaeda' campaign, regardless of any responsibility of Afghanistan.
I am not sure how much we can extrapolate from this. The existence of a campaign or the presence of an attack as understood above from non-state groups requires demonstration, to give rise to a right of self-defence at all. There was evidence of complicity of the Ugandan authorities with the hijackers in the Entebbe incident. To demonstrate the necessity of action against the territory of another state not directly responsible for the acts of the non-state group requires, inter alia, the demonstration that there is no other means of meeting the attack (and that this way will do so). The state potentially under threat might be persuaded to co-operate in the face of a legitimate threat to its territory but, save for the most compelling emergency, the territorial state is surely entitled to proceed first in its own way against an identified group on its territory. The Security Council might authorise action. It seems to me likely that satisfying the necessity test will be rare.
One of the criticisms levelled against the US after its attacks on the Taliban and Al-Qaeda is that the government of Afghanistan was not responsible for the actions of terrorist groups based on its territory. The question is, “at what point do we decide that the conduct of a terrorist group is attributable to that of the state upon whose territory the group is based?” Those who deny the legality of Operation Enduring Freedom argue that the Taliban lacked control over Osama bin Laden. In the Nicaragua Case, the ICJ maintained that the USA was not responsible for the breaches of international humanitarian law by the Contras since ‘it had not directed and controlled the individual operations giving rise to these breaches’.114 Applying this precedent to the case of Afghanistan, it is argued that the Taliban’s involvement in the attacks on the United States of 11 September 2001 was too distant for it to be held responsible for the attack.
The implication of this is that the US was not legally justified in using force against Afghanistan, even though it had knowingly harboured the terrorist group that launched such a devastating attack upon the territory of the USA. This is not a realistic position, and the support given to the US position by Security Council resolutions and many other states, demonstrated that international society acknowledged that there was a legal basis to use force in self-defence against attacks such as those launched on 11 September 2001.
Sir Michael Wood
Yes (provided, of course, that the other conditions for self-defence are met). Compare the “unable or unwilling” test for intervention to protect nationals. In many cases, however, where a state is unable it will consent to action so there will be no need to have recourse to the right of self-defence.
Case references and Resolutions
Exchange of letters between US Secretary of State Daniel Webster and Lord Ashburton, Foreign Secretary of Great Britain, relating to the case of the SS Caroline, 1837; 29 BFSP 1137-1138; 30 BFSP 195-196; See Jennings (1938) 32 AJIL 82 and Rogoff and Collins (1990) 16 Brooklyn JIL 493.
International Court of Justice, Advisory Opinion of 9 July 2004 on “Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory”, ICJ Reports (2004).
Separate Opinion of Judge Higgins on “Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory”, ICJ Reports (2004).
Separate Opinion of Judge Buergenthal on “Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory”, ICJ Reports (2004).
International Court of Justice, Order of 2 June 1999 on Request for Provisional Measures “Legality of Use of Force (Yugoslavia v. United States of America/Spain)” ICJ Reports (1999) 124.
International Court of Justice, Advisory Opinion of 8 July 1996 on “Legality of the Threat or Use of Nuclear Weapons”, ICJ Reports 1996, 226; 35 ILM 809 and 1343.
International Court of Justice, Judgment of 25 September 1997 in the “Case Concerning the Gabcikovo-Nagymaros Project (Hungary v. Slovakia)” ICJ (1997).
International Court of Justice, Judgment of 6 November 2003, “Case Concerning Oil Platforms (Islamic Republic of Iran v. United States of America)”.
International Court of Justice, Preliminary Objections, “Case Concerning Oil Platforms (Islamic Republic of Iran v. United States of America”, ICJ Reports (1996) 803.
International Court of Justice, Judgment (Merits) of 27 June 1986, “Military and Paramilitary Activities in and Against Nicaragua (Nicaragua v. United States of America)”, ICJ Reports 1986, 14.
United Nations Security Council Resolution 1373 “Threats to international peace and security caused by terrorist acts”, S/RES/1373 (2001).
United Nations Security Council Resolution 1368 “Threats to international peace and security caused by terrorist acts”, S/RES/1368 (2001).
United Nations General Assembly Resolution 3314 (XXIX) “Definition of Aggression” (1974).
Bowett, D. “Reprisals Involving Recourse to Armed Force”, 66 AJIL 1 (1972).
Brunnée, J and Toope, S (2004) “The Use of Force: International Law After Iraq”
ICLQ 53: 785-806.
Byers, M. “Terrorism, the Use of Force and International Law after 11 September” (2002) 51 ICLQ 401.
Byers, M. “Preemptive Self-defense: Hegemony, Equality and Strategies of Legal Change”, 11 Journal of Political Philosophy, November 2003, pp.171-190.
Cassese, A. “Terrorism is also Disrupting some Crucial Legal Categories in International Law” (2001) 12 EJIL 993.
Charney, J. “The Use of Force Against Terrorism and International Law” (2001) 95 AJIL 835.
Corten, O. and Dubuisson, F. “Opération ‘liberté immuable’: une extension abusive du concept de légitime défense” (2002) 106 RGDIP 51.
Dinstein, Y. (2001) “War, Aggression and Self-Defence”, (3rd ed.) Cambridge University Press: CAMBRIDGE.
Eick, C. “‘Präemption’, ‘Prävention’ und Weiterentwicklung des Völkerrechts”, (2004) ZRP, 200.
Evans, G. “When is it Right to Fight? Legality, Legitimacy and the Use of Military Force”, 2004 Cyril Foster Lecture, Oxford University, 10 May 2004.
Franck, T.M. (2002) “Recourse to Force: State Action against Threats and Armed Attacks”, Hersch Lauterpacht Memorial Lectures Series, Cambridge University Press: CAMBRIDGE.
Franck, TM (2001) “Terrorism and the Right of Self-Defence”, 95 AJIL 839.
Franck, T.M. (1998) “Fairness in International Law and Institutions”, Clarendon Press: OXFORD.
Gray, C. (2000) “International Law and the Use of Force: Foundations of Public International Law”, Oxford University Press: OXFORD.