David Kretzmer is Professor Emeritus, Hebrew University of Jerusalem and Professor of Law, Sapir Academic College.
Many thanks to the editors of the EJIL for selecting my article for discussion on the blog and to Noam, Gina, Thomas and Mary Ellen for their thoughtful and perceptive comments. These comments provide me with the opportunity of clarifying some of the points I raised in the article and expressing my view about issues that I failed to consider.
The discussion in my article was confined to use of force in exercise of a state’s inherent right to self-defence, recognized in article 51 of the UN Charter. I did not consider humanitarian intervention, nor use of force authorized by the Security Council under article 42 of the Charter. However, Gina is quite right in concluding that my analysis of unilateral use of force by states implicitly rules out unilateral humanitarian intervention. Any decision on such intervention must be a collective one taken by the SC under Chapter VII. ( I shall not discuss the controversial view of the Independent International Commission on Kosovo that there may an intervention which while unlawful is nevertheless legitimate.) While article 42 speaks of forcible action “as may be necessary to maintain or restore international peace and security” I fully accept Mary Ellen’s view that such action must also meet the demands of proportionality.
What is the place of the “narrow proportionality” test in jus ad bellum? Thomas points out that while intimating that this test does indeed have such a place I neglected to develop the issue. Following the line of just war theory, Mary Ellen argues that the very essence of proportionality in jus ad bellum involves “weighing the cost of resort to military force in terms of lives lost and property destroyed relative to the value of the legitimate military end.” While Thomas mentions that there is little, if any, authority on which one can “conclude that the law on the use of force already includes a ‘narrow proportionality’ criterion” it seems to me that such a criterion is inherent in the very notion of proportionality. Hence, as in other contexts in which the means-end proportionality test is employed, some “cost-benefit” analysis must indeed be part of the jus ad bellum test too.
It must be recalled, however, that we only reach this part of the test after having decided that the means employed were necessary to achieve the legitimate ends. In this context “necessary” can mean either a rational connection between the means and ends, or an assessment that there were no less drastic means available to achieve those ends. Means that are unnecessary are ipso facto disproportionate. If regarded as necessary, the narrow proportionality test requires an assessment of their marginal costs and benefits in terms of the legitimate ends of using force in self-defence, which I discuss in the article. It is important to keep in mind that in addressing this question we look at what Noam calls “the big picture, not an ongoing inspection of single operations.” The latter are to be examined under jus in bello. And, of course, addressing “narrow proportionality” of the big picture is likely to be even more difficult than examining the proportionality single attacks.
Assessment of proportionality must initially be made before the resort to force, but it does not end there. Unless we accept Dinstein’s view that once a war has started the victim state may fight until the enemy is vanquished or surrenders, when the conflict continues an ongoing assessment of the costs and benefits of further use of force is required. And it is clear that at this stage that assessment must include weighing the marginal self-defence benefits to the state involved against the marginal costs to all affected parties. The problem is that in making this assessment we once again face the wide division of opinion on what these “self-defence benefits” include. Unless we extend the meaning of the words “halting and repelling” so that they become too open to be of any real help, we have to decide to what extent aiming to prevent (and possibly deter) further attacks by the party responsible for the armed attack is a legitimate aim. If we answer in the affirmative, furthering this aim will be part of the self-defence calculus.
Mary Ellen comments that I did not mention the link between the requirements of necessity and proportionality and the requirement of an armed attack in Article 51. I thought I had made it quite clear that the right to use force may only be exercised when an armed attack occurs, and that any such use of force is subject both to necessity and proportionality (p. 242). True, as opposed to Mary Ellen, I take the widely-accepted view that states may use force to thwart an imminent armed attack. But I mention both that this view does not seem compatible with the text of article 51 and that the opposing view (taken by Mary Ellen) “is not devoid of logic, and may be regarded as consistent with the collective security policy adopted in the Charter.” (p. 248). Among others, Mary Ellen cites Dinstein as a scholar who takes a strict view of the actual armed attack requirement in article 51. It should be recalled, however, that Dinstein also gives a very wide interpretation to the notion of an armed attack occurring. He argues that it includes a situation in which a state has embarked “on an irreversible course of action” towards an attack. I assume that it was with this view in mind that Christopher Greenwood remarks that “some commentators who reject the notion of anticipatory or pre-emptive self-defence adopt an approach to the question of when an armed attack begins which means that their application of the law of self-defence to the facts of many cases is very close to that of the proponents of some variations of the anticipatory self-defence theory”. (“Self-Defence” in Max Planck Encyclopedia of Public International Law, para. 42).
One major problem with the notion of “halting and repelling” as the sole legitimate aim of using force against an armed attack is that it implies that if the armed attack is over before the victim state can respond (9/11, a one-time missile attack) that state may not use force in self-defence. But, as we all know, in its Resolutions 1368 and 1373 of September 2001 relating to the 9/11 attacks on the US, the UN Security Council expressly recognized and reaffirmed the inherent individual or collective right to self-defence in accordance with the Charter. Implicit in this reference was recognition of the right of the US to use force in self-defence even though there was no longer an armed attack occurring. Furthermore, even such a strict interpreter of article 51 as Olivier Corten concedes that a state does not lose its right to use force in self-defence merely because the armed attack is over before it can respond. As I mention in the article, it is difficult to see how an armed attack that is no longer going on can be either halted or repelled. So either we have to say that “halting and repelling” doesn’t really mean what we think it means, or to accept that legitimate use of force after an armed attack has occurred may involve something other than halting and repelling that attack.
In the article I write that use of force may be disproportionate in jus ad bellum even if all specific attacks meet the proportionality demand of jus in bello. Noam raises an interesting question to which I gave some thought when writing the article: what about the opposite situation? Can use of force be proportionate under jus ad bellum if the conflict is waged in violation of jus in bello? Perhaps the answer was provided by the ICJ in its Nuclear Weapons Advisory Opinion, when it stated that “[a] use of force that is proportionate under the law of self-defence, must, in order to be lawful, also meet the requirements of the law applicable in armed conflict which comprise in particular the principles and rules of humanitarian law.” (Ibid, para. 42). Following this dictum I would argue that the issue Noam raises is not really one of proportionality. Clearly the fact that there are some violations of jus in bello will not of itself lead to the conclusion that the very use of force was unlawful. However, if the whole campaign rests on a strategy that is incompatible with jus in bello, there will be a violation of jus ad bellum too. Thus, for example, if a state that has been subject to an armed attack defends itself by bombing the civilian population of the attacking state, it will be liable not only for the violations of the IHL prohibition on making civilians the object of an attack, but also for violation of jus ad bellum.
As to the question of armed reprisals raised by Noam, my aim was to examine why such reprisals are regarded as incompatible with the Charter regime on use of force. I agree, of course, that the fact that some states carry out reprisals but try to dress them up as use of force in self-defence does not mean that they are lawful. On the other hand, when, as I mention in the article, the international community condones actions that look very much like armed reprisals, or at the very least refrains from condemning them, one might ask why the particular actions were not regarded as illegitimate uses of force. The answer to this question is highly relevant in determining the legitimate ends of force used in self-defence. Do these ends include deterrence against further attacks by the same party? Or possibly harming the military potential of that party so as to prevent further acts at some undefined time in the non-immediate future?
Thomas poses the question of whether when choosing between different means of achieving a legitimate end, the party involved must consider less harmful means even if they will not be as effective in achieving those ends. This question has cropped up in Israeli jurisprudence in which the three-pronged Verhältnismäßigkeit test has become the standard for examining the legality of security measures. In one case the Israel Supreme Court intimated that the question raised by Thomas may be tackled in two different ways. The first would be to say that if the reduction in effectiveness by choosing the less harmful means is not appreciable, the more harmful means could not be regarded as necessary. The other would be to examine this question under the “narrow proportionality” test: does the more drastic harm outweigh the marginal benefit to be achieved by employing the more harmful means? This seems a reasonable approach to the question.
Thomas’ perceptive distinction between epistemic and normative uncertainty in resolving questions of “narrow proportionality” emphasizes the inherent difficulties involved in applying the test in real live cases. In stating that courts and other decision-making bodies do not seem equipped to decide between conflicting perspectives I did not intend to say that they should not make an attempt to do so. But we should not be surprised if they try to avoid having to tackle the issue by dealing with necessity rather than balancing the competing interests. As I noted in the article most of the academic commentators who discussed Israel’s 2006 campaign in Lebanon also preferred to push the narrow proportionality question aside when addressing the issue of jus ad bellum. And the one commentator who did discuss the question only discussed one side of the balance, namely the harm caused. Given this commentator’s narrow view of the legitimate aims of force he did not examine the security benefits that were sought when causing this harm. As I noted in the article, it does not seem to me that one can say that harm was disproportionate unless one says to what it was disproportionate.
Finally Thomas’ point about theoretical situations in which a state that has been subject to an armed attack may not be able to use force in self-defence, because any effective use of force by it would be disproportionate. I would think that given the wide range of attacks that may meet the definition of an armed attack, such a situation is indeed conceivable. Thomas cites the Nuclear Weapons Advisory Opinion in which the ICJ mentioned that use of nuclear weapons may possibly be lawful “in an extreme situation of self-defence, in which its very survival would be a stake.” As Thomas remarks, this could certainly be regarded as relating to jus ad bellum. The conclusion would seem to be that in assessing use of force in self-defence we must ask in which way the state’s security was threatened by the attack. The answer to this question could be highly relevant in discussing whether its use of force was proportionate. There may conceivably be situations in which an isolated armed attack does not really threaten serious security interests of the victim state, and that given the inevitable harm that would be caused by any forcible reaction, any such reaction would be disproportionate. It is more difficult to conceive of a state being prevented from any use of force in self-defence when its territory has been invaded or it is the victim of a massive armed attack.
All the commentators mention issues which I failed to discuss in the article. I certainly did not deal with all aspects of the use of force in self-defence or its proportionality. Many isssues remain open for further discussion and research, which I hope that both the article itself and the present debate will stimulate. Here I am reminded of a saying attributed to the second century sage Rabbi Tarfon: “It is not thy duty to complete the work, but neither are thou free to desist from it.”(Aboth: Sayings of the Fathers, 2, 21).