¹ Gardam treats this matter as an ‘area of uncertainty’: Gardam, Judith, Necessity, Proportionality and the Use of Force by States (CUP 2004) 115CrossRefGoogle Scholar. Shany suggests that state practice indicates that risk-avoidance by armies is acceptable, even when it infringes some humanitarian concerns: Shany, Yuval, The Principle of Proportionality under International Law (Israel Democracy Institute 2009) 63–64Google Scholar and accompanying notes (in Hebrew). Rogers, Anthony PV, ‘Zero-casualty Warfare’ (2000) 82 International Review of the Red Cross 165, 178Google Scholar, quotes the previous British Ministry of Defence, British Defence Doctrine (JWP 0-01) (1996) which stipulated that ‘there may be occasions when a commander will have to accept a higher level of risk to his own forces in order to avoid or reduce the collateral damage to the enemy's civil population’. Interestingly, the 2004 manual opts for a different stipulation: ‘Sometimes a method of attack that would minimise the risk to civilians may involve increased risk to the attacking forces. The law is not clear as to the degree of risk that the attacker must accept. The proportionality principle does not itself require the attacker to accept increased risk. Rather, it requires him to refrain from attacks that may be expected to cause excessive collateral damage. It will be a question of fact whether alternative, practically possible methods of attack would reduce the collateral risks. If they would, the attacker may have to accept the increased risk as being the only way of pursuing an attack in a proportionate way’: The Joint Service Manual of the Law of Armed Conflict (Joint Service Publication 383, 2004), para 2.7.1. cf Final Report to the Prosecutor by the Committee Established to Review the NATO Bombing Campaign Against the Federal Republic of Yugoslavia’ (2000) 39 International Legal Materials 1257, para 21: ‘[a military commander] is entitled to take account of factors such as … risks to his own forces’ (making reference to Rogers, ibid).

² Protocol Additional to the Geneva Conventions of 12 August 1949 and relating to the Protection of Victims of International Armed Conflicts (entered into force 7 December 1978) 1125 UNTS 3 (‘Additional Protocol I’ or ‘AP I’); Sandoz, Yves, Swinarski, Christophe and Zimmermann, Bruno (eds), Commentary on the Additional Protocols of 8 June 1977 to the Geneva Conventions of 12 August 1949 (International Committee of the Red Cross, Martinus Nijhoff 1987) (‘ICRC Commentary’) 681–82Google Scholar.

³ AP I, art 57(2)(b). But cf the Rome Statute of the International Criminal Court (entered into force 1 July 2002) 2187 UNTS 90 (‘ICC Statute’) art 8, which uses a different formula (‘clearly excessive in relation to the concrete and direct overall military advantage anticipated’). The phrase ‘overall military advantage’ seems broader than AP I, though arguably still falls short of referring to the whole conflict: Dörmann, Knut, Elements of War Crimes under the Rome Statute of the International Criminal Court: Sources and Commentary (International Committee of the Red Cross, CUP 2003) 170–71CrossRefGoogle Scholar. The ICC formula also sets a ‘clearly excessive’ standard which, according to Fletcher, should lead to conviction only in cases of extreme disproportion: Fletcher, George, ‘Reclaiming Fundamental Principles of Criminal Law in the Darfur Case’ (2005) 3 Journal of International Criminal Justice 539, 561CrossRefGoogle Scholar.

⁴ AP I, art 52(2) defines military objectives as ‘those objects which by their nature, location, purpose or use make an effective contribution to military action and whose total or partial destruction, capture or neutralization, in the circumstances ruling at the time, offers a definite military advantage’.

⁵ This article does not enter into the debate over the implications of placing military objectives in the vicinity of civilian installations or residential areas, which revolves around interpreting AP I, art 51(7) (‘The presence or movements of the civilian population or individual civilians shall not be used to render certain points or areas immune from military operations, in particular in attempts to shield military objectives from attacks or to shield, favour or impede military operations’) and Convention (IV) relative to the Protection of Civilian Persons in Time of War (entered into force 21 October 1950) 75 UNTS 287, art 28 (‘The presence of a protected person may not be used to render certain points or areas immune from military operations’). See, for example, Dinstein, Yoram, The Conduct of Hostilities under the Law of International Armed Conflict (CUP 2004) 129–30CrossRefGoogle Scholar.

⁶ AP I, art 51(2) enunciates that ‘[t]he civilian population as such, as well as individual civilians, shall not be the object of attack’. Article 52(1) similarly prohibits attacks against civilian objects. Notably, the principle of proportionality, discussed below, is a testament to the distinction that IHL makes between targeting civilians intentionally and causing civilian casualties inadvertently.

⁷ AP I, art 51(4) and (5). It is assumed that all of the requirements regarding indiscriminate attacks are satisfied, save for art 51(5)(b), which can be interpreted as setting a proportionality test (‘excessive’) as part of the definition of an indiscriminate attack. Notably, IHL does not draw a distinction between the level of protection which should be afforded to enemy civilians and to civilians of a third party; this article nonetheless primarily uses the term ‘enemy civilians’ in juxtaposition with ‘compatriot soldiers’.

⁸ ‘Expected’ should be read as referring to the ‘expected value’ in probability theory terms. The ‘expected number of deaths’ can be explained intuitively as the average number of deaths had the same operation been carried out a very large number of times under the exact same circumstances.

⁹ The expected loss of civilian lives may reflect markedly divergent maximal and minimal poles: for instance, two military operations may be expected to cause 10 civilian deaths; the first operation has a 10 per cent probability of causing 100 deaths and a 90 per cent probability of causing zero deaths, whereas the second operation has a 50 per cent probability of causing 15 deaths and a 50 per cent probability of causing 5 deaths. Distinguishing between the two scenarios may be normatively significant; nonetheless, it has no direct bearing on the claims examined in this article, since the normative dilemma may equally arise whether or not the military operations pose risks to the attacker's soldiers.

¹⁰ See the analysis of the different proportionality tests in text at nn 19–24.

¹¹ It is assumed that, if the loss of some compatriot soldiers' lives reduces the attacker's military advantage, then the reduced military advantage is still proportionate (not excessive) in relation to c2. Importantly, this assumption should be distinguished from the assessment below of whether the possible additional military advantage (the averted risk to compatriot soldiers) is proportionate to the consequent additional risk to enemy civilians.

¹² Walzer, Michael, Just and Unjust Wars: A Moral Argument with Historical Illustrations (Basic Books 1977)Google Scholar advocates retaining the distinction between jus ad bellum and jus in bello, whereas McMahan, Jeff, ‘The Ethics of Killing in War’ (2004) 114 Ethics 693CrossRefGoogle Scholar, argues that the distinction is unsustainable in his account of the ‘deep morality of war’. However, McMahan concedes that the ‘laws of war’ must diverge significantly from the ‘deep morality of war’: ibid 730. cf Sloane, Robert D, ‘The Cost of Conflation: Preserving the Dualism of the Jus ad Bellum and the Jus in Bello in the Contemporary Law of War’ (2009) 34 Yale Journal of International Law 47Google Scholar. Sloane asserts that, on several occasions, jus ad bellum determinations have had an influence on the application of jus in bello rules. For instance, when sentencing two leaders of the Civil Defence Forces (CDF), a warring party in Sierra Leone's civil war, the Special Tribunal for Sierra Leone mentioned, as a mitigating factor, CDF's alleged legitimate cause ‘to restore the democratically elected Government of President Kabbah’: ibid 48–50. Similarly, the international assessment of the propriety of NATO's high-altitude bombing of Serbia during its 1999 campaign – a method designed to avoid risk to NATO soldiers – was arguably affected by the ‘just cause’ of preventing the continuation of ethnic cleansing of Albanian Kosovars, a jus ad bellum matter.

¹³ This appears to be the prevailing view, primarily because permitting the overall effect of a single military operation undertaken during an armed conflict on the outcome of that conflict to be considered as part of the proportionality analysis arguably ‘collapses’ proportionality jus in bello into proportionality jus ad bellum. Similarly, if jus in bello proportionality assesses even partially the just causes of war, it cannot be assessed independently of jus ad bellum considerations, and especially of the moral importance of the conflict's causes. See Hurka, Thomas, ‘Proportionality in the Morality of War’ (2005) 33 Philosophy and Public Affairs, 34, 44CrossRefGoogle Scholar. See also Sloane (n 12) 50 and his references in fn 13.

¹⁴ The article refrains from navigating the contentious terrain of determining whether some provisions of AP I have crystallised into norms of customary international law. See, for example, the manual published by the International Committee of the Red Cross: Henckaerts, Jean-Marie and Doswald-Beck, Louise, Customary International Humanitarian Law, Vol I: Rules (CUP 2005)CrossRefGoogle Scholar.

¹⁵ AP I, art 57(2)(a)(ii).

¹⁶ AP I, art 57(2)(b).

¹⁷ The ICRC Commentary does not offer clear guidelines as to the relevant considerations entailed by ‘feasible precautions’, though it seems to support a cautious approach; it is noted that, in the negotiations, some delegations suggested that ‘feasible’ means ‘everything that was practicable or practically possible, taking into account all circumstances at the time of the attack, including all those relevant to the success of the military operations’. It is then opined that the latter criterion seems to be too broad since there might be reason to fear that, by invoking the success of military operations in general, one may end up neglecting the prescribed humanitarian obligations: ICRC Commentary (n 2) 681–82.

¹⁸ cf Margalit, Avishai and Walzer, Michael, ‘Israel: Civilians and Combatants’ (14 May 2009) The New York Review of Books 56(8) 21–22Google Scholar. While Margalit and Walzer argue that soldiers should assume risks in order to reduce risks to enemy civilians, they qualify this duty in cases of exposure to extreme personal risk, and argue that soldiers ‘don't have to take suicidal risk’.

¹⁹ For a detailed description of the proportionality sub-tests, see Regina v Oakes [1986] 1 SCR 103 [70] (Canada). Domestic courts have used proportionality in an IHL context: for example, HCJ 2056/04 Beit Sourik Village Council v Israel and Israeli Defence Forces Commander in the West Bank, 2004 PD 58(5) 807, ILDC 16 (IL 2004), 30 June 2004, Sup Ct (as High Court of Justice) (Israel) [40]–[41].

²⁰ AP I, art 57(3) stipulates that ‘[w]hen a choice is possible between several military objectives for obtaining a similar military advantage, the objective to be selected shall be that the attack on which may be expected to cause the least danger to civilian lives and to civilian objects’. This stipulation may literally refer to a choice between two separate objectives, the destruction of which is expected to result in a similar military advantage (eg two ammunition facilities in different locations). The ICRC Commentary, however, offers the following example by way of explanation: when attacking railways, an attacker should refrain from attacking railway stations, which are usually located in towns, and instead hit railway lines at crucial points, but away from inhabited areas: ICRC Commentary (n 2) 687. This example illustrates that the provision may be applied analogously to choices that are made between two methods of attacking the same objective, since the distinction between two separate objectives and two methods of attacking the same objective is sometimes blurry.

²¹ See AP I, arts 57(2)(a)(iii) and 57(2)(b).

²² The question whether proportionality ought to be assessed from a military commander's viewpoint or from that of an average (or ‘reasonable’) person is hotly debated; dissensus also persists as to whether a different threshold should be set for determining that an attack contravened IHL when an individual is tried for alleged war crimes. See, for example, ICC Statute (n 3) art 8(2)(b)(iv). However, resolving these matters has no direct bearing on this article's analysis.

²³ There is a conceivable scenario according to which loss of lives of enemy civilians itself reduces the military advantage that is gained by the destruction of the military objective: namely, that there is a ‘hearts and minds’ effect. However, this scenario does not pose a special difficulty to the model of this article since its ramification is that the additional military advantage gained by the aerial method is smaller – ie the averted risk to compatriot soldiers minus the military disadvantage of enemy civilian casualties (in terms of the ‘hearts and minds’ effect).

To isolate the question of the value of compatriot soldiers and for the sake of simplicity, the scenario devised by this article assumes there is no ‘hearts and minds’ effect. Nevertheless, as shown, the ‘hearts and minds’ effect can easily be reintegrated in the proportionality analysis.

²⁴ If force protection of itself cannot be included in the military advantage side of the proportionality calculus, the ramifications are that when the military objective is force protection (for example, when a unit of compatriot soldiers is about to be attacked by a remote controlled missile that is launched from within the enemy civilian population, and an attack aimed at intercepting the missile entails risking enemy civilians) the interception cannot be undertaken, and inaction (resulting in risk to compatriot soldiers) must be accepted as the only choice. In other words, since destroying the missile entails no military advantage other than force protection, the interception would be disproportionate (excessive).

This choice between inaction (resulting in risk to compatriot soldiers) and action (averting such risk, at the cost of enemy civilians) raises the same dilemma, and hence fits the analytic model with regard to the choice between the aerial method (resulting in lesser risk to compatriot soldiers and greater risk to enemy civilians) and the ground method (resulting in greater risk to compatriot soldiers and lesser risk to enemy civilians). In other words, choosing the aerial method means choosing to save compatriot soldiers' lives. Furthermore, the missile interception scenario is in fact a special case of the more general case exemplified by the aerial or ground method scenario: in this special case (ie the missile interception) inaction cannot be construed as the least harmful means, since it does not achieve any military end, and the analysis will inevitably funnel into the last proportionality stricto sensu sub-test. Therefore, the more general scenario is assumed, according to which, unlike in the missile interception scenario, the military advantage entailed by an attack (namely, the destruction of the storage facility) justifies it independently of force protection considerations. Hence, both the ground and the aerial methods are legal (proportionate), and the military commander is not required to consider inaction.

²⁵ Walzer (n 12) 156. Blum challenges the combatants/civilians status-based dichotomy, and calls for the introduction of (individualised) risk and least harmful means criteria for assessing the propriety of attacks directed against soldiers, so that not all soldiers will be ‘fair game’: Blum, Gabriella, ‘The Dispensable Lives of Soldiers’ (2010) 2 Journal of Legal Analysis 69, 71–74CrossRefGoogle Scholar.

²⁶ Walzer (n 12) 157.

²⁷ See text at n 18.

²⁸ Smith, Thomas, ‘Protecting Civilians … or Soldiers? Humanitarian Law and the Economy of Risk in Iraq’ (2008) 9 International Studies Perspectives 144, 146–7CrossRefGoogle Scholar. Notably, if one accepts that instrumental considerations are legitimate, Smith's view could lead to adopting Position B, discussed below. It seems, however, that Smith favours Position A.

²⁹ Regarding the third reason, see Reisman, W Michael, ‘The Lessons of Qana’ (1997) 22 Yale Journal of International Law 381, 395–97Google Scholar, who argues that democratic societies involved in elective combat operations (such as interventions in foreign countries to shield indigenous populations from massacre) will have little tolerance for high numbers of casualties among their own soldiers. The argument can be applied mutatis mutandis to ongoing armed conflicts when public opinion may not be willing to accept a high casualty toll.

³⁰ See text at n 21.

³¹ Notably, however, Kasher's article assumes a state fighting ‘terrorism’ emanating from a territory outside its control. It is unclear whether Kasher intended that the article's rationale would be applied in any armed conflict: Kasher, Asa, ‘The Principle of Distinction’ (2007) 6 Journal of Military Ethics 152, 166CrossRefGoogle Scholar.

³² Benvenisti, Eyal, ‘Human Dignity in Combat: The Duty to Spare Enemy Civilians’ (2006) 39 Israel Law Review 81, 90CrossRefGoogle Scholar.

³³ ibid 108. Benvenisti notes that this conclusion does not preclude the possibility that compatriot soldiers have a moral duty to consider taking some risks to reduce harm to enemy civilians: ibid 90.

³⁴ ibid 83–84. Benvenisti argues that regardless of whether international human rights law should be directly applicable in armed conflicts, the principles that inspire the law on human rights should have bearing on the interpretation and application of IHL. Benvenisti refers to the Universal Declaration of Human Rights, which recognises ‘the inherent dignity and … equal and inalienable rights of all members of the human family’: Preamble to the Universal Declaration of Human Rights, UNGA Res 217 A(III), 10 December 1948, as well as the Preamble to the International Covenant on Civil and Political Rights stating that the rights in the covenant ‘derive from the inherent dignity of the human person’ (entered into force 16 March 1976) 999 UNTS 171: Benvenisti (n 32) 86.

³⁵ For example, the prohibition on causing unnecessary suffering to enemy soldiers, enunciated in AP I, art 35(2), and in Convention (IV) respecting the Laws and Customs of War on Land and its annex: Regulations concerning the Laws and Customs of War on Land (entered into force 26 January 1910) art 23(e). See also Common Article 3 of the Geneva Conventions (as well as AP I, art 75) proclaiming with regard to persons who are in the power of a party to the conflict that, as a minimum, ‘the following acts are and shall remain prohibited at any time and in any place whatsoever … outrages upon personal dignity’. See also Gardam (n 1) ch 3. Blum, Gabriella, ‘The Laws of War and the “Lesser Evil”’ (2010) 35 Yale Journal of International Law 1, 57Google Scholar quotes the canonical 1899 Martens clause, and suggests that general principles of military necessity and humanity mandate that some respect for the well-being of soldiers, even when they are actively engaged in the war effort, is warranted. Blum further posits that IHL should be conceived as treating the death of soldiers as a ‘lesser evil’ rather than a goal to be promoted.

³⁶ Dworkin, Ronald, Taking Rights Seriously (Harvard University Press 1977) 24, 26–27Google Scholar. Similarly, Alexy distinguishes between rules and principles, suggesting that rules are definitive, non-balanceable norms, whereas principles are competing optimisation requirements which should be realised to the ‘greatest possible extent’, applying proportionality: Alexy, Robert, A Theory of Constitutional Rights (OUP 2002) 47Google Scholar.

³⁷ Notably, this duty cannot be derived from IHL; rather, it can be considered as part of a ‘social contract’ between states and soldiers fighting on their behalf. Similarly, Kasher argues that states have a moral obligation to their soldiers qua citizens to justify their military activities: Kasher (n 31) 155.

³⁸ For descriptions of three major justifications for the IHL principle of distinction, see Blum (n 25) 86–92.

³⁹ cf Blum who argues that, for a country at war, protecting its soldiers is as strong (and sometimes even immediately stronger) an interest as protecting its civilians, since soldiers are the government's war machine and their success will determine the fate of the government and the country: Blum (n 35) 59.

⁴⁰ A consequentialist account seems to be more compatible with IHL provisions concerning unintentional incidental damage. cf Blum, ibid 41–42, arguing that under a deontological paradigm it is hard to account even for some of the absolute IHL prohibitions.

⁴¹ In a semi-formalised way, this argument may be presented as follows:

⁴² Ironically, perhaps, while Position A seems, at first glance, to be more restrictive from the attacker's point of view, a logically consistent application thereof suggests otherwise: since, according to Position A, the value of compatriot soldiers cannot be taken into account, the military advantage of the ground method for the purpose of the proportionality analysis is greater than its actual military advantage. Accordingly, in the context of proportionality stricto sensu, it justifies greater risk to enemy civilians than would be justified by the actual military advantage that it achieves.

⁴³ Kamm, Frances, ‘Failures of Just War Theory: Terror, Harm, and Justice’ (2004) 114 Ethics 650, 674;CrossRefGoogle Scholar Hurka (n 13) 61. See also the discussion in Lefkowitz, David, ‘Partiality and Weighing Harm to Non-Combatants’ (2009) 6 6 Journal of Moral Philosophy 298CrossRefGoogle Scholar.

⁴⁴ cf Hurka (n 13) 62.

⁴⁵ cf Moore, Michael S, Act and Crime: The Philosophy of Action and its Implications for Criminal Law (OUP 2010) 34CrossRefGoogle Scholar; Moore, Michael S, Placing Blame: A Theory of Criminal Law (Clarendon Press 1997)Google Scholar 270. It may be contended that the use of domestic analogies as a ground for moral principles when interpreting IHL provisions is problematic; however, the partiality argument and the offsetting counter-argument, above, are equally susceptible to a disanalogy objection.

⁴⁶ It is nevertheless acknowledged that the moral relevance of this fact is closely related to the question of the moral relevance of the distinction between action and omission. Consequentialist theories view normative evaluations as dependent consequences. As such, at least act-consequentialism would not hold the distinction between acts and omissions to be morally significant, cf Ben, Piers, Ethics (UCL Press 1998) 74Google Scholar.

⁴⁷ Christopher, rejecting Position D, posits that it is the nature of soldiers to take risks, and so the risk to the lives of soldiers should not be weighed equally against the risk to the lives of civilians: Christopher, Paul, The Ethics of War and Peace: An Introduction to Legal and Moral Issues (Prentice Hall 1994) 165Google Scholar.

⁴⁸ In a semi-formalised way, this argument may be presented as follows:

⁴⁹ cf Kasher (n 31) 166.

⁵⁰ ibid.

⁵¹ Hurka (n 13) 66. Nonetheless, Hurka's position is qualified: he asserts that if, according to the politicians' assessment, declining public opinion is likely to result in defeat, choosing to reduce risk to compatriot soldiers may be proportionate. Nevertheless, the line which Hurka draws seems to be quite vague; the better view is that considerations which cannot be directly related to the operation's result should not form part of the equation.

⁵² Hurka (n 13) 64, who posits that they are of ‘approximately equal weight’.