Israel Law Review

Articles

The Principle of Proportionality: ‘Force Protection’ as a Military Advantage

Robin Geissa1

a1 University of Potsdam (email: robin.geiss@uni-potsdam.de).

Abstract

‘Force protection’ is a primary concern of every military commander. Undoubtedly, it is an important and legitimate factor in the planning of every attack. However, when it comes to the humanitarian proportionality principle there is considerable controversy over the question to what extent ‘force protection’ can be factored into the humanitarian proportionality calculus as a relevant military advantage to be weighed against expected civilian casualties, injuries and damage. This question is pursued in this article.

(Online publication March 05 2012)

Keywords

  • principle of proportionality;
  • military advantage;
  • risks for soldiers

Footnotes

Warm thanks are due to Cordula Droege, Nils Melzer, Yaël Ronen, Henning Lahmann and the anonymous reviewers for their critique and helpful comments.

1. Introduction

Minimising risks for one's own forces has always been an essential component of war fighting. From the bow and arrow via gunpowder to today's drones, moving soldiers out of the danger zone has always preoccupied military strategists. Autonomous combat robots will perhaps one day mark the epitome of this evolution.1 Proponents of autonomous weapon systems argue that, precisely because the life of one's own soldiers will no longer be at stake when machines do the fighting, robots could be programmed to behave more carefully in zones of conflict than humans ever could.2 But, for the time being, humans remain directly involved in the conduct of hostilities, and risk and casualty aversion remain a predominant concern for any military commander. Asymmetric conflicts, indistinguishable enemies and combat operations in unclear urban settings are bringing this issue to the fore.3 The ‘fog of battle’ is particularly thick in these situations. There are high risks for both soldiers and civilians. Military commanders are regularly faced with the question of how to minimise risks for their forces as well as for the civilian population.

The problem is that these two factors often seem to be diametrically opposed. Minimising risks for one's own forces may imply increased risks for the civilian population and vice versa. High altitude bombing, for example, is relatively risk free for the pilots and fighter planes, but the imprecision that usually accompanies high altitude operations is often at the cost of the civilian population. Conversely, targeted ground operations could be carried out with greater precision and less impact on the civilian population, but they entail heightened risks for the soldiers who are fighting their enemies at close range.

This raises the question of how far international humanitarian law (IHL) allows for the consideration of ‘force protection’. The question has different facets. Are soldiers under a general obligation to accept risks for themselves in order to prevent harm to the civilian population?4 Are probable casualties among one's own forces a relevant factor in determining whether an alternative operation would bring about a ‘similar military advantage’ in the sense of Article 57(3) of Additional Protocol I?5 And how does ‘force protection’ relate to the obligation to take ‘all feasible precautions … with a view to … minimizing, incidental loss of civilian life’ entailed in Article 57(2)(a)(ii) of Additional Protocol I?6 Yet, arguably, the most important and certainly the most controversial aspect is to what extent ‘force protection’ may be factored into the proportionality calculus as a relevant military advantage to be weighed against expected civilian casualties, injury and damage. This particular question will be the focus of this article. The question is often approached from a moral or ethical perspective.7 This article, in contrast, pursues a doctrinal approach. It aims to analyse possible interpretations and interpretative limitations of the humanitarian proportionality principle in relation to the issue of ‘force protection’.

2. The Spectrum of Opinion

The question to what extent ‘force protection’ amounts to a ‘military advantage’ that may be taken into consideration when assessing the proportionality of an attack is surrounded by considerable complexities; views on the matter are split and are not always clearly expressed. In military practice the standard stipulation on ‘force protection’ is something like this: In taking care to prevent harm to civilians, combatants are required to ‘assume additional risk to minimize potential harm’.8 According to the British Defence Doctrine ‘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 collateral damage to the enemy's civil population’.9 That sounds right. But what precisely is meant by ‘assum[ing] additional risk to minimize potential harm’ or ‘occasions when a commander will have to accept a higher level of risk to his own forces’ remains somewhat elusive.

For a significant number of authors it appears to be self-evident that ‘force protection’ amounts to a proportionality-relevant military advantage.10 According to Dinstein, for example, ‘force protection is a valid concern that can be factored in any attack’.11 Schmitt explains that ‘an attack in which the personnel or equipment are lost is self-evidently not as advantageous as one in which they survive to fight again’.12 In fact, there are a number of writers who would agree that ‘the correct statement of the law is that the security of the attacking force is a factor when determining what would be proportional collateral damage’,13 and that ‘in an effort to improve security of the attacking force, increased collateral damage is permissible’.14

But there are also emphatic voices to the contrary.15 According to some writers, ‘force protection’ should be excluded from the proportionality assessment altogether. They fear that the consideration of ‘force protection’ as a relevant military advantage would blur and decisively undermine the proportionality analysis.16 Fenrick writes: ‘Military casualties incurred by the attacking side are not a part of the [proportionality] equation’.17 Oeter acknowledges that ‘force protection’ amounts to a military advantage but cautions that ‘[t]his, however, blurs decisively proportionality. Minimizing your own losses then justifies more and more collateral damages under civilians – a clear case of a sliding scale’.18 This concern seems to be shared by Solis who argues19 that

[f]orce protection is not a concrete and direct military advantage that allows proportionality to be disregarded or slighted. Were it otherwise, an attacker with superior arms would be free to annihilate all opposition with overwhelming firepower and call all civilian casualties collateral.

An expert meeting held in Geneva in 2005 also concluded that ‘the protection of one's own forces must never be conducted at the cost of the civilian population’.20 It is not fully clear whether these latter two statements were indeed meant to exclude per se any consideration of ‘force protection’ as a proportionality-relevant military advantage or whether they simply serve as reminders of the undisputed fact that ‘force protection’ is not an overriding concern that could set aside all other considerations when assessing the proportionality of an attack.

Be that as it may, the above-cited statements show that a number of authors are concerned about including ‘force protection’ in the proportionality calculus and that opinions expressed in the literature diverge considerably. In this regard it is telling that in 2000 the Final Report to the Prosecutor by the Committee established to review the NATO bombing campaign against the Federal Republic of Yugoslavia declared that the question ‘[t]o what extent […] a military commander [is] obligated to expose his own forces to danger in order to limit civilian casualties or damage to civilian objects’ was unresolved.21 Eleven years later it seems that this is still the case.

3. ‘Force Protection’ – A Controversial Issue Already During the Negotiations of Additional Protocol I

Controversy over the issue of ‘force protection’ is not new. Already in the 1970s, during the negotiations that led to the adoption of Additional Protocol I, state representatives were aware of the risk that ‘force protection’ could potentially set aside the protection of the civilian population. Although the negotiating history of Articles 51 and 57 of Additional Protocol I is silent on the matter, during the drafting process the notion of ‘force protection’ had seeped into the text of what was later to become Article 52(3) of Additional Protocol I. The draft text (of then Article 47 of Additional Protocol I) read: ‘In case of doubt whether an object which is normally dedicated to civilian purposes, such as a house or other dwelling or a school, is being used to make an effective contribution to military action, it shall be presumed not to be so used [except in contact zones where the security of the armed forces requires a derogation from this presumption]’.22 By the time the Working Group submitted the article for official voting, the final phrase of paragraph 3 had disappeared. According to the US delegate, the bracketed text had given rise to differences of opinion in the Working Group.23 Apparently, ‘certain delegations had feared that that provision might authorize infantry in the front line to attack civilian objects located in a zone of military operations, while others had considered such an exception necessary for the safety of the soldiers’.24

The deletion of the bracketed text in the final version of today's Article 52(3) of Additional Protocol I arguably allows for some tentative conclusions. First, the use of the wording ‘security of the armed forces’ as it was envisaged in the bracketed text – in a provision that also used (and still uses) the wording ‘military advantage’ (see today's Article 52(1) of Additional Protocol I) – shows that at least the drafters of Additional Protocol I considered these notions to be distinct. But, admittedly, this is at best a possibility and by no means a compelling conclusion. After all, the bracketed text was ultimately not adopted. Secondly, the deletion of the bracketed text indicates serious concern that the ‘security of the armed forces’ could all too easily be invoked to set aside civilian protection. It seems safe to say that ultimately the concern over the protection of the civilian population prevailed. Thus, today's Article 52(3), as well as its ‘pendant’ Article 50(1) second sentence of Additional Protocol I, show that at least in certain instances the law requires soldiers to accept certain risks for the sake of the protection of the civilian population. The presumption of civilian status in case of doubt about a person's status (Article 50(1) of Additional Protocol I) or an object's status (Article 52(3) of Additional Protocol I) is potentially dangerous for the soldiers involved.25

Do these provisions thus allow the deduction of a general obligation on soldiers to accept higher risks for themselves in order to avoid harming the (enemy) civilian population, with the consequence that ‘force protection’ would have to be excluded from the proportionality analysis from the outset? The answer is no. Both of these provisions have remained controversial and they have not (yet) acquired the status of customary international law.26 Upon ratification of Additional Protocol I, the United Kingdom as well as France explained that the presumption of civilian status does not override a commander's duty to protect the safety of troops under his command or to preserve their military situation, in conformity with other provisions of Additional Protocol I.27 Moreover, even if these provisions could be accorded customary law status – quid non – one could not necessarily draw the general conclusion from these specific provisions that in all possible contexts soldiers would have to accept higher risks for themselves vis-à-vis civilians. Indeed, one could also draw the exact opposite conclusion that precisely because the acceptance of higher risks for soldiers is inherent in only a few (non-customary) specific provisions, there is no such obligation under the general law. Against this background and in view of their specific function, Articles 52(3) and 50(1) of Additional Protocol I cannot be invoked to construe a general obligation on soldiers to accept higher risks vis-à-vis civilians.

4. ‘Force Protection’ as a Concrete and Direct Military Advantage?

In turning to the controversial question as to what extent ‘force protection’ amounts to a relevant military advantage when assessing the proportionality of an attack, it seems helpful to distinguish the following questions: (1) Does the protection of one's own forces qualify as a military advantage at all? (2) If so, under what circumstances is it considered to be sufficiently direct and concrete to be factored into the proportionality analysis?

4.1 ‘Force Protection’ as a Military Advantage

Unequivocal indications of what constitutes a ‘military advantage’ in the legal sense and the factors that determine its relative value are largely absent.28 State practice, however, supports a relatively broad interpretation of the notion of the ‘military advantage’ that includes also the protection of one's own forces. There is wide agreement that the notion of the military advantage comprises all sorts of tactical gains and military considerations, and that different advantages that need not necessarily derive from the destruction of the specific object under attack may be considered cumulatively.29 The historical example typically provided to support such an extended interpretation is the Allied landing in Normandy during the Second World War.30 Prior to this event the Allies had carried out extensive air raids in other areas of France in order to mislead the enemy as to the actual location of their intended landing.31 As part of a broader ruse strategy the military advantage did not derive from the military objectives under attack during these air raids, but from the purely tactical advantage of having confused the enemy.32 Moreover, the United States Department of Defense has stated that ‘“[m]ilitary advantage” is not restricted to tactical gains, but is linked to the full context of a war strategy’.33 Similarly, the commentary on the recently released HPCR Manual on International Law Applicable to Air and Missile Warfare emphasises that ‘[t]he term “military advantage” must not be too narrowly construed, for instance by restriction to ground gained or weakening the enemy armed forces’.34 This being said, there is also agreement that only such advantages that are military in nature may be considered within the proportionality assessment. Other advantages, namely those of a political or moral nature, must be excluded from the proportionality calculus.35 The HPCR Manual stipulates the general view on the matter, namely that ‘[m]ilitary advantage does not refer to an advantage which is solely political, psychological, economic, financial, social, or moral in nature’.36

These stipulations confirm an accumulative benefits approach that looks beyond the actual object under attack and allows consideration also of other (tactical) military advantages.37 Statements on the matter, however, are not always fully conclusive with regard to the question whether ‘disadvantages avoided’ (that is one's own forces that have been protected from harm) could also be subsumed under the notion of the military advantage.38 Still, there is state practice in support of such a view. Three states – Australia, Canada, and New Zealand – upon ratification of Additional Protocol I made explicit declarations to the effect that they interpret the term ‘military advantage’ to include the security of the attacking forces.39 Whilst it is not a party to Additional Protocol I, the US has included a similar stipulation in The Commander's Handbook on the Law of Naval Operations, which provides that ‘[m]ilitary advantage may involve a variety of considerations, including the security of the attacking force’.40 These stipulations appear to be in line with the modern-day formulation of the principle of military necessity according to which military necessity permits ‘that degree and kind of force, not otherwise prohibited by the law of armed conflict, that is required in order to achieve the legitimate purpose of the conflict, namely the complete or partial submission of the enemy at the earliest possible moment with the minimum expenditure of life and resources’.41 Military necessity thus is understood to justify not only ‘what is required to win the war, but also what reduces the risks of losses or costs of the war’.42

From the above it follows that states apply an accumulative benefits doctrine when assessing the military advantage of an attack.43 In addition to the military value of the specific target under attack, other considerations – as long as they are military and not purely political – are also taken into account. This is coherent with the interpretation by states that the military advantage is determined not only by insular aspects of an attack but by the attack considered as a whole.44 State practice is arguably somewhat less clear with regard to the question whether the accumulative benefits doctrine allows consideration of the military disadvantages avoided in addition to the advantages gained from an attack. As far as can be seen, no state has ever excluded such an interpretation of the military advantage notion, whereas a number of states have made explicit declarations in support of such an extended understanding of the term.45 As long as the latter is the prevailing standard – and in the contemporary practice of states it seems safe to say that it is – ‘force protection’ qualifies as a relevant tactical gain to be considered as a military advantage within the proportionality equation.

Thus, generally speaking, nothing precludes ‘force protection’ from being considered as a military advantage when conducting a proportionality assessment prior to an attack. However, whether in a given scenario ‘force protection’ can be considered in a particular proportionality analysis depends on whether the anticipated ‘force protection’ is sufficiently ‘concrete and direct’ within the meaning of Articles 51(5)(b) and 57(2)(a)(iii) of Additional Protocol I. Thus, the actual issue with ‘force protection’ is not whether it can be considered as a military advantage – for clearly it can – but when it may be factored into the proportionality assessment as a military advantage that is sufficiently ‘concrete and direct’.

4.2 When is the Military Advantage Anticipated from the Protection of One's Own Forces Sufficiently ‘Concrete and Direct’?

A general argument against the consideration of ‘force protection’ as a relevant military advantage could be that ‘force protection’ is simply too abstract to be considered as a ‘concrete’ and ‘direct’ military advantage within the meaning of Articles 51(5)(b) and 57(2)(a)(ii) of Additional Protocol I. Clearly these adjectives were intended to limit the range of military advantages that may lawfully be weighed against expected collateral damage. But there has always been a debate as to the precise scope of these qualifiers, and on where to draw the line between ‘concrete and direct’ military advantages on the one hand and ‘abstract and indirect’ military advantages on the other.

The 1923 Hague Rules of Air Warfare – the first codified reference to the military advantage notion – spoke of ‘an obvious military advantage’.46 A 1969 Resolution issued by the Institut de Droit International spoke of ‘a substantial, specific and immediate military advantage’.47 And the ICRC commentary to Additional Protocol I states:48

The expression ‘concrete and direct’ was intended to show that the advantage concerned should be substantial and relatively close, and that advantages which are hardly perceptible and those which would only appear in the long term should be disregarded.

But it has rightly been questioned whether the word ‘substantial’ as it appears in the ICRC commentary may be seen as a synonym for the word ‘concrete’.49 Nothing in the law requires that the military advantage anticipated be ‘substantial’ or ‘significant’. Moreover, it has been pointed out that even long-term effects may be considered to be ‘direct and concrete’.50 Article 8(2)(b)(iv) of the Rome Statute of the International Criminal Court (‘ICC Statute’) explicitly mentions the ‘overall’ military advantage anticipated. Footnote 36 to the Elements of War Crimes, which accompany Article 8, explains51 that

[t]he expression ‘concrete and direct overall military advantage’ refers to a military advantage that is foreseeable by the perpetrator at the relevant time. Such advantage may or may not be temporally or geographically related to the object of the attack.

Of course, Elements of War Crimes in the realm of international criminal law do not automatically influence the interpretation of IHL in general.52 However, almost all NATO member states – including the United States and also states such as New Zealand, Australia, and Nigeria53 – when ratifying respectively signing Additional Protocol I, made almost identical declarations54 according to which

[i]n applying the rule of proportionality in Article 51 and Article 57 [of Additional Protocol I], ‘military advantage’ is understood to refer to the advantage anticipated from the attack considered as a whole and not only from isolated or particular parts of the attack.

From this it may be inferred that states perceive temporally or geographically distant military advantages as sufficiently concrete and direct.55 In a large-scale attack it is not the single components that need to be assessed; rather a large-scale attack may be considered and assessed as a single ‘complex mosaic’.56 However, if the adjectives ‘concrete and direct’ are not to be deprived of all meaning, at least such military advantages as are merely speculative or hypothetical must be excluded.57

Therefore, the remaining ambiguities notwithstanding, it seems possible to sketch out the following points of reference regarding the interpretation of the words ‘concrete and direct’. First, just like the overall aim of winning the war, the overall aim of protecting one's own forces would indeed be too abstract to be considered when assessing the anticipated military advantage of an attack.58 In such an abstract state it would simply trump all other considerations and therefore cannot be considered in the proportionality analysis.59 But that, of course, does not necessarily mean that destroying a concrete military target or, for that matter, protecting a specific military asset would automatically be excluded from the proportionality equation. In the case of the destruction of a tank that is just driving through a village without posing any acute risk to one's own forces, the ‘force protection’ that is gained – namely this tank will be excluded from attacking one's own forces in the future – is too indirect and abstract, and is therefore irrelevant. In such an abstract scenario it would be impossible to measure the relative military value of the ‘force protection’ gained, given that it is not (yet) clear which concrete forces or military assets the tank might have attacked in the future. Even the proponents of the most lenient interpretations of the limiting qualifiers ‘concrete and direct’ would seem to accept that such a hypothetical military advantage (that is preventing future attacks by that tank) must not be factored into the proportionality equation.60

Conversely, in the following scenario (Scenario 1) the military advantage gained from the protection of one's own forces would be sufficiently concrete and direct, and could thus be considered in the proportionality analysis.

Scenario 1 A is attacking the ground forces of B. Civilians are in the vicinity. B calls in aerial support in order to stop A's attack which would otherwise result in the destruction of B's ground forces. B's planes destroy the attacker A. As a result B's ground forces are saved.

In this scenario the destruction of A will prevent the imminent annihilation of the ground forces of B. The protection gained from the attack against A will be both temporarily and geographically ‘relatively close’. Thus, it would satisfy even the stricter interpretation of the wording ‘concrete and direct’ as it is to be found in the ICRC commentary cited above.61 Most importantly, in this constellation the relative value of the force protection gained is tangible and not only hypothetical. It carries the relative value that is to be accorded to the ground forces of B (comprising persons and military assets) that are saved. If they were to be defeated this would carry a specific military advantage for A; if they are protected from defeat this carries a corollary military advantage for B.

The fact that, in practice, no coherent standard for the measurement of such relative values exists does not counter the argument that ‘force protection’ – like other military considerations – may qualify as a ‘concrete and direct’ military advantage. It is an inherent deficiency of the humanitarian proportionality principle as a whole that there simply is no arithmetical precision to measure the relative value of military advantages – either of those that are derived from destroying a military target or of those derived from protecting a certain military asset. As Walzer has aptly observed, ‘political and military leaders [ought] to worry about costs and benefits. But they have to worry, they can't calculate’.62

4.2.1 ‘Force Protection’ – A Justification to Extend the Limits of What May be Considered ‘Concrete and Direct’?

A particular problem with the issue of force protection is that whenever it is at issue, self-preservation is at stake.63 Inherently there is thus a strong inclination to argue in favour of a heightened military advantage to be gained from protecting one's own forces, to invoke ‘just causes’ to this end, and generally to accord priority to the lives of one's own citizens and soldiers over those of the enemy. Time and again there have been attempts to invoke extrinsic considerations of ‘justness’ so as to shift the proportionality balance in favour of one's own forces.64 Kasher and Yadlin make a slightly different – albeit related – argument on the basis of a conception of priorities of duties of states.65 They argue66 that

[a]ccording to our norm of priorities on grounds of duties, the state should give priority to saving the life of a single citizen, even if the collateral damage caused in the course of protecting that citizen is much higher, which may seem unacceptable. […] Consideration of accumulative benefits will eliminate the apparent disproportionality.

This is a just cause argument that is used to justify the consideration of hypothetical military advantages.67 Taking into consideration ‘accumulative benefits’ in the manner that is suggested by the authors – namely by considering that the attacked person will be prevented from carrying out further (undetermined) attacks in the future68 – would mean to factor hypothetical and purely speculative military advantages into the proportionality calculus. This exceeds the limits of what may be considered ‘concrete and direct’ under the humanitarian law proportionality principle. Here the accepted ‘accumulative benefits doctrine’ is further developed into an ‘accumulative hypothetical benefits doctrine’. Of course, experience tells us that every tank or enemy soldier in an armed conflict will be used to carry out future attacks without any possibility to know what exactly the quality or quantity of these attacks will be. Thus, if the accumulative benefits doctrine applied also to such hypothetical events, every tank or soldier that is attacked in an armed conflict could ultimately be said to carry a very high military advantage given that, unless destroyed, it is likely to be used in the future to carry out further attacks. Consequently, with every attack – even against prima facie low-level military targets – one could ultimately justify considerable civilian casualties or, in the words of Kasher and Yadlin, ‘eliminate apparent disproportionality’.

To ‘eliminate apparent disproportionality’ by allowing hypothetical factors into the proportionality equation, however, is the slippery slope feared by those authors who deny that force protection could ever be considered as a relevant military advantage.69 An attack that to all observers would be grossly disproportionate could be rendered lawful just as the result of a mind game. One could endlessly add hypothetical military advantages to the proportionality equation until even an otherwise disproportionate attack is rendered proportionate. Ultimately, any number of civilian casualties and any amount of damage would be justifiable. Thus, if ‘apparent disproportionality’ could be eliminated, the standard of excessiveness would be eliminated too. This, however, is contrary to the law and it cannot be justified by relying on a conception of priorities of state duties.

A state's undisputed duty to protect its own citizens – civilians and soldiers alike – cannot be invoked to alter the fundamental structure of the humanitarian proportionality assessment. The proportionality principle is predicated on this duty. The principle of proportionality allows belligerents to subject their duty to respect enemy civilians to their conflicting duty to protect their own nationals in striving to overcome the enemy. Thus, to a certain degree the humanitarian proportionality principle already accords precedence to a state's duty to protect its own citizens and soldiers. But, in order to avoid unlimited war, the proportionality principle also stipulates a limit.70 The limit is excessiveness. Any reasoning that does away with this fundamental barrier of excessiveness – by according categorical precedence to certain military advantages, by allowing an infinite number of hypothetical military advantages into the equation, or by invoking extrinsic just causes – cannot be maintained under the law of war as it stands.

5. Protection of the Attacking Forces

‘Force protection’ amounts to a relevant military advantage in a scenario where the aim of the attack is to save one's own forces from an imminent attack by the enemy (Scenario 1).71 However, what remains rather controversial is the question of whether the protection of the attacking forces – that is the specific forces that are actually carrying out a given attack – may also be factored into the proportionality assessment of that particular attack.

If it is accepted that force protection can be subsumed under the notion of the military advantage – and state practice clearly supports this view – then it cannot be categorically excluded that in certain scenarios the protection of the attacking forces may also be factored into the proportionality calculus. For example, B is under attack from A; in fending off the attack, B in its counter-attack may take into account the military advantage of destroying A and of saving itself from A's attack. The scenario is similar to Scenario 1.72 As a result B's military advantage, consisting not only of A's destruction but also of B's protection, is increased. Of course, the military advantage is not endlessly extendable. The maximum relative value of the military advantage gained from B's protection is the value of the forces that are directly and concretely being protected against the threat posed by A. While this could indeed justify increased collateral damage, there would be no slippery slope leading towards unlimited civilian casualties and damage. As far as the application of the proportionality principle is concerned, this scenario does not pose any particular problems.

In practice, however, the protection of the attacking forces is often invoked in a different way and under different circumstances, namely when a choice has to be made between different methods of attack – for example, between the employment of ground forces or distance weapons, between low- and high-altitude aerial operations, and generally between heavier and less heavy forms of attack. The following scenario (Scenario 2) may help to illustrate the way in which the protection of the attacking forces is often invoked.

Scenario 2 A plans to attack B. It is evident that civilians and enemy fighters are in the vicinity of the target. A decides to rely exclusively on air power involving high-altitude attacks instead of ground forces, arguing that by doing so the risks for A's attacking forces are minimised, and that therefore high(er) numbers of civilian casualties can be justified.

5.1 Distinguishing ‘Force Protection’ and the Abstract ‘Protectiveness’ of the Means and Methods Used

‘Force protection’ must not be confused with the abstract protectiveness of the means and methods used to attack. The focus must always be on the forces that are actually and verifiably being protected. Only in that case it is possible to discern a tangible military advantage that is sufficiently concrete and direct for the purposes of the proportionality calculus. Of course, there can be no doubt that an aerial attack usually involves fewer risks for the attacker than a ground operation in which soldiers would have to fight the enemy at close range. The aerial attack is safer. But the problem is that ‘protection’ or ‘safety’ in and of themselves – that is without any fixed point of reference of what is actually and verifiably being protected (directly and concretely) – have no allocable military value. ‘Protection’ is relative. It can only be higher or lower in comparison to something else. A high-altitude attack, for example, is more protective than a low-altitude attack, and an even higher altitude attack could be argued to be even more protective, and so on. Ultimately, ever greater protection would grant an ever greater military advantage and could justify ever more civilian casualties, irrespective of whether any forces are actually being protected, directly and concretely. The proportionality assessment, however, requires fixed points of reference. Otherwise it would be impossible to carry out the balancing judgment that the proportionality principle requires. Thus, where ‘force protection’ is at issue, the focus must be on the forces that are in fact being protected, and the maximum value that may be considered when conducting a proportionality assessment is the relative military value of these forces.73

5.2 Hypothetical Points of Reference Must be Excluded When Assessing the Proportionality in a Specific Attack

As was mentioned earlier when setting out the various opinions regarding force protection, it is argued that ‘an attack in which the personnel or equipment are lost is self-evidently not as advantageous as one in which they survive to fight again’.74 This, of course, is indisputable. Yet, despite the ostensibly striking conspicuousness of this example it does not necessarily prove that the protection of the attacking forces generally amounts to a proportionality-relevant military advantage. Just because something is self-evidently ‘advantageous’ does not mean that it automatically amounts to a ‘concrete and direct’ military advantage relevant to the proportionality analysis. The problem is that in many instances – albeit not necessarily always – the advantages associated with the protection of the attacking forces are merely hypothetical: they are too abstract and indirect to be considered for the purposes of the proportionality assessment. This is precisely the problem in Scenario 2.

In Scenario 2 A opted for an aerial attack instead of sending in ground forces, arguing that this would decrease the risks for its soldiers. This risk assessment was correct. But that does not mean that the alleged advantage of the aerial attack could automatically be factored into the proportionality calculus. The problem is that the alleged advantage of the aerial attack becomes apparent only when it is compared to the hypothetical less secure ground operation that was never carried out. A argues that, by having chosen an aerial attack, it protected its forces against the threat that would have emerged if it had opted for an attack by ground forces. The protection gained was the protection of the soldiers and military assets that would have presumably been lost in a ground operation. The same logic is reflected in the above-cited statement which compares ‘an attack in which the personnel or equipment are lost’ to an attack ‘in which they survive to fight again’. The problem is that the ground operation remains entirely hypothetical; only the aerial attack was carried out. It thus remains unclear how many forces would have been involved in a ground operation, what the actual threat would have been and, most importantly, it remains indeterminable how many – if any – forces would have been lost or captured if the ‘less protective’ attack had in fact been carried out.

On this basis it could always be argued ex post facto that a given attack was chosen over a hypothetical, less secure attack and that, by having done so, a military advantage was gained. Thus, if an aerial attack was carried out at a certain altitude it could always be argued that it was chosen over an attack at a lower altitude. If an attack was carried out with 20 soldiers, a commander could always argue that an attack with 10 soldiers would have been less secure and so on. In all of these examples – and in fact in every attack – the military value of the hypothetically protected forces could be brought into the proportionality equation. Categorically, in every attack the military advantage would consist of (a) the advantage gained from the actual target's destruction and related tactical gains, as well as (b) the protection of those forces that would have been lost in a hypothetical less protective attack. Evidently, if the protection of the attacking forces could be invoked in this way, there would indeed be a slippery slope and proportionality could be manipulated one-sidedly.75 This is why it is widely agreed that hypothetical advantages must not be factored into the proportionality calculus. Therefore, A's approach in Scenario 2 of accumulatively invoking hypothetical disadvantages avoided cannot be maintained. A cannot justify higher numbers of civilian casualties for the sole reason that it has opted for a more secure aerial operation instead of a less secure ground operation.

Strictly speaking, the advantage that is claimed by A in Scenario 2 does not even derive from the aerial attack itself. It derives from the strategic decision not to attack with ground forces. The forces that would presumably have been lost in a ground operation are protected not because an aerial attack is opted for, but because the ground operation is not carried out. In and of itself the aerial attack does not save any of A's forces. They are not imminently threatened until the attack is commenced. There would only have been a threat for A's ground forces if A had carried out a ground operation. The decision to carry out an aerial attack rather than a ground operation, however, precedes the actual attack. If the advantage that derives from this decision could be factored into the proportionality assessment the attacker would have gained a military advantage prior to the attack simply by having chosen a ‘safer’ way to attack. Even if an attack is aborted, the attacker would have gained a military advantage by having chosen a high-altitude operation rather than a less advantageous low-altitude sortie or ground operation. However, an advantage that derives from a decision that precedes the actual attack is not an advantage that could be ‘anticipated’ from the attack as is required by Articles 51(5)(b) and 57(2)(a)(iii) of Additional Protocol I. Thus, while every military commander will, of course, have considered various methods of attacking prior to an attack, including less harmful alternatives, these considerations – while fully legitimate and common military practice – are not an adequate point of reference for the proportionality assessment of the concrete attack that is ultimately chosen. Therefore A, in Scenario 2, cannot invoke a heightened military advantage because of force protection. The military advantage anticipated from the chosen aerial attack has to be assessed in isolation from hypothetical alternative attacks.

6. Conclusion

‘Force protection’, like other military considerations, may generally be considered when determining the military advantage that is anticipated from an attack. This, however, does not mean that force protection categorically factors into the proportionality assessment of each and every attack. Only where one's forces are being protected directly and concretely – as in Scenario 1 above – can ‘force protection’ amount to a concrete and direct military advantage that may be factored into the proportionality calculus. In such instances, notwithstanding the impossibility of being able to measure military advantages with any degree of arithmetical precision, at least in theory it is possible to accord a certain military value to the fact that one's own forces (comprising a specific number of persons and military equipment) are protected. In contrast, in Scenario 2 the advantage gained and the corresponding disadvantage avoided are merely hypothetical. They emerge only by comparing a given attack with a hypothetical, less protective form of attack. In such a case, the protection that derives from having chosen a different form of attack is simply too vague and hypothetical to be factored into the proportionality calculus. In a strict sense the protection does not even derive from the actual attack but from a tactical decision that precedes the attack. While such tactical decisions are, of course, perfectly legitimate and common to the planning of every attack, they cannot be factored into the proportionality calculus.

There is no doubt that the proportionality principle suffers from various deficiencies, that it leaves a considerable margin of ambiguity, and that in certain grey zones it may not provide conclusive answers. But there is also no doubt that it must have some limiting effects on how attacks may be carried out. Any argument that eventually rules out the correlation of excessiveness altogether is contrary to the law. With the consideration of ‘force protection’ as a relevant military advantage this is not automatically the case. But given that ‘force protection’ inherently denotes the fundamental interest of self-preservation, the risk is particularly high that the fine distinctions between merely hypothetical and concrete military advantages are blurred, or that ‘force protection’ is simply being accorded a superior military value irrespective of the actual military importance of the forces that are being protected.

The problem is that whenever ‘force protection’ is at issue, self-preservation is at stake. Therefore, even the basic limitations imposed by the humanitarian proportionality principle may be perceived as overly rigid or counter-intuitive. The following example illustrates this dilemma.

An enemy sniper has been spotted. He is surrounded by a very large number of civilians. The soldier who spotted the enemy sniper has only heavy weaponry available. He decides that the injury to civilians as a result of an attack against the sniper would be grossly disproportionate and that the attack must not be carried out. Seconds later, however, he realises that the sniper is aiming at him. Leaving aside the urgency of the moment and possible criminal law exculpations, the question is whether the attack that was previously assessed as being grossly disproportionate and prohibited under IHL could now be viewed as proportionate and therefore legitimate.

The answer is no, on the assumption that in the scenario described the accumulative ‘military value’ of the ‘soldier's self-preservation’ and the sniper's neutralisation still does not outweigh the expected civilian casualties. Thus, in this scenario, given that the expected civilian casualties would be excessive, the proportionality principle bars the soldier from attacking the sniper. This would even hold true if the attack against the sniper was the only way to save the soldier's life. Therein lies the dilemma: attacks that cause excessive civilian casualties and damage must not be carried out even in a situation in which the attack would save one's own forces. This is a logical and indisputable consequence of the humanitarian proportionality principle. But it is counter-intuitive and it explains attempts to somehow increase the military advantage that is to be gained in such a scenario – either by allowing just causes to enter the equation or by allowing for the consideration of additional hypothetical military advantages – in order to render the attack legitimate.

The issue of ‘force protection’ may in many ways be seen as pars pro toto for the various problems, controversies, and ambiguities that surround the humanitarian proportionality principle in general, and the notion of the military advantage in particular.76 The split of opinions regarding the rather specific issue of ‘force protection’ shows that not only is there no arithmetical precision when comparing such unequal factors as military advantages and the loss of civilian lives and property, it is also not even agreed what precisely is to be put in each respective scale of the proportionality balance.77 It is often said that despite the well known deficiencies of the humanitarian proportionality principle, there simply is no viable alternative. This may be true. But certainly there are a number of aspects of the humanitarian proportionality principle that still could be worked out more precisely than they have been to date.

Notes

1 Beard, Jack M, ‘Law and War in the Virtual Era’ (2009) 103 American Journal of International Law 409, 420, 428–42, 444 [OpenURL Query Data]  [Google Scholar].

2 Ronald C Arkin, ‘Governing Lethal Behavior: Embedding Ethics in a Hybrid Deliberative/Reactive Robot Architecture’, GVU Technical Report GIT-GVU-07-11 (GVU Center, Georgia Institute of Technology, 2007), available at http://www.cc.gatech.edu/ai/robot-lab/online-publications/formalizationv35.pdf.

3 Geiss, Robin, ‘Asymmetric Conflict Structures’ (2006) 864 International Review of the Red Cross 757 [OpenURL Query Data]  [Google Scholar]; Schmitt, Michael N, ‘Asymmetrical Warfare and International Humanitarian Law’ in von Heinegg, Wolff Heintschel and Epping, Volker (eds), International Humanitarian Law Facing New Challenges (Springer 2007) 11 [OpenURL Query Data]  [Google Scholar].

4 See, for example, Benvenisti, Eyal, ‘Human Dignity in Combat: The Duty to Spare Enemy Civilians’ (2006) 39 Israel Law Review 81, 81, 82, 90 [OpenURL Query Data]  [Google Scholar].

5 Article 57(3) of 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’) provides that ‘when 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’ (emphasis added).

6 The debate regarding art 57(2)(a)(ii) of Additional Protocol I is mainly focused on the question whether the wording ‘feasible’ encompasses also military considerations. Such a reading is supported, for example, by the declaration made by the UK at the time of ratification (28 January 1998). The text of the declaration is available at http://www.icrc.org/ihl.nsf/NORM/0A9E03F0F2EE757CC1256402003FB6D2?OpenDocument. See also Protocol on Prohibitions or Restrictions on the Use of Mines, Booby-Traps and Other Devices (CCW Protocol II) (entered into force 2 December 1983) 1342 UNTS 137, art 3(4); Protocol on Prohibitions or Restrictions on the Use of Incendiary Weapons (CCW Protocol III) (entered into force 2 December 1983) 1342 UNTS 137, art 1(5); and Protocol on Prohibitions or Restrictions on the Use of Mines, Booby-Traps and Other Devices as amended on 3 May 1996 (‘Amended CCW Protocol II’) (entered into force 3 December 1998) 2048 UNTS 93, art 3(10), which define feasible precautions as ‘those precautions which are practicable or practically possible taking into account all circumstances ruling at the time, including humanitarian and military considerations’ (emphasis added). This is also the understanding adopted in the ICRC's Interpretive Guidance: see Melzer, Nils, Interpretive Guidance on the Notion of Direct Participation in Hostilities (ICRC 2009) 75 [OpenURL Query Data]  [Google Scholar].

7 Benvenisti (n 4) 81–109; Walzer, Michael, Just and Unjust Wars: A Moral Argument with Historical Illustrations (Basic Books 1977) 151 [OpenURL Query Data]  [Google Scholar].

8 US Counterinsurgency Field Manual, FM 3–24 MCWP 3-33.5 (Headquarters Department of the Army, December 2006) 162, note 7-30, available at http://www.fas.org/irp/doddir/army/fm3-24.pdf. See also Rogers, Anthony PV, ‘Conduct of Combat and Risks Run by the Civilian Population’ (1982) 21 Military Law & Law of War Review 293, 310 [OpenURL Query Data]  [Google Scholar].

9 British Defence Doctrine (JWP 0-01) issued by the British Minister of Defence, 1996, cited in Rogers, Anthony PV, ‘Zero-casualty Warfare’ (2000) 82 International Review of the Red Cross 165, 178 (note 41) [OpenURL Query Data]  [Google Scholar].

10 For Kasher and Yadlin ‘[J]eopardizing combatants rather than bystanders during a military act against a terrorist would mean shouldering responsibility for the mixed nature of the vicinity for no reason at all’: Kasher, Asa and Yadlin, Amos, ‘Military Ethics of Fighting Terror: An Israeli Perspective’ (2005) 4 Journal of Military Ethics 3, 18 [OpenURL Query Data]  [Google Scholar].

11 Dinstein, Yoram, The Conduct of Hostilities under the Law of International Armed Conflict (CUP 2010) 141 [OpenURL Query Data]  [Google Scholar] (emphasis added).

12 Schmitt, Michael N, ‘Precision Attack and International Humanitarian Law’ (2005) 87 International Review of the Red Cross 455, 462 [OpenURL Query Data]  [CrossRef]  [Google Scholar]. According to Schmitt, the ‘survival of the military personnel and equipment is an appropriate consideration when assessing the military advantage of an attack in the proportionality context’; ibid.

13 Henderson, Ian, The Contemporary Law of Targeting: Military Objectives, Proportionality and Precautions in Attack under Additional Protocol I (Martinus Nijhoff 2009) [OpenURL Query Data]  [CrossRef]  [Google Scholar] 62, 205.

14 ibid; Stephens, Dale and Lewis, Michael W, ‘The Law of Armed Conflict – A Contemporary Critique’ (2005) 6 Melbourne Journal of International Law 55, 72 [OpenURL Query Data]  [Google Scholar].

15 Solis, Gary D, The Law of Armed Conflict: International Humanitarian Law in War (CUP 2010) 283 [OpenURL Query Data]  [CrossRef]  [Google Scholar], 285. See also Bring, Ove, ‘International Humanitarian Law after Kosovo: Is Lex Lata Sufficient?’ (2002) 71 Nordic Journal of International Law 39, 47 [OpenURL Query Data]  [Google Scholar].

16 Oeter, Stefan, ‘Collateral Damages – Military Necessity and the Right to Life’ in Tomuschat, Christian, Lagrange, Evelyne and Oeter, Stefan (eds), The Right to Life (Martinus Nijhoff 2010) 167 [OpenURL Query Data]  [Google Scholar], 167, 185.

17 Fenrick, William J, ‘Attacking the Enemy Civilian as a Punishable Offense’ (1997) 7 Duke Journal of Comparative and International Law 539, 548–49 [OpenURL Query Data]  [Google Scholar].

18 Stefan Oeter, ‘Is the Principle of Distinction Outdated?’ in Wolff Heintschel von Heinegg and Volker Epping (n 3) 58. Oeter goes on to argue that ‘[the] more technological possibilities you have as a belligerent to avoid direct risk exposure of your own military personnel to enemy force, the more collateral damage you will tend to find justified and proportionate if it helps to spare the lives of your own soldiers’.

19 Solis (n 15) 285. According to Sandoz ‘If the price to absolute security of one's own soldiers is heavy casualties among civilians, this price is too high’: Sandoz, Yves, ‘Commentary’ in Wall, Andru E (ed), Legal and Ethical Lessons of NATO's Kosovo Campaign, International Law Studies, vol 78 (Naval War College, Newport RI 2002) 273 [OpenURL Query Data]  [Google Scholar], 277.

20 Report of the Expert Meeting, ‘Targeting Military Objectives’ (University Centre for International Humanitarian Law 2005) 17, available at http://www.adh-geneva.ch/docs/expert-meetings/2005/1rapport_objectif_militaire.pdf (emphasis added).

21 See 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, 1273. According to the report ‘[t]he questions which remain unresolved once one decides to apply the principle of proportionality include the following: a) What are the relative values to be assigned to the military advantage gained and the injury to non-combatants and or the damage to civilian objects? b) What do you include or exclude in totaling your sums? c) What is the standard of measurement in time or space?, and d) To what extent is a military commander obligated to expose his own forces to danger in order to limit civilian casualties or damage to civilian objects?’.

22 Official Records of the Diplomatic Conference on the Reaffirmation and Development of International Humanitarian Law Applicable in Armed Conflicts, Geneva (1974–1977), vol XV (Federal Political Dept 1978) CDDH/III/224, at 331 (emphasis added; brackets appeared already in the Working Group's original text).

23 Official Records of the Diplomatic Conference on the Reaffirmation and Development of International Humanitarian Law Applicable in Armed Conflicts, Geneva (1974–1977), vol XIV (Federal Political Dept 1978) CDDH/III/SR.24, at 220 (Mr Aldrich, USA).

24 ibid (emphasis added).

25 As far as objects are concerned, the presumption of civilian character in case of doubt is also contained in Amended Protocol II to the Convention on Certain Conventional Weapons, art 3(8)(a). The ICRC's Interpretive Guidance on the Notion of Direct Participation in Hostilities (n 6) 74, provides that ‘[i]n case of doubt, the person must be presumed to be protected against direct attack’. For a critique of the distribution of risks inherent in this approach see Schmitt, Michael N, ‘Deconstructing Direct Participation in Hostilities: The Constitutive Elements’ (2010) 42 NYU Journal of International Law and Politics 697, 737, 738 [OpenURL Query Data]  [Google Scholar]. Schmitt suggests that there should be ‘a presumption that questionable activities qualify as acts of direct participation’ and that ‘[g]ray areas should be interpreted liberally, i.e., in favor of finding direct participation’, n 123 citing Schmitt, Michael N, ‘Direct Participation in Hostilities and 21st Century Armed Conflict’ in Fischer, Horst and others (eds), Crisis Management and Humanitarian Protection: Festschrift für Dieter Fleck (Berliner Wissenschafts-Verlag 2004) 509 [OpenURL Query Data]  [Google Scholar].

26 With regard to military objects the ICRC Study of Customary International Humanitarian law (‘ICRC Study’) concludes that ‘the issue of how to classify an object in case of doubt is not entirely clear’: see Henckaerts, Jean-Marie and Doswald-Beck, Louise, Customary International Humanitarian Law, Vol I: Rules (ICRC, CUP 2005) 35 [OpenURL Query Data]  [Google Scholar], Rule 10. See also US Department of Defense, Final Report to Congress on the Conduct of the Persian Gulf War, Appendix O, The role of the law of war (1992) 31 International Legal Materials 612–44, 627. According to the Report, ‘[t]his language, which is not a codification of the customary practice of nations, causes several things to occur that are contrary to the traditional law of war. It shifts the burden for determining the precise use of an object from the party controlling that object (and therefore in possession of the facts as to its use) to the party lacking such control and facts, i.e. from defender to attacker. This imbalance ignores the realities of war in demanding a degree of certainty of an attacker that seldom exists in combat. It also encourages a defender to ignore its obligation to separate the civilian population, individual civilians and civilian objects from military objectives, as the Government of Iraq illustrated during the Persian Gulf War’ (emphasis added).

27 See declarations and reservations made by France and the UK upon ratification of Additional Protocol I, available at http://www.icrc.org/ihl.nsf/NORM/D8041036B40EBC44C1256A34004897B2?OpenDocument. Against this background the ICRC Study (n 26) concludes ‘… that, in case of doubt, a careful assessment has to be made under the conditions and restraints governing a particular situation as to whether there are sufficient indications to warrant an attack’. The ICRC Study also points out that ‘[i]n the case of non-international armed conflicts, the issue of doubt has hardly been addressed in State practice,…’ (n 26) 36, 24.

28 Benvenisti (n 4) 93.

29 See, for example, Program on Humanitarian Policy and Conflict Research at Harvard University (HPCR), Commentary on the HPCR Manual on International Law Applicable to Air and Missile Warfare (The President and Fellows of Harvard College 2010) 44–5, available at http://ihlresearch.org/amw/Commentary%20on%20the%20HPCR%20Manual.pdf.

30 Dinstein (n 11) 95.

31 ibid.

32 Commentary on the HPCR Manual on International Law (n 29) 45.

33 US Department of Defense, ‘Conduct of the Persian Gulf War: Final Report to Congress’ (April 1992) 699, available at http://www.ndu.edu/library/epubs/cpgw.pdf.

34 Commentary on the HPCR Manual on International Law (n 29) 45. With regard to the more restrictive approach that is reflected in the ICRC Commentary the HPCR commentary states: ‘A better approach is to understand military advantage as any consequence of an attack which directly enhances friendly military operations or hinders those of the enemy. This could, for example, be an attack that reduces the mobility of the enemy forces without actually weakening them, such as the blocking of an important line of communication.’

35 Dinstein (n 11) 93. Of course, the question where to draw the line between relevant military advantages and irrelevant political advantages has at times been controversially discussed. A certain degree of ultimately unavoidable peripheral vagueness has always remained. After all, military ends are never entirely autonomous in themselves and war itself ultimately remains a political act: see Carnahan, Burrus M, ‘Lincoln, Lieber and the Laws of War: The Origins and Limits of the Principle of Military Necessity’ (1998) 92 American Journal of International Law 213, 213, 231 [OpenURL Query Data]  [CrossRef]  [Google Scholar]; Geiss, Robin, ‘Military Necessity: A Fundamental “Principle” Fallen into Oblivion’ in Fabri, Helen Ruiz, Wolfrum, Rüdiger and Gogolin, Jana (eds), Select Proceedings of the European Society of International Law (Hart 2008) 554 [OpenURL Query Data]  [Google Scholar], 557.

36 Commentary on the HPCR Manual on International Law (n 29) 45.

37 According to the US Counterinsurgency Field Manual: ‘In conventional operations, proportionality is usually calculated in simple utilitarian terms: civilian lives and property lost versus enemy destroyed and military advantage gained’ (n 8) n 7–32 (emphasis added).

38 The US Counterinsurgency Field Manual (n 8) 162, note 7–30, for example, refers only to the ‘concrete and direct military advantage expected to be gained’. Similarly, Israel's manual on the law of war refers to ‘harm on the civilian population that is disproportionate to the expected military gain’: Israel, Law of War in the Battlefield, Manual (Advocate General Headquarters, Military School 1998) 40 (emphasis added). In her dissenting opinion in the Nuclear Weapons case before the ICJ in 1996, Judge Higgins stated: ‘Even a legitimate military target may not be attacked if the collateral civilian casualties would be disproportionate to the specific military gain from the attack’: Legality of the Threat or Use of Nuclear Weapons, Advisory Opinion of 8 July 1996 [1996] ICJ Rep 226, Dissenting Opinion of Judge Higgins [20] (emphasis added).

39 Australia, upon ratification, declared: ‘In relation to paragraph 5(b) of Article 51 and to paragraph 2(a)(iii) of Article 57, it is the understanding of Australia that references to the “military advantage” are intended to mean the advantage anticipated from the military attack considered as a whole and not only from isolated or particular parts of that attack and that the term “military advantage” involves a variety of considerations including the security of attacking forces. It is further the understanding of Australia that the term “concrete and direct military advantage anticipated” used in Articles 51 and 57, means a bona fide expectation that the attack will make a relevant and proportional contribution to the objective of the military attack involved.’ The declaration is available at http://www.icrc.org/ihl.nsf/NORM/10312B4E9047086EC1256402003FB253?OpenDocument.

40 The Commander's Handbook on the Law of Naval Operations, NWP 1-14M/MCWP 5-2.1/COMDTPUB P5800.7 (Department of the Navy, Office of the Chief of Naval Operations and Headquarters, US Marine Corps, and Department of Transportation, US Coast Guard, October 1995), para 8.1.1 (emphasis added).

41 See Solis (n 15) (emphasis added).

42 Blum, Gabriella, ‘The Dispensable Lives of Soldiers’ (2010) 2 Journal of Legal Analysis 69, 78 [OpenURL Query Data]  [CrossRef]  [Google Scholar], available at http://dash.harvard.edu/bitstream/handle/1/4324405/dispensiblelives.pdf?sequence=1.

43 The term is borrowed from Kasher and Yadlin (n 10). For a critique of their approach, however, see Section 4.2.1.

44 See, for example, The Commander's Handbook (n 40) which states that the term military advantage ‘refers to the advantage anticipated from the military operation of which the attack is a part, taken as a whole, and not from isolated or particular parts of that operation’, para 8.1.2.1.

45 See nn 39 and 40.

46 The Rules concerning the Control of Wireless Telegraphy in Times of War and Air Warfare (The Hague, February 1923), art 24(1) stipulates: ‘An air bombardment is legitimate only when it is directed against a military objective, i.e. an objective whereof the total or partial destruction would constitute an obvious military advantage for the belligerent’ (emphasis added). The rules are available at http://www.icrc.org/ihl.nsf/FULL/275?OpenDocument.

47 Institut de Droit International, Resolution entitled ‘The Distinction between Military Objectives and Non-Military Objectives in General and Particularly the Problems Associated with Weapons of Mass Destruction’ in Résolutions de l'Institut de Droit International 1957–1991 (Edition Pedone 1992) 67 [OpenURL Query Data]  [Google Scholar], 69 art 2, cited in Dinstein (n 11) 93.

48 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 (ICRC/Martinus Nijhoff 1987) [OpenURL Query Data]  [Google Scholar] 677, 684.

49 Dinstein (n 11) 134.

50 ibid.

51 Dörmann, Knut, Elements of War Crimes under the Rome Statute of the International Criminal Court: Sources and Commentary (CUP 2003) 161 [OpenURL Query Data]  [CrossRef]  [Google Scholar], n 36.2.

52 In 1999 the ICRC submitted a paper to the Working Group on Elements of Crimes of the Preparatory Commission for the ICC, emphasising that ‘[t]he addition of the words “clearly” and “overall” in the definition of collateral damage [in Article 8(2)(b)(iv) of the 1998 ICC Statute] is not reflected in any existing legal source. Therefore, the addition must be understood as not changing existing law’: see ICRC, Paper relating to the Crimes listed in Article 8, paras 2(b)(i), (ii), (iii), (iv), (v), (vi), (ix), (xi), (xii) of the Statute of the ICC annexed to UN Doc PCNICC/1999/WGEC/INF.2/Add.1, 30 July 1999, 29.

53 See Aldrich, George H, ‘Prospects for United States Ratification of Additional Protocol I to the 1949 Geneva Conventions’ (1991) 85 American Journal of International Law 1 [OpenURL Query Data]  [CrossRef]  [Google Scholar]. Despite the refusal to ratify Additional Protocol I, the US claims to adhere to Article 57 as a matter of customary law: see Fenrick, William J, ‘The Rule of Proportionality and Protocol I in Conventional Warfare’ (1982) 98 Military Law Review 91 [OpenURL Query Data]  [Google Scholar]; Taft, William H IV, ‘The Law of Armed Conflict after 9/11: Some Salient Features’ (2003) 28 Yale Journal of International Law 319, 322 [OpenURL Query Data]  [Google Scholar].

54 The text of this and of the other parallel declarations are available at http://www.icrc.org/ihl.nsf/WebSign?ReadForm&id=470&ps=P. Moreover, the ICC Statute, art 8(2)(b), criminalises such attacks only when committed in the knowledge ‘that such attack will cause incidental loss of life or injury to civilians or damage to civilian objects or widespread, long-term and severe damage to the natural environment which would be clearly excessive in relation to the concrete and direct overall military advantage anticipated’ (emphasis added). See also ICRC Study (n 26) 50.

55 Zimmermann, Andreas, ‘The Second Lebanon War: Jus ad Bellum, Jus in Bello and the Issue of Proportionality’ (2007) 11 Max Planck Yearbook of United Nations Law 99, 99, 132 [OpenURL Query Data]  [Google Scholar].

56 Dinstein (n 11) 94, citing Oeter, Stefan, ‘Methods and Means of Combat’, in Fleck, Dieter (ed), The Handbook of International Humanitarian Law (OUP 2008) [OpenURL Query Data]  [Google Scholar] 186.

57 Dinstein (n 11) 93.

58 ibid, 95.

59 Even ‘an attack as a whole’ is a finite event. In this context, Dinstein criticises, as a gross exaggeration, the finding of the Eritrea-Ethiopia Claims Commission which held that ‘a definite military advantage must be considered in the context of its relation to the armed conflict as a whole at the time of the attack’ and ‘not simply in the context of a specific attack’: see Eritrea-Ethiopia Claims Commission: Partial Award – Western Front, Aerial Bombardment and Related Claims, Eritrea's Claims 1, 3, 5, 9–13, 14, 21, 25, 26 (2005), (2006) 45 International Legal Materials 396, 418. According to Dinstein, ‘the admissible mise-en-scène is only “an attack as a whole” and not “the armed conflict as a whole”’ (n 11) 95.

60 But see the interpretation by Kasher and Yadlin of the accumulative benefits doctrine in relation to the fight against terrorists, Section 4.2.1.

61 Sandoz and others (n 48).

62 Michael Walzer, ‘Justice and Injustice in the Gulf War’ in David E Decosse (ed) But Was it Just? – Reflections on the Morality of the Persian Gulf War (Bantam Doubleday Dell Publishing Group 1991) 7.

63 This may seem odd given that IHL, unlike the law enforcement paradigm, is not threat-dependent. A military objective may be attacked at any time irrespective of whether it poses an imminent threat to one's own forces or not. However, where the justification of civilian casualties and damage is at issue, the threat posed by the enemy starts to matter because it may have an influence on the relative value of the military advantage that is to be gained from an attack. This is a logical consequence of regarding ‘force protection’ as a relevant military advantage. Destroying an enemy tank that currently attacks will be perceived as a more important military advantage than the destruction of the same tank while it is just driving through a village. In the former scenario the military advantage gained consists of the destruction of the enemy tank plus the protection of the forces that were attacked by that tank, whereas in the latter scenario the military advantage gained consists only of the destruction of the enemy's tank.

64 See Sassòli, Marco, ‘Ius Ad Bellum and Ius in Bello – The Separation between the Legality of the Use of Force and Humanitarian Rules to be respected in Warfare: Crucial or Outdated?’ in Schmitt, Michael N and Pejic, Jelena (eds), International Law and Armed Conflict: Exploring the Faultlines, Essays in Honour of Yoram Dinstein (Martinus Nijhoff 2007) 241–64, 246–54 [OpenURL Query Data]  [CrossRef]  [Google Scholar]; Gardam, Judith G, ‘Proportionality and Use of Force in International Law’ (1993) 87 American Journal of International Law 391, 410 [OpenURL Query Data]  [CrossRef]  [Google Scholar].

65 Kasher and Yadlin (n 10) 14, 15.

66 ibid, 20.

67 Of course, Kasher and Yadlin's argument is specifically designed to apply to the fight against terrorism. However, it is unclear how they would apply a different reasoning in the context of a ‘regular’ armed conflict. State duties vis-à-vis their own citizens remain the same.

68 Similarly, the US Counterinsurgency Field Manual (n 8) 162, note 7–32, provides: ‘In COIN environments, the number of civilian lives lost and property destroyed needs to be measured against how much harm the targeted insurgent could do if allowed to escape’ (emphasis added).

69 See Oeter (n 16); Fenrick (n 17).

70 According to Benvenisti, ‘[w]hen armies attack they face two conflicting obligations. The first is the obligation to ensure the rights of their own nationals, and the second is the obligation to respect enemy nationals by not targeting them and by striving to reduce the harms the army inflict on them’: Benvenisti (n 4) 89.

71 See Section 4.2.

72 ibid.

73 Criteria such as the size, importance or strategic location of the protected forces may be taken into consideration in order to determine their military value. All other things being equal, protecting only five soldiers is evidently of a lower relative value than the protection of 500 soldiers. As Ian Henderson has stated: ‘Crudely, certain elements of a military are more valuable to it in terms of their military capabilities and replaceability than other parts’: Henderson (n 13) 205. See also Greenwood, Christopher, ‘The Twilight of the Law of Belligerent Reprisals’ (1989) 20 Netherlands Yearbook of International Law 35, 45 [OpenURL Query Data]  [CrossRef]  [Google Scholar].

74 Schmitt (n 12) 462.

75 Given that proportionality is necessarily assessed ex ante, the military and humanitarian effects of the attack are always to some degree speculative and ultimately depend on subjective risk assessments: see Amichai Cohen and Yuval Shany, ‘A Development of Modest Proportions: The Application of the Principle of Proportionality in the Israeli Supreme Court Judgment on the Lawfulness of Targeted Killings’ (Research Paper No 5–07, April 2007), available at http://www.ssrn.com/abstractid=979071. But the uncertainty that inherently lies in every ex ante decision is different from the speculation that derives from comparing a future attack (and its anticipated consequences) with a hypothetical attack. In the latter case the uncertainty is not about what is most likely to happen if an envisaged attack is carried out, but what would have happened if a hypothetical attack – that exists only in the strategic thinking of military commanders – had been executed.

76 The little clarity that exists derives primarily from the limiting adjectives ‘concrete’ and ‘direct’ rather than the ‘military advantage’ notion itself.

77 A genuine balancing judgment is possible only if there is at least a theoretical possibility to fill both scale pans with the same amount of things. Thus, if the military advantage notion is interpreted broadly, as a logical consequence a similarly broad approach should be employed with regard to civilian casualties and damage. Otherwise the proportionality assessment is distorted from the outset. This is supported by the wording of arts 51(5)(b) and 57(2)(a)(iii) of Additional Protocol I. Whereas the military advantage is limited by the adjectives ‘concrete and direct’, no such limiting qualifiers apply with respect to civilian damage. The word ‘incidental’ is certainly broader than the requirements of being ‘concrete and direct’. Therefore, at least foreseeable long-term repercussions on the civilian population have to be taken into consideration when determining what civilian damage may be expected from an attack. See Geiss, Robin, ‘The Conduct of Hostilities in Asymmetric Conflicts – Reciprocity, Distinction, Proportionality, Precautions’ (2010) 23 Journal of International Law of Peace and Armed Conflicts 122, 130 [OpenURL Query Data]  [Google Scholar].