Collateral Damage: A Dangerous Omission in the Law of Armed Conflicts

  1. Introduction

Hon. Judge Pillay is a distinguished judge, with an impeccable record of service at the ICTR and ICC. She is a committed human rights lawyer. Judge Pillay is in a unique position, with the necessary institutional support, to initiate a process of bringing the discredited concept of ‘collateral damage’ within the legal framework. This will make it legally possible to bring perpetrators to account by indicting and prosecuting them for killing civilians, by the use of disproportionate force. The exercise will further provide support for the on-going fight against impunity. In this short paper, I provide a brief commentary on the concept of ‘collateral damage’ in the context of the Law of Armed Conflict and recommend that ‘collateral damage’ be criminalized and designated an international crime.

  1. An Overview of Collateral Damage

The Law of Armed Conflicts does not expressly address the legal character of collateral damage. There is no mention of collateral damage in the Geneva Conventions of 1949 and its Additional Protocols I and II of 1977. Similarly, there is no mention of ‘collateral damage’ in the 1899 and 1907 Hague Conventions or in customary international law.

It is because ‘collateral damage’ is not necessarily a war crime under the Laws of Armed Conflict, among other things, that has led to non-prosecution of persons responsible for excessive deaths of civilians during international and non-international armed conflicts.

The term ‘collateral damage’ refers to the killing of persons, who are civilians or non-combatants. The civilians killed are those who do not threaten combatants. The killing of these civilians, including combatants who are hors de combat do not serve any military purpose. In a legal sense, it is tempting to pin every killing on intention. However, in addressing responsibility arising from the killing of civilians in the context of armed conflicts, proof of the perpetrator’s intention is problematic and needs some consideration. The Law of Armed Conflict draws a distinction between legitimate and illegitimate targets via the notion of a threat. Combatants are ordinarily armed and threatening, non-combatants are not. Non-combatants include the civilian population (women and children) and all combatants who are hors de combat, prisoners of war, or the sick, wounded or shipwreck. These categories of non-combatants are not a threat, and therefore not legitimate military targets. The killing of any civilian or non-combatant is either accidental or intentional. Accidental killing lacks intent. To that extent, the perpetrators may bear no criminal responsibility for collateral damage.

Military ethics attempts to explain collateral damage by analogy to the doctrine of double effect. That doctrine applies to situations in which it may be necessary or unavoidable to take an action that will cause the loss of innocent life in order to achieve some greater good. The doctrine holds that such an action should be performed only if the intention is to bring about the good effect and the bad effect will be an unintended or indirect consequence – and therefore characterized as collateral damage.

According to Ronald Munson, as a general rule, four conditions must be satisfied before the doctrine of double effect is applied. First, it must be objectively demonstrated that the action that results in a collateral damage itself is morally indifferent or morally good. Second, the bad effect is the means by which the good effect is achieved. Third, the motive is the achievement of the good effect only and, fourth, the good effect is at least equivalent in importance to the bad effect.[1]  Therefore, in theory at least, whereas intent to kill civilians is never permissible, according to the doctrine of double effect, foreseeing civilian deaths as an effect of a permissible action (such as aiming at a military target but striking at a civilian one) is not prohibited.[2] Such analogy tends to suggest that collateral damage, even when many civilians are killed, will not necessarily result in the commission of a war crime. In other words, it is an accidental but not intentional killing.

The issue of motive, as stipulated in the third requirement, suggests that lack of motive may be sufficient to exonerate the perpetrator from criminal responsibility.[3]  However, as persuasively argued by Walzer, not to intend to kill non-combatants is, by itself, not to show enough restraint. Walzer articulates his arguments as follows:

Simply not to intend the death of civilians is too easy; most often, under battle conditions, the intentions of soldiers are focused narrowly on the enemy. What we look for in such cases is some sign of a positive commitment to save civilian lives … And if saving civilian lives means risking soldiers’ lives, the risk must be accepted. But there is a limit to the risks that we require… We can only ask soldiers to minimise the dangers they impose …on civilians.[4]

Steps required of military commanders and their subordinates to ‘minimise the dangers’ to non-combatants include measures such as ensuring proper intelligence-gathering about targets so as to enable combatants not to deliberately destroy homes of non-combatants just because leaving the houses standing makes it more difficult to access military targets. It is not acceptable when a superior side in an asymmetric armed conflict uses advanced technology and weapons to fight from such a distance that the risk to its own combatants is minimal but the risk to non-combatants on the enemy side is unnecessarily increased.[5] Minimising the danger to non-combatants must be the concern of all belligerents. This requirement means that in some cases, fighter pilots must be ordered by their superior officers to fly low enough to clearly identify targets and to avoid ‘carpet bombings’ that will inevitably result in large numbers of civilian deaths. The use of unmanned drones in contemporary armed conflict is unfortunate, and needs to be criminalized, as the drones often miss their intended targets and kill civilians who are protected persons under the Geneva Conventions of 1949 and Additional Protocols I and II of 1977.

Air combat, viewed in the context of the doctrine of double effect, and the resulting ‘collateral damage’ raises difficult questions in determining the norms and rules on the protection of civilians, civilian population and civilian objects in warfare. Air combat often results in ‘accidental’ killing of civilians, and the destruction of their property, which in turn is conveniently described as ‘collateral damage’. That term – collateral damage – is not a legal term, but a term of art, describing the striking of civilian targets without necessarily determining criminal responsibility of the perpetrators. Therefore, in one sense, it may be argued that collateral damage is an accident caused, for example, by a bomb that has gone astray. On the other hand, it could be a deliberate act of the enemy combatant, who uses the concept of ‘collateral damage’ as a cover for an essentially unlawful and criminal act.

Therefore, when a bomb misses a military target, and hits a civilian one, what criteria must be used in determining whether the act is intentional or, accidental?  To attempt to provide an answer, we need to examine the circumstances under which the decision to bomb the intended target was taken by the superiors, how the targets were selected, the manner in which the target was communicated to the fighter pilots, and what precautions, if any, were taken by the fighter pilots to avoid or minimize civilian deaths before dropping the bomb. It is therefore necessary, for example, to determine whether the bomb was dropped in a crowded civilian target or a mixed civilian and military target so as to make reasonable inference based on the facts of each case.

  1. Whether Collateral Damage is Accidental or Intentional

In assessing the situation, at least two things must be considered. First, it is important to determine, as a matter of fact, whether the strike was accidental. If it was not accidental, then the possibility that it was intentional may not be ruled out. Therefore further steps should be taken to verify independently and objectively whether or not the dropping of the bomb was not intentional. Best practice would require third parties to conduct the investigation as they are more likely to be objective, transparent and fair as compared to belligerents. Rules for such determination must exist prior to the commencement of hostilities. It is against the principle of natural justice for a suspect to investigate his own conduct that forms the basis of the allegation. In most of the armed conflicts in Africa and elsewhere, when government fighter pilots kill civilians, it is the same  military, controlled by the civilian leaders who run the government that insists, and always conducts, the investigation and reports to itself.  It is not surprising that the result of these investigations nearly always vindicate the responsible members of the armed forces for any wrong-doing. Such practice goes against the principle of natural justice and must be rejected.

Second, the process, quality and the threshold of the investigation and evaluation of evidence collected should be transparent and public. Ordinarily one would expect investigators to be objective and fair. However, in the absence of a prior agreement by the parties, the problem is that the findings of the investigation may depend on who takes the key decisions in appointing members of the panel of investigators, the terms of reference, the target of investigation and what the motives of the appointing authority are. The process of investigation may be subjective with the sole objective of protecting the perpetrators by shielding them from possible criminal prosecution.

There are, for example, three scenarios under which killings of civilians may be alleged to constitute ‘collateral damage’. The first scenario is when an act or acts that caused collateral damage is an isolated one; the second is when an air strike is a result of a series of accidents that may or may not depend on poor communications, or a break in the chain of command; and the third is when the act or acts is a result of a consistent pattern of conduct.  I now examine these three different scenarios.

First, in determining whether an act that resulted in ‘collateral damage’ is an isolated one, it is important to ascertain factually, objectively and transparently if the military ‘attack’ or ‘operation’ was an isolated act. If the answer is ‘yes’, then it may be concluded that it is an accident. A finding of accidental killing of civilians or destruction of civilian objects leads to the conclusion that the combatant, or in this case, a fighter pilot and all persons in the chain of command, leading to the highest level, are responsible for the resulting ‘collateral damage’, but bears no individual criminal responsibility. As a matter of law, not all accidental killings automatically translate into criminal acts.  An example of an isolated act resulting in ‘collateral damage’ was when NATO fighter pilots, during the Yugoslavia war, accidentally bombed the Chinese Embassy in Belgrade killing four diplomats.  Under this type of situation, compensation may be paid to the victims without necessarily admitting criminal responsibility for the unlawful bombings and the resultant deaths of civilians.

Second, where ‘collateral damage’ is caused by a series of military ‘attacks’ or ‘operations’, and over a prolonged period of time, the civilian population and civilian objects continue to be targeted, then it must be factually, objectively and transparently determined whether these series of military strikes are accidental or deliberate. Once evidence of regular striking of the civilian targets is established, the evidential burden shifts to the offending party. In this case, the burden of proof is not beyond a reasonable doubt, but on a balance of probability. If it is determined that it was accidental, no criminal responsibility results, notwithstanding the high numbers of civilian deaths, or the frequency with which the civilian targets were hit.

If, on the other hand, the allegation is that the civilians’ targets were hit intentionally, then the party making the allegation has a higher threshold for proof, though not beyond a reasonable doubt. The burden of proof is limited to satisfying the prima facie requirement that the fighter pilot and the superiors in the chain of command have a case to answer. Therefore, under these circumstances, the resulting ‘collateral damage’ may or may not be a war crime. This level of burden of proof is intended to prevent the making of wild allegations without sufficient supporting material.

Third, where collateral damage is caused by a consistent pattern of military attacks or operations, targeting civilians, such actions tend to rise to the level of war crimes because it is illegal, under the Geneva Conventions of 1949 and Additional Protocols I and II of 1977, to deliberately and intentionally target civilians. However, to prove that combatants intentionally targeted civilians is problematic. Even combatants, who deliberately and intentionally target civilians, often deny doing so. To that extent, not much credibility may be attributed to regular press statements issued by parties to armed conflict re-stating their compliance with the Law of Armed Conflicts and denying attacking civilians. On the contrary, one should treat such statements with caution and use a reasonable man’s (or woman’s) standard in assessing the available evidence prior to determining whether the attack was accidental or intentional.  The test therefore depends on the facts on the ground as factually, objectively and transparently determined and whether a reasonable superior or commander, with sufficient knowledge of the factual situation, would order a fighter pilot to conduct an attack or operation when fully aware that civilians would be killed.

Acts described as collateral damage resulting in deaths of civilians during the First World War led to an early, but unsuccessful, attempt to regulate air warfare.[6]  Use of air warfare, with its increasing ‘collateral damage’ continued during the Second World War. In less than three decades air warfare became an important military strategy for many States engaged in international and non-international armed conflicts. The Vietnam War (known to the Vietnamese as the ‘American war’) ending in 1975, the Arab-Israeli six-day war of 1967 and a second Arab-Israeli war of 1973, the Iraq wars of 1991 and 2003, the Palestinian-Israeli and Israeli-Hezbollah-Lebanon wars and the current war in Afghanistan are evidence of the immense gap between theory and practice of the law of air warfare as more civilians continue to get killed and combatants take no criminal responsibility for the resulting ‘collateral damage’.

During the Nuremberg trials, at the instigation of the victorious Allied Powers, the violation of the principle of discrimination was excluded from the charges brought against the accused persons.[7] The prosecutors at the Nuremberg Tribunal did not consider indiscriminate killing of civilians  and the civilian population resulting in ‘collateral damage’ as war crimes and there were no criminal prosecution of fighter pilots and their superiors in the chain of command for the indiscriminate killing of civilians at Hiroshima, Nagasaki, Kobe, Humbug or Dresden. At Hiroshima and Nagasaki, the two cities where the United States used atomic bombs against a civilian population and objects that resulted in the death of thousands of Japanese for no apparent military reason at a time when Japan was all but defeated is one such example of indiscriminate bombing. At Kobe in Japan, Hamburg and Dresden in Germany, the Allied Air force destroyed cities resulting in the deaths of civilians therein at a time when the Second World War was, for all practical purpose, over. No person was prosecuted for these horrendous crimes.

In the post Second World War period, in international and non-international armed conflicts, the killing of civilians took place on a massive scale, while air warfare was taken to horrifying excesses, consigning almost to oblivion, the basic principle that the civilian population should be protected as far as possible.[8] The indiscriminate use of the air force made terrorising the civilian population the dominant purpose of air warfare. The practice threatened to bury all restrictions on the use of force as biological weapons were introduced in warfare. In Vietnam for example, the United States army used ‘agent Orange’ indiscriminately resulting in the deaths of many civilians. Many children of the survivors of ‘agent Orange’ were born with varying degrees of deformity.

At the instigation of the Allied Powers, the main question concerning the regulation of weapons, means and methods of warfare, including adjustment of warfare to the principles of military necessity and humanity, were subsequently excluded from the work of the Red Cross meetings which preceded the Geneva Conventions of 1949. However, the Geneva Red Cross Conventions of 1949 made efforts to regulate most of the violations that had arisen in the wake of the Second World War and to fill the gaps, which had become obvious in earlier treaties.[9] These efforts were unsuccessful. Thus, the Geneva Conventions of 1949 are silent on the legal nature of ‘collateral damage’.

However, the Fourth Geneva Convention developed a legal regime that dealt with the problems of military occupation, but the questions of possible limits on air warfare, and on arms that cause unnecessary suffering were left within the scope of customary international law as codified in the 1899 and 1907 Hague Conventions. The Diplomatic Conference on the Reaffirmation and Development of International Humanitarian Law Applicable in Armed Conflict held in Geneva from 1974 until 1977 considered regulating the killing of civilians as a result of ‘collateral damage’ in a wider context under the general principles of protecting civilians and civilian population.[10] Nonetheless, no specific provision proscribing the killing of civilians was adopted.

With the adoption of Additional Protocol I of 1977, the provisions of Article 51 on the protection of the civilian population and Articles 52 to 56 on the protection of civilian objects, an attempt was nevertheless made to achieve such a specification of customary rules, although it remains an open question as to what extent these provisions constitute a pure codification of pre-existing customary international law and to what extent they constitute further development or even creation of new rules.

The provisions regulating how combatants may treat the civilian population was heavily criticised by some military experts, and – in connection with the absolute prohibition against reprisals against the civilian population contained in Additional Protocol I – are probably the main reason important military powers still refuse to ratify Additional Protocol I of 1977.

It is against this background that the killing of civilians is conveniently described as ‘collateral damage’, notwithstanding the fact that the phrase is of no legal consequence. Further, the Law of Armed Conflicts continues to be complicated because when the basic underlying rules that regulate the protection of civilian is breached by combatants, whether accidentally or intentionally, once described as ‘collateral damage’, the phrase tends to acquire a new meaning, that is, of a combatant having killed a person by accident. It is this fact that makes it imperative that the phrase ‘collateral damage’ is brought within a legal framework, properly defined and the resulting acts criminalized.

In a broad sense, the Law of Armed Conflicts addresses not only cases of international wars in a formal sense, but also other types of armed conflicts which are not classical wars. This approach is in conformity with the legal expressions in the Geneva Conventions of 1949[11] and the general principles of the ‘Law of Geneva’.[12] The Law of Armed Conflicts covers traditional wars regulated by the laws and customs of war as incorporated in the Hague Conventions of 1899 and 1907 (the Law of The Hague).[13]  The primary objective of the normative ‘Law of Geneva’ and the ‘Law of the Hague’ is to protect combatants, non-combatants, including civilians, women and children.  The adoption of a provision criminalizing ‘collateral damage’ would not therefore alter major provisions of the Law of Armed Conflicts to the detriment of the civilian population. On the contrary, as I discuss below, the Law of Armed Conflicts sufficiently regulate the means and methods of conducting wars and provides sufficient protection to the civilian population. What is lacking is the criminalization of ‘collateral damage’.

The jurisprudence of the Law of Armed Conflicts makes a distinction between the laws governing resort to force (jus ad bellum), and laws regulating wartime conduct (jus in bello).[14] Examples of jus ad bellum include the General Treaty of Renunciation of War of 1928, (the Kellogg-Brian Pact, also known as the Pact of Paris) and the Charter of the United Nations, particularly Articles 2(4) and 51. The former condemns the use of war as an instrument of national policy[15] and the latter prohibits the threat or use of force against any State except in self-defence.[16]

Jus in bello is further divided into the Law of Geneva and the Laws of The Hague. The Law of Geneva extends protection to the wounded, sick and shipwrecked, prisoners of war, civilians and civilian populations.  The Law of The Hague, on the other hand, regulates the means and methods of conducting armed conflicts. Failure to comply with the Law of The Hague by belligerents can have great impact on the lives of combatants as well as non-combatants. Jean Pictet correctly observed that the objective of the Law of The Hague is to humanize war by balancing the means and methods of warfare under the principle of military necessity with the idea of respect for humanity.[17]

The main difference between the Law of Geneva and the Law of The Hague is that the Law of Geneva are characterized by strict non-derogable prohibitions while the Law of The Hague comprise of vaguely worded provisions with limited and discretionary procedures for its implementation and enforcement by belligerents. Moreover, the generalities encountered in many of the provisions of the Law of The Hague make it easier for States to introduce and use new weapons that often cause unnecessary suffering to the civilian population, particularly women and children as witnessed in the recent and current armed conflicts, for example, in Palestine, Iraq and Afghanistan.


In conclusion, it is my submission that, in the interest of humanity, it is imperative that ‘collateral damage’ is brought within the legal framework and criminalized. It is only then will adequate protection be extended to the civilian population, particularly women and children in situation of armed conflicts and the perpetrators, are prosecuted and held to account.


[1]  Munson, R, ‘An Overview of Aquinas’ Natural Law Theory,’ reprinted in George R. Lucas et al., Ethics for Military Leaders (Boston: Simon and Schuster Custom Publishing, 1998), and p.397.

[2] Lichtenberg, J, ‘The Ethics of Retaliation,’ Philosophy and Public Policy Quarterly 21.4 (Fall 2001) p.7.

[3] French, S.E, ‘Murders, Not Warriors: The Moral Distinction between Terrorists and Legitimate Fighters in Asymmetric Conflicts’ Terrorism and International Justice, edited by James P. Sterba (Oxford University Press, 2003), p.39.

[4] Walzer, M Just and Unjust Wars. (New York: Basic Books Inc., 1977), p.155-156.

[5] Michael Ignatieff, Virtual War: Kosovo and Beyond (New York: Picador USA [Metropolitan Books, Henry Holt and Company], 2000), pp214-215.

[6] Parks, W.H, ‘Air War and the Law of War’ Air Force Law Review 32 (1990) 1-225 at p.25-36.

[7] Pictet, J.S, (ed.,) Development and Principles of International Humanitarian Law, Geneva (1985); Ph. Piccigallo, The Japanese on Trial: Allied War Crimes (1979).

[8] Best, G, Humanity in Warfare: The Modern History of the International Law of Armed Conflicts (1980).

[9] Kunz, J.L, ‘The Chaotic Status of the Laws of War and the Urgent Necessity for their Revision’ AJIL 45(1951) 37-61; and ‘The Laws of War’ AJIL (1956) 313-37.

[10] Bothe, M; Partsch K.J, and Solf,  W.A,; New Rules for Victims of Armed Conflict, The Hague/Boston /London (1982); see also Solf, W.A  ‘Protection of Civilians against the Effects of Hostilities under Customary International Law and under Protocol I’ The American University Journal of International Law and Policy I (1986) 107 135.

[11]Article 2 common to all the four 1949 Geneva Conventions provide:

‘In addition to the provisions which shall be implemented in peacetime, the present Convention shall apply to all cases of declared war or of any other armed conflict which may arise between two or more of the High Contracting Parties, even if the state of war is not recognized by one of them

The Convention shall also apply to all cases of partial or total occupation of the territory of a High contracting party even if the said occupation meets with no armed resistance. Although one of the Powers in conflict may not be a party to the present Convention, the Powers who are parties thereto shall remain bound by it in their mutual relations. They shall furthermore be bound by the Convention in relation to the said Power, if the latter accepts and applies the provisions thereof.’(Emphasis added).

[12]The ‘Law of Geneva’ is represented by the four 1949 Geneva Conventions. Throughout this work, the First Geneva Convention shall refer to the Geneva Convention for the Amelioration of the Condition of the Wounded and Sick in Armed Forces in the Field; the Second Convention shall refer to the Geneva Convention for the Amelioration of the Condition of the Wounded, Sick, and Shipwrecked Members of the Armed Forces at Sea; the Third Convention shall refer to the Geneva Convention Relative to the Treatment of Prisoners of War; and the Fourth Convention shall refer to the Geneva Convention Relative to the Protection of the Civilian Persons in Time of War.

[13]The ‘Law of the Hague’, in a nutshell, is the law regulating the conduct of hostilities. It encompasses not only the Conventions concluded at The Hague, but also customary international law. The Law of The Hague, inter alia, lays down the rights and duties of belligerents and limits the method of warfare. See generally J.Pictet, ‘The Need to Restore the Laws and Customs Relating to Armed Conflicts ‘Review of the International Commission of Jurists, (March, 1969) at p.23.

[14] Bailey, S.D, Prohibitions and Restraints in War (1972); Best, G, Humanity in Warfare: The Modern History of the International Law of Armed Conflicts (1983).

[15]The General Treaty of Renunciation of War (the Kellog-Briand Pact or the Pact of Paris), 94 L.N.T.S.57, see particularly Art.1, 46 stat.2343.

[16]Article 2(4) of the UN Charter which provides: ‘All Members shall refrain in their international relations from the threat or use of force against his territorial integrity or political independence of any state, or in any other manner inconsistent with the Purposes of the United Nations.’

[17]Pictet, J, Development and Principles of International Humanitarian Law (Nijhoff Publishers (1985)).

Leave a Reply

Your email address will not be published. Required fields are marked *