Reckless endangerment warfare

Civilian protection in international humanitarian law

International Humanitarian Law provides for the protection of civilians through the legal principles of distinction, military necessity, proportionality, and feasible precaution. ¹ Distinction requires combatants to distinguish at all times between civilians and combatants and between military objectives and civilian objects in planning and executing attacks (Protocol, 1977: Articles 48–49). Military necessity stipulates that the use of force by combatants must be strictly limited to that which is necessary to compel the submission of the enemy with the least possible expenditure of time, life, and resources (Lieber, 1863: Article 14). Legitimate targets are limited to those that make an effective contribution to military action and whose destruction or neutralization offers a definite military advantage in circumstances ruling at the time (Protocol, 1977: Article 52).

The conflict between military advantage (the expected benefit from a military attack) and civilian immunity creates an uneasy tension that is mediated by the principles of proportionality and feasible precaution. Proportionality holds that in all circumstances, the anticipated military advantage of an attack or action must be balanced against the probable or expected civilian losses. Such losses cannot be ‘excessive’ in relation to said advantage gained from the attack (Protocol, 1977: Articles 51 (5b) and 57 (2iii)). Combatants must make specific calculations concerning the effects of their attacks on civilian lives and objects and then balance anticipated military gains with expected civilian losses prior to launching a strike. Thus, combatants may not attack even a legitimate military target if a commander can reasonably anticipate that the collateral damage to civilians is likely to be disproportionate to the specific military gain from the attack (International Court of Justice, 1996: 587). While high civilian losses in and of themselves may not indicate a violation of the distinction principle, they may very well suggest a disproportionate attack. In other words, a target may be legal but the effects of an attack on it disproportionate.

Proportionality is the most difficult of the four principles to assess in practice since it requires balancing two incompatible values, civilian casualties and military advantage, both of which require subjective evaluations. Since there is no precise objective way of evaluating the value of a military target vis-à-vis the collateral damage it would likely cause, military commanders are provided with a loophole for engaging in attacks that they know will result in high civilian casualties by claiming that military advantage outweighed the expected losses. Thus, Alexander Downes (2008: 53) holds that the greater the military advantage associated with an object, the more collateral damage is legally permissible in an attack. Taken to its logical conclusion, this approach can be exploited to justify almost any type of attack, so long as commanders determine that the military value of the target is high enough.

In addition, the proportionality rule contains at least two other subjective elements that are difficult to weigh and easy to exploit. First, combatants are required to evaluate the anticipated military advantage against expected civilian losses. Determining what is anticipated and expected can be manipulated by those with an interest in a particular outcome. At the same time, there are limits to how far military commanders can go in doing so. Like all areas of law, there is an assumption of reasonableness that provides boundaries for even subjective judgments. Thus, for example, the International Criminal Tribunal for the Former Yugoslavia (2003) ruled that when determining whether an attack was proportionate, ‘it is necessary to examine whether a reasonably well-informed person in the circumstances of the actual perpetrator, making reasonable use of the information available to him or her, could have expected excessive civilian casualties to result from the attack’. This assumption is a crucial element in defining depraved or reckless behavior (see below).

Second, although civilian losses may not be ‘excessive’, there is no firm measure for what this constitutes. As International Criminal Court Chief Prosecutor Luis Moreno-Ocampo (2006) held, ‘Under international humanitarian law and the Rome Statute, the death of civilians during an armed conflict, no matter how grave and regrettable, does not in itself constitute a war crime. International humanitarian law and the Rome Statute permit belligerents to carry out proportionate attacks against military objectives even when it is known that some civilian deaths or injuries will occur.’ I will address these issues in more detail below.

If combatants adhere to these three principles, they are not held legally liable for ‘incidental’ or ‘inadvertent’ civilian casualties, nor are they responsible for the unintentional destruction of civilian objects that may occur during lawful actions. This is the collateral damage exemption in the law of armed conflict. Under IHL, such casualties are considered to be inadvertent if they are the result of an unanticipated or accidental mishap, for example a weapons or equipment malfunction, faulty information, or other types of human or mechanical error (Schmitt, 2005: 457). Civilian casualties are incidental if they are the unintended (although likely anticipated) spillover effects of an attack that is executed according to plan that abides by the above principles of IHL.

Moreover, the responsibility of the attacker for civilian losses can be mitigated by two factors: the failure of the defenders to clearly separate military from civilian objects, and the use of civilian facilities for military purposes. According to IHL, the defending state has an obligation to protect its own populations from an attack by removing civilians from the vicinity of military objectives and avoiding the placement of military facilities and personnel near populated civilian areas (Program on Humanitarian Policy and Conflict Research, 2009: 20–21). Failing to do so could place at least some responsibility for civilian losses on the defending state. This is known as the principle of Passive Precaution. In addition, objects that serve a primarily civilian function can become military objectives when, by their nature, location, purpose or use, such objects make an effective contribution to military action. This could theoretically include purely civilian facilities such as apartment buildings if they are used by combatants as staging areas or attack positions.

The primary criterion for determining whether such effects are collateral (as opposed to intentional or depraved) is the degree to which the attacker conformed to the principles of distinction, military necessity, and proportionality. Thus, an attack on an arms factory that kills civilian workers inside is legal because the facility was a legitimate military target and the civilians were not the object of the attack, but attacking factory workers on their way home from work is a violation of IHL because they were.

At the same time, neither the failure of a defender to clearly separate civilian objects from military facilities nor the presence of military personnel or equipment within the civilian population relieves the attacker of its obligation to adhere to the principle of proportionality (Protocol, 1977: Article 50, (3)). Moreover under IHL, military organizations are also required to take affirmative steps to minimize collateral damage, not just avoid committing blatant violations. This principle of ‘feasible precaution’ places limits on the collateral damage exemption. It goes beyond simply balancing civilian casualties with military advantage; it involves choosing the means and methods that are expected to avoid or minimize the loss of civilian life and property (Protocol, 1977: Article 57 (2) (a)(iii)), and implementing preventive measures that are ‘practicable or practically possible, taking into account all circumstances ruling at the time including humanitarian and military considerations' (Bothe, Partsch & Solf, 1982: 362). Under this principle, states are required to choose the means and methods of attack that minimize incidental loss of life, injury to civilians, and damage to civilian objects (Protocol, 1977: Article 57(2)(a)(ii)). If there is any question as to the status of an object, individual, or facility, the attacker must assume it to be civilian and therefore off limits (Protocol, 1977: Article 52(3)).

This means that attackers are required under IHL to exercise discretion before deciding to target a military objective if it would put civilians at risk even if they decide that such civilian losses would be proportionate to the military advantage (Rogers, 2004: 128; Boothby, 2009: 43). Moreover, attackers may not strike a military objective if there are other viable targets that offer a similar military advantage with less danger to civilian lives and objects; they must also choose among tactics and weapons that are least likely to cause incidental loss of civilian life, injury to civilians, and damage to civilian objects (Protocol, 1977: Article 57(3)).

War crimes

A violation of any of the above principles constitutes a breach of IHL. ‘War crime’ is a broad term that can encompass a wide variety of practices, from a soldier abusing a prisoner to a military unit massacring an entire town. However, although a wide range of actors – from individual soldiers to commanding officers – can commit such violations, it is the conduct of governments and the public officials that comprise them that has the greatest impact on how combatants fight. It is also this conduct that poses the greatest threat to the bedrock principle of civilian immunity. While violations by individual combatants can certainly have grave consequences for both civilians and other combatants, the most serious and wide-ranging breaches are usually the result of specific war-fighting strategies pursued by states. I argue that we can roughly divide strategies that threaten the lives and welfare of civilians (protected under IHL) into three categories: civilian targeting, depraved indifference warfare, and reckless endangerment warfare.

Civilian targeting is a calculated strategy to inflict suffering on the general populace by deliberately focusing attacks on population centers and public facilities such schools, commercial areas, hospitals, and refugee camps (Downes, 2008; Pape, 2008: 18). Such attacks are executed with the belief that mass killing, destitution, and terror are appropriate – or at least necessary – responses to military and political problems (Slim, 2008: 121). This type of warfare flouts the principle of distinction by making the civilian population, individual civilians, non-defended localities, and demilitarized zones the object of attack (Protocol, 1977: Articles 85(3)(a) and (d)).

This brazen violation of IHL has been employed in a variety of conflicts over the past two decades. In conflicts such as Bosnia, for example, civilian casualties were not collateral damage, but rather the primary goal of the Serbian forces attacking defenseless towns and cities in Bosnia-Herzegovina (see Glenny, 1996; Kaldor, 1999: 59). Similarly, the widespread indiscriminate attacks against major towns and cities employed by Russian forces during the Second Chechen War in 1999 were aimed at terrorizing the population and punishing them for supporting secessionist groups (Council of Europe, 2000; Faurby, 2002). These types of violations have been the primary focus of research by academics and journalists, as well as disciplinary action taken by international organizations against violators (see, for example, Dworkin, Gutman & Rieff, 2007).

A second type of practice that threatens civilian immunity is one that I call Depraved Indifference Warfare (DIW). In US criminal law, depraved indifference is defined as an act that presents a very high risk of death to others, and is ‘so wanton, so deficient in a moral sense of concern, so lacking in regard for the life or lives of others, and so blameworthy as to warrant the same criminal liability as that which the law imposes upon a person who intentionally causes a crime’ (People v. Suarez, 2005). Unlike intentional murder, an act of depraved indifference is defined by the risk that is intentionally created by the perpetrator’s conduct, rather than a specific malevolent intent or the injuries actually resulting from the action. Belligerents engage in DIW when they execute massive attacks on towns and cities for the purpose of progressively degrading the adversary’s infrastructure and political and economic institutions. In doing so, the attackers threaten further destruction if the adversary does not capitulate.

Unlike civilian targeting, which violates the principle of distinction, DIW involves attacks that are widespread and thus cannot be reliably directed at a specific military objective, which violates the prohibition against indiscriminate attacks (Protocol, 1977: Article 51(4)). By launching such massive attacks on heavily populated towns and cities, excessive civilian casualties are both foreseeable and inevitable, suggesting a wanton indifference to the lives of non-combatants, and violating the principle of proportionality (Protocol, 1977: Article 51(4)).

This type of strategy reflects an extreme application of contemporary theories of military force that stretch the principle of military necessity beyond that which is clearly intended in IHL. As stated above, the principle of distinction limits targets to those that provide a definite, direct, and concrete military advantage. On the surface, this would presumably enjoin combatants from using military power to pursue objectives that are not directly related to the war effort, particularly if they adversely affect civilian life and welfare. Yet over time, combatants have expanded the definition of military necessity in ways that allow for far broader goals. For example, the St Petersburg Declaration (1868: preamble) – which helped lay the foundation for the development of contemporary IHL – stated that ‘the only legitimate object which states should endeavor to accomplish during war is to weaken the military forces of the enemy’. Yet, contemporary military doctrines have greatly enlarged the definition of military advantage to include not only degrading an enemy’s war-fighting capabilities, but also undermining their leadership and demoralizing the population in an effort to destroy their capacity to sustain the conflict (Parks, 2007: 100; Department of the Air Force, 1997: 51; Dunlap, 2000: 9).

For example, US Air Force Colonel John Warden’s widely-adopted Five Strategy Rings theory argues that the massive use of air power should be employed to paralyze each of the adversary’s five centers of gravity, which he portrays as a series of concentric circles: military forces, population, supporting infrastructure, ‘system essentials’ such the national economy, and political leadership. According to this approach, the military forces of the adversary (the outer ring) are less important than the social and political centers of gravity (the inner rings), and, therefore, technologically advanced states should skip the outer ring and focus their attacks on the political leadership, infrastructure, and foundations of popular support (Warden, 2002).

Similarly, the US strategy of Rapid Dominance (from which the phrase ‘shock and awe’ emerged) that was developed by military planners in 1996 involves ‘the capability to deny an opponent things of critical value … Shutting the country down would entail both the physical destruction of appropriate infrastructure and the shutdown and control of the flow of all vital information and associated commerce so rapidly as to achieve a level of national shock akin to the effect that dropping nuclear weapons on Hiroshima and Nagasaki had on the Japanese’ (Ullman & Wade, 1996: 13). This would entail disrupting ‘means of communication, transportation, food production, water supply, and other aspects of infrastructure’ (Ullman & Wade, 1996: xxvii). The appropriate balance of shock and awe ‘must cause the perception and anticipation of certain defeat and the threat and fear of action that may shut down all or part of the adversary’s society or render his ability to fight useless short of complete physical destruction’ (Ullman & Wade, 1996: 82). This comes very close to engaging in ‘acts or threats of violence the primary purpose of which is to spread terror among the civilian population’, prohibited under IHL (Protocol, 1977: Article 51(2)).

While neither of these theories specifically advocates either violating IHL or deliberately targeting civilians, they clearly threaten to render the concept of military objective almost meaningless, since as Stephen Oeter (2007) points out, if one gives up the requirement of ‘direct contribution to military efforts’ as a basis for distinguishing between civilian and military targets, there is no longer a clear criterion of distinction. Any strategy that is based on targeting ‘population, supporting infrastructure, system essentials such the national economy, and political leadership’ threatens the principle of proportionality inasmuch it is foreseeable that launching attacks against targets that are deeply imbedded in civilian areas could result in excessive civilian casualties relative to the military (although not necessarily political) advantage attained. For this reason, the principle of proportionality requires combatants to carefully calculate the expected civilian losses when planning both broad military campaigns and specific attacks. Although large numbers of civilian casualties in and of themselves do not demonstrate either intentional targeting or an intent to violate this principle, attacks that are designed to destroy a nation’s infrastructure – including large numbers of civilian structures and support facilities – do offer strong evidence of a depraved indifference to civilian lives.

For example, Amnesty International’s (2006: 4) investigation of the 2006 Israeli–Hezbollah war in Lebanon suggest an Israeli war-fighting strategy that was calculated to punish both the Lebanese political leadership and the civilian population for the aggressive and provocative actions of Hezbollah. Over a two-day period Israel launched more than 7,000 air attacks and 2,500 naval bombardments on heavily populated towns throughout the country – including the capital city of Beirut – killing approximately 1,000 civilians. ² According to Amnesty (2006: 2), ‘The widespread destruction of apartments, houses, electricity and water services, roads, bridges, factories and ports, in addition to statements made by Israeli officials, suggests a policy of punishing both the Lebanese government and the civilian population in an effort to get them to turn against Hezbollah’. These findings were corroborated by a study conducted by Human Rights Watch (2006a: 3), which concluded that Israeli forces launched massive artillery and air attacks that demonstrated ‘limited or dubious military gain but excessive civilian cost’. The fact that Hezbollah targeted Israeli towns with indiscriminate rocket attacks – clearly violating the principle of distinction – does not release Israel from its own obligations to protect civilians under IHL.

Reckless endangerment warfare

Both Civilian Targeting and Depraved Indifference Warfare involve serious violations of IHL by states that intentionally bend, flout or ignore the rules. More subtle, and therefore perhaps more insidious, are strategies that are pursued by states who are generally committed to following the laws and customs of war, but who push the boundaries of IHL by manipulating the collateral damage exemption. ³ As Thomas Smith (2008: 145) argues, civilian casualties often flow from policy preferences in predictable ways, and therefore when states act in this manner, collateral damage becomes a calculated strategy. States engage in such behavior when they employ overwhelming force under conditions that are likely to adversely affect the civilian population on a massive scale, in violation of the principle of feasible precaution. I refer to this practice as Reckless Endangerment Warfare (REW). In criminal law, reckless endangerment is defined as acts that create a substantial risk of serious physical injury or death to another person, where there is no specific intent to cause harm, but the action demonstrated a reckless disregard for the foreseeable consequences of the actions (New York State Penal Law, Article 120). Unlike DIW, such actions are reckless and irresponsible rather than wanton and depraved.

The use of overwhelming military force against an adversary is obviously not in and of itself an issue. The laws of armed conflict were never designed to ensure a ‘fair fight’ between belligerents (Schmitt, 2007: 27), and no one expects combatants to forgo their strategic, tactical, or technological advantages. IHL provides wide latitude for combatants to exploit such advantages. IHL was not designed to shield political leaders from the consequences of war, nor are government or economic institutions that directly contribute to the war effort immune from attack. At the same time, the employment of overwhelming force even against legal targets can threaten the principles of feasible precaution and proportionality when it is widespread, devastating, and directed at infrastructure and dual-use facilities that are necessary for the well-being of the civilian population. ⁴ States engaged in REW therefore push the boundaries of legality by manipulating the collateral damage exemption to IHL.

Prior to launching any attack, commanders are required to ask two questions: is the target a lawful military objective and is there an excessive risk of disproportionate collateral damage (Downes, 2008: 36)? States can seek to comply with the letter of IHL but still manipulate its intent by expanding the definitions of military objective and military advantage in order to render civilian deaths as collateral damage rather than intentional or anticipated casualties. They often do so by broadening the definition of military advantage to include ‘degrading the resources of the adversary’, even if such resources do not directly contribute to the war effort. Thus, Lieutenant General Michael Short mused during NATO’s 1999 air war against Serbia, ‘I felt that on the first night (of the attack) all of the power should have gone off, the major bridges around Belgrade should have gone into the Danube, and the water should be cut off so the next morning the leading citizens of Belgrade would have got up and asked, “Why are we doing this?”’ (Whitney, 1999: A1).

To the extent that the sought-after effects of armed force by many Western armies are often primarily political rather than military, Western military doctrine often equates political expediency with military necessity. This leads to a reinterpretation of what constitutes a legitimate military target and what counts as ‘excessive’ in determining proportionality. However, IHL does not permit the use of military force to achieve diplomatic or political ends.

Additionally, Western doctrine often encourages attackers to declare dual-use facilities to be legal targets, without placing the operation of these facilities within any kind of specific context. For example, W Hays Parks (2007: 100) argues that the following are ‘always military objectives’ regardless of the circumstances: power sources, communications networks and equipment, industry (not only military, but also those that are geared toward export), railroads, waterways, and highways that are known or suspected enemy avenues of approach or withdrawal. British Major General and Director of British Army Legal Services APV Rogers (2004: 37) lists other examples of military objectives, including ports, airfields, main roads, tunnels and canals, television stations, and telephone networks that could be used for military communications. This approach was used extensively during NATO’s air war against Serbia; of the 802 target groups developed by the commanders during the second month of the campaign, only 40% were military facilities or air defenses. The remaining 60% were factories, infrastructure, oil installations, roads, bridges, railroads, and other dual-use facilities (Cordesman, 2001: 130).

Both Parks and Rogers are correct that dual-use facilities could be legitimate targets if such facilities directly contribute to the war effort. A bridge that is used for moving soldiers and weapons to a conflict zone, for example, is clearly a legal military target. However, the question of whether a target or series of targets is legitimate requires a contextual evaluation rather than a blanket policy that all dual-use facilities are always legal military objectives, particularly if the real purpose of the attacks is to disrupt the political and social life of population. Rogers (2004: 83–84) acknowledges this, adding that the ‘mere fact that an object … is in the list does not mean that it is necessarily a military objective. It must make an effective contribution to military action and its neutralization must offer a definite military advantage’ (emphasis added).

Even with an expanded definition of military objective, combatants are still prohibited from attacking dual-use facilities if they are indispensable to the survival of a civilian population, unless they are being used specifically for military purposes (Protocol, 1977: Article 54(2)). When military lawyers argue that ‘attacks on an enemy’s electrical generating and distribution systems are legitimate’, since ‘the nature of national power grids means the entire system becomes a military objective’ (Downes, 2008: 62) they may be technically correct in terms of defining military advantage. However as the US Air Force itself officially acknowledges, when attacking power sources, transportation networks, and telecommunications systems, distinguishing between the military and civilian aspects of these facilities is virtually impossible (Waxman, 2000: 22). Such an approach to total war is likely to produce high levels of incidental loss of civilian life, injury to civilians, and damage to civilian objects; under these conditions, it is reckless inasmuch as it does not permit commanders to take feasible precautions to avoid such casualties.

States also act recklessly, in violation of the spirit of IHL, when they manipulate the principle of proportionality by narrowing the evaluation of expected collateral damage to each specific attack rather than also making an assessment of the likely cumulative effect on civilians and civilian objects of the entire campaign. ⁵ Typically, states committed to following the laws of armed conflict evaluate the legality of targets on an individual basis. The United States and Israel, for example, regularly include judge advocates (military lawyers) in meetings to determine target selection. These lawyers are charged with making judgments regarding the legality of specific targets (Kahl, 2005; Arkin, 2007: 125). During the 2003 US invasion of Iraq, for example, the Pentagon developed a list that detailed each potential target and submitted this list to military lawyers for review. Accompanying this was another list that specified objects, buildings, and infrastructure that would be off limits to attack (Komarow, 2003: 9A). Clearly, the Air Force had made an effort to ensure that the individual targets were legal. At the same time, there was no apparent effort to consider either the collective impact or long-term consequences of what was planned as an extensive series of widespread and devastating attacks. As Judge-Advocate and Lieutenant-Colonel Joseph Holland (2004: 62) argues, there are too many variables to consider when calculating long-term or future collateral damage from a particular attack because long-term effects are too remote and therefore not the attacker’s responsibility.

This is ironic, since many Western states define military advantage, not only by the immediate effects of each individual attack, but from the ‘advantage anticipated from the attack considered as a whole and not from isolated or particular parts of the attack’ (see, for example, the statements submitted by the United States, Belgium, Canada, Federal Republic of Germany, and the United Kingdom in Henckaerts & Doswald-Beck, 2005: 329; emphasis added). Apparently only one side of the proportionality equation requires long-term analysis.

When taking this approach toward assessing the legality of targets, the calculation of expected civilian losses – a requirement under the principle of proportionality – is highly skewed. The spirit of the proportionality principle suggests that the attacker should consider any civilian losses that are foreseeable, and this includes those that are likely to occur beyond an immediate attack. This interpretation is supported by at least several international jurists; the International Criminal Tribunal for the Former Yugoslavia (2000) ruled that ‘although single attacks might not per se be illegal, their cumulative effect might render them so … In case of repeated attacks, all or most of them falling within the grey area between indisputable legality and unlawfulness, it might be warranted to conclude that the cumulative effect of such acts entails that they may not be in keeping with international law. Indeed, this pattern of military conduct may turn out to jeopardize excessively the lives and assets of civilians, contrary to the demands of humanity.’ Similarly, military lawyer Michael Schmitt (2008) contends that it ‘appears well settled that any consequences, even long-term ones, which are foreseeable and likely in light of the information reasonably available to the attacker, must be factored into the proportionality equation’.

For example, during the Gulf War of 1991, the US-led coalition targeted and destroyed or disabled most of Iraq’s electrical power facilities throughout the country. Although the attacks were ostensibly designed to eliminate Iraq’s military command and control capability, the destruction of these facilities also shut down most of the nation’s water purification and sewage treatment plants. As a result, epidemics of gastroenteritis, cholera, and typhoid broke out, leading to an estimated 100,000 civilian deaths and a doubling of the infant mortality rate (Waxman, 2000: 22). As a report issued by United Nations Under Secretary General Marti Artisaari (United Nations, 1991: Section II/8) following a field investigation concluded, air attacks by the allies ‘wrought near-apocalyptic results upon the economic infrastructure of Iraq’ and ‘the country was relegated to a pre-industrial age’. The attackers may not have intended to cause this calamity, but it was their legal and moral responsibility to consider the high probability that such an outcome would likely occur. An honest consideration of the principle of feasible precaution should have demanded it. Anything less demonstrates a reckless disregard of one’s obligation to take precautionary measures when dealing with civilian populations and assets.

When states launch massive attacks in heavily populated areas over a relatively short period of time, it is foreseeable that this will likely produce significant civilian casualties even if they only attack legitimate military objectives. Under such conditions, the principle of feasible precaution requires the commanders to consider the degree to which such actions are consistent with the obligation to minimize incidental damage. For example, NATO dropped 23,000 bombs during its 1999 air campaign against Serbia, mostly within the heavily-populated cities of Belgrade and Nis. In a campaign of this magnitude, even with the use of precision-guided munitions it is very difficult for the attackers to exercise ‘feasible precaution’ to minimize collateral damage. In fact the use of precision or ‘smart’ weapons may actually exacerbate this problem inasmuch as it encourages commanders to increase their target sets to include objects that are deeply embedded in civilian areas, greatly increasing the risk to non-combatants. Regardless of how accurate a delivery system may appear, under these conditions, small miscalculations or mishaps can have great consequences for civilians located near the intended targets. Such outcomes are both foreseeable and likely, as demonstrated by experience. ⁶ Thus, this practice is reckless – a dereliction of the attacker’s duty to perform precautionary duties.

Another practice that often produces an excessive amount of collateral damage is one that Martin Shaw (2005) labels ‘risk transfer’, whereby a belligerent minimizes the risk to its soldiers by increasing the dangers to the civilian population. Since the war in Vietnam, governments in Western democracies have been highly sensitive to casualty rates among their armed forces and have consequently gone to great lengths to minimize them. As a result, risk transfer has become a common aspect of the ‘Western way of war’ (Shaw, 2005). During NATO’s air campaign against Serbia, for example, pilots were ordered to fly at 15,000 feet to avoid any possibility of being hit by Serbian anti-aircraft fire. However at that height, the pilots could not easily evaluate the presence of civilians near their targets and as a result, many civilians were inadvertently killed during bombing raids over Serbian cities. NATO officials acknowledged as much, saying that at that altitude they would only be able to identify whether the objective was the intended one, but would be unable to tell whether civilians were in the vicinity (Amnesty International, 2000: 16).

Certainly soldiers are expected to protect themselves and minimize their risk when engaging with enemy combatants. However, the principle of feasible precaution prohibits them from trading the lives of soldiers for those of civilians except in cases where there is no other alternative. Risk-transfer inverts the rights of civilians vis-à-vis the rights of soldiers, thereby violating the basic norm that combatants are to assume greater risks than non-combatants (Smith, 2008: 147). Former Director of Army Legal Services APV Rogers (1982: 310) acknowledges that ‘in taking care to protect civilians, soldiers must accept some element of risk to themselves’. Thus, if they assess that (a) the risk of getting close enough to a target to identify it properly is too high, (b) that there is a real danger of incidental death, injury or damage to civilians or civilian objects because of lack of verification of the target, and (c) they or friendly forces are not in immediate danger if the attack is not carried out, there is no need for them to put themselves at risk to verify the target. Quite simply, the attack should not be carried out (Rogers, 2005).

At the same time, Rogers also argues that ‘the rule (of proportionality) is unclear as to what degree of care is required of a soldier and what degree of risk he must take. Everything depends on the target, the urgency of the moment, the available technology and so on.’ This, of course, provides just enough of a loophole to justify most types of risk transfer, possibly violating Article 57 of the First Protocol, which mandates that attackers shall at all times choose the course of action which may be expected to cause the least danger to civilian lives and civilian objects.

States also threaten the principle of feasible precaution when they recklessly employ weapons that have been demonstrated to produce high rates of both accidents and collateral damage. Under international law, the only weapons that are absolutely prohibited are those that: (1) are specifically banned by treaty (such as chemical or biological weapons), (2) cannot be directed at a specific target and are therefore indiscriminate (such as German V-2 or Hezbollah’s Katyusha rockets), or (3) cause superfluous injury and/or unnecessary suffering (such as blinding lasers) (Boothby, 2009: 78–81). Yet a weapon or delivery system can be legal but its use a violation of feasible precaution if its employment in conflict has a demonstrated history of excessive civilian casualties.

One example of such as weapon is cluster bombs. Cluster bombs are munitions containing up to two hundred unguided, impact-fused ‘bomblets’ that are widely disbursed (up to 100 square yards) on contact. Each bomblet combines blast and fragmentation to produce a military effect. They are generally used to attack area targets such as concentrations of military personnel, vehicles or armor (Boothby, 2005). The US Air Force argues that use of cluster munitions reduces the risks to aircrews and equipment by reducing the number of sorties required to effectively attack such military objectives (United States Air Force, 2002: 296). At the same time, according to Human Rights Watch (2012), ‘cluster munitions are inaccurate and unreliable weapons that by their very nature pose unacceptable dangers to civilians. They pose an immediate threat during conflict by randomly scattering thousands of submunitions or “bomblets” over a vast area, and they continue to take even more civilian lives and limbs long after a conflict has ended, as hundreds of submunitions may fail to explode upon impact, littering the landscape with landmine-like “duds”.’ Thus, for example, five per cent of the cluster bombs dropped by NATO during the Kosovo conflict failed to explode immediately, however up to 54 civilians later died when the duds detonated after the conflict ended (Amnesty International, 2000: 18).

Finally, those engaged in REW often attempt to exploit the collateral damage exemption by manipulating the principle of passive precaution. It has become commonplace in asymmetric conflicts for the weaker state to co-mingle military facilities and civilian objects for the purpose of discouraging their vastly stronger adversaries from launching attacks that could result in significant civilian casualties, a practice Jefferson Reynolds (2005: 4, 78–79) labels ‘concealment warfare’. Stronger states, in turn, often take advantage of the unwillingness, difficulty or inability of a weaker adversary to clearly separate military from civilian objects by shifting the blame for collateral civilian deaths to that adversary. This enables the vastly stronger attacker to avoid calculating expected collateral damage into their proportionality equation. For example, Reynolds (2005: fn. 332) holds that the ‘majority of responsibility to minimize collateral damage must remain with the adversary who governs the civilian population and controls the military’.

One can certainly understand the frustration held by military officials in these circumstances. However, this situation arises precisely because in virtually all asymmetric conflicts (that is, those in which the resources of the warring parties are vastly uneven) most if not all of the combat occurs solely on the territory of the weaker side. This not only relieves the stronger side from having to separate its own population from its military facilities, but it places a great burden on the weaker side to move its military facilities when they are located too close to civilian areas. Typically, states in peacetime do not consider the need to maintain the required separation, since they are not prepared for an armed conflict on their territory. It is not unusual for governments to locate military objectives in or near populated areas, since such facilities typically need to take advantage of infrastructure, access to workers, and support institutions. In the United States, for example, National Guard armories and weapons factories are located within densely populated cities. Once a conflict begins, it is usually not feasible to relocate such facilities. This issue is exacerbated when states broaden the definition of military targets, to include government buildings, economic installations, communications centers, bridges, and infrastructure, most of which are usually located in towns and cities that are densely populated by civilians. Clearly the adversary cannot be expected to relocate these away from population centers even if it could.

In addition, in cases where the entire battle is fought in the towns and cities of the weaker party – and when the attacker enjoys complete, uncontested control of the airspace above – the defenders have few places to safely gather their own forces. Thus, the very advancements in military technology that have given stronger states the ability to conduct up-to-the minute surveillance and launch precision attacks virtually anywhere in a particular town or city, have also encouraged the weaker parties to avoid massing in open areas (which are easy targets for attack) and to avoid fixed positions. Defenders are thus often confronted with the choice of co-mingling their military and civilian objects or leaving themselves exposed to certain attack in an open area.

This dilemma is exacerbated by the fact that in contemporary armed conflicts, the battlefield is usually not well defined and attackers often declare large parts of the territory to be conflict zones, even those within areas where both combatants and civilians live. Thus, by evoking the passive precaution principle, technologically dominant countries like the United States, Russia, Great Britain, and Israel can declare attacks on civilian areas to be legal, so long as there are military objectives somewhere in the vicinity. Certainly states that engage in concealment warfare bear considerable responsibility for deliberately putting civilians in harm’s way; however, this does not relieve the attacker of its own obligation to minimize civilian casualties even if this means altering its attack strategy. Military convenience is not the same as military necessity.

Conclusion

The primary purpose of IHL is to protect non-combatants and the facilities that sustain their endurance and survival. This goal is universally accepted by all states and international organizations, even though there are obviously some differences of interpretation and outright violations in particular cases. Any action that threatens the principle of civilian immunity violates both the spirit and intent of the laws of armed conflict. This article has sought to expand our understanding of the causes of violence against civilians by highlighting an aspect that has been woefully understudied: incidental non-combatant casualties and damage to civilian objects inflicted by states committed to the principle of civilian immunity. These violations are particularly insidious precisely because they are shrouded in the appearance of legality.

While political leaders and media organizations express horror at flagrant violations of civilian immunity by states that deliberately target non-combatants, the subtle exploitation of the collateral damage exception by law-abiding states has undergone far less scrutiny (see, for example, Downes, 2007, 2008; Eck & Hultman, 2007). In this sense, research on what scholars call ‘one-sided violence’ is limited by the narrow conception of what constitutes attacks against civilians during armed conflicts. Yet even if one focuses solely on fatalities (that is, excluding injuries and damage to facilities vital to the survival of the population), the level of collateral damage in contemporary international armed conflicts is troubling, if not alarming.

For example, an estimated 11,516 civilians were killed by US-led coalition forces from 2003 to 2008 during the invasion and occupation in Iraq (Hicks et al., 2011: Table 1). During the 34-day war between Israel and Hezbollah in Lebanon in 2006, Israeli attacks resulted in approximately 1,183 civilian fatalities (Amnesty International, 2006: 2; Human Rights Watch, 2006b). In the course of the two-month war that led to the overthrow of the Taliban in 2001, attacks by NATO forces produced between 1,000 and 1,300 civilian deaths (O’Hanlon, 2002: 55; Conetta, 2002). NATO’s two-week air war against Serbia in 1999 left between 500 and 1,500 civilians dead (Ralston, 1999; Arkin, 2000). According to the most recent estimates, about 3,600 Iraqi civilians were killed by coalition forces over the six-week Persian Gulf War of 1991 (Daponte, 1993: 65). Most recently, between 75 and 115 civilians were killed by NATO airstrikes in Libya, a campaign specifically designed to protect civilians from attack (Amnesty International, 2012). In all cases, the casualties and damage were inflicted by military organizations that were committed to the principle of civilian immunity and to the protection of non-combatants imposed by the laws and customs of war.

These practices are particularly onerous because they are usually employed by states that already hold a substantial advantage in war-fighting capability and therefore have many more options for pursuing their military goals. Quite simply, technologically dominant states are in a far stronger position to adhere to the principle of feasible precaution. This much is admitted by at least some military lawyers. Air Force Major Jefferson Reynolds (2005: 21) said, ‘A military commander from a depressed, under-developed state with little access to the resources necessary to make the most prudent command decisions cannot be held to the same standard of reasonableness as commanders from highly-advanced military states with extensive resources, information, technology, and high situational awareness of the battle space’.

Certainly the laws protecting civilians place greater limits on military organizations than they might prefer, inasmuch as the rules require them to be much more prudent in employing massive force than they would be in the absence of international law. And it is also true that a more scrupulous adherence to the principle of feasible precaution could prolong a conflict that could otherwise be ended much earlier with a less restricted use of force. Moreover, it is also the case that the use of concealment warfare by weaker adversaries in asymmetric conflicts can frustrate the militaries of stronger states and place them in a very difficult position in terms of employing their advantage while adhering to the IHL. None of this, however, is relevant in terms of meeting one’s legal and political obligations. Civilian immunity has become deeply ingrained not only in international law, but also in diplomatic practice and domestic military law. The global consensus over the protection of civilian lives and welfare is overwhelming and is firmly embedded in international treaties, customary law, domestic public opinion, media scrutiny, and international organizations.

This may mean that states will have to be more prudent in deciding whether to initiate a conflict in the first place. If a war cannot be fought according to the agreed upon restrictions, political leaders may have to choose other means of promoting their interests. This is particularly the case for states that hold an overwhelming advantage vis-à-vis their adversaries and for which the conflict is one of choice rather than survival, territorial security, or sovereignty.

Greater adherence to legal, political, and normative obligations could indeed inhibit states from prosecuting the type of conflicts that they are technologically capable of doing. Yet, while the types of practices discussed above may be convenient, they are reckless, irresponsible, and often illegal and should be discontinued.