Collateral kids: Weighing the lives of children in targeting

Abstract

The principle of proportionality under international humanitarian law prohibits an attack if the expected harm to civilian persons and objects is excessive in relation to the anticipated concrete and direct military advantage. In this article we argue that, when applying the principle of proportionality, the incidental harm to a child must be given a higher value as compared to incidental harm to an adult. This reflects the broader framework of international humanitarian law, which creates stratifications amongst different groups of civilians and provides special protection for children in times of war. This aligns with the practice of many militaries, which tends to implicitly assign a heightened worth to the lives of children due to moral and political considerations. Such reasons stem from the perceived vulnerability of children as well as their moral innocence reflecting harmlessness and blamelessness. Indeed, harm to children’s lives tends to generate a greater backlash among the community to which they belong and, as a result, a military disadvantage. We argue that the greater weight assigned to the lives of children in proportionality assessments is not simply a matter of morality or strategic calculations, but in fact a requirement from a more wholistic interpretation of international humanitarian law.

1. Introduction

Thanks to an intense intelligence operation over many months, the leader of a non-state armed group engaged in a non-international armed conflict has been located in an isolated house deep in enemy-controlled territory. Telemetry confirms the presence in the house of a phone used by the leader just a few minutes ago, and no one has since left the building. Infrared imagery from a drone indicates eight individuals in the house, but no information is available as to who the others are, although some likely would be the leader’s bodyguards. The drone is equipped with Hellfire missiles, ready to fire. Under international humanitarian law (IHL), the leader is clearly a military target that can be directly attacked. If no conclusive evidence exists to the contrary, others in the house must be presumed civilians, even if it is suspected that some might be combatants. At this point, the mission commander must assess whether the incidental harm to civilian persons and objects would be excessive to the anticipated concrete and direct military advantage. What difference, if any, would it make to this calculation to learn that six individuals in the house are children under the age of ten, gathered for a birthday party? Is the difference, if it exists, purely one of morality, or does international law demand that the incidental killing of civilian children be weighed more heavily than that of adult civilians? We argue that this difference matters not only in morals but also in how international law is understood and practiced on the battlefield.

The lynchpin of the calculation just described is the notion of proportionality, enshrined in Article 51(5)(b) of Additional Protocol I to the Geneva Conventions (Protocol I). ¹ This Article states that ‘an attack which may be expected to cause incidental loss of civilian life, injury to civilians, damage to civilian objects, or a combination thereof, which would be excessive in relation to the concrete and direct military advantage anticipated’, is to be considered disproportionate. ² On its face, the provision refers to loss of life and injury to civilians in a general manner to calculate whether such harm is excessive in relation to the direct military advantage anticipated: no specific class of civilian is explicitly given special consideration in the provision. Nothing in the travaux préparatoires suggests that this issue was raised during the negotiations of Protocol I. ³ This has generally been taken to imply that ‘[t]he formal rule of proportionality in IHL does not distinguish between different types of civilians; the value of the life of a young man or an old man is no different from the value of the life of a woman or a child’. ⁴ This is normally taken as a given, without any significant engagement with the question of whether children can or should be treated differently in proportionality calculations. ⁵ We argue to the contrary that Article 51(5)(b) should be understood in the context of the larger IHL framework, to demand that a higher value be allocated to children’s lives within the requirement of proportionality.

The general framework of IHL posits a fundamental distinction between civilian lives and those of combatants, but its protection regime also creates further stratifications among different classes of civilians that fall within its purview. Protections granted to civilians thus can vary depending on their nationality, gender, occupation, as well as age. In this article, we investigate specifically one implication of the higher safeguards provided to children under IHL. In general, the interpretation of IHL safeguards is informed by military practice, which does assign a heightened worth to lives of children. This practice reflects the perceived vulnerability of children, their presumed innocence, and lack of agency projecting an image of harmlessness and moral blamelessness. As a result, harm to children triggers more intense social and political backlash and attracts a higher scrutiny from the local population and from the international community, including by international tribunals. We show that, in their practice, militaries attribute a greater weight to harm to children in assessing collateral damage, and we critically appraise how such heightened worth attributed to children’s lives can be accommodated within the proportionality principle in IHL.

Our claim is based on an approach which examines IHL provisions within, rather than decoupled from, the moral and political contexts in which they are interpreted and applied. Law is not the only, or even necessarily the dominant, source of restrain on the behaviour of combatants in war. ⁶ Combatants are influenced by their social, economic, political, and religious affiliations, in addition to legal norms governing warfare. While militaries tend to be hierarchical and highly sectarian communities, setting their own paradigms of lawful behaviour, even the boundaries set up by such sectarian communities are porous, given that they seek validation and legitimacy from the normative world outside their core. ⁷ Where the law allows space for interpretation, as in the case of determining the excessiveness of an attack, commanders are likely to be influenced by common sense, ethics, and morality. ⁸ The adoption of the Martens Clause, which first appeared in the preamble to the 1899 Hague Convention II with Respect to the Laws and Customs of War on Land, reflects the need to use normative frameworks not exclusively within the confines of the law. ⁹

The manner in which militaries provide heightened attention to protecting the lives of children in proportionality assessments suggests a resort to morality and public conscience. In such cases, commanders appear to act on the basis of an ‘instinctive bias’ favouring protection of children. ¹⁰ For example, some military manuals provide that a pattern of life, indicating the potential for harm to children in the target area, creates a ‘high risk’ situation. ¹¹ Furthermore, the tools and processes that support commanders in making their determinations of excessiveness require ‘a thorough study of the total operating environment’ including political and cultural factors. ¹² This indicates that cultural approaches towards children and the social and political backlash that may arise from the deaths of children as collateral damage can affect how militaries conduct their proportionality calculations. Thus, how children are taken into account in proportionality calculations reflects not only strict adherence to the law but also ethical considerations and the factual appreciation of the impact of an attack. ¹³

In our claim that lives of children are considered to weigh more than those of other civilians in proportionality calculations, two distinct elements can be differentiated, addressed in successive parts of this article. The first element involves the process of identifying an individual as a child. This speaks both to the justification of treating children differently than adults in proportionality calculations and to the clarity of the concept of the child in this context, including the material possibility of carrying out such an identification in the messy reality of the battlefield. Second, the calculation requires putting into relation elements often characterized as incommensurable and incomparable: the lives of children and the anticipated military advantage. ¹⁴ We argue that even if there is a commensurability and comparability gap between military advantage and collateral damage, differentiating between harm to children and harm to adult civilians can inform proportionality calculations.

At the outset, we note that our focus is on civilian collateral damage, with a specific focus on peaceful civilian children, rather than on discussions surrounding the targetability of children taking direct part in hostilities. ¹⁵ We acknowledge that this separation between protected civilian children and children taking direct part in hostilities is complicated by lack of clarity, both conceptually and practically, in the definition of the child and the blurred lines delineating when a child can be considered directly targetable. This highlights the inevitability of how IHL practice needs to coexist with a degree of uncertainty that can be common on the battlefield, especially in dynamic targeting. ¹⁶

In carrying out this analysis, we rely on descriptions of state practice that have been collated and commented upon in academic discussions in law and other disciplines. In addition, we carried out ten semi-directed interviews with current or former members of the armed forces of Canada, India, Israel, The Netherlands, and Sri Lanka. These interviews were carried out on the basis of anonymity, and informants are identified here using aliases. We approached serving or retired members of the militaries of several other countries, including the United States, France, the United Kingdom, and Australia. However, we encountered a generalized reluctance to meaningfully engage with the issue, whether dictated by the chain of command or reflecting more personal unease of would-be informants. For us, perhaps paradoxically, this reluctance strengthened the sense that this is an issue that needs a fuller exploration.

2. Looking for children on the battlefield

The way a society mistreats children is often invoked as a clear signal of its barbarity, justifying external criticism or delegitimizing whatever other actions it is taking. ¹⁷ To some degree, these judgements on how a child is to be treated reflect biases about what an ideal childhood is meant to look like. To add to this, not all children’s lives are measured with the same scale, the precarity of some lives appearing to be accepted more easily than those of others. The idea of childhood overlaps with other factors such as race, nationality, religion, poverty, etc. which may impose intersecting barriers on how some children get to access certain rights and protections attached to childhood. It is taken as a given that the category of ‘children’ is not monolithic. However, the fact remains that media images of dead or wounded children in war are some of the most powerful indictments of the justness of a military campaign. ¹⁸ There is a consistent pattern, in the media, in governmental reporting, and in academic discussions of specific cases of targeting in war, of listing children (sometimes women as well) as a distinct subcategory of civilian casualties. ¹⁹ Even in the internal, informal discussions of targeting within the military, if children are present in the strike zone, that fact will most often be noted. We found in our interviews that a striking point of convergence in the otherwise varied practice of targeting was that no one seems to be indifferent to the presence of children among collateral victims of an attack.

To note that no one is indifferent does not necessarily imply a normative claim. Indeed, the impact of the known presence of children among anticipated collateral victims of a strike appears to vary widely. At one end of the spectrum, especially in deliberate (as opposed to dynamic) targeting in counter-insurgency operations, the presence of children in the strike zone can block the attack altogether. Thus, a military lawyer indicated that, in the Israel Defence Force, ‘[f]or targeted killings, there was a basic rule that you cannot kill children’, although the intensity and nature of the conflict that started in and around Gaza in October 2023 have altered proportionality assessments, including the acceptability of collateral harm to children. ²⁰ At the other end of the spectrum, the leaked video of a US strike in Baghdad on 12 July 2007 in which a four-year-old girl and an eight-year-old boy were killed by helicopter machine gun fire suggests that no duty of restraint flows from the presence of children. The transcript includes an exchange between the pilots of the two helicopters commenting ‘Well, it’s their fault for bringing their kids into a battle’ ‘That’s right’. ²¹

Beyond the obvious and essential differences in actors and contexts, the range of reactions attached to the presence of children reflects the apparent silence of IHL on the matter. Neither the legal definition of civilian nor the codification of the principle of proportionality in Protocol I, the only treaty in which these concepts are articulated, includes any indication that children are a distinct class of civilians for proportionality assessments. Most legal experts seem to take it for granted that this silence is an affirmation that ‘a life is a life’, that all civilians should weigh the same in proportionality calculations, and that no difference should be made between the collateral destruction of a prison and a kindergarten. ²² Military doctrine, in the form of manuals or other directives given to troops, likewise remains silent on the question of the possible impact of the presence of children among the anticipated collateral victims of a strike. As will be elaborated in Section 2.1, those who argue against a hierarchy of valuing lives present substantive moral and ethical claims to defend their positions. These moral considerations and the fears of comparing the value of one life against another may explain the silence that IHL maintains as to whether to treat children differently than adults in proportionality calculations.

The absence of evidence of a difference, however, does not necessarily imply evidence of absence of a difference. The law’s silence vis-à-vis children and targeting, replicated in military doctrine, reflects a broader ‘black box’ approach to proportionality. The net result is that law fails to guide behaviour, even in ways that perhaps could fulfil humanitarian imperatives while respecting military necessity. As bemoaned by Daniel Reisner, former head of international law in the Israel Defence Force:

Proportionality does not tell us what to include in the calculation, what is the equation and what is the exchange rate … How do you count the death of children? Does one child equal one dead grownup, or does he equal five grownups? As a lawyer, I need numbers to work with. I need thresholds in order to instruct the soldiers. Any number could become a useful benchmark. But when the ground of the law is shaking, I am also unstable. ²³

There is an evident disconnect between, on the one hand, a consensus that the presence of children on the battlefield matters and, on the other hand, the law’s apparent silence on the issue. In part, it may signal a gap between moral intuition and legal standards, begging the question of why children should be taken to belong to a distinctive class of civilians in targeting decisions. It may also reflect the particular dynamics of armed hostilities and their transformative impact on notions such as childhood.

2.1 Why should children be considered differently?

In the narrative construction of childhood in war, three attributes contribute to a heightened sense that they call for special protection against the harmful impact of hostilities on children: vulnerability, innocence, and the capacity to evoke pity. The ideas of victimization and vulnerability are closely intertwined: sick, old, or very young people correspond the most obviously to the notion of ‘ideal victims’. ²⁴ Attributing a heightened value to specific victims on account of their vulnerability offers a seemingly simple and coherent framework to navigate complex issues of political and moral victimhood. ²⁵ While we do not uncritically embrace the notion of an ‘ideal victim’, we recognize that these social constructs of victimhood can influence how persons are treated during wartime including within legal provisions regulating warfare.

The degree of protection provided by IHL is often directly related to how weak or vulnerable the individuals belonging to a group are perceived to be. Vulnerability as a basis for protection under IHL can be related to an individual’s capacity to harm the enemy. Likewise, vulnerability in the context of war can pertain to the lack of safety derived from not bearing arms. Combatants are empowered by their right to bear weapons and target the enemy. Therefore, they are provided limited legal protection unless they are hors-de-combat, as a result of being wounded, sick or shipwrecked, or when they are captured by a hostile party (Article 41, Protocol I). Civilians, in general, are more vulnerable as compared to combatants, given that they do not bear arms (Articles 50 and 51(3), Protocol I). Conversely, the point at which a civilian participates in hostilities and thereby poses a threat or causes harm to the enemy corresponds to the moment when the protective framework of IHL ceases to shield them against direct attacks.

Against this backdrop, children’s deaths incite more moral indignation compared to other civilian deaths due to their perceived vulnerability. In both legal and moral regimes, vulnerability is an attribute of childhood. ²⁶ Within the class of civilians, children explicitly are provided special protections under IHL given their heightened vulnerability (Article 77, Protocol I). For children, lack of safety is compounded by their dependency on adults as well as their relative inability to deal with unfolding and oftentimes life-threatening situations. While women also have been considered a vulnerable category in IHL, the idea that women should be seen as lacking in agency and dependent on others has been debunked as one rooted in patriarchy. ²⁷ On the contrary, the vulnerability and dependency of children are often considered to be relevant factors in their victim status. ²⁸ Connecting children’s vulnerability to their immaturity is often taken for granted, as can be seen in the preamble to the Convention on the Rights of the Child which invites us to bear in mind that ‘the child, by reason of his physical and mental immaturity, needs special safeguards and care, including appropriate legal protection, before as well as after birth’. ²⁹ While adults who are wounded or sick as well as older persons become vulnerable and dependent due to their physical incapacity, they remain mentally mature. Children, on the other hand, are perceived as being vulnerable on both counts of being physically weak and mentally immature. Studies of blast injuries indicate that not all bodies are equally vulnerable when a bomb is dropped: children tend to be injured differently and more severely. ³⁰ In war, the universal assumption of children’s vulnerability aligns with the fact that commanders commonly appear more willing to spare the lives of children as compared to the lives of other civilians. In the context of targeting, the greater vulnerability of children implies that they are at greater risk than other civilians.

While the conventional meaning of ‘innocence’ often relates to the internal morality of a person, in war, combatants are not able to ascertain the personal moral guilt of each person they target by peering into their subjective state of mind. As such, moral guilt typically is not a realistic legal marker to distinguish between legitimate and illegitimate targets in the planning of an attack. ³¹ Therefore, IHL attributes innocence not based on the moral guilt of an individual but on the threat they visibly represent. Non-combatants are deemed innocent because they do not participate directly in the war effort and cannot injure, whereas combatants are deemed to have this capacity. ³² In other words, in IHL, the notion of civilian is imbued with innocence to the extent that the notion of a ‘guilty civilian appears oxymoronic’. ³³

Children’s presumed lack of agency makes it challenging to attribute moral guilt to them. Instead, their lack of moral agency and their resulting ‘pureness’ make it easier to portray them as blameless victims. ³⁴ The term ‘innocence’ as it relates to childhood has religious origins in which lack of moral maturity connotes an incapacity to decide between good or evil. ³⁵ As David Mastey notes, ‘[t]his vision of the innocent child is a vital facet of the complicated (and frequently contested) discourse on childhood in the Western world’. ³⁶ There are echoes of these conceptions of childhood innocence in IHL. In general, IHL reflects the notion that certain activities that constitute modern warfare are antithetical to childhood, leading to prohibit children from being made direct participants in war. ³⁷ Thus, as per Article 77 of Protocol I, children under 15 must not be enlisted in the armed forces and all possible measures must be taken in practice to ensure that they do not take part in hostilities. Without necessarily endorsing this image of the child as endowed with minimal agency, we acknowledge that this image of children exists and has shaped how children’s victimhood in war is framed.

Perceptions of the agency of victims influence the degree to which they evoke pity or sympathy. ³⁸ A reason why certain issues gain more attention than others is how easily they can be framed ‘so as to resonate with policymakers and publics’. ³⁹ Literature on victimization points out that the image of victim, rather than being objective, is constructed at the individual and social levels. ⁴⁰ Rebecca DiBennardo highlights the media’s significant impact on shaping public opinion of victims and offenders, which in turn can influence how the law is applied in practice. ⁴¹ For example, child victims tend to receive more media attention partly due to their perceived blamelessness, a quality not always equally attributed to adult victims. ⁴² This is true also in the case of international atrocities where the victimization of children has been instrumentalized by various actors with the intention of reaching public sensibilities. ⁴³

Militaries are cognizant of the backlash and moral indignation that harm to the lives of children can generate, and indeed appear influenced by it when engaging in the assessment of the proportionality of an attack. Likewise, US military manuals admit that being perceived as using force disproportionately can adversely affect support from their military partners and the civilian population. ⁴⁴ Therefore, even where targeting operations may be lawful, commanders are expected to minimize any negative impact to operational or strategic objectives, which means that additional restraints may be placed beyond the law’s requirements in order ‘to avoid collateral damage that might bolster the local population’s support for the enemy, decrease its support of US involvement, or degrade the US population’s support for the operation’. ⁴⁵ US military doctrine further notes that failure to observe such measures could result in ‘global criticism’ of US military efforts. ⁴⁶ The significance of this dimension of proportionality is captured in a Tactical Directive issued in 2009 by General Stanley McChrystal, commander of NATO forces in Afghanistan: ‘We must avoid the trap of winning tactical victories – but suffering strategic defeats – by causing civilian casualties or excessive damage thus alienating the people.’ ⁴⁷ Such considerations have led militaries to take special measures to protect the lives of children even where the law is unclear as to whether they are required to do so. Thus, on a different but related issue, the 2001 Canadian law of armed conflict manual requires that even children under 15 who do take part in hostilities remain protected. ⁴⁸

All these elements suggest that, when it comes to collateral harm to children, military practice does reflect ethical and political considerations. In assessing collateral damage and calculating proportionality, harm to the lives of children seems to hold a high value in military narratives. What reality these narratives correspond to, and to what extent they track legal categories under IHL, are aspects that must in turn be unpacked.

2.2 The elusive figure of the child in war

A finding that children occupy a distinctive place on the battlefield in the minds of soldiers making targeting decisions raises at least two, interconnected questions. First, does childhood in such a context correspond to a concept that is well identified and agreed upon among different armed forces of the world? Second, if children are indeed a different kind of civilian, how does a commander go about determining whether children are present within a strike zone? As we will see presently, there floats a haze around the idea of the child in the context of collateral damage estimates, and the practices and tools used to determine the presence of children will in some ways produce a war-specific concept of childhood.

At dawn on 21 February 2010, a combined US and Afghan force was preparing an action near the village of Khod, Shahidi Assas District, Uruzgan province. ⁴⁹ Intercepted radio chatter suggested that Taliban fighters had been called to assemble to attack the coalition force. Shortly thereafter, three pickup trucks carrying about 30 individuals were spotted approximately a dozen kilometers away by a US MQ-1 Predator drone. The drone was flown by a pilot and a sensor operator sitting in Creech Air Force Base, Nevada. Data collected by the drone sensors were analysed by two screeners, operating from Hurlburt Field, Florida. The force commander and a Joint Tactical Air Controller on the ground in Afghanistan, the drone crew in Nevada, and the screeners in Florida were all in communication, with several others monitoring the operation. Initial surveillance led the screeners to identify a possible weapon with ‘military-age males’. A few minutes later, the screeners reported that a child or two children were near the vehicles:

Operator: ‘bullsh*t, where?’

Pilot: ‘at least one child … Really? Assisting the [military age male], uh, that means he’s guilty//Yeah review that (expletive deleted) … why didn’t he say possible child, why are they so quick to call (expletive deleted) kids but not to call (expletive deleted) a rifle.’

Operator: ‘I really doubt that children call, man I really (expletive deleted) hate that.’ ⁵⁰

There ensued a discussion in which the commander asked whether they were talking ‘toddlers or teens’, and the screeners changed their assessment to ‘adolescents’:

Controller: ‘Is adolescent a different call then [sic] child or children?’

Second Screener: ‘I think it varies from Screener to Screener. One Screener may be more comfortable with calling out adolescent. It is very difficult to tell. I personally believe an adolescent is a child, an adolescent being a non-hostile person’ ‘[I] believe an adolescent [is] 9-14 years-old.’

First Screener: ‘[I] believe an adolescent to be 7-13 years-old … in a war situation they’re considered dangerous.

Some time later, the order to engage is passed from the drone team to two US helicopters that were called in support, with no mention of possible children. A number of missiles and rockets are fired at the three pickup trucks, resulting in the death of between 16 and 23 individuals. Twelve individuals were also injured, including children aged five, 12, and 14. A military investigation concluded that everyone in these vehicles was a civilian.

This incident is not necessarily given as representative of the way militaries generally consider children in collateral damage estimates; indeed, the investigation that followed concluded that several standards of the US armed forces had not been respected. The Uruzgan incident does signal, however, that the difficulty in considering the weight of children on the battlefield reflects both an uncertainty around the concept of a child in this context and the fragmented and conflicted ways of seeing the battlefield. As mentioned in the introduction, it is important to reiterate that the focus of the present inquiry is the peaceful child civilian, as children who are combatants or who are civilians directly participating in hostilities can be directly targeted by an attack and would not be included in the estimate of collateral harm for proportionality calculations. ⁵¹

There is a stark contrast between the hesitant description of what is an ‘adolescent’ by the protagonists of the Uruzgan strike, offering a range from seven to 14 years old, and the accepted definition of the child under public international law. In IHL, parties to both international and non-international armed conflicts have a general obligation to protect children under the age of 15. ⁵² There are also a number of provisions in the Geneva Conventions and Protocols guaranteeing special protection for those under the age of 18. ⁵³ In international criminal law, although Article 26 of the Rome Statute limits the jurisdiction of the International Criminal Court to those who were 18 at the time of the commission of the crime, the child in the context of the crime of forced recruitment and use has been defined using the standard from IHL to correspond to a person aged 15 or younger. ⁵⁴ In international human rights law, the 1989 Convention on the Rights of the Child, after a protracted debate on the question, favoured defining the child as a person below the age of 18 years old, although Article 1 opens a door to national law setting a younger age of adulthood. In the context of recruitment and direct participation in armed conflict, the Convention on the Rights of the Child in Article 38 reflects the IHL norm of 15 years old, but its Optional Protocol on Children in Armed Conflict raised the standard to 18. ⁵⁵ The conclusion to draw from this overview is that the traditional definition of a child in IHL referred to anyone under the age of 15, but that there has been a clear evolution under the influence of human rights law to raise that bar to 18 years old. ⁵⁶ It therefore seems justified for the ICRC, in its study on customary IHL, to define a child mainly as anyone under the age of 18. ⁵⁷

What are the implications of defining a child as anyone younger than 18 years old for the suggestion we have made that giving more weight to children in collateral damage estimates is consistent with the proportionality principle in IHL? The law generally operates by way of the identification of analytical categories, to which factual elements are associated. While the normative and the factual elements of legal reasoning are typically represented as autonomous (rules are rules, and facts are facts), a compelling argument can be made that they are to a significant extent mutually constitutive. In Clifford Geertz’ pithy formula, ‘the legal representation of fact is normative from the start’, in the sense that a fact like the presence of a child does not arise, as it were, by itself; it corresponds to a social construction that associates a norm defining children to an observable situation, an operation in which the norm of what is a child is partly shaped by what is observed, and the perception of what is observed is partly shaped by the norm defining children. ⁵⁸ In that sense, childhood is appropriately described as a ‘technology of governance’ to situate individuals within a social framework. ⁵⁹ What all of this suggests for the application of the proportionality principle to targeting involving children is that the norm defining a child will necessarily be modulated by the context of collateral damage estimates. Concretely, IHL may define a child as anyone under 18, or even 15, for the purpose of the prohibition of the recruitment and use of child soldiers, or for the protection of detained children, but this does not transpose to targeting without incorporating the specific reality of that operation. Thus, while certain means may be available or necessary to verify the age of new recruits or detainees, including verifying identification papers and interviewing possible children and their relatives, the means available to identify the presence of children in the context of collateral damage estimates typically will be reduced to visual clues. This is indeed the way that militaries seem to approach the question of children in the target zone. As put by Advocate ‘Giti’ of the Israel Defence Force: ‘In the IDF, the question asked is “does that look like a child?”, so that 16–17-year-olds would be seen as looking like adults, while 11-12-year-olds would be seen as non-combatants.’ ⁶⁰

The fact that assessing the presence of children in mapping collateral damage is essentially a visual exercise does not set it apart from many other rules of IHL. This is especially so for the law of armed conflict dimension of IHL, regulating the conduct of hostilities, in which visibility often plays a central role. An important example is the key function of uniforms as a device to remotely identify those who qualify as combatants and to which party to the conflict they belong. ⁶¹ This also reflects the requirement under Article 4(A)(2) of the Third Geneva Convention that, in order to benefit from combatant privilege, members of irregular armed groups must have a distinctive sign ‘recognizable at a distance’ in addition to ‘carrying arms openly’. ⁶² These norms underscore the visual nature of the general obligation for combatants ‘to distinguish themselves from the civilian population’ (Article 44(3), Protocol I). IHL addresses only one side of the distinction coin and imposes no corresponding obligation on civilians to distinguish themselves from combatants. It goes without saying that children have no duty to distinguish themselves from combatants, or from adult civilians. How, then, is a belligerent supposed to identify the presence of children on the battlefield, to factor this element into proportionality calculations? There is no question that it is often extremely difficult to identify children on the battlefield. Indeed, this was an objection formulated to us by an Indian officer, who offered that ‘it is not possible to look out for kids, especially when bullets are flying around’. ⁶³ In some circumstances, such as fluid and intense ground combat operations, it may indeed not be possible to identify children, or apply the principle of proportionality in any sophisticated manner. This nevertheless leaves a wide range of situations in which attacks are decided with a degree of deliberation that opens a door to the identification of children as part of collateral damage estimates.

One significant practice that reflects an age-based approach to determining the category to which individuals on the battlefield belong is the US label of ‘military-age male’. The US military claims that this expression does not correspond to a legal category, and more contentiously that it has no normative implication, but that it is merely a practical short-hand categorization of some individuals. ⁶⁴ Indeed, the 2023 revisions to the US Law of War Manual clarifies that:

although an individual’s age and gender may be relevant in determining whether a person is a military objective, the mere fact that a person is a military-aged male with no additional information would be speculative and insufficient to determine that person to be a military objective. ⁶⁵

The claim that this is not a legal concept should not be taken to imply that there is no consequence to using the expression. For example, the New York Times in 2012 reported that the Obama administration had adopted a method for counting casualties from drone strikes according to which all ‘military-age males’ in a strike zone were counted as combatants. ⁶⁶ The practice in Afghanistan was described by one officer involved in targeting in the following way:

If there is no evidence that proves a person killed in a strike was either not a MAM, or was a MAM but not an unlawful enemy combatant, then there is no question. They label them EKIA [enemy killed in action]. ⁶⁷

The reasoning is essentially combatancy by association, based on the assumption that non-state armed groups or terrorist organizations do not mingle casually with civilians, so that if a ‘military-age male’ stands with a verified combatant, then they are presumed to be one as well. ⁶⁸ The momentous significance of this approach becomes clear when considered in light of the fact that, in some contexts, nearly 90 per cent of ‘military-age males’ killed by strikes were incidental ‘enemy killed in action’ rather than ‘jackpots’ – the intended, identified target. ⁶⁹

What exactly is ‘military age’ in this context? How does this concept interact with the legal definition of childhood in international law? Consistent with a view of the concept of ‘military-age male’ as a technique to assess data rather than as a norm, there appears to be no formal or informal policy in the US armed forces establishing a specific ‘military age’. While some take 16 years of age as a baseline, this seems both subjective and fluid. Sarah Shoker, in her in-depth analysis of the concept, reports how one informant addressed the issue of age at the heart this practice:

The idea of the military-age male is a cultural one. It’s completely cultural … there are cultures out there that when a male is considered a man, they’re given a rifle and told to go fight. A MAM is when a boy no longer remains a boy, when he can’t hang out with women and sit with them. He has to go and stay with the men. When this happens over there [Afghanistan], it’s around 12 years old. I don’t remember how old it was in Iraq, but this idea is widespread across Middle Eastern culture. ⁷⁰

What seems to be suggested by this statement and the more general US practice is that military age is a factor constituted by its environment. It is an age that reflects both the general context corresponding to a specific culture as imagined or perceived by those making targeting decisions, and the more immediate surroundings of a given individual. Biological age certainly does seem to play a role but, as earlier illustrated in the discussion among screeners during the Uruzgan strike, it seems to set parameters that can diverge widely from the IHL standard of children being 15 years old or younger (one screener mentioning seven years old as the lower threshold of ‘dangerous adolescents’). More generally, it appears to create a factual age-based presumption of combatancy that is inconsistent with IHL. Unsurprisingly, the concept of the ‘military-age male’ has not been widely endorsed in the practice of other state armed forces, although some non-US informants will admit to the fact that they rely on some of its underlying elements in their targeting decisions. ⁷¹ Significantly for our purpose, the notion of ‘military-age male’, for all its flaws, comes to reinforce the idea that age matters in targeting assessments and that it is feasible to carry out age assessments on the battlefield. Of course, in the US practice just described, this does not go towards balancing against military advantage in proportionality calculations but, on the opposite, it is a technique used to exclude ‘military-age males’ from the civilian collateral casualties category. We argue, on the opposite, that non-military age persons, a.k.a. children, should be identified and more fully protected in proportionality assessments.

It is important to note, when evoking a notion like the ‘military-age male’, the obviously gendered construction of combatancy that it embodies. There is no corresponding ‘military-age female’ in US practice. This echoes a more general critique of IHL first articulated three decades ago by feminist scholars like Judith Gardam, arguing that the principle of distinction between civilian and combatant is in its essence gendered, opposing the construct of the male warrior to ‘his essential foil, the weak and powerless “feminine” civilian in need of protection’. ⁷² What the ‘military-age male’ adds to this critical reading of IHL is an age dimension, in a manner that does not detract from the implicitly gendered construction of combatancy. To put it differently, gender and age are two vectors of identity that interact in constituting the figure of the combatant on the battlefield, the corresponding reflection of which is the civilian. If the ‘military-age male’ embodies might and threat, innocence and vulnerability converge in the female child. ⁷³

Beyond the use of a construct like the ‘military-age male’, the pervasive penetration of technology on the battlefield has had an impact of the belligerents’ ability to ‘see’ children when assessing the anticipated collateral harm caused by a strike. The fog of war, whether or not deliberately created by the enemy, has always posed enormous obstacles to the establishment of collateral damage estimates in proportionality calculations. Very often, as was the case for the Uruzgan strike, attacks result in greater civilian harm than anticipated because information relied upon in deciding to authorize fires was inaccurate or incomplete. It is important to note that, in such situations, the attack would not be disproportionate, if better or more accurate information was not reasonably available to the person making the proportionality assessment. ⁷⁴ The transformative effect of technology in general, and of drones in particular, is said to ‘make the terrain of war hyper-visible’. ⁷⁵ This was linked directly to the principles of distinction and proportionality by John Brennan, then a high-ranking official in the US Government:

Targeted strikes conform to the principles of distinction, the idea that only military objectives may be intentionally targeted and that civilians are protected from being intentionally targeted. With the unprecedented ability of remotely piloted aircraft to precisely target a military objective while minimizing collateral damage, one could argue that never before has there been a weapon that allows us to distinguish more effectively between an al-Qaida terrorist and innocent civilians.

Targeted strikes conform to the principle of proportionality, the notion that the anticipated collateral damage of an action cannot be excessive in relation to the anticipated military advantage. By targeting an individual terrorist or small numbers of terrorists with ordnance that can be adapted to avoid harming others in the immediate vicinity, it is hard to imagine a tool that can better minimize the risk to civilians than remotely piloted aircraft. ⁷⁶

The point then made by Brennan related to civilians as an undifferentiated group, but one may claim that any increased visibility on the battlefield allowing to better distinguish civilians from combatants, by the same token, increases the ability to distinguish between adults and children among civilians.

Sharper conceptual clarity of the notion of childhood and increased visibility of the battlefield suggest the possibility of more effectively identifying the presence of children in strike zones and, in turn, factoring any anticipated harm to these children in a specific manner for proportionality calculations. That said, to return to a point made earlier, these two vectors do not exist in isolation one from the other. As aptly put by Christiane Wilke, ‘seeing is not simply a biological process but a social and situated activity’. ⁷⁷ Although we typically do not formulate it as such, the action of seeing always implies asking the question ‘what are we looking at?’. The process of visibilizing occurs through a socio-cultural prism that is produced collectively, collaboratively, and discursively. ⁷⁸ Placing individuals in categories like combatants, civilians, and children is all at once a factual claim, a cultural claim, and a normative claim. While there is no explicit legal norm requiring parties to distinguish between children and adult civilians in proportionality calculation, technology does make it a more practical endeavour. This is in addition to the fact that moral and ethical considerations also allow for such an interpretation of the proportionality principle without contradicting the foundational framework of IHL.

3. Weighing the lives of children: Incommensurability meets incomparability

Having noted the moral and political factors that influence combat operations which tend to place a higher weight on children’s lives, in this section we seek to demonstrate that the principle of proportionality as set out in IHL can also accommodate an elevated consideration of children’s lives. It is our claim that such an interpretation of excessiveness would not contradict the purpose of the proportionality principle, which is the protection of civilian populations.

The principle of proportionality lies at the heart of the tension that frames the humanitarian project, mediating between a desire to shield civilians from harm caused by war and the accepted necessity of effectively pursuing the aims of war. This mediation, by way of a balancing exercise, implies both a measurement (of civilian harm and military advantage) and a comparison, raising possible obstacles of incommensurability and incomparability. ⁷⁹ The ICRC Commentary to Additional Protocol I states that the importance of Article 51 rests on the fact that it demands that ‘innocent civilians be kept outside hostilities as far as possible’. ⁸⁰ The proportionality principle acknowledges that harm to civilians can at times become inevitable in war, marking the limit of the overarching principle in IHL that civilians cannot be directly targeted. The purpose of the proportionality principle is not to give belligerents a carte blanche to harm civilians. Rather, it provides a relative safety net for civilians in conflict zones by placing limits on such harm. Article 51 of Protocol I thus stands in contrast to Article 15 of the 1863 Lieber Code, which held that ‘[m]ilitary necessity admits of all direct destruction of life or limb of armed enemies, and of other persons whose destruction is incidentally unavoidable in the armed contests of the war’. ⁸¹ The Lieber Code did not devise a principle of limitation to incidental and unavoidable harm to non-combatants; the notion of ‘excessiveness’ in Article 51, on the contrary, is meant for this purpose. ⁸²

To measure the ‘excessiveness’ of an attack, values need to be attributed to expected civilian harm as well as to the anticipated concrete and direct military advantage. However, Article 51 does not offer a scientific method or a set of clear guidelines on how to measure and compare these two different categories. ⁸³ As noted in the discussion of how proportionality could be applied by autonomous weapon systems, no ‘objective’ algorithm is available that would arithmetically compare such apples and oranges. ⁸⁴ Indeed, the haziness of the concept in IHL is captured in the fact that Protocol I does not include the word ‘proportionality’, but instead prohibits attacks that are ‘excessive’. ⁸⁵ While ‘excessiveness’ is not easily measurable, in military practice, methodologies have been devised in order to tabulate collateral risk assessments. For example, in US- and EU-led targeting operations, the expected collateral damage in an attack is estimated by way of a Joint Methodology for Estimating Collateral Damage for Conventional Weapons, Precision, Unguided, and Cluster. ⁸⁶

Although moral and philosophical arguments can be made against quantifying the worth of life, proportionality calculations necessarily attribute value to an anticipated military advantage and compare it against expected civilian harm. ⁸⁷ Accordingly, decisions on proportionality entail conducting policy decisions on how military targets and civilian harm are to be valued through calculations. This determination of excessiveness yields itself, at least to some extent, to a subjective evaluation, if not at the level of the commander, then at a national military policy level. ⁸⁸ This is offset by the embrace of fixed methodologies relying on multilevel operations administered in a bureaucratic fashion, corresponding to a routinization of procedures in which a heavy reliance on neutral-sounding acronyms results in what Jones describes as the ‘adiaphorization of killing’. ⁸⁹ Our claim that IHL can accommodate a proportionality calculation that allocates more weight to children’s lives is an argument about the measurement of one variable of this exercise.

The turn to the (implied) language of proportionality in IHL standards is an attempt to grapple with the messy, disparate, and fluid effect of an attack that affects the civilian population. It belongs to the broader phenomenon of the ‘mathematization of life’, in which the language of the most exact science is invoked to dress in objective language morally fraught decisions. ⁹⁰ One does hear echoes of the semantic turn to science in assessing the excessiveness of an attack in some comments by US military lawyers that they can apply IHL ‘with almost mathematical precision’ by way of processes that are ‘very scientific and involve a lot of math’. ⁹¹ Much criticism can be levelled at a vision of any measurement as the simple overlaying of a scale onto a pre-existing set of facts, to suggest instead that it corresponds to the forcible collation of otherwise disparate elements into a created set. ⁹² Frederick Schauer captures the complexity of the phenomenon with the example of the famous statement by William Blackstone that ‘better that ten guilty persons escape, than the one innocent suffer’, contrasting it with the same sentiment expressed by Benjamin Franklin but with a ratio of one hundred to one. ⁹³ It is evident that this exercise involves much more than simply the counting of the innocent and the guilty, to instead amount to the construction of the principle of the rule of law, the value of liberty, the purpose of justice, and the legitimacy of the criminal law. When measurement systems like the collateral damage estimate methodology are propagated through a community of experts, the political and ideological choices that underpin them become invisible – even to the experts themselves – so that these measurement systems become simply natural. ⁹⁴ Thus one may feel that civilian harm is comparatively easier to quantify than military advantage, even concrete and direct, in the sense that the civilian dead and injured can be counted. But being able to count the dead does not tell us anything about their weight. We are sent back to Blackstone and Franklin to wonder what the appropriate ratio is, a question we have seen to be part of the process of constituting not only what is being counted but also the world into which it is situated. The emergence of AI-driven software to automatize proportionality calculations, seemingly an embodiment of that unreachable algorithm, far from resolving this dilemma only hides more effectively the choices that underpin it. ⁹⁵

All this could be taken to be profound, or confused, but in any case not a terribly helpful attempt to more effectively protect the victims of war, and in particular children. The point of underscoring the constructed nature of the building blocks of proportionality calculations in IHL is not to resolve the problem that it requires discrete judgments that remain unstructured and opaque despite a semantic turn to science. Rather, more modestly, it seeks to take advantage of the fact that the proportionality assessment requires belligerents to embrace some form of methodology. Within that methodology, civilian harm as a concept becomes a cypher of commensurability, forcing into a single class elements that are disparate. To ascribe a greater weight to children does not disrupt a system of values that is objective, balanced, and transparent but rather signals that, even if the parameters of proportionality are ‘always relative, situational and immanent’, the fact that children are among collateral victims per se augments the collateral damage estimate. ⁹⁶ Values may be unfixed and yet have relative values.

In state targeting practice, differentiated values are attributed in proportionality assessments to each target when measuring the military advantage gained in an attack. Accordingly, certain targets require special care because a failed or improper engagement can lead to major adverse consequences. ⁹⁷ Such targets are often characterized as ‘sensitive’ and targeting them may require co-ordination with and approval from higher authorities. ⁹⁸ Especially in instances where the commander determines that the effects from striking the target may have adverse political ramifications, national level approval is required. ⁹⁹ The determination of possible political ramifications of an attack, in order to decide whether to carry it out or not, requires an inherently subjective and qualitative assessment of the value of its target. Value-based differentiation applies to how damage to civilian objects is calculated in proportionality assessments. To calculate whether the anticipated military advantage would be excessive, a value or weight must be assigned to the incidental harm against which it would be balanced. ¹⁰⁰ In assessing such weight or value, parties to a conflict can take account of the cultural value provided to the civilian object. ¹⁰¹ In general, US forces divide civilian objects into two categories based on their sensitivity when conducting a collateral damage assessment. ¹⁰² The manner in which the US military manual frames these categories demonstrates that they were formulated to reflect subsets of civilian objects. Although these military practices do not equate to legal rules, one cannot deny their normative power, meaning that they do dictate how decisions are made on the battlefield.

IHL rules themselves reflect a hierarchy of how civilian objects are to be valued when targeting decisions are made. Thus, while the general rule of distinction dictates that all civilian objects are protected from attack, some objects are provided higher levels of protection within IHL (Article 50, Protocol I). This is the case for medical units and transports which are provided special protection given their importance in providing assistance to victims of armed conflict. Some civilian property is considered worthy of higher level of protection given its importance to humankind in general. Cultural property, which includes property of great importance to the cultural heritage of every people, is protected not only against direct attack but also against incidental harm if it is excessive in relation to the concrete and direct military advantage anticipated. ¹⁰³ Thus, elevated protections are provided to objects based on a qualitative and symbolic value attached to them. These values factor in assessing whether an attack is excessive in terms of the proportionality calculation. For example, in the Prosecutor v. Prlić et al., the ICTY Trial Chamber found that the fact that the Old Bridge of Mostar was of ‘great symbolic value’ was a factor when determining the proportionality of an attack. ¹⁰⁴ Save for the situations in which cultural property is used for military purposes, such property is otherwise considered civilian. Under IHL, its nature as cultural property calls for protection greater than that afforded other civilian objects, as evidenced in the specific provisions and instruments dedicated to this class of objects. Accordingly, collateral harm to a cultural icon like the Old Bridge of Mostar should weigh more in proportionality calculation than collateral harm to a non-descript bridge with purely utilitarian value, a position embraced in the revised US Law of War Manual. ¹⁰⁵

The classification of civilian objects into varying levels of protection has not stirred controversy, but such gradations become contentious when applied to weighing the value of human lives. As mentioned as the outset, there is a general understanding that IHL rules on the conduct of hostilities make no explicit distinction among civilians. ¹⁰⁶ Accordingly, some experts argue that all civilian lives weigh the same in terms of incidental loss under IHL. According to that view, considering the ‘person’s role or “value” in the eyes of society’ (a cultural or spiritual leader) or ‘moral considerations (such as a kindergarten as against a prison) … could lead to a slippery slope towards considering that the lives of some civilians “weighed less”’. ¹⁰⁷

The debate on whether the notion of force protection, which comprise of ‘endeavors made by a party to a conflict to prevent its own … forces from being lost’, ¹⁰⁸ coming within ‘anticipated military advantage’ in the proportionality calculation, also points to the contested nature of deciding whose lives are given what value during an attack. ¹⁰⁹ Australia and New Zealand for example made declarations when ratifying Protocol I that ‘the term “military advantage” involves a variety of considerations including the security of attacking forces’. ¹¹⁰ Scholars argue further that a state should attempt to minimize the losses of its own troops. ¹¹¹ Asa Kasher and Amos Yaldin claim it is immoral to make the duty to minimize casualties among combatants the last on the list of priorities of a state: ‘A combatant is a citizen in uniform. In Israel, quite often, he is a conscript or on reserve duty. His state ought to have a compelling reason for jeopardizing his life.’ ¹¹² Despite these arguments, the notion of force protection has failed to gain traction among many IHL practitioners. ¹¹³ The concern is that the concept of ‘force protection’ would see the weight of combatant life increased, thus offsetting one of the primary purposes of the proportionality principle: adopting a balanced approach to the protection of civilian life even in cases where some civilian harm is inevitable. Clearly, affording greater weight to the lives of children does not raise a concern similar to the notion of force protection, which tends to tilt the scales away from protection of civilian lives. On the contrary, providing a higher value to the lives of children would enhances civilian protection rather than offset the balance by giving more consideration to military advantage.

Another argument that can be brought against giving a higher value to children’s lives in proportionality calculations is that it discriminates against other civilians whose lives are not so prioritized, such as adult women and especially men, thereby producing a scenario of adverse distinction. ¹¹⁴ One of the basic principles set out in the Geneva Conventions is that persons protected therein shall be treated without any adverse distinction. ¹¹⁵ Likewise, Article 12(2) of the First Geneva Convention dictates that the wounded and sick must be cared for ‘without any adverse distinction founded on sex, race, nationality, religion, political opinions, or any other similar criteria’. Rather than prohibiting all distinctions, IHL only prevents ‘adverse’ distinctions based on prohibited criteria. ¹¹⁶ The effect of providing certain groups of civilians with a higher value in terms of the proportionality calculation does not amount to adverse distinction. This is evident when Article 12(2) is read in light of Article 12(4). While Article 12(2) prohibits adverse distinction based on sex, Article 12(4) allows for favourable distinction to cater to specific needs that may arise due to sex. ¹¹⁷ Therefore, in IHL, providing favourable treatment to some groups of civilians in a given context is not necessarily antithetical to treating all civilians with equality. This is so as providing elevated protections to certain groups of civilians does not in any way diminish the general protections granted to all civilians. It simply increases the number of attacks that would have to be reconsidered.

Granting higher value to children in proportionality calculations would appear consistent with the fact that not all persons, indeed not all civilians, are provided the same protections under IHL. The Four Geneva Conventions, in which different groups of people are granted different protections based on the situation in which they find themselves, is a testament to this. Within the Fourth Geneva Convention, which deals with the broad category of civilians, there are groups that are recognized as deserving of additional or special protections. Accordingly, women must be treated with all consideration due to their sex ¹¹⁸ and privileged treatment may be accorded by reason of age. ¹¹⁹ Likewise, children are the object of special protection: parties to a conflict are required to provide them with the support and the aid that they need by reason of their age or any other reason. ¹²⁰

Such elevated protections granted to children extend to how they are treated during conduct of hostilities. For example, Article 14 of the Fourth Geneva Convention dictates that children under 15 shall be admitted in hospital and safety zones and localities which have been established to protect certain categories of the population from the effects of war. Children are also to be removed from besieged or encircled areas by concluding agreements to that effect, as per Article 17 of the Fourth Geneva Convention. Moreover, according to Article 4 of Protocol II, whenever possible or necessary, and subject to the consent of the parents or persons who have been entrusted with the principal responsibility by virtue of the law or custom, children are to be temporarily evacuated from where hostilities are taking place to a safer sector of the same country.

Most significantly, the more general requirement to take precautions in an attack codified in Article 57 of Protocol I must be modulated to factor in the presence of children if (and only if) it is ascertainable. This implies a certain duty on the part of belligerents to: seek information specifically about the presence of children in a strike zone, reflecting a duty to ‘do everything feasible to verify that the objectives to be attacked … are not subject to special protection’ as stated in Article 57(2)(a)(i) of Protocol I; choose a means or weapon that inflicts comparatively less harm on children if one is available and produces the same military advantage; determine the timing of an attack to lessen the likelihood of child collateral victims, for instance in relation to school hours. The duty to take precautions in attacks in Article 57 of Protocol I covers some autonomous elements, but it is in significant ways a clarification of some implications of the principle of proportionality, the requirements of which are indeed repeated (twice) in this provision. Likewise, there are implications for what is described as ‘passive precaution’ codified in Article 58 of Protocol I, requiring a belligerent to remove civilians under its control from the vicinity of military objectives, which would apply more stringently with respect to children.

In general, under IHL, parties to a conflict are expected to make additional efforts to ensure that children do not become victims of attacks. Our focus here was on the impact of the presence of children in a strike zone on proportionality calculations, but a claim that children weigh more in IHL reverberates on other norms as well. Seen in this light, the general framework of IHL is committed to provide special protection for children to the extent that it is achievable. Likewise, children are treated as a distinct class of civilians in targeting operations, and the legal recognition of this distinctive status by assigning them a greater weight in proportionality assessments would align with the general approach to children under IHL.

4. Conclusion

In the practice of many states, whether collateral victims of an attack are children or not appears to be a factor that influences proportionality calculations under IHL. The distinctiveness of the civilian child in the assessment of collateral harm reflects not only moral and ethical considerations, but also broadly formulated legal norms in IHL that call for special care to protect children against the effects of war. While it is often not possible or practical to determine whether children are expected to be harmed by an attack, there are circumstances in which the attacking side knows or could reasonably know that some of the civilians expected to be harmed by a strike are children. In such a situation, there is a legal obligation to accord greater weight to the harm to children than to adults in assessing whether this collateral harm is excessive in relation to the concrete and direct military advantage anticipated.

The interpretation of IHL defended here, whereby civilian children must be considered to weigh more than civilian adults in proportionality calculations, does not wash away the more general normative haze shrouding the principle of proportionality in that field of international law. To return to the factual scenario with which we introduced the question, the assessment of the lawfulness of a strike targeting the leader of a non-state armed group with the anticipated collateral death of six civilians must factor the identity of these civilians as adult or children, if that fact is reasonably ascertainable in the circumstances. What has been resolved is an issue of incommensurability, because the lives of civilian children and adults can be put on the same scale of unit of value, with the life of a child being of greater value than the life of an adult. What has not been resolved is an issue of incomparability, allowing us to put civilian lives in an evaluative comparative relation to the overall military advantage anticipated. In other words, even if we know that the civilians in the house are children rather than adults, resulting in the need for a correspondingly greater military advantage in order to conclude that the strike is lawful, that knowledge does not provide a clear answer as to whether a strike would comply with IHL in this context. The answer to that question turns on myriad specific factual elements such as the strategic significance of this rebel commander, the availability of alternate modes of attack or capture, the chance that this target may be hit in a different location, and so on.

There may be some reasonable unease with an argument that values differently the lives of individuals. At one level, it may be taken to go against the grain of IHL’s steadfast refusal to engage in moral judgment to differentiate between different classes of participants in armed hostilities, for instance refusing to provide a lesser protection to those who illegally trigger an armed conflict, or indeed even to combatants who commit war crimes. ¹²¹ Equality is perceived as a bedrock of IHL if it is ever to succeed in stemming somewhat the horrors of war. The claim advanced here is limited to requiring that special attention be paid to the presence of civilian children in the battlefield. It does not create the child as a truer civilian than others, nor does it open the door to any claim that there are correspondingly lesser civilians like the elderly. On the contrary, what it does do is ensure that proportionality under IHL aligns with the moral intuition of many combatants and the practice of many belligerents, resulting in a measure of improved protection for children as victims of armed conflicts.