1 The Nature of the Problem

In 1994, prompted by a growing awareness of the degree to which children were being forced to participate in armed conflicts, the General Assembly of the United Nations commissioned an expert study on the problem.1 The expert - Graça Machel - was asked to report on methods of preventing this evil and the measures needed to promote the recovery and social reintegration of affected children. In 1996 she presented her report.2 The introduction - with no sense of hyperbole - was entitled "The Attack on Children". It declared that:

"[millions of children are caught up in conflicts in which they are not merely bystanders, but targets. Some fall victim to a general onslaught against civilians; others die as part of a calculated genocide. Still other children suffer the effects of secual violence or the multiple deprivations of armed conflict that expose them to hunger or disease. Just as shocking, thousands of young people are cynically exploited as combatants."

The last sentence is the subject of this monograph.

Non-combatants have always fallen victim to armed conflict, and the Machel Report confirmed that throughout the 20th century civilian casualties have been increasing at a disturbing rate. A frequently cited progression of figures shows that in the First World War civilians represented only 5% of the casualties; in the Second World War this figure had risen to 50%; in the Vietnam War it rose to 80%.3

Of all the victims of armed conflict, children are the most vulnerable, and the Machel Report revealed that they are now especially at risk.4 The principal reason is the nature of contemporary conflicts. Most are no longer international in the sense that they involve cross-border attacks by one state against the territory and citizens of another. Instead, for reasons owing much to the artificiality of colonial borders and the forces of ethnicity and religion, conflicts occur within states. These struggles drag on for long periods of time, without any decisive beginnings or ends.5

Traditionally, humanitarian law (the rules designed to regulate the conduct of hostilities) assumed that armed conflict would be between states. The four key treaties on the subject - now simply known as the Geneva Conventions - which emerged shortly after the Second World War in 1949 were based on the same assumption. These treaties or conventions therefore placed a primary obligation on states to take responsibility for the actions of their military personnel.

When the power and authority of the state is at issue, however, governments are in no position to carry out their duties. Hence, what is perhaps the most basic principle of international humanitarian law - that civilians should be spared military attack - is being ignored. According to the UN Secretary-General, the level of adherence to humanitarian norms in crisis situations has been deteriorating dramatically and unacceptably. Civilians, especially women and children, are now becoming the main targets of hostilities.6

The ferocity of civil wars is due in large part to the fact that battles are fought by people who know one another. In spite of the natural ties of kinship or neighbourhood, communities are polarised into hostile camps by propaganda appeals to differences of language, religion and ethnicity. Brutalities are committed to deepen these divisions, to give a sense of bridges burned. Attacks provoke counter-attacks and a cycle of vengeance and terror. In these circumstances, traditional restrictions on the conduct of hostilities - not only those derived from international humanitarian law but also local custom - are cast aside. The result is an "ethical vacuum".7

It is, of course, in the nature of a civil war to draw in the entire population, and, when guerrilla battles are being fought in remote rural areas or in crowded ghettos, it becomes physically impossible to distinguish combatants from non-combatants. Nonetheless, in modern conflicts blurring the distinction between soldiers and civilians also happens to be a deliberate tactic. In consequence, far from being kept out of the war zone, civilians find themselves at its centre - and they become the first casualties. This was particularly so in cases such as Bosnia and Rwanda, where ethnic cleansing was the purpose of the struggle.

In the countries afflicted by these ruthless battles for power, social and economic infrastructures are usually weak or underdeveloped and governments are unable to maintain even basic public services. When, in addition, family and community structures have disintegrated, children have no sources of shelter and protection, even if peace returns. Deprived of their normal bases of support, children become homeless refugees. Hence, although many children are killed in fighting, many more succumb to the disease, malnutrition, starvation and social disruption that follow in the aftermath of war.

Children are not simply the helpless victims of conflict but also its active perpetrators. Although this phenomenon has appeared as a serious social issue only within the last two decades, it is widespread. In countries as diverse as Afghanistan, El Salvador, Cambodia, Iran and Liberia, children have been coopted into military units to become active combatants, and, as several recent studies have shown, under-age troops may constitute a significant percentage of the total armed force.8

Children are never an obvious source of manpower, for they have neither the intellectual nor the physical maturity to make them reliable fighters. In fact, because children are so unpredictable, they may be as much a threat to their own side as to the enemy. Nevertheless, military leaders may deliberately recruit under-age troops, because they find them more malleable and thus easier to exploit than adults. During Iran’s war with Iraq and the Khmer Rouge insurgency in Cambodia, for example, children were said to make particularly effective fighters. They were supposed to have "no fear" - although the state of fearlessness was often induced by drugs, alcohol and brainwashing - and, since the production of cheaper and lighter weapons, children can kill as easily and efficiently as adults.9

Whatever the special merits of the child soldier, the fundamental reason for using under-age recruits is shortage of manpower. Once a country has been debilitated by war, poverty and disease, children provide a last reserve of able-bodied combatants. Irregular and guerrilla forces are the first to resort to this option, because they have no administrative means for conscripting members of the general population, but, when governments are desperate to replenish their fighting forces, they too yield to the same temptation.

Many children are forcibly conscripted. The law might require citizens over a prescribed age to perform military duties, but, if administrative systems are defective or if the adult reserve has been exhausted, lawful conscription can degenerate into lawless press-ganging. In the Democratic Republic of the Congo, for instance, although the state is formally committed to recruiting only adults, the government army (the Forces Armées Congolaises) has been openly enrolling children. Similarly, in Ethiopia in the 1980s, the government of the Dergue regime periodically took to seizing youngsters from city streets, schools and marketplaces with no particular regard for their age.10

Because rebel forces do not enjoy access to the propaganda of state media or the coercive power of state law, they tend to resort to the simple expedient of abduction. During Mozambique’s civil war, for example, RENAMO forces regularly kidnapped children from their homes, gave them basic military training and then sent them into battle against the Frelimo government. The same technique is currently being used by Joseph Kony’s "Lord’s Resistance Army" (LRA) in the north-eastern region of Uganda. Although the exact number of combatants fielded by the LRA is uncertain, it is clear that most are children, mainly between the ages of 12 and 16. Approximately 10 000 have been kidnapped from their schools, farms, villages and families, moved across the border into Sudan, and terrorised into accepting LRA discipline.11

Not all children are forcibly conscripted. In fact, the majority enlist more or less of their own volition. Sometimes, as in the Ethiopia-Eritrean and Hutu-Tutsi conflicts, the inducement to take up arms is a sense of patriotism or ethnic grievance. Usually, however, the motive is more basic: becoming attached to a military force is the best chance of surviving a situation of violence and social chaos. Hence, children may actively seek to be recruited to arm themselves against an adversary or simply to secure food and shelter.

Case studies in Sierra Leone, for instance, revealed that most of the child volunteers had lost their homes, families and friends. They joined up in order to find security and protection. By playing on a sense of national duty, young military personnel and other child soldiers might have persuaded some of the recruits to enlist, and a significant number were attracted "by the sheer ‘fun and adventure’ of wearing a military gear and carrying an AK47 around".12 Similar studies in Liberia indicated that children were looking for an opportunity to avenge the deaths of family members or were seeking fulfilment of false promises: "that they would be paid in US dollars, that they would get a house in Monrovia, or cars."13

Once inducted into military units, especially those of non-governmental rebel groups, children face a life of violence and danger for which they are ill-prepared. As combatants, they stand to lose their civilian status and thus the protection that they would normally enjoy under humanitarian law. More serious is the fact that they are presented to the enemy as legitimate targets of war. The child soldier, like any other member of an armed force, is exposed to the traditional aims of warfare: the death and disablement of enemy personnel.

Children are not only liable to attack by the enemy but to exploitation by their own side. Through beatings and torture children are forced to perform acts that regular troops consider too degrading or too dangerous to do themselves. Alternatively, a child might be spared participation as an armed combatant. He or she might be given various support tasks to perform, whether as a guard, messenger or bearer of food and munitions. Girls do not escape this regime. They too are expected to go into battle, but, in addition, they are regularly subjected to the miseries of sexual assault, rape and prostitution. At best a girl might find that she is given as a "wife" to a military commander.

The outcome of these practices is death, permanent injury, psychological trauma and social dislocation. Children who are lucky enough to survive grow up without the benefit of their fundamental rights to food, clothing, shelter, education, healthcare and a secure family background. Of the ones who are rescued, some may be reunited with their kin; others may find places in orphanages or refugee camps; the luckless have little option but a life on the streets. In all cases, rehabilitation into a normal social life is a long and difficult process.14

If children are to be spared this fate, they must be excluded from any form of participation in armed conflict. As will become apparent below, an extensive repertoire of rules is already in place to achieve this aim. The sources for these rules lie in humanitarian and human rights law. While the former has a long tradition of protecting civilians (and by implication children) from the effects of war, its field of application is restricted to armed conflict. Human rights law has the advantage of being more general, since it is applicable at all times and in all places, but until recently it was not specifically concerned with the problem of child soldiers.

Both humanitarian and human rights law are branches of international law, as opposed to municipal or national laws (which are the legal systems applied internally by states to those living within its borders). International law is designed to govern the relationship of states, an underlying purpose that poses the most serious obstacle to effective application of a rule intended to secure the interests of individual human beings.15 Those most likely to be responsible for recruiting child soldiers are rebel groups. Rules of international law, however, are addressed to states, with the result that non-governmental entities, such as rebels, irregular militias and armed insurgents, are not bound. And the beneficiaries of the rules - children - have no enforceable rights.

2 Existing Protections Under Humanitarian Law

Since early medieval times, general principles of humanitarian law have restricted the ways in which warfare may be conducted. Thus a principle of humanity limited the choice of means to the most humane; necessity permitted only a minimal amount of force to be used and discrimination stipulated that only military targets could be attacked. Children, together with the sick, the elderly and mothers of children - an especially vulnerable category of civilians - derived their greatest measure of protection from the requirement that civilian and military targets were to be distinguished. Even in 17th-century Europe, therefore, the established custom was to keep children under 12 out of areas of conflict.16

In 1949, at a major conference in Geneva, the assortment of customs, treaties and declarations governing the conduct of hostilities was defined, extended and codified in four conventions. These and their subsequent Protocols (which collectively are known as "the Geneva rules") are now synonymous with humanitarian law. The Geneva rules did not set out to guarantee non-combatants immunity from the violence of war. Rather, the aim was to shield persons belonging to one belligerent party from the arbitrary exercise of power by another party. Hence, the principal purpose of the four Geneva Conventions was to protect certain categories of people who fell into the hands of an adversary during an armed conflict.

The Fourth Geneva Convention (1949) was concerned, as its title declared, with the Protection of Civilian Persons in Time of War. Parts II and III contain an extensive catalogue of rules intended to protect civilians from the physical and psychological suffering of hostilities, mainly by isolating them from the conflict. Part II makes certain protections available to all civilians, whether or not they happen to be nationals of states party to the Convention.17 Of these, some are explicitly aimed at children, such as those allowing the creation of hospital and safety zones18 and those providing for free passage of relief consignments.19 The most comprehensive safeguards are found in art 24, which provides that children under 15, who are orphaned or separated from their families, should not be left to their own devices. States party are obliged to facilitate their maintenance, education and exercise of their religion.

Part III of the Convention contains an even more extensive set of protections, but it has a narrower field of application, since it applies only to "protected persons". This category is defined to mean those individuals who "find themselves, in case of a conflict or occupation, in the hands of a Party to the conflict or Occupying Power of which they are not nationals".20 Again, certain provisions are specifically aimed at children, although the definition of "children" varies. One class, without any age specification, is entitled to maintenance and education21 and is exempt from service in the armed forces of an occupying power.22 Another class is defined as children under the age of 15; they are entitled to "preferential treatment".23 Yet another class includes those children who are under 18; they may not be compelled to work24 or be subjected to the death penalty.25

The Fourth Convention is applicable in any case of war or armed conflict and in situations of partial or total occupation.26 But it applies only to conflicts of an "international character", with the implication that the safeguards outlined above do not apply to "non-international" or internal conflicts. Given the fact that most conflicts since the Second World War have been "non-international", this provision severely restricts the ambit of the Geneva regime.

Even when the Geneva Conventions were being drafted, it was appreciated that internal conflicts could not be left unregulated. Article 3, which is common to all four of the Conventions, therefore extends certain minimum protections derived from customary international law to non-combatants and to the sick and the wounded. Experience of the type of conflicts that proliferated after 1949, especially the war in Vietnam, however, demanded an even more comprehensive set of regulations.

Accordingly, in 1977 two additional Protocols were adopted in order to supplement the four Conventions. Protocol I applied to certain types of internal armed conflict, namely those in which people were fighting against colonial domination, alien occupation and racist regimes to assert their right to self-determination.27 In this instrument the first attempt was made to address the problem of children participating in hostilities. The Fourth Convention had prohibited a belligerent from recruiting "protected persons" (which included children) of an adversary to its armed forces,28 but this was an unexceptional provision taken from customary international law. The Protocol went a significant step further by requiring states to refrain from recruiting children who were their own nationals.

Article 77 of Protocol I provides that:

"(2) The Parties to the conflict shall take all feasible measures in order that children who have not attained the age of 15 years do not take a direct part in hostilities and, in particular, they shall refrain from recruiting them into their armed forces. In recruiting among those persons who have attained the age of 15 years but who have not attained the age of 18 years, the Parties to the conflict shall endeavour to give priority to those who are oldest.

"(3) If, in exceptional cases, despite the provisions of paragraph 2, children who have not attained the age of 15 years take a direct part in hostilities and fall into the power of an adverse Party, they shall continue to benefit from the special protection accorded by this Article, whether or not they are prisoners of war."29

In several respects, advocates of stricter controls on recruitment were to find this article a disappointment.30 The phrase "take all feasible measures" was a diplomatic compromise that allowed states party considerable freedom to evade the general prohibition. During the drafting process, the International Committee of the Red Cross had recommended a more trenchant expression - "take all necessary measures" - but this was not accepted.31 (The Red Cross had also recommended banning the voluntary enrolment of children, but this proposal was also dropped.)

No minimum age limit was attached to the term "children" in art 77(2). Admittedly, this omission was deliberate, partly to evade debate about what the minimum age for recruitment should be and partly to accommodate the diversity of national laws defining the concept of childhood. Nonetheless, the absence of any definition of childhood left a potentially troublesome area of ambiguity that states could exploit to their own advantage.

Finally, parties to the Protocol were obliged only to ensure that children did not take a "direct" part in hostilities. Qualifying the nature of participation in this manner had the immediate effect of opening up a debate about what constituted "direct" or "indirect" participation, with the corollary that states could allow children to take part in an indeterminate range of "indirect" activities.

The term "direct", which is repeated elsewhere in the Protocols,32 was no doubt intended to denote some form of active engagement in hostilities alongside the regular armed forces. According to the International Committee of the Red Cross, it meant a causal connection between the act of participation and its immediate result in military operations.33 Hence, "direct" participation would imply any attempt to kill, injure and capture enemy soldiers or any attempt to damage their matériel and installations. It would probably also include conveying arms and equipment to regular troops, artillery spotting, spying and sabotage. "Indirect" participation, on the other hand, would probably denote support activities, such as gathering and transmitting information, manufacturing munitions or performing minor service tasks, such as cooking food and cleaning. The broad range of different activities entailed by any military operation, however, fosters areas of ambiguity and invites subjective interpretation.

What was more to the point was the futility of prohibiting only direct participation. Recent evidence shows that children who start their military engagement in a support role usually graduate to becoming active combatants.34 Besides, permitting any degree of participation undermines the principle at stake: for their own safety, individuals who are especially vulnerable should be excluded from hostilities. Whether children participate directly or indirectly, they are placed in danger. In the first place, even a low level of involvement as a messenger or menial camp attendant, exposes the individual to attack by the enemy. In the second place, association with a war effort means that, if captured, a child could be treated as a spy, saboteur or illegal combatant.35

It was partly to cater for the last possibility that paragraph (3) was added to art 77. This unusual provision seeks to regulate what was already seen as being a likely infraction of the general rule. If children under 15 were in fact to participate directly in hostilities, they would stand to lose their entitlement to prisoner of war status under the Third Geneva Convention. Article 77(3) therefore provided that, for purposes of the Fourth Convention, such children would still be deemed "protected persons".

Additional Protocol II, like Protocol I, was intended to supplement the common art 3 provisions in the four Conventions on "non-international" armed conflicts, but the second Protocol is applicable to conflicts not covered in Protocol I. These are defined as conflicts between state armies and "organised armed groups" which operate under a responsible command structure and exercise sufficient control over a portion of a state’s territory to enable them to carry out sustained military operations.36 Although regulating the most serious and nowadays prevalent type of internal conflict, Protocol II does not apply to lesser forms of disorder, namely, "internal disturbances and tensions, such as riots, isolated and sporadic acts of violence".37 Thus, the Protocol still leaves gaps: for example, it was arguably inapplicable in Lebanon and Somalia, because the conflicts there were not between governments and dissident groups.

Protocol II listed certain fundamental protections for non-combatants including what are probably still the most comprehensive regulations on the recruitment of children. Article 4(3) provides that:

"Children shall be provided with the care and aid they require, and in particular:...

(c) children who have not attained the age of 15 years shall neither be recruited in the armed forces or groups nor allowed to take part in hostilities;

(d) the special protection provided by this Article to children who have not attained the age of 15 shall remain applicable to them if they take a direct part in hostilities despite the provisions of sub-paragraph (c) and are captured."

A number of the delegations that worked on the drafting of this article considered the age of 15 too young, but again the divergence of national laws made agreement on the higher age of 18 impossible. In other regards, however, Protocol II represented a significant advance on all the earlier treaties. It gave states no measure of discretion in prohibiting recruitment nor did it permit indirect participation. By providing simply that children may not "take part in hostilities", Protocol II proscribed all forms of participation, together with voluntary enlistment. Unfortunately, several states have not ratified the Protocol.38

3 International Human Rights Law

(a) The Convention on the Rights of the Child

The Geneva Conventions and their Protocols have developed, and to a large extent continue to develop, independently of human rights law. From their inception, the two branches of law were conceived on different principles. Humanitarian law was concerned mainly with a state’s treatment of enemy persons in times of armed conflict, while human rights law was concerned with the relationship between a state and its own nationals in times of peace.

Admittedly, differences between the two systems are gradually diminishing. The Fourth Convention, for instance, protects the individual rights of protected persons and certain articles of the two Geneva Protocols were derived from the International Covenant on Civil and Political Rights (1966). Nevertheless, humanitarian instruments failed to incorporate the rapid advances made in children’s rights during the 1980s.39 The reasons were twofold. First, because the Geneva Protocols were drafted 12 years before a separate treaty on children appeared, they could obviously not reflect principles encoded in the 1989 UN Convention on the Rights of the Child (abbreviated below as the CRC). Secondly, because the drafters of the Geneva Conventions and Protocols did not foresee children becoming the intended victims of armed conflict, they saw no need to legislate specially for the problem.40

From a child’s point of view, therefore, humanitarian law has several defects. First, it does not give priority to the child’s interests, which is a cornerstone of human rights law. Article 3 of the CRC declares that:

"In all actions concerning children, whether undertaken by public or private social welfare institutions, courts of law, administrative authorities or legislative bodies, the best interests of the child shall be a primary consideration."

If children’s interests were to be given first consideration, there could be no question of justifying recruitment of children on grounds that their interests should be subordinated to a greater war effort, military necessity or state security.

Secondly, the principle of equal treatment, another cornerstone of human rights law,41 finds no place in humanitarian law. According to the former, all children, regardless of differences in circumstance or social status, deserve equal protection. According to the latter, protection may depend on the nationality of the child and its parents or their relationship to one of the parties to a conflict. Yet, if a child’s safety is the prime concern, protection should be available regardless of the legal or political character of a conflict; it is, after all, irrelevant to an affected child whether a particular dispute is designated international or non-international.

Thirdly, and leaving aside for a moment the question of derogation, children’s rights are always applicable, whether in times of war or peace. Although a child’s participation in hostilities normally becomes an issue only during periods of armed conflict, there are good reasons for maintaining a prohibition on recruitment even during peace. Not only may the process of recruiting begin well in advance of an outbreak of hostilities but it may also be difficult to determine when fighting begins and ends.

The UN Convention on the Rights of the Child (1989) is an almost universally accepted human rights instrument, ratified to date by all but two of the world’s states (the United States and Somalia). The purpose of the treaty is to secure special care and protection for children. Thus, as we have seen, art 3(1) provides that the child’s interests are to be given primary consideration and art 6(2) obliges states parties to ensure "to the maximum extent possible the survival and development of the child". Other articles recognise a child’s right to health and care services,42 a standard of living adequate for its development43 and a right to education.44

Certain provisions deal specifically with the position of children caught up in armed conflicts. These were an innovation for human rights law, because previous treaties in this field had included special clauses designed to incorporate the Geneva rules by reference. The CRC, however, directly obliges states party to respect the rules of humanitarian law relevant to children45 and to promote the "physical and psychological recovery and social reintegration" of children who have been war victims.46

Article 38(2) has a special provision on using children in armed conflict:

"States Parties shall take all feasible measures to ensure that persons who have not attained the age of 15 years do not take a direct part in hostilities."47

Simply because it happens to be contained in a human rights treaty, this article transcends the technical distinctions bedevilling application of the Geneva Protocols. Thus art 38(2) applies whether a situation qualifies as an international conflict, an internal conflict for the right of self-determination (Protocol I) or a high-intensity conflict between a government and organised armed groups (Protocol II).

Nonetheless, art 38(2) can hardly be considered a full or satisfactory answer to the problem of children becoming involved in hostilities. It merely repeats the terms of art 77(2) of Geneva Protocol I, with all the ambiguities and failings of that article.48 While the CRC was being drafted, attempts were made to improve on the Geneva Protocol, both by raising the minimum age to 18 and by requiring states to take "all necessary measures" or "legal, administrative and other measures" to prevent children participating in hostilities. These proposals were defeated by the need to achieve consensus and by the United States’ argument that the CRC was not the proper vehicle for rewriting humanitarian law.49

One of the most contentious provisions in art 38(2) is the age for determining childhood. Although, for all other purposes in the CRC, a child is defined as any person below the age of 18, art 38(2) specifies 15. The final version of this article was the result of protracted debate.50 When the working group began to discuss proposals in 1986, no age was mentioned, for it was assumed that the general age of 18 would be applicable. Some delegations, including the British, Soviet, Canadian and American, argued for 15, because it was in line with humanitarian law and their national legal systems. Others, notably Sweden, Switzerland and the Red Cross, argued for 18 to keep the article in harmony with human rights. From the final (and thoroughly confused) session of the working group, the age of 15 eventually emerged.

States in favour of 18 are free, of course, to interpret art 38(2) in light of art 41 of the CRC, which provides that any conflict between the provisions of the Convention and obligations under municipal or international law must be settled in favour of whichever rule gives a child the greatest protection. To this end, when certain states, such as Argentina and Austria, ratified the CRC, they made special declarations that they would enforce the age of 18.

Two years after the CRC came into force, the Organisation of African Unity promulgated a Charter on the Welfare and Rights of the African Child (1991). This instrument met some of the objections to art 38(2). Article 22(2) of the Charter requires parties simply to "take all necessary measures to ensure that no child shall take a direct part in hostilities and refrain in particular from recruiting any child". Although the restriction of "direct" participation was preserved, children were deemed to be all human beings under the age of 18.51 Unfortunately, the Charter is a regional treaty open to members of the OAU and it is not yet in force.

Two final concerns about the CRC involve the possibility of states derogating from its provisions in times of internal unrest and the problem of enforcing the Convention against rebel groups. In the CRC itself, formal provision is made for derogating from the freedoms of expression, religion and association in the interests of national security, public safety or public order.52 On an analogy with systems of constitutional law, application of other articles might also be suspended for similar reasons. Once it is conceded that governments may order limitation of the rights enshrined in the Convention, they have a justification for starting to recruit child soldiers.53

Another problem is posed by a general breakdown of state authority. If through external aggression or internal unrest a government loses its ability to maintain law and order, it can no longer be held responsible for failing to enforce all the prescriptions of human rights law. In particular, governments cannot be considered responsible for the actions of rebel forces. These are groups that by definition place themselves beyond the reach of the state power.

Humanitarian law, by contrast, is intended to operate in these circumstances. During internal conflicts both governments and non-governmental entities are bound by the two Protocols to the Geneva Conventions.54 Human rights treaties are different. Because international law generally allows only states to become parties to a convention, non-state entities incur no rights or duties.

(b) The draft optional protocol on the Recruitment of Children into Armed Forces

In 1992, the UN Committee on the Rights of the Child convened a Theme Day on Children in Armed Conflict. At this gathering a proposal was made to remedy the shortcomings and to extend the ambit of art 38(2) of the CRC by drafting an optional protocol on the Recruitment of Children into Armed Forces. With the support of UN specialised agencies and several NGOs, the UN Commission on Human Rights established a working group to proceed with the project.

So far, the most interesting innovation - for international human rights law at least - is an attempt to reach non-governmental rebel groups. A draft clause provides that:

"States Parties shall take all feasible measures, including any necessary legislation, to prevent recruitment of persons under the age of [18] years [of minors] subject to their jurisdiction by non-governmental armed groups [which are parties to] [involved in] an armed conflict."

While the expectation seems to be that the protocol can be made to work in the same way as humanitarian treaties, the chance of this happening is remote.

In the first place, the reason why both government and non-governmental entities obey the Geneva Rules is due to a particular tradition and a unique enforcement machinery (notably the institution of an independent "protecting power" and, of course, the Red Cross). In the second place, it is doubtful whether governments themselves will accept an arrangement such as that contemplated in the protocol. They are generally reluctant to negotiate with rebel groups, since mere recognition of the existence of these groups tends to give them political legitimacy. In the third place, governments are unlikely to implement this type of provision. If a state actually succeeds in apprehending members of a rebel group, they will be prosecuted for treason or crimes against state security rather than a lesser offence of recruiting under-age children.55

Rebel forces are simply not amenable to the ordinary methods of enforcing international law and any attempt to get them to respect children’s rights will depend entirely on their willingness to cooperate. In the turmoil resulting from a civil war, therefore, the best that can be hoped for is that a rebel group will decide voluntarily to abide by the CRC. To this end, the Machel Report urged non-state entities to agree to comply with the standards of the Convention. It is encouraging to note that, in 1995, certain groups in Sudan had in fact done so.56

The various drafts of the protocol to the CRC offer few meaningful improvements on existing law. Parties will probably be obliged, by the now familiar formula, to do no more than "take all feasible measures" to ensure that children do not participate in hostilities.57 Debate continues as to whether the qualification "direct" participation in hostilities should be preserved and, of course, whether the age limit of 15 should be raised. Although there seems to be general agreement that those under 18 should not be subject to compulsory conscription, certain interest groups want to lower the age for voluntary enlistment to 16 and to make special provision for allowing enrolment of even 15-year-old children into educational institutions operated by armed forces.

Notwithstanding considerable international support for the draft protocol, the effort being expended seems profitless. Apart from the age limit, it will probably give children less protection than the provisions of the statute proposed for the international criminal court, because the draft protocol at present offers no advance on the notoriously weak enforcement procedures contained in the CRC (which will be considered below). The publicity surrounding the protocol has doubtless done much to heighten consciousness about the plight of child soldiers, but, as one disillusioned commentator asks, "is the power of repetition so great that it can make a significant improvement?"58

4 Problems of Enforcement

(a) Enforcement of international human rights

The CRC encoded an impressive array of rules designed to protect children from human rights abuses. That nearly all the world’s states decided to ratify this treaty is testimony to a rare degree of consensus on its importance. However, in spite of the clear obligation expressed in art 2 that all parties must respect and must ensure respect for rights contained in the CRC, few states have actually implemented the rights, let alone enforced them.

The Democratic Republic of Congo, for example, is party to the Convention and is therefore bound to implement art 38(2). As it happens, Congo has municipal legislation specifying a minimum age of 18 for conscription.59 Yet, in spite of the country’s formal legal commitments, it has become apparent that in the current struggle between Laurent Kabila’s government and the rebel movement, Rassemblement Congolais pour la Démocratie, both sides are using "kidogo" ["little ones" in Swahili]. UNICEF reported that the Congolese army (the Forces Armées Congolaises), was not only enrolling below-age volunteers but also rounding up children in Kinshasa. (Predictably, the children were offering their services in exchange for food and money.) What is more, the army remobilised between 400 and 500 former child soldiers from a transit centre near Kisangani and another 100 children from Bukava.60 The situation was such that, on August 31, 1998, when the Security Council issued a statement on the situation in Congo, it condemned inter alia "the recruitment and use of child soldiers".61

What action can be taken to halt such a patent violation of treaty obligations? This question cuts to the quick of international human rights law, for it will become evident that the highly developed code of substantive law is not matched by comparable enforcement measures - a deficiency that must ultimately be ascribed to the nature of the international legal system. States are the primary subjects of international law and only they can be held responsible for refusing or neglecting to discharge international obligations. The intended beneficiaries of the law (and the agents responsible for carrying it out) are human beings; but, because they are legal objects, they cannot function as right- or duty-bearers.

According to the orthodox philosophy of international law, therefore, human beings bear no responsibility for violating the rules,62 nor do they have any legal means of protecting their interests. It follows that members of Kabila’s government who are enlisting Congolese children cannot, as individuals, be held to account. In times of peace, a particular child (represented of course by an appropriate adult) could possibly resist conscription by challenging government action in local courts on the basis of Congo’s municipal legislation. If the government decides to flout its own laws on grounds of state security or military necessity, however, such a challenge would be futile.

Individual children have even less chance of successful legal action against states that neglect to implement provisions of art 38(2) of the CRC. Most systems of municipal law provide that treaty obligations become enforceable by individuals in state courts only when the treaty is made part of municipal law by an appropriate legislative act. Thus, if the CRC had not been incorporated into domestic law, the Convention alone would afford individuals no actionable rights.

Could another state party to the CRC take a government, such as the Democratic Republic of Congo, to task for failing to carry out its duties? Non-Congolese citizens who were forcibly conscripted could request their states of nationality to take action on their behalf against Kabila’s government. In this case, the right of action would be derived from breach of a customary-law duty requiring states to accord aliens certain internationally prescribed standards of treatment. The right of "diplomatic protection" vests in the state, not the individual, however, and whether it would be exercised lies entirely within the protecting state’s discretion.

The range of options available to states party to human rights conventions for protesting against another party’s violation of treaty provisions is equally limited. For ordinary types of treaty, such as commercial or trading agreements, a typical reaction to a serious breach would be to take countermeasures by way of reprisals. Thus the aggrieved party could retaliate against the other party’s failure to carry out obligations by suspending the treaty or by refusing to implement provisions operating in favour of the offender. But, where the purpose of a treaty is to benefit human beings rather than states, this course of action would be morally indefensible and it is specifically barred by international law.63

A less severe countermeasure to breach of an international duty would be an act of retorsion, a lawful but nevertheless harmful act. In this case, the aggrieved state could impose some form of legally permissible but economically or politically damaging sanction on the offending party. For instance, trade boycotts have been imposed on states accused of employing child labour in breach of the Convention Concerning the Minimum age for Admission to Employment.64 Again, whether it will be expedient to exercise this option must be carefully weighed, for where the purpose of a treaty was to protect human beings, an act of retorsion could damage rather than protect their interests.

In the final analysis, diplomatic protest is the only viable method of expressing disapproval of breaches of a human rights treaty. Given states’ sensitivity about their sovereignty, however, even a mere verbal criticism could be construed as unwarranted intervention in domestic affairs.

Concerned states must therefore rely upon the enforcement provisions made available in CRC itself. These provisions, like those written into other human rights treaties, are decidedly weak. Compliance with the Convention is supervised by a Committee on the Rights of the Child,65 a body authorised to receive reports on measures that states party have taken to implement provisions of the Convention. The Committee’s powers are limited to hearing communications from states - not affected individuals - and, in case of a breach of the CRC, it may do no more than issue observations and general recommendations. Thus, even such an obvious offender as the Congolese government can expect nothing worse than a public reprimand.

When the agent responsible for recruiting children is a rebel group, such as the Rassemblement Congolais pour la Démocratie, the possibility of enforcing art 38(2) of the CRC becomes even more remote. Were the Congolese government both capable and willing, it could of course take action for breach of its municipal law, as could the individual child. But, if the government were reluctant to implement (or had not enacted) the relevant municipal law, an affected child could not compel the government to do anything, for, as an individual, the child acquires no direct rights from human rights treaties. Nor could other states party to the Convention hold the government accountable, because it is not responsible for the actions of groups that, by their very nature, lie beyond its control.

The difficulties of enforcing international human rights obligations demonstrate a major limitation in the international justice system as a whole: the lack of a central authority with the political power and legal authority to force offenders to submit to its jurisdiction. If states are not prepared to protect their nationals or to punish offenders, victims are left helpless and wrongdoers escape prosecution.

(b) Enforcement of humanitarian law: individual responsibility

The enforcement measures available under humanitarian law are stronger than those under human rights law, in part because humanitarian law has a longer history and in part because violations of the rules tend by their very nature to be taken more seriously.

Before the Second World War, offenders were usually prosecuted by one of the belligerents under its municipal law. In this event, an accused could escape liability by pleading a defence of superior orders. This defence implied that, if the accused had been instructed to commit a wrongful act by a military commander or if the act had been more generally sanctioned by the accused’s system of municipal law, the prior command excused the conduct complained of.

From the end of the First World War, however, a principle started to develop that individuals should bear personal responsibility for infringing the precepts of international humanitarian law. In 1945, this principle was given concrete expression when the Allies created an ad hoc International Military Tribunal in Nuremberg - and a year later a similar tribunal in Tokyo - dedicated to prosecuting Axis war criminals. The Charters for these Tribunals gave them jurisdiction over three broad categories of offence: crimes against peace (preparing, initiating or waging a war of aggression), crimes against humanity (murder, enslavement, deportation and other inhumane acts committed against civilian populations) and war crimes (violations of rules governing the conduct of warfare, which today would typically be violations of the Geneva Conventions).

As it happened, the majority of the Axis war criminals were tried not by the International Military Tribunals but by local state courts, where accused persons were charged with breaches of municipal criminal laws. These state prosecutions nevertheless followed Nuremberg and Tokyo principles, in particular that individuals were personally responsible for their acts (with the result that the defence of superior orders no longer held good). Through this consistent pattern of state practice, the offences created by the Charters for the International Tribunals were transformed from crimes stipulated by specific treaties to crimes under customary international law.66 The implication of an act being deemed criminal in customary law was to allow any state to exercise "extraordinary" jurisdiction over an offender caught within its jurisdiction.

At the time of the trials in Nuremberg and Tokyo, involving children in hostilities was not thought to be an important issue deserving special attention. (Nor, for that matter, were other offences directed primarily against children.) It followed that the Charters for the International Military Tribunals did not contain specific crimes of recruiting children or using them to participate in hostilities, and, in consequence, these activities were omitted from the three categories of offence that came to be accepted as part of international customary law.

Prosecution of international criminals by municipal courts was a right vesting in states, which suggested that they had a discretion whether or not to take action. In 1949, the Geneva Conventions converted this discretion into an obligation. The Conventions obliged parties to enact legislation providing penal sanctions for persons who had committed or had ordered commission of certain violations of the Conventions - what are termed "grave breaches". Parties were further obliged to search for offenders (regardless of their nationality) and to try them locally or (depending upon extradition obligations) to hand offenders over for trial by another party.67 These provisions were repeated in Protocol I68 but not in Protocol II.

The "grave breaches" referred to in the Geneva Conventions and Protocol I included "compelling a protected person to serve in the forces of a hostile Power",69 which is a widely accepted offence derived from customary law.70 The Fourth Convention did not stipulate a more general crime that would cover recruitment of child soldiers nor, of course, did customary international law. Protocol I did regulate recruitment of children,71 but this activity was not mentioned in the description of "grave breaches".72 As a result, although using children as combatants was clearly in breach of Protocol I, it was not a criminal action attracting individual liability.

The Democratic Republic of Congo is party to the Fourth Geneva Convention and Protocol I, but for at least two reasons the Congolese government would have no duty to prosecute individuals responsible for recruiting children. The first is obviously because the treaties do not contemplate this practice as a "grave breach" and therefore an international crime. The second is because neither the Convention nor Protocol I applies to the type of internal struggle being waged within the country. The Fourth Convention is applicable to "international" conflicts and Protocol I to struggles for self-determination.

Whether or not states have made the Geneva Conventions and Protocols part of their municipal law, they are still entitled by customary law, as we have seen, to exercise an "extraordinary" jurisdiction over persons alleged to have committed serious violations of their municipal law.73 In effect, international law has delegated the task of law enforcement. Simply giving states an authority to prosecute, however, is not conducive to the regular enforcement of justice.

Certain states are dilatory in prosecuting or extraditing suspects at large in their territories. (After the Second World War Argentina and Paraguay were notoriously delinquent in this regard.) There is, of course, no international police force with authority to apprehend offenders sheltering in such states, and the principle of territorial sovereignty precludes any other concerned state from effecting the arrest itself.74 Although individuals who have suffered the effects of breaches of humanitarian law have a clear interest, they are not entitled to proceed directly against a state that fails to prosecute, because states are the primary right-bearers and only they would be permitted a right of action.

The abductions and forced conscriptions perpetrated by Joseph Kony’s "Lord’s Resistance Army" in north-eastern Uganda and southern Sudan give a factual context to these problems. Although Sudan denies that it is supporting Kony, the LRA operates from bases in that country, where it enjoys a safe haven from the Ugandan army. Not only does the Sudanese government lack the political will to take action against Kony, but it has no legal obligation to do so (not least because the criminal acts were committed mainly in Uganda against Ugandan citizens). For its part, Uganda might be prepared to prosecute Kony for crimes against children - although it too has no duty in this respect - but it may not enter Sudan to arrest him (because to do so would infringe Sudan’s territorial sovereignty).

Sudan and Uganda are both parties to the CRC. Sudan is also party to the Fourth Geneva Convention and Uganda is party to the Fourth Convention and the two Geneva Protocols. With the exception of the Fourth Geneva Convention, all these treaties forbid forcible conscription of children, but none regards it as an international crime. Finally, and just to demonstrate how complex enforcement could be in these circumstances, Protocol I is arguably inapplicable, because the conflict is not a struggle for self-determination.

(c) The movement towards establishing an international criminal court

States’ jealous protection of their sovereignty has been the main stumbling block to the punishment of serious breaches of human rights and humanitarian law. Only when a permanent international tribunal acquires jurisdiction over offenders, regardless of where they happen to be sheltering, will an international criminal justice system be realised.

This idea was mooted shortly after the foundation of the UN in the context of the Convention on the Prevention and Punishment of the Crime of Genocide (1948). Throughout the 1950s it continued to receive attention from the General Assembly and the International Law Commission. Further progress was then cut short by the politics of the Cold War, and the project was shelved. It was revived in the 1980s, and in 1993 the International Law Commission submitted a draft statute for a tribunal. When a comprehensive text eventually appeared in 1994, the General Assembly established a committee to consider the issues involved.

This initiative was overtaken by events in Bosnia-Herzegovina and Rwanda. In response to programmes of "ethnic cleansing" in these countries, the UN Security Council resolved to create ad hoc international tribunals to try specific offenders. In 1993, acting under Chapter VII of the Charter,75 the Council established a tribunal with jurisdiction over persons responsible for committing serious violations of humanitarian law in the former Yugoslavia.76

The powers of this body, although limited geographically to offences committed within the territory of former Yugoslavia and temporally to those committed after 1991, covered grave breaches of the Geneva Conventions, violations of the law and customs of war and crimes against humanity.77 In fact, the tribunal has had only limited success. To effect arrests it has to rely on UN peacekeeping forces or it has to wait until offenders strayed outside Serbia into neighbouring states. Three years after the creation of the tribunal, only 74 persons had been charged and most of these were still at large.

The situation in Rwanda posed an even more formidable undertaking.78 More than 80 000 people were accused of committing genocide in 1994. Acting again under Chapter VII of the Charter, the Security Council constituted an international criminal tribunal. This body differed from its Yugoslav counterpart in at least two respects. First, because the conflict in Rwanda had been purely internal, the technical legal question whether the tribunal’s jurisdiction extended to crimes committed in the context of an international armed conflict was irrelevant.79 Secondly, the Rwandan government actively supported the project.

While these tribunals were being established, proceedings for the creation of a permanent court continued. In 1995 a draft statute was prepared, outlining the court’s jurisdiction, composition and procedure, and these proposals went forward to a UN diplomatic conference, which was held at Rome in June and July of 1998. Here a statute for an International Criminal Court (abbreviated as ICC) was in principle adopted.80 The conference decided that a permanent Court should be set up with authority to exercise its functions on the territory of any state party.81 The Court’s jurisdiction will be over "the most serious crimes of international concern",82 namely genocide, crimes against humanity and crimes against the laws and customs of war.83 Persons who commit these crimes will be personally responsible and liable for punishment.84

(d) Recruitment of children under the draft statute for the ICC

Two offences in the draft statute are of particular significance for this monograph. The tribunal’s jurisdiction over war crimes85 includes "committing rape, sexual slavery, enforced prostitution"86 and "conscripting or enlisting children under the age of 15 years into the national armed forces or using them to participate actively in hostilities".87

The draft statute attempts to dispose of the preliminary question whether an armed conflict is international or internal. Although the offences mentioned above must be committed in the context of an "international armed conflict",88 special provision is made for internal conflicts. Thus, in the case of armed conflicts "not of an international character",89 although the offences of sexual violence remain the same, the term "national" in "national armed forces" is deleted.90

This modification makes it clear that non-governmental rebel groups or other irregular militias are also capable of committing the offence of involving children in hostilities. Non-international armed conflicts are defined - in terms echoing Geneva Protocol II - to mean those "that take place in the territory of a state when there is protracted armed conflict between governmental authorities and organised armed groups or between such groups."91 The draft statute excludes "disturbances and tensions, such as riots, isolated and sporadic acts of violence or other acts of a similar nature".

While these clauses expand the scope of existing law considerably, some imaginative work on the part of the future Court will be required to relate the offences described in the draft statute to the targeted practices. One set of crimes - rape, sexual slavery and enforced prostitution - involves acts of sexual violence that are already regularly prosecuted in systems of municipal law. Interpretation and application of this category of offences should therefore be straightforward.

"Conscripting or enlisting children ... or using them to participate actively in hostilities", however, presents a new type of offence. It is true that recruiting under-age children constitutes criminal wrongdoing in the laws of many states, but cases of prosecution are virtually unknown.92 Hence, when interpreting these offences, the Court will have no body of precedent to rely on.

By presenting the terms "conscript" and "enlist" as alternatives, the draft statute suggests two different activities.93 "Conscript" would imply some form of forced participation, although it is far from clear what the wrongful act should be. The draft statute might contemplate the formal call-up of children, the process of training them as soldiers or subjecting them to military discipline - or all three of these activities. The common element in the targeted practices, however, which vary from official acts of conscription, to press-ganging to abduction, is simply making under-age persons members of an armed force against their will.94

By contrast, "enlist" would suggest a child’s voluntary enrolment, an interpretation that is borne out by art 51 of the Fourth Geneva Convention (which forbids any pressure or propaganda aimed at securing "voluntary enlistment"). The criminal act would presumably be similar to that contemplated in the crime of conscription, with one difference: that any volition on the part of a child would not be permitted to function as a justification or defence.

There are good reasons for discounting the fact of a child’s consent. If consent were taken into account in determining guilt, state agencies and military commanders would be afforded an easy defence to the crime. Besides, the circumstances and motives for children becoming attached to armed forces - which extend from a desperate need for food and security to propaganda and trickery - are bound to complicate any inquiry into whether the consent was fully informed. The principle that all under-age children must be deemed incapable of forming a proper consent is straightforward. After all, most systems of municipal law refuse children the capacity to give valid consent to legal transactions without their guardians’ approval.

Conscription and enlistment are supplemented by a third offence: using children to "participate actively in hostilities". (The existence of this crime would in practice render the specific offences of conscription and enlistment redundant, for it is the more general.) Unlike the previous crimes, using children to participate in hostilities suggests the absence of any formal induction into a military unit. Hence, it would be unnecessary to prove that a child was put into uniform, subjected to military discipline, made to bear arms or subjected to any of the traditional means of marking an individual as a soldier rather than a civilian. The criminal act would therefore be employing a child in hostilities regardless of what tasks the child had to perform.

Instead of speaking of an "armed conflict" - which was the phrase used in the Geneva Conventions and Protocol I - "hostilities" was chosen for the draft statute (recalling art 4(3) of Geneva Protocol II). Originally, "armed conflict" was decided upon to avoid the limitations and technicalities that had earlier plagued use of the word "war". Accordingly, it has been argued that the phrase is not a legal concept, but is purely descriptive of a factual situation. In the same spirit, "hostilities" seems to have been intended to denote the actual state of fighting.95

The requirement that participation in hostilities be "active" has no immediate reference to existing concepts of international law. Although the drafters’ intention might have been to avoid narrowing the scope of the crime to "direct" participation, the future Court will probably take as its point of departure the current distinction between "direct" and "indirect". Hence, active participation would clearly entail actually arming a child and sending him or her into battle, whereas ordering the child to clean billets or prepare food would not. Transporting munitions, gathering information or guarding bases might again fall into a grey area, although it could be argued that deliberate avoidance of the term "direct" widens the range of the offence to include these activities.

Merely allowing a child that has volunteered services to remain at a military camp should clearly not be included in the criminal offences of enlistment or use for active participation. Such an act (or omission) is potentially humane rather than wrongful. Nonetheless, does a military commander incur any responsibilities for children who have sought shelter and protection? It could be argued, in line with the Fourth Geneva Convention, that commanders are at least obliged to segregate children from combatants. The short-term aim would be to afford children protection from attacks by the adversary; the long-term aim would be to return them to their families and communities.

The draft statute leaves certain issues, notably the elements of crimes, open.96 These will be established after further deliberation. Similarly, the law to be applied by the Court in order to interpret the crimes is still to be decided, although the draft statute provides that it must apply: whatever provisions are contained in the statute dealing with procedure, evidence and the elements of crimes, failing which, applicable treaties and the principles and rules of international law (including those governing armed conflict), failing which, general principles derived from municipal laws.97 In all cases, the tribunal is obliged to apply and interpret laws consistently with the principle of non-discrimination.98

In the event of a direct contradiction between the draft statute and another treaty - and the CRC is a likely candidate - the Court would be bound, in the first instance, to apply the terms of its own statute. (This situation could arise if and when the additional protocol to the CRC comes into force, for it sets the minimum age for recruitment at 18.) Only in the absence of any provisions in the statute would the Court be required to take account of the CRC. One conflict of this nature has been preempted. The draft statute makes provision for adopting a special procedure when persons accused of crimes against children are being tried.99 The purpose, which is consonant with the CRC’s insistence on a child’s particular needs in the criminal justice system,100 is to protect child complainants, especially those in sexual offences, from the trauma of confronting their assailants in open court.

5 Conclusion: Taking Stock

An array of treaty and draft treaty provisions restricts the participation of children in armed conflict. These various instruments weave an intricate fabric of regulation, albeit a loose network rather than a close-knit cloth. The Geneva Protocol I and the CRC require states party to "take all feasible measures" to prevent the "direct" participation of children in hostilities. The draft statute for the ICC is worded in similar terms. By omitting the qualifications "feasible" and "direct", Geneva Protocol II and the draft protocol to the CRC extend the ambit of regulation.

Prohibiting an activity - a matter of substantive law - and ensuring that the prohibition is actually implemented - a matter of enforcement and procedure - pose two quite different problems. Obtaining agreement on a substantive prescription is an exercise in verbal persuasion, to be pursued in the comfortably abstracted circumstances of diplomatic conferences. Putting enforcement machinery into operation is infinitely more demanding, costly and time-consuming.

Compared with the many deficiencies in the existing enforcement provisions for human rights and humanitarian law, the proposed ICC offers the more effective means of realising a prohibition on involving children in hostilities. Particular offenders can be held personally accountable for infractions of the rule; commanders of rebel groups cannot hide behind the technicality that they are not bound by international law; and criminals will no longer find it as easy to shelter in sympathetic states.

While it is tempting to regard the ICC as a panacea, its draft statute could not be expected to realise a perfect solution. One ideal - giving individuals enforceable rights - had to be sacrificed. Any system of criminal law allows only the prosecutor, who represents a community interest, the power to institute proceedings. The people who are supposed to profit from these proceedings - in this case children, together with their parents and guardians - have no powers or entitlements.101 A child’s presence in the ICC will thus be necessary only to give evidence.

A second concern is the somewhat limited scope of the Court’s jurisdiction. Three provisions establish the superior position of the Court in relation to states that become party to the statute: they must cooperate with the Court,102 the Court is entitled to try crimes committed within their territories or by their nationals103 and the prosecutor may initiate investigations of his or her own accord.104 Offsetting these provisions, however, is a general principle that the jurisdiction of municipal courts is not to be usurped. Thus, unlike the International Military Tribunals and those for Yugoslavia and Rwanda,105 the future Court’s jurisdiction will not be primary or overriding.106

The Court has exclusive competence where the prosecutor initiates proceedings - presuming that an interested state has declined to prosecute - or where a state party to the statute refers a matter to the Court.107 Otherwise, cases will be inadmissible before the Court if states that have jurisdiction in the circumstances are themselves investigating or prosecuting. It could then happen that a state will attempt to stall proceedings in the Court indefinitely by claiming to be searching for an accused or processing a claim.108 The draft statute seeks to provide for this eventuality by allowing the Court to assume jurisdiction if it can be proved that the state is unwilling or genuinely unable to carry out the investigation or prosecution.109

The third, more significant shortcoming in the draft statute goes to the substance of the offences proposed. So far, those responsible for drafting the document missed an opportunity to improve the level of protection for children. They could have designed a far more comprehensive article by repeating the terms of art 4 of the Geneva Protocol II and by incorporating amendments proposed in the draft optional protocol to the CRC. Both instruments avoid the restrictive qualification of "direct" or "active" participation - they simply prohibit the use of children in hostilities - and, of course, the draft optional protocol aims to raise the relevant age to 18.

The age limit has been the single most divisive issue in the many attempts to secure agreement on the substantive law. Although a survey of more than 100 systems of municipal law indicated that more than two thirds accepted 18 as the minimum age for compulsory recruitment,110 certain countries, notably the United States, have held out for a lower age. Oscillating between the extremes of 15 and 18 does, of course, produce glaring contradictions. For instance, while 18 is in line with the generally accepted definition of childhood and is the age specified for purposes of controlling child labour,111 a child may begin working in an army at 15. This type of contradiction is apparent in the ICC statute itself: it sets a military recruitment age of 15 but bars the trial of persons below 18.

Dispute about the appropriate age will persist, notwithstanding the good arguments in favour of raising the age limit to 18. Nonetheless, the time may now have come to move on to other issues. One reason for accepting the lower age is the fact that the most common offenders - non-governmental rebel groups - recruit children even younger than 15, and raising the age limit to 18 is obviously not going to solve this problem.112 At a more principled level, however, it must be appreciated that the conception of childhood varies according to economy and culture: an extended period of childhood dependency is a luxury unavailable in the less affluent economies of developing countries.

Moreover, implementing a recruitment age of 18 depends on an efficient machinery for identifying below-age children. While departments of state might be expected to institute proper inquiries into the age of recruits, their capacity to do so presupposes a reliable system for documenting births. In countries such as the Democratic Republic of Congo or in areas such as southern Sudan, this type of administrative infrastructure functions spasmodically or not at all.113 And, even in countries with developed administrative systems, children are unlikely to be carrying documentation in times of war or civil upheaval.

Despite the criticisms above, acceptance of a permanent international criminal court marked a turning point in the enforcement of international human rights and humanitarian law. Now is perhaps the time to take stock of the process that has culminated in this event. We should begin by recalling that at the end of the Second World War the tendency was to regard a child’s engagement in the war effort as nothing less than heroic self-sacrifice. (In South Africa, the 1976 Soweto schools uprising was still being driven by this perception.) Hence, when the Geneva Conventions of 1949 were established, exploiting the services of children in armed conflict was not deemed a particularly harmful and therefore illegal act. In the eyes of many offenders, it is not necessarily morally or legally wrong. Some still consider it a public duty for children to take up arms, often claiming that this duty is endorsed by cultural tradition.

The first step in any strategy to eliminate a social vice must therefore be a campaign to transform perceptions. What might originally have been perceived as permissible (and even laudable) must now be seen as inherently destructive. In this respect, remarkable progress has been made over the last 50 years. But deeming a previously legitimate activity wrongful is only the first stage in a longer process.

The second step is to generate an awareness of just how serious the practice is. This is hardly a simple task, given the fact that using children as combatants is only one evil of the many that are perpetrated in times of war. In 1977, when the Protocols to the Geneva Conventions came into force, recruiting children was considered wrongful but not a "grave breach" of humanitarian law, attracting criminal liability. Recent studies undertaken by NGOs and the UN have done much to focus attention not only on the immediate sufferings of children but on the long-term social consequences of involving children in a war effort. (The image of the "lost generation" has been especially apt in this regard.) Current efforts marshalled around the draft protocol to the CRC, together with the appointment of a special representative of the UN Secretary-General for Children in Armed Conflict,114 have done much to continue publicising the issue.

Merely prohibiting the military exploitation of children, however, is no more than a declaration of public outrage. The third step is to give effect to the prohibition. Given the limited enforcement procedures in human rights law, the best way of accomplishing this goal was to bring criminal prosecutions against specific offenders. Thereafter, in order to ensure consistent enforcement, the initiative had to be removed from individual states - few of which were likely to have any interest in pursuing offenders - and vested in a permanent international tribunal.

Securing the arrest of offenders is the fourth and last step. The obvious purpose of instituting an international criminal court is to bring present offenders to book and thereby to deter future offenders - the enduring aim of all criminal justice systems. Deterrence depends on successful arrest being a probability, not a mere possibility. In the case of the ICC this goal may be achieved indirectly, for, once the Court is operating, fear that it might supplant municipal prosecutions could have the effect of spurring states into taking independent action.115

The future ICC therefore represents the most promising way of suppressing the practice of recruiting children. Nevertheless, the ICC is far from being operational,116 and, even when it is, the Court is no proof against the offence persisting. The draft statute itself may serve to encourage persistence. Admittedly, the instrument allows no reservations,117 but it does allow states on becoming party to declare that, for a period of seven years after their accession to the statute, they will not accept the jurisdiction of the Court over crimes referred to in art 8.118

Dissatisfaction with this staying clause (and with other aspects of the draft statute) has provoked thought about an ad hoc tribunal - possibly run on a regional basis119 - dedicated to prosecuting the crime of exploiting children in armed conflicts. Even states that might be hesitant about accepting the ICC will probably be more willing to support the cause of children’s rights. (The CRC, for instance, attracted almost immediate and universal backing.) With governmental support, the idea of a specialised international tribunal would be viable, especially if different interest groups, notably those working in human rights and humanitarian law, combine forces to back the project.

Endnotes

General Assembly Resolution A/RES/48/157.



The Impact of Armed Conflict on Children, A/51/306, UN Department of Information, New York, (1996).



See G Van Bueren, The International Law on the Rights of the Child, Kluwer, Netherlands, 1995, p.328.



A finding that is borne out by numerous other studies. See, for instance, Van Bueren, ibid., on child casualties in Africa.



Paragraphs 22 and 23 of the Machel Report, op. cit.



Report of the Secretary-General to the Security Council on Work of the Organisation, April 13, 1998, A/52/871, paragraph 49.



Statement made by Olara A Otunnu, Special Representative of the Secretary-General of the UN for Children and Armed Conflict, at an open debate in the Security Council on Children Affected by Armed Conflict, 29 June 1998, New York.



This and other information included in the Machel Report was drawn from a study by R Brett et al, Children: The Invisible Soldiers, Report on the Participation of Children in Armed Conflict and Internal Disturbances and Tensions, Quaker UN Office and the International Catholic Child Bureau, Geneva, 1996.



Paragraph 27 of the Machel Report, op. cit.



Vocational Training Systems of the International Labour Organisation, The Reintegration of Young Ex-Combatants into Civilian Life, Expert Meeting on the Design of Guidelines for Training and Employment of Ex-Combatants (Africa Region), ILO, Geneva, 1995, Appendix 3. And see Brett et al, op. cit., p.88.



UNICEF has been active in creating awareness of these abductions and the sufferings of the victims. Thanks in part to UNICEF’s publicity campaign, the UN Human Rights Commission passed a resolution on April 22, 1998 condemning the "abduction of children from northern Uganda".



Vocational Training Systems of the ILO, op. cit., p.5



Human Rights Watch/Africa & Human Rights Watch Children’s Project, Easy Prey. Child Soldiers in Liberia, Human Rights Watch, New York, Washington, LA, London, Brussels, 1994.



Paragraphs 29ff and 49ff of the Machel Report, op. cit.



Notwithstanding a dynamic interplay between international and municipal law, most systems of municipal law do not ascribe rights derived from international law to individuals, unless the rights were specifically incorporated into the municipal system.



Islamic law, too, forbids children under 15 from participating in a jihad: Van Bueren, op. cit., p.334.



Article 13.



Article 14.



Article 23.



Article 4.



Article 50.



Article 51.



Hence, under art 38(5), they "shall benefit ... to the same extent as the nationals of the State concerned".



Article 51.



Article 68.



Article 2.



Article 1(4). Certain combatants fighting in internal conflicts may also benefit from protections offered by the Protocol under art 44(3).



Article 51.



Other provisions in Article 77(1) require "special respect" for children and protection "against any form of indecent assault". If arrested or detained, art 77(4) and (5) require that children must be held in quarters separate from adults and those who are under 18 may not be subjected to the death penalty.



See H Mann, International Law and the Child Soldier, International & Comparative Law Quarterly, 36, London, 1987, p.32 at pp.388ff for a history of the drafting of these provisions.



Y Sandoz et al (eds), Commentary on the Additional Protocols, ICRC, Martinus Nijhoff, Geneva, 1987, pp.900-1.



In arts 43(2) and 51(3) of Protocol I and art 13(3) of Protocol II. The latter two articles provide that civilians enjoy protection unless and until they take a direct part in hostilities.



See Mann, op. cit., pp.45-7 for the following text.



R Brett, Child Soldiers: Law, Politics and Practice, International Journal of Children’s Rights, 4, Kluwer, Netherlands, 1996, p.115 at p.124.



For these reasons, the International Commission of the Red Cross proposal omitted the term "direct". See Sandoz et al, ibid., p.901.



Article 1(1).



Article 1(2).



As of 31 December 1997, 188 states had ratified the Conventions, 148 Protocol I and 140 Protocol II.



Notably the Geneva Declaration of the Rights of the Child (1924), the UN General Assembly Declaration on the Rights of the Child (1959) and its Declaration on the Protection of Women and Children in Emergency and Armed Conflict (1974).



See C Hamilton & T Abu El-Haj, Armed Conflict: The Protection of Children under International Law, International Journal of Children’s Rights, 5, Kluwer, Netherlands, 1997, p.1 at p.25.



Contained in art 2 of the CRC.



Article 24.



Article 27.



Article 28.



Article 38(1).



Article 39.



See S Detrick (ed), The United Nations Convention on the Rights of the Child: a guide to the "travaux préparatoires", Kluwer, Netherlands, 1992, pp.502ff for the drafting history of this provision.



If a state is also party to the Geneva Protocol II, however, the Protocol obligations prevail over art 38, since art 41 of the CRC provides that: "Nothing in the present Convention shall affect any provisions which are more conducive to the realisation of the rights of the child ...."



See Hamilton & Abu El-Haj, op. cit., p.36 and Brett, op. cit., pp.115ff.



See L J LeBlanc, The Convention on the Rights of the Child: UN Lawmaking on Human Rights, University of Nebraska Press, Lincoln and London, 1995, pp.150ff.



Article 2.



Articles 13, 14 and 15, respectively.



Although some commentators have argued that, in times of armed conflict, only art 38 applies, the Committee on the Rights of the Child declared that the whole Convention remains applicable. See Hamilton & Abu El-Haj, op. cit., pp.37-8.



And the common art 3 in the Four Conventions applies to "non-international" conflicts.



Brett, op. cit., pp.126-7.



Paragraph 222 of the Machel Report, op. cit.



The drafts are presented in J van der Schatte Olivier et al, (eds), Child Soldiers, Proceedings of a Conference organised by the Dutch Coalition for the Rights of the Child in Armed Conflict, November 1997, Save the Children, The Hague, 1998, Annex 2.



Van Bueren, op. cit., p.335.



Article 30 of the Constitution of the former Zaire, February 15, 1978, as amended by Law 90-002 of 5 July 1990.



Reuters News Service, August 14, 1998.



Xinhua News Agency, August 31, 1998.



Exceptions dating from the 18th century were made for pirates and slave-traders: individuals guilty of these offences were held personally responsible.



Article 60 of the Vienna Convention on the Law of Treaties (1969) prohibits the suspension of humanitarian treaties in reprisal for prior breaches.



International Labour Organisation Convention No. 138 (1973). Thus, in 1993, the United States boycotted products from Bangladesh on the grounds that they had been made by children under the age of 12 in contravention of the ILO Convention.



Created by art 43 of the Convention.



Customary law evolves from the practice of states, ie a constant and uniform usage over a period of time, coupled with a sense that the usage is legally obligatory, not a prescription of mere etiquette or morality.



Articles 49, 50, 129 and 146 of the four Conventions.



Article 85.



Article 147 of the Fourth Convention.



It had been encoded in art 23 of the Hague Convention IV (1907).



Article 77.



Article 85 of Protocol I.



The jurisdiction is "extraordinary" because a state may prosecute offenders for committing crimes outside its borders against persons who were not its nationals. In a Canadian case, R v Finta, (1994) 1 Supreme Court Reports, Ottowa, p.701, for example, the accused was charged with committing war crimes in Hungary in 1944. The court held that extension of its normal jurisdiction was permissible, because international law recognised the special competence of states over this category of crime.



Israel’s spectacular abduction of Eichmann from Argentina to stand trial for war crimes in Israel (Attorney-General of the Government of Israel v Eichmann, reported in International Law Reports, 36, Cambridge, 1961, p.5) involved violation of Argentine sovereignty, for which Israel duly apologised.



Which involved a finding that the situation in Bosnia-Herzegovina constituted a threat to international peace and security. For a general history of the Yugoslav and Rwandan tribunals, see the special issue of the International Review of the Red Cross, 321, Geneva, 1997, pp.601ff and for the relevant instruments see the Criminal Law Forum, 5, Camden New Jersey, 1994.



By Resolution 808 of 1993. Because the Council’s resolution was taken under Chapter VII of the Charter, states were prima facie obliged to accept the tribunal. Nevertheless, the competence of the Security Council to create such an institution was challenged (unsuccessfully) in Prosecutor v Tadic (Jurisdiction), International Law Reports, 105, Cambridge, 1997, p.419.



Articles 2, 3 and 5 of the tribunal’s statute, respectively.



See P A Khaven, The International Criminal Tribunal for Rwanda: The Politics and Pragmatics of Punishment, American Journal of International Law, 90, Washington DC, 1996, p.501.



Because the "grave breaches" provision in the Geneva Conventions was not applicable, art 4 of the tribunal’s statute specifically included violations of art 3 common to the Geneva Conventions and Protocol II.



1998 A/Conf.183/9, reproduced in International Legal Materials, 37, Washington DC, 1998, p.999. 120 states voted in favour, seven against and 21 abstained.



Article 4(2). To avoid undue conflicts with national courts, it was generally agreed before the Rome conference that the jurisdiction of the Court should be limited to only serious crimes in which the international community as a whole had an interest.



Article 1.



Article 5(1).



Article 25(2).



War crimes are defined in art 8(2)(a) to include grave breaches of the Geneva Conventions.



Article 8(2)(b)(xxii). These offences are defined in more detail in art 7, which deals with "crimes against humanity". Thus art 7(1)(g) prohibits "rape, sexual slavery, enforced prostitution, forced pregnancy ... or any other form of sexual violence of comparable gravity".



Article 8(2)(b)(xxvi).



Article 8(2)(b).



Article 8(2)(c) provides that the relevant offences in this situation are serious violations of art 3 common to the Geneva Conventions.



Article 8(2)(e)(vii).



Article 8(2)(f).



Coalition to Stop the Use of Child Soldiers, Stop Using Child Soldiers!, Rädda Barnen, on behalf of Save the Children Alliance, London, 1998, p.18.



The term "recruit" was omitted, perhaps unfortunately, since it would imply both voluntary and compulsory enrolment. See M T Dulti, Captured Child Combatants, International Review of the Red Cross, 278, Geneva, 1990, p.421 at p.424.



When interpreting this offence, account should be taken of art 12 of the CRC which provides that: "States Parties shall assure to the child who is capable of forming his or her own views the right to express those views freely in all matters affecting the child ...."



See discussion by the International Commission of the Red Cross, The Involvement of Children in Armed Conflict, International Review of the Red Cross, 322, Geneva, 1998, p.105 at pp.117-19.



Article 9(1).



Article 21(1).



Article 21(3).



Thus, art 68(1) provides that the Court must take "appropriate measures to protect the safety, physical and psychological well-being, dignity and privacy of victims and witnesses". In so doing, the Court must take into account factors of age and gender, especially "where the crime involves sexual or gender violence against children". Article 68(2) therefore allows Chambers of the Court to conduct proceedings in camera.



Articles 37(c) and 40(3) of the CRC, for example, lay down the principle that children charged with offences should be treated "in a manner which takes into account the needs of persons of [their] age".



Individual interests are not entirely ignored, however, since art 75 of the draft statute requires the Court to establish principles allowing victims to be given restitution or compensation for their injuries.



Article 86. Moreover, under art 88, states must arrange their municipal laws to ensure cooperation.



Article 12(2).



Article 15.



Articles 9 and 8, respectively, of the instruments constituting these tribunals, declared that they had primacy over national courts. This principle was confirmed in Prosecutor v Tadic (Jurisdiction), International Law Reports, 105, Cambridge, 1997, p.419 at pp.479-84.



See W A Schabas, Justice, Democracy and Impunity in Post-genocide Rwanda: Searching for Solutions to Impossible Problems, Criminal Law Forum, 7, Camden New Jersey, 1996, p.523 at pp.554-5.



Article 13(a) and (c), as read with art 14. States’ jurisdiction is also superseded, under art 13(b), if the Security Council directs prosecution in terms of Chapter VII of the United Nations Charter.



The situation that arose in the case of Libya v USA, reported in International Legal Materials, 37, Washington DC, 1998, p.587. The United States called on Libya to surrender two suspects who had been indicted in the USA and Scotland for causing the destruction of an American aircraft over Lockerbie. Libya resisted the claim on the grounds that the Montreal Convention for the Suppression of Unlawful Acts against the Safety of Civil Aviation (1971) entitled it to try the offenders itself.



Article 17(1)(a). Article 17(2) provides that, in order to determine "unwillingness" in a particular case, the Court must consider whether independent and impartial proceedings were actually being undertaken, whether the accused person was being shielded and whether there was an unjustifiable delay.



G S Goodwin-Gill & I Cohn, Child Soldiers: The Role of Children in Armed Conflict, Henry Durant Institute, Geneva, and Oxford University Press, 1994, pp.53-75.



Under art 51 of the Fourth Geneva Convention and art 3 of the ILO Convention No.138 (for working with hazardous materials).



Brett, op. cit., p.123.



Although arts 7 and 8 of the CRC oblige states party to register births and provide identification.



Namely, Olara A Otunnu. See Report of the Secretary-General to the Security Council on Work of the Organisation, 13 April 1998, A/52/871, paragraph 52.



T Meron, The International Criminalising of Internal Atrocities, American Journal of International Law, 89, Washington DC, 1995, p.554 at p.555.



According to art 126 of the draft statute, the treaty will come into force when 60 ratifications have been made.



Article 120.



Under art 124, provided a crime is alleged to have been committed by one of its nationals or on its territory. What is more, art 11 provides that the Court’s jurisdiction will be prospective, namely applicable only to crimes committed after the treaty enters into force for a state ratifying the statute.



The Charter on the Welfare and Rights of the African Child, for instance, could provide the foundation for an African tribunal operating under the auspices of the Organisation of African Unity.