About the author(s):
Laura Inigo Alvarez’s book ‘Towards a Regime of Responsibility of Armed Groups in International Law‘ emerges out of the observation that while there has been significant focus in recent years on the primary obligations of armed groups, there remains quite a degree of uncertainty regarding the secondary rules by which actions can be attributed to armed groups. In the panel that was convened by ALMA to mark the launch of her book, I was asked to look at questions relating to rebel governance. With that instruction in mind, in this blog post I will look at some legal questions that arise out of the factual reality of armed groups controlling territory and providing services to the civilian population.
Structural disparities between armed groups
One major strand of enquiry in Laura’s book relates to figuring out how the structural disparities of different armed groups should be taken into account, when addressing this question of attribution. Laura correctly indicates that although there have been a number of shorter studies on this topic, no research has been done that connects the responsibility of armed groups with an analysis of their organizational structure. In addressing the question of structural disparity, Laura points out that there are groups that adopt a state-like structure and those that follow looser structures (p122-3). She comments that armed groups that adopt state-like structures are characterized by hierarchy and centralized command. Groups with a looser configuration follow non-centralised structures. She indicates that there is much work to be done in studying the characteristics of these different kinds of groups, not only so that clarity can be gained on how international legal norms can be applied to them but also because the structure of a group is often an ‘important determinant of their behaviour’ (ICRC 2018).
Rebel governance and primary norms
Scenarios where armed groups control territory and exercise governance – i.e. the scenarios that I’ve been asked to address – are in some ways the most straightforward when it comes to legal analysis. They are the easiest for international lawyers to address by way of international norms. The five point organization criteria put forward by the Boskoski trial judgment and the case law from the Kosovo conflict emerged out of groups with vertical structures. In these instances, it is fairly easy to show that an armed group is a single entity rather than a band of loosely aggregated individuals, not least because the strategic make-up of the group is designed to achieve the aim of centralization and to project external unity. This is far from the situation with non-vertical or networked groups where elaborate steps may be taken by the group to prevent outsiders discerning unity between what appear to the outside world as atomized units.
Rebel governance and secondary norms
From a secondary norm perspective too, groups controlling territory allow a logical analogy to be made with the Articles on State Responsibility for Internationally Wrongful Acts. When groups carry out governance activities in the spaces in which they control, they have most similarities with States. They control territory, they have a population, they administer justice and they provide public services, such as healthcare and education. But although these similarities might seem to suggest a straightforward comparison, it in fact leads to one potential legal obstacle that Laura points out in her book and which I would like to focus on here.
How to conceptualise armed groups for the purposes of attribution?
In her analysis of attribution, Laura points out that the first task of any attribution exercise is to determine the concept of armed group. To guide this analysis, she utilises the term organized armed group noting that it is generally understood to refer exclusively to the armed or military wing of a non-state party to a non-international armed conflict (p123). It has been said not to include those segments of the civilian population that are supportive of the non-state party such as its political wing (definition provided in the How Does Law Protect in War glossary of terms under ‘organised armed group’). I think that it can be agreed that this definition comes mainly from the ICRC’s DPH Study in 2009, where it took pains to define the term ‘organised armed group’ so that it could later define its membership for the purposes of targeting (DPH Study 2009, p32).
To add further support to her use of this definition in her study, Laura refers to scholarship indicating that non-international armed conflicts are fought between armed forces of states (‘states’) and the military wings of non-state armed groups (‘non-state armed groups’ or ‘armed groups’) or between armed groups’ (Sivakumaran, p3). Stated in these terms, one may easily conclude – as Laura seems to – that the military wing of an armed group is the non-State party to a non-international armed conflict. Although Laura does not draw on these sources, such an approach could feasibly also be supported by an analysis of the ‘organization requirement’ under international humanitarian law in international criminal law case law. This takes little account of an armed group’s civilian infrastructure (probably because civilian infrastructure is generally not necessary for a group to adhere to international humanitarian law).
If one has the DPH definition of OAG in hand, one might find further support for such a position in the text of Additional Protocol II. The term ‘organised armed group’ (a term which as already stated is defined in the DPH study as referring only to the military wing) is found Article 1 of Additional Protocol II which states that the Protocol applies to armed conflicts that ‘take place in the territory of a High Contracting Party between its armed forces and dissident armed forces or other organized armed groups under responsible command, exercise control over a part of its territory etc’. If Article 1 of APII identifies the parties to a non-international armed conflict – and the term ‘organised armed group’ is understood as referring only to the military wing – then this Article adds further weight to the idea that the military wing (so defined) is indeed the party.
Some possible problems with this approach
But if this is a good reading of APII, it leads to a significant inconsistency between international armed conflicts and non-international armed conflicts. For while the parties to international armed conflicts are High Contracting Parties i.e. States, the parties to a non international armed conflicts are only the armed or military wing of the armed group. This reading produces several legal problems, at the level of primary norms. It would be strange to consider that only a State’s armed forces were bound by the Geneva Conventions or Additional Protocol I or II. I would like to suggest that it is equally strange to consider that only the military wing of the armed group is bound by the norms contained in Additional Protocol II, in circumstances where such a civilian wing exists. Such a reading of APII raises questions about how and whether the civilian wing of an armed group is bound by the provisions of Additional Protocol II, if they are not part of the legal entity that constitutes the party that is bound by it. This question is particularly important when one examines provisions in APII like Article 3(a) which places an obligation upon the parties of the armed conflict to provide an education, a service that would tend not to be provided by members of the military wing (and is not necessarily thought to be equally binding upon the group’s individual members). It also sits uncomfortably with provisions relating to detention and prosecution, as it is known that the courts and prisons of some armed groups are run by their civilian wings – even when they are detaining or prosecuting people for reasons connected to the armed conflict.
At the level of secondary norms, such a reading ushers in the attribution problem that Laura hints at in her book. If the party to an armed conflict is a State/ High Contracting Party, that overarching entity can bear responsibility for acts or omissions carried out by its military, administrative and legislative organs that constitute a violation of the Conventions (i.e. consider for example a failure to legislate in a manner that provides effective penal sanctions for grave breaches). But if the ‘party’ to the armed conflict is only the military wing of the armed group, it becomes uncertain whether that military entity can bear responsibility for acts carried out by its civilian wing (when it has one). In other words, if an armed group has a sophisticated ministry of justice that is separate from the military wing – could the ministry of justice be considered an organ of that group, such as its violations of IHL could be attributed to it as its organ? Equally, could the acts of a civilian employed by that civilian wing and violating IHL be attributed to the organized armed group, if it is so narrowly defined? A flipped version of the same problem may also emerge at the end of the conflict. When the military wing has been extinguished, could the remnant civilian wing be able to held responsible for the violations of the military wing – if it is treated as a separate legal entity that was never included in conceptual parameters of the original non-State party to the non-international armed conflict?
Room for a different approach?
I would like to suggest that we should be taking a different approach. A different approach could be achieved by understanding the term ‘organised armed group’ as having a different definition when it comes to thinking about ‘parties’ and the direct participation of hostilities debate where its parameters were first clarified as pertaining only to the military wing. This would provide room for an understanding that the parties to a non international armed conflict comprise of (i) the State party and (ii) the armed non State actor (comprising both military and civilian wing, if it exists) or two armed non State actors. Such an approach brings the advantage of closing the potential accountability gaps identified above.
It is notable that there are some indications that this is what the ICRC may have even meant in their DPH study. In that study, the text states: ‘organised armed groups belonging to a non-State party to an armed conflict include dissident armed forces and other organized armed groups’. The study also emphasizes the importance of distinguishing between a non-State party to a conflict (e.g. an insurgency, a rebellion or a secessionist movement) and its armed forces (i.e. an organized armed group). It also makes an explicit reference to “political and humanitarian wings” as being part of non-State parties (p33). So indeed, this suggests that the ICRC might agree with me. Though if so, it is unfortunate the term ‘organised armed group’ was used in the DPH study, because there is no doubt that its parallel use in Article I of APII creates potential for confusion on this point. I would be intersted to hear Laura’s thoughts on this issue?