17 Aug Emerging Voices: The Role of Attribution Rules Under the Law of State Responsibility in Classifying Situations of Armed Conflict
[Remy Jorritsma (LL.M.) is a lecturer in public international law at the Department of International and European Law of Maastricht University. In September 2015 he will join the Max Planck Institute Luxembourg as a Research Fellow/PhD candidate. Contact at firstname.lastname@example.org.]
Armed conflicts involving e.g. Ukraine/Russia, Israel/Palestine, and the self-proclaimed Islamic State demonstrate legal ambiguities with regard to State responsibility as a result of the State exercising control over organized armed groups. Under customary international law an act by a non-State actor is attributable to a State if, inter alia, the latter exercises a certain level of control over the former.
However, it is unclear how much control is required for attribution. Equally unclear is the exact function of attribution in relation to the application of international humanitarian law (IHL). At stake is the issue whether the secondary rules of attribution may assist in classifying the armed conflict, thereby determining the framework of primary rules in which hostilities take place: the rules of international armed conflicts (IACs) or non-international armed conflicts (NIACs). The line of case law responsible for this debate is often characterized as a conspicuous example of “fragmentation” of international law (see here, here; but see with more nuance here).
In the Nicaragua case the ICJ assessed whether the acts of the contras could be attributed to the US for the purpose of State responsibility. The Court set the required level of control as follows:
‘[Even decisive participation] in the financing, organizing, training, supplying and equipping of the contras, the selection of its military or paramilitary targets, and the planning of whole of its operation, is still insufficient [. I]t would in principle have to be proved that that State had effective control of the military or paramilitary operations in the course of which the alleged violations were committed.’ (§115, emphasis added.)
The Nicaragua test did not convince the ICTY Appeals Chamber when it had to ascertain the nature of the armed conflict between Bosnia and Herzegovina and the Federal Republic of Yugoslavia, acting after 19 May 1992 through the Bosnian Serb Army. The Appeals Chamber considered when an organized armed group fighting in a prima facie NIAC acts on behalf of another State – or ‘belong[s] to a Party’ in the sense of Article 4(a)(2) of Geneva Convention III – with the result that the conflict is internationalized and thus subject to IAC law. It held this to be the case if that State exercises…
‘overall control, going beyond the mere financing and equipping of [organized armed] forces and involving also participation in the planning and supervision of military operations. [It is not required] that such control extend to the issuance of specific orders or instructions relating to single military actions.’ (§145)
The next round of this ‘dialogue des sourds’ (Simma 2009, p.280) came when the ILC cited with approval the Nicaragua test in its Commentary (§5) to Article 8 of the Articles on State Responsibility, apparently discarding the overall control test. Later in the Lubanga Confirmation of Charges Decision the ICC approved (§211, later endorsed in the Lubanga Judgment at §541) the overall control test for the purpose of determining the nature of the conflict. Finally, the stamp of approval for the effective control test comes when the ICJ in Bosnian Genocide explicitly rejects (§403) the Tadić test of overall control for the purpose of State responsibility. Oddly, and perhaps even merely as a ‘gracious concession to the ICTY’ (Cassese 2007, p.651), the Court in Bosnian Genocide somehow purports (§405) to conceal this conflict by suggesting that the level of control required for the internationalization of a NIAC can ‘without logical inconsistency’ differ from the degree of control required for attribution in terms of State responsibility. The Court thus leaves open the possibility that the less-demanding test of overall control may be used to bring about an international(ized) armed conflict.
This line of case law reveals a fundamental difference of opinion as to if, and how, questions on State responsibility and conflict classification can be answered through a process of attribution of conduct. Two specific points of conflict can be observed: (1) disagreement as to the level of control required for State responsibility, and (2) disagreement as to whether conflicts can be classified by “borrowing” from the law on State responsibility. Bosnian Genocide creatively seeks to avoid the appearance of an all-out confrontation with the ICTY by acknowledging its subject-matter expertise and by approving of its approach in matters of conflict situation. However, what was meant to avoid or minimize the appearance of fragmentation actually ended up exacerbating it. There are three possible ways of looking at the overall/effective control debate (cf. Report of the Study Group on Fragmentation, §43-50):
- As involving a conflict between two different interpretations of international law: effective control versus overall control for State responsibility;
- As involving a conflict between the general law and a particular rule that claims to be the lex specialis exception to it: effective control for State responsibility but overall control for State responsibility in situations of armed conflict;
- As involving no conflict at all, because the cases can be distinguished on basis of facts: Nicaragua/Bosnian Genocide concerned State responsibility, whereas Tadić concerned classification of armed conflict.
Particularly the ICJ’s disconnection between State responsibility and classification of conflict by differentiating between various functions of attribution is problematic for a number of reasons.
First, the ICJ’s suggestion that State responsibility is conceptually distinct from conflict classification appears to depart from its earlier practice in Nicaragua. Here the Court held that the actions of the contras in relation to Nicaragua, not imputable to the US, were subject to the law of NIACs, whereas the actions of the US itself (e.g. issuing a guerrilla warfare manual to the contras) were subject to the law of IACs. Without explicitly saying so, Nicaragua suggested a close link to exist between attribution of conduct and classification of armed conflict, making it unlikely to maintain that these questions are very different in nature. The lack of attribution through effective control meant that the conduct of the contras was de facto and de jure their own. Conversely, should the Court have found that the contras’ acts were legally attributable to the US, it is expected that their actions be assessed in light of the rules of IACs (cf. § 215 and 254).
Second, the primary rules of IHL also demonstrate a clear connection between attribution, State responsibility, and classification of conflict. An IAC subject to the rules of IHL exists whenever there is an armed conflict between States, whereas NIACs are armed conflicts in which at least one of the belligerent parties is not a State. In the Bemba Confirmation of Charges Decision the International Criminal Court held (§223) that an IAC exists ‘in case of armed hostilities between States through their organs or other actors acting on behalf of the State.’ Having in mind that conflicts are distinguished by the parties involved (Zegveld 2002, p.136), it becomes clear that ‘a determination of attribution will affect the classification of conflict’ (Somer 2006). The legal process of attribution in armed conflict situations defines who the belligerent parties are, and this in turn determines the applicable law in light of which the lawfulness of the belligerent parties’ behaviour must be assessed. This interpretation of IHL by reference to general rules of international law (Art. 31(3)(c) VCLT) arrives at a solution that is the most coherent when looking at IHL as such, and at IHL within the wider system of public international law.
Third, more generally, attribution rules are relevant to ‘define the conditions upon which the primary rules applies’ (Gaja 2014, p. 989). They are ‘transsubstantive’ rules (Caron 1998, p.128) which permeate the content and scope of primary rules. Primary rules may be addressed to specific legal subjects only and secondary rules of attribution rules then serve to determine whether the conduct complained of can be imputed to the addressee of the norm. This, of course, only unless a special law determines otherwise (cf. Art. 55 ARSIWA). If a State exercises control over a non-State actor this triggers the application of fields of law made to regulate State behaviour.
The ICJ’s separation of control for conflict classification from control for State responsibility rigidly adheres to a strict separation between primary and secondary rules. By suggesting to agree on a minor point with the ICTY the ICJ actually obscured the bigger picture and created more legal uncertainty. Also, a less-demanding test for conflict classification being distinct from State responsibility opens the door for States to fight a proxy IAC without being responsible towards the victims for possible violations committed in the course of that conflict. This is unacceptable in light of IHL’s emphasis on responsible command and at odds with the general structure of the law on State responsibility which logically places attribution before establishment of a breach or legal consequences.
At the end of the day, the ICJ and ICTY/ICC’s diverging approach towards the function of State control over a non-State actor may just as well be regarded as a case of concealed confrontation, or at the very least a judicial dialogue that feigns some agreement in order to downplay the actual extent and form of fragmentation.