Indirect Occupation and the ICTY: It’s Not So Complicated

[Radhika Kapoor is a Harvard Kaufman Fellow at the Public International Law and Policy Group, Washington, D.C.].

The notion of belligerent occupation is of fundamental importance to international humanitarian law in its role as a threshold requirement for relevant provisions of the Geneva Conventions of 1949, which contain the standards for humanitarian treatment during conflict. Traditionally, a territory was considered ‘occupied’ when a foreign power exercised effective control over it. However, in contemporary conflicts, countries often act indirectly, exerting force over foreign territories through non-state agents. While the law of occupation would continue to be applicable to these cases, the contours of such occupation by proxy – or ‘indirect occupation’ – where a foreign power occupies territory in another state through an intermediary group, are unclear in international law. In particular, there is uncertainty regarding the level of control necessary to demonstrate the existence of a state of indirect occupation. Delineating the precise scope of indirect occupation can be crucial to demystifying the legal status of indirect foreign control in multiple modern contexts, including the presence of Russia in Moldova’s Transdniestria, Armenia in Nagorno-Karabakh, and Russia in South Ossetia and Abkhazia.

The approach of the International Criminal Tribunal for the Former Yugoslavia (ICTY) is exceptionally relevant to comprehending indirect occupation; during the Yugoslav wars, Croatia and the Federal Republic of Yugoslavia were alleged to have regularly implemented military operations in Bosnia and Herzegovina through armed groups titled the Croatian Defense Council (HVO) and the Bosnian Serb Army respectively. Nonetheless, the ICTY stopped short of pronouncing a definitive test for determining the existence of a state of indirect occupation. Even when squarely confronted with the issue, the Appeals Chamber refrained from unambiguously asserting the constitutive elements of indirect occupation, listing instead a “number of factors [indicating] that Croatia through the HVO had actual authority over the relevant municipalities.”

Although the jurisprudence of the ICTY Trial Chamber has been similarly hazy, I submit that it does provide instructive starting points which can assist in articulating a clear test for indirect occupation. While assessing whether a state of occupation existed in various municipalities of Bosnia and Herzegovina, the Prlić Trial Chamber explicitly endorsed the signifiers of control outlined by the Naletilić Trial Chamber (Naletilić Factors) as indicative of the actual authority of an occupying power over a territory. However, in arriving at its conclusion, the Trial Chamber stated that “occupation by the HVO [could] be established, inasmuch as Croatia/the HV wielded overall control over the HVO.” This statement is erroneous because the Trial Chamber was required to consider whether there was occupation by Croatia – and not HVO. Further, this conclusion makes no reference to the Trial Chamber’s findings regarding the Naletilić Factors, which it deemed determinative of occupation.

I suggest that the assessment of indirect occupation be divided into two key components: determining whether a state of occupation exists on the ground, and then, uncovering the identity of the occupying power. Although these are entirely separate questions, the distinction between them in extant ICTY jurisprudence is blurry. In Naletilić, for instance,the Prosecution relied on the following passage from the Blaškić Trial Decision to support its claim that the overall control test formulated by the ICTY Appeals Chamber in the Tadić Appeal  could be used to assess the level of control necessary for occupation:

 “…by using the same reasoning which applies to establish the international nature of the conflict, the overall control exercised by Croatia over the HVO means that at the time of its destruction, the property of the Bosnian Muslims was under the control of Croatia and was in occupied territory.”

The Naletilić Trial Chamber rejected this argument, holding that occupation required a higher degree of control than ‘overall control’, and finding that the Blaškić Trial Chamber erred in employing Tadić’s overall control test in its evaluation of occupation. In particular, it emphasized that there was an essential distinction between the determination of occupation and that of the existence of an international armed conflict, and that Tadić’s overall control test ought to be applied only to the latter.

In fact, a comprehensive analysis of the existence of a state of indirect occupation requires the employment of both, the Naletilić Factors and Tadić’s overall control test. A contextual appraisal of the above extract from the Blaškić decision makes it evident that the Blaškić Trial Chamber did not actually apply the overall control test to determine whether a state of occupation existed. In fact, Tadić’s overall control test was used to identify who the occupying power was. In the relevant portion of the decision, the Trial Chamber’s conclusion was that “Croatia played the role of occupying power through the overall control it exercised over the HVO.” To arrive at this inference, the Trial Chamber referred to the property of the Bosnian Muslims as lying on ‘occupied territory’ from the outset, indicating that ‘overall control’ here was not utilized to assess the existence of a state of occupation (which was treated somewhat as a foregone conclusion), but to unearth the identity of the occupying power. A similar approach was employed, for instance, in DRC v. Uganda, where the International Court of Justice sought to determine whether Uganda, through its control over active armed groups, was the relevant occupying power.

Support for applying the overall control test to determine the identity of the occupying power can also be located in the Trial Chamber’s approach in Rajić. Since this decision preceded the Tadić Appeal, the overall control test did not exist in its present form. Nevertheless, it is noteworthy that the Rajić Trial Chamber used separate tests to determine: (a) the level of control exercised by Croatia over the Bosnian Croats, and (b) whether the level of control exercised over the territory of Stupni Do was sufficient to constitute “occupation”. Rajić used a rudimentary form of Tadić’s overall control test to determine the former, and criteria similar to the Naletilić Factors to determine the latter.

Accordingly, with a view to disentangling the approaches variously taken in relation to the issue of occupation, the optimum way to address indirect occupation would be to explicitly adopt a two-step analysis:

Step One: Is the control exercised over the territory sufficient to constitute occupation? Although the Geneva Conventions do not define the term ‘occupation’, a definition can be found in Article 42 of the Hague Regulations, referring to the exertion of actual authority by a hostile group over a territory. The Naletilić Factors identified and endorsed by the ICTY Trial Chamber may be used to determine whether such authority has actually been established over the concerned territory.

Step Two: Who is the occupying power? Where an armed group under the overall control of a foreign State occupies territory, the occupation must be attributed to the foreign State. Admittedly, Tadić’s overall control test is less rigorous than Nicaragua’s ‘effective control’ test for determining the degree of control exercised by a foreign State over an armed group. Tadić propounded a lower threshold because it found that international humanitarian law was not “grounded on formulistic postulates”and that internal groups could not be permitted to violate humanitarian law norms merely because they did not hold formal positions of authority. Indeed, there could be instances where private individuals/ groups acted on behalf of a foreign State and wielded de facto power. Where such private individuals constituted an organized and hierarchically structured group, the overall control of a foreign State over the group would be sufficient to attribute the group’s acts to the State, and to conclude that the group acted on behalf of the State. This was reiterated in Aleksovski, where it was found that the “activities” of a group under the overall control ofa State could be attributed to the State, regardless of whether any specific instructions were supplied to members of the group. In the context of occupation, therefore, where a foreign State acts de facto through an internal group (by exercising overall control over it) to occupy territory, it must not be permitted to disassociate itself from the occupation: the occupation must be attributed to the foreign State, which must then be considered the true occupying power.