For a civilian caught up in the crossfire of an international armed conflict, determining protected status under international humanitarian law (‘IHL’) represents more than a theoretical exercise of distant intellectualisation by jurists. It takes centre stage in the trenches of suffering, and is arbiter in deciding who receives the full protection of the Fourth Geneva Convention (‘GC IV’) or who is limited to the residual guarantees of Additional Protocol I reserved to those “who do not benefit from more favourable treatment under the Conventions” (Art. 75(1)).
GC IV’s text stipulates clearly that its protection applies to nationals of the adversary party to the conflict (Art. 4). Yet applying this provision in practice has curiously been anything but straightforward. The International Criminal Tribunal for the Former Yugoslavia’s (‘ICTY’) jurisprudence in the context of the Balkan wars, uncritically restated by the International Criminal Court (‘ICC’), introduced an expanded test of ‘allegiance’. It determines protected status not merely based on formal bonds of nationality, but rather based on a person’s demonstrable allegiance to a party through a range of other factors, significantly stretching GC IV’s scope of protection. Many authoritative commentators accepted this expansive interpretation without attempting to clarify its limits and interplay with the nationality test in the Convention’s text (Triffterer, Ambos 325; Dörmann 28-29; Schabas 240). Very few authors critically assessed the challenges of legal status and scope of the allegiance test (Martínez 13).
In the relentless quest to expand protections against horrendous crimes, there are those who view judicial creativity as being at the heart of international criminal judicial function (Powderly), while others consider it so misguided it vitiates international justice’s essence (Corsi). What is certain is, where the law’s frontiers are sought to be shifted, the ramifications only rear their head long after robes are removed, courtroom doors are closed, and ink has dried. Progressive jurisprudence, after all, does come with its perils.
Reconceptualising preconditions for protected status exemplifies this problem. Attempts to extend protection to novel situations can have the unintended consequence of wrenching it from others.
Consider Ukrainians in occupied territories. Does seeming allegiance to Russian authorities deprive these civilians of full IHL protection? There is inherent absurdity in having to assess the possible allegiances of millions of Ukrainians living under Russian occupation. Not to mention there is no threshold for what clearly establishes allegiance. Perhaps clear for enthusiastic collaborators, what about farmers who simply want to survive and sell their produce under occupant’s regulations? Or civil servants who accept roles in essential services provided by occupying institutions?
On the other end, this would mean collaborators in a state’s own territory (e.g., Ukrainian nationals in Ukrainian-controlled territory) are to benefit from IHL’s full protection. This is hardly a sensible application of the law, considering states patently did not want such a class of persons to be covered by GC IV (Darcy 119; ICRC 46).
Concept of ‘allegiance’
GC IV grants protection to civilians who find themselves in the hands of a Party to the conflict of which they are not nationals (Art. 4). Applying this test in the Balkan wars’ context, the ICTY faced a dilemma. As an international armed conflict was occurring between the forces of fledgling Bosnia and Herzegovina, Yugoslav federal forces and ethnic Serbian non-state groups under overall Yugoslav control – GC IV applied. However, since both Bosnian Muslim civilians and their Bosnian Serb persecutors possessed identical Bosnian nationality, conventional application of GC IV would have excluded civilians from full IHL protection.
Trying to resolve the dilemma, ICTY Chambers in Čelebići and subsequently in Tadić ruled that, instead of legal nationality, broader bonds of ‘allegiance’ (e.g., ethnic) with a warring party may be decisive in determining protected status. Tadić Appeals Chamber (‘AC’) (§164-169) started its reasoning by referring to GC IV’s drafting history, where “part of the drafters” intended to extend protected status to refugee nationals of an occupying power they previously fled (e.g., German Jews in occupied France). The AC considered this to indicate that even at the drafting stage, “the legal bond of nationality was not regarded as crucial and allowance was made for special cases [emphasis added]”. The AC corroborated this conclusion by Articles 44 and 70 GC IV protecting refugees’ right to asylum.
For the AC, in modern international armed conflicts where new states are often being created in the course of fighting and/or states intervene indirectly through control over non-state actors, nationality does not necessarily adequately define the scope of protection while other allegiance bonds, e.g., ethnicity, “may become determinative”. According to the AC, this reading was supported by GC IV’s object and purpose to “ensure protection of civilians to the maximum extent possible”.
For a long time, Tadić AC’s position was uncontested by many commentators (Hoffmann 502, fns. 20-22), other ICTY Chambers (Čelebići AC, §81-84; Blaskić AC, §172-182), and the ICC (Lubanga PTC, §277-280; Katanga PTC, §289-293). In Katanga, the ICC Pre-Trial Chamber (‘PTC’) determined an international armed conflict existed between the Democratic Republic of the Congo and Uganda (via the latter’s influence over non-state armed groups in Ituri province). In an analysis nominally identical to the prior Lubanga PTC, the Katanga PTC cited Tadić AC in supporting the conclusion that legal nationality is not a “crucial” or “definitive” test for determining civilians’ protected status, although the link between nationality and allegiance “remains an important factor”. As an example, the PTC provided a scenario when a warring party gradually gains control of an adversary’s village: civilians “automatically become” protected persons unless they claim allegiance to this attacking party.
Legal incoherency of the ‘allegiance’ requirement
Many readers will recall this is far from the first instance Tadić AC’s progressive reasoning and holdings are questioned. Notably, the International Court of Justice (‘ICJ’) in the Bosnian Genocide case was critical of Tadić AC’s attempt to impose its reinterpretation of the notion of state control over an armed group (i.e., ‘overall control’ test) into the state responsibility realm (§403-407). The ICJ particularly noted that Tadić’s judicial creativity was not required by “logic” and that it stretched central legal requirements “too far, almost to breaking point”.
Tadić AC’s replacement – in whole or in part – of the nationality test with the allegiance test seemingly represents the same problem. However, unlike the aforementioned example, it seems to have passed unnoticed by most.
It does not appear the allegiance test has solid basis in any IHL rules, preparatory works or subsequent practice. Tadić AC based its conclusions on the intention by “part of the drafters” to expand GC IV’s protection to refugees possessing an occupying power’s nationality. However, it is unclear how certain drafters’ views can be conclusive proof of the test’s status in existing law, especially with Article 4’s ordinary meaning clearly suggesting otherwise. Certain authors today claim the ICTY “misconstrued an object and purpose to fit their own needs and to justify their otherwise baseless theory [of allegiance]” (Martínez 13-14; Hoffmann 510, 514-517).
Moreover, Tadić AC’s reliance on the Convention’s Articles 44 and 70, which extend certain limited protections to refugees’ right to asylum, is unconvincing. Exceptions accepted by drafters are not evidence of a broader scope of application. Rather, explicit addition of exceptions to the broader rule suggests, given the very nature of exceptions, that they should be applied narrowly (Hoffmann 511).
The understanding demonstrated by Tadić AC of GC IV’s object and purpose also seems unreliable. A more nuanced reading of the Convention indicates ensuring protection of the population “to the maximum extent possible” was never the drafters’ intention (Martínez 13; Hoffmann 514-517). By introducing a rule on ‘protected persons’, drafters sought to limit protection “to a few special cases that merit extra care, while avoiding stronger burdens regarding the vast majority of civilians in armed conflicts” (Martínez 13). For example, states rejected the Convention’s early drafts aimed at guaranteeing protection to all civilians in order to relieve such burden and concentrate only on foreigners’ protection (Martínez 13).
Even contextually, Tadić seemed misguided. The assertion that modern international armed conflicts increasingly featured non-national, e.g., inter-ethnic, dimensions and associated emergence of new states, which were previously much rarer, seems a self-serving falsehood. The 20th century was replete with international conflicts whereby the emergence of new states led to armed forces sharing the same nationality but displaying different allegiances. The Russian Empire’s dissolution and subsequent conflicts in newly independent Ukraine and the Baltic states (1917-1920), the first Indo-Pakistani War (1947-1949), the Vietnam War (1955-1975), the Bangladesh Liberation War (1971), and the First Nagorno-Karabakh War (1988-1994) are primary examples of such conflicts – something Geneva Convention drafters would have been well aware of and that is by no means a “modern” phenomenon.
In any case, Tadić AC, subsequent case-law, and legal authorities failed to clearly set the contours of the allegiance test’s application, the necessary threshold for it to be established, and the relevant factors to be considered and possible exceptions. For instance, to what extent a state national’s real or perceived involvement with the enemy affects allegiance? Is this determination to be made separately for every individual civilian? Is there a presumption of allegiance to one’s state of nationality to be rebutted to lose protected status? In leaving aside these crucial questions, Tadić AC and subsequent affirming jurisprudence bequeath a test with dubious applicability.
The relationship between ‘nationality’ and ‘allegiance’ requirements
The Tadić allegiance test has doubtful roots and legal status. This calls, at minimum, for a significant reconfiguring of the test’s place within GC IV’s framework, if not outright repudiation. Focus will be placed on the former, as a full removal of the test seems is unlikely in practice.
Meaningful reconciliation is only possible considering several objective conclusions flowing from the jurisprudence and associated context. First, the allegiance test does not displace or replace the nationality test. Even where both the ICTY and ICC discussed the allegedly declining significance of the nationality test, both never claimed allegiance takes over nationality completely. Moreover, both recognised that nationality may still remain important to determining allegiance.
Second, if the allegiance test is to be applied, it must be reserved – as the ICTY itself highlighted – as an “allowance” for “special cases” where reference to nationality is no longer meaningful, such as international conflicts where new states are being created and/or states intervene indirectly through control over non-state armed groups, such that armed groups and civilians share the same nationality (and thus other considerations, e.g., ethnic divisions, prevail). This was the case in Bosnia, where despite common nationalities, victims and perpetrators professed allegiance to different sides. The situation in the ICC’s Katanga was similar, given that the hostilities in this international armed conflict were mostly waged by ethnic militias in a region where civilians typically all held the same Congolese nationality (§13-14).
Allegiance thus may be an exception in special cases, but not a general rule. Where conflicts are waged between “well-established States”, as flows from the ICTY’s own pronouncement (Tadić AC, §166), nationality remains the primary governing rule. The ICTY reserved this possibility by claiming that, depending on circumstances, allegiance may be dictated by other grounds than nationality (but not necessarily is), e.g., ethnicity “may become determinative of national allegiance [emphasis added]” (Tadić AC, §166) (but not necessarily does).
The only logical way forward is that the nationality test remains both sufficient and governing in the determination of protected status, particularly in armed conflicts between established states. The allegiance test subsists as a supplement. It applies where an international armed conflict coupled with the formation of new states or the indirect intervention of states through militias causes a significant portion of armed groups and civilians to share the same nationality – thus rendering protection based on nationality useless, and the only applicable dividing lines are ethnic, political or other.
Thus, in inter-state armed conflicts devoid of the considerations presented above, the nationality test continues to apply, while ethnic or political affiliations do not affect one’s protected status. This renders the allegiance test inapplicable, e.g., to the armed conflict between Russia and Ukraine.
Concerns linked to changes in nationality
A more complicated question concerns the legal implications of residents in occupied territories changing their nationality from occupied to occupying state (e.g., Ukrainian in occupied territories receiving a Russian passport). The question will arise whether these persons continue to be protected by GC IV as Ukrainian nationals or whether protection ceases with receipt of Russian citizenship.
IHL does not provide any straightforward guidance, and the determination will necessarily depend on the context of every individual case. While forcible and coercive imposition of citizenship in violation of international law (being void) will unlikely affect protected status, voluntary change of nationality can – theoretically – deprive a person of GC IV’s protection and move them within the realm of domestic law applicable in occupied territories. Yet, the line between these cases is often blurred and difficult to draw. In circumstances such as Russian occupation policies creating an environment where procuring citizenship is necessary to obtain basic services essential for survival, acquisition of that citizenship cannot imply renunciation of another. The presumption then must stay that GC IV’s protection continues to apply, unless the protected person clearly and out of genuine will declares their intention to remove themselves from protection by renunciation of their prior Ukrainian citizenship.
This reading is in line with the fundamental obligations of the occupying power vis-à-vis the population of occupied territories, namely to respect the laws in force in the occupied country (Hague Regulations, Art. 43) and to ensure protected persons in occupied territory are not deprived of the benefits of GC IV by any change resulting from occupation, by the institutions of the occupied territory or by any annexation (GC IV, Art. 47).
To apply the nationality test in a manner legitimising an occupant’s policies on citizenship imposition (and thus depriving civilians in occupied territories of protected status by their simple acquisition of a passport necessary to access basic needs) will run contrary to these provisions, as well as “affect the legal status” of the occupied territory in breach of Additional Protocol I, Article 4.
Conclusion
In presenting and justifying a sounder application of protected status under IHL, this article may serve as a cautionary tale for both those who seek to advance the law and those who subsequently endorse these efforts. There is no question that international law is far from static, but restraint is paramount. Advancements must be substantiated with unimpeachable reasoning and those considering endorsement need to scrutinise these efforts. This is how we may seek to rightfully forge a path forward without jumping out the frying pan into the fire.
