A few weeks ago, my great friend Elvina Pothelet analysed, on this blog, the decision of the International Criminal Court (ICC) Prosecutor’s to request authorization to investigate, inter alia, acts of ill-treatment of detainees allegedly committed since 2002 by the CIA in black sites in Poland, Romania and Lithuania, in connection with the armed conflict occurring in Afghanistan. Elvina affirmed that there may be an added value in qualifying these alleged behaviours as war crimes, but she also hinted that such qualification might support the idea that International Humanitarian Law (IHL) applies globally, even outside the borders of the States where active hostilities take place. In this post I will argue that a wide geographical scope of application of the IHL rules on the treatment of detainees — especially those contained in common Article 3 to the Geneva Conventions and reflected in customary international law — does not necessarily imply an equally wide applicability of the rules on the conduct of hostilities.
To put my intervention in context, I should recall the obvious: a war crime presupposes a serious violation of an IHL rule. And for a rule of IHL to be applicable, there must be a sufficient link of correlation (so-called ‘nexus’) between the behaviour in question and an armed conflict (see ICTY AC, Kunarac, § 57 ff., referring to acts that are ‘closely related to the armed conflict’; see also Cassese). Although these sources refer to international criminal law (ICL), they build on the principle that IHL only applies to conducts and events which are sufficiently related to an armed conflict, as recognized e.g. in the ICRC Introduction to IHL, at pp. 28 and 59 (see also Practitioners’ Guide to Human Rights Law (HRL) in Armed Conflict, § 4.23). When such behaviour occurs outside the theatre of hostilities — e.g. where acts of torture were allegedly perpetrated in Poland/Romania/Lithuania, but the supposedly related hostilities took place in Afghanistan — one should ask whether such ‘sufficient nexus’ exists and, additionally, whether are there any geographical limitations to its establishment. In other words, is IHL applicable to conduct or an event as soon as it is sufficiently connected to an armed conflict, regardless of the territory where it took place (as contended, e.g., by Lubell-Derejko)? Or should the applicability of IHL be limited only to behaviour occurring in the area where active hostilities are being fought, or in the territory of a State party to the conflict (as deemed preferable by the ICRC in its 2015 report, at p. 15)?
Like ‘global battlefield’ theorists, I am convinced that geographical considerations per se do not necessarily limit the applicability of IHL. But, as also accepted by Lubell and Derejko, I believe that they are a fundamental factor to be taken into account when assessing the existence of the necessary nexus between an event under scrutiny and an armed conflict. Geographical distance from the actual conflict may be an indication that the relevant conduct or event is sufficiently ‘closely related to the hostilities’. And that is where I think the difference between the rules on the treatment of detainees and other IHL rules (especially those on the conduct of hostilities) lies.
Indeed, the nexus between an individual’s detention and a certain armed conflict is usually very strong — be it because such individual is deemed to belong to one of the parties to the conflict, or is believed to possess information related to the conflict, etc — and thus always ‘closely related to the armed conflict’. In this respect, the circumstance that an individual was captured anywhere for any reason related to the conflict ipso facto implies the protection of the applicable conventional and customary IHL rules on treatment of detainees in international and non-international armed conflicts. And serious violations of these rules would amount to a war crime. Needless to say, such nexus is even clearer whenever the individual was first captured in the theatre of operations and subsequently transferred to a different State where no active fighting is taking place. To deny the applicability of IHL in this latter case would be non-sensical, as it could deprive an individual of protection by his/her simple transfer to a different location (see the OTP Request, § 251, citing the 2016 ICRC Commentary to common Article 3; see also Pejic at p. 15).
More generally, the wording of common Article 3 — prohibiting certain behaviours against individuals who find themselves in the enemy’s hands ‘at any time and in any place whatsoever’ — confirms that, once an individual has been captured for conflict-related reasons, no geographical consideration may break the nexus between such detention and the relevant armed conflict. Such conclusion is even more justified if one recalls that common Article 3 is designed to act as a minimum set of protective rules for victims of armed conflict, applicable regardless of whether any other legal protection applies as well.
Thus, common Article 3 protects detainees no matter where they have been captured, no matter when, no matter the (un)lawfulness of the detention, no matter what other rules are applicable at the same time, as long as they are detained for a reason related to an armed conflict. Consistently with this approach, the ICC Prosecutor’s insists on the nexus between the situations of detention under scrutiny and the armed conflict in Afghanistan, rather than on the territory where such detention was carried out (see OTP Request, §§ 246-252).
The said presumption that a sufficient nexus exists, however, does not cover other situations, especially the issue of targeting, on which I will focus here. Indeed IHL rules on targeting of individuals are among those that are “clearly bound up with the hostilities” (ICTY AC, Tadić 1995, § 68) and their primary purpose is to be applied in areas of active fighting. If anything, there should be a presumption that an attack occurring outside the area of active hostilities is not sufficiently connected to those hostilities, and thus not governed by IHL. The farther an attack is from such area, the stronger the presumption of non-applicability, hence the stronger would the evidence of a sufficient nexus be for IHL to apply. Whilst, for the rules on the treatment of detainees, a sufficient nexus originates ipso facto from the fact of detention, a similar nexus is not necessarily evident when it comes to the rules on targeting. With regard to the latter, and the existence of sufficient nexus would need to be demonstrated in a compelling manner case-by-case. I concede, for instance, that a sufficient nexus might materialize outside the area of active hostilities when it can be clearly demonstrated that the target is actively participating in the ongoing hostilities. Crucially, even in this case, compliance with other rules of international law would still be necessary. At the very least, these include the rules of jus ad bellum. In my opinion, they also include rules of international human rights law, which are more fitting for contexts where active hostilities are not occurring and which impose stricter standards for the use of lethal force (see ICRC 2015 report, p. 15-16; and Practitioners’ Guide to HRL in Armed Conflict, principles 4.33 to 4.35).
In a comment to her blog post, Elvina warned that the debate on the geographical scope of IHL application should not be framed in “humanitarians vs warmongers” terms. However, it is hard to deny that each legal theory proposed in scholarly writings heavily depends on the scholar’s philosophical and political ideas, and favourite outcome. I believe that expressing the philosophical/political/ideological reasons behind a legal argument does not necessarily weaken it, as long as the relevance of those reasons is coherently explained. Therefore, I think it is necessary to acknowledge that my suggestion relies on two ideological assumptions: first, that it is desirable to apply IHL rules on the treatment of detainees as widely as possible; second, that it is desirable to make the IHL rules on targeting applicable as narrowly as possible. For the sake of intellectual honesty, I would like to briefly explain my rationale.
As to the first assumption, I believe that — when it comes to their treatment — individuals detained anywhere in connection with an armed conflict should benefit as a minimum from the protection of IHL, even considering the fact that international human rights law (IHRL) concurrently affords them a greater protection. As already mentioned by Elvina, a few States still deny that they are bound by any human rights obligations outside of their own national territory. Additionally, the very idea that non-state entities (like armed groups party to a NIAC) may bear human rights obligations on their own is contested. I am convinced that human rights obligations persist even extraterritorially (a belief shared inter alia by the Human Rights Committee, by the European Court of Human Rights in Al-Skeini and by eminent doctrine), and that certain non-state entities do have human rights obligations (more here and here), but I find it useful that — at the very least — the guarantees listed in common Article 3 are seen as binding on all States and on all non-state actors which are party to an armed conflict, even when they mistakenly assume that human rights law does not bind them. Moreover, common Article 3’s guarantees, unlike certain human rights provisions on the right to liberty and security or to the right to a fair trial, are not subject to any possibility of derogation.
Additionally, from an accountability perspective, the applicability of IHL makes the legal regime of war crimes applicable even when it may be hard to establish the necessary requirements for other categories of crimes, e.g. crimes against humanity. The possible qualification of certain conducts as war crimes not only matters for questions of ICC jurisdiction (as it is the case for the OTP request that originated this blog post), but brings into play well-established procedural rules such as the duty to investigate such conducts, or the right to establish national jurisdiction over alleged perpetrators.
As to the second assumption, it is premised on the fear that a broader application of the IHL rules on targeting, relatively permissive, may lead to an unnecessary escalation of violence and legitimize more frequent recourse to lethal force (a similar concern is expressed e.g. by Horowitz, describing IHL as a Trojan horse; contrary, see Haque here, explaining that IHL rules are prohibitive in nature). Such risks would not materialize by simply accepting a wider geographical reach of the rules on the treatment of detainees — an expansion which per se does not pass judgment on whether such detention would be lawful or not, but takes the protracted circumstance of detention as a fait accompli and tries to make it more humane. True, some may argue that, whenever these two bodies of law are applicable at the same time, IHRL could be interpreted in light of IHL and, consequently, the approach I propose may lower the standard of protection for detainees. However, IHL and IHRL do not necessarily modify one another and, especially when their provisions are compatible (like in the case of the treatment of detainees in NIACs), they could concurrently apply side by side. If one of these two bodies of law (e.g. IHRL), when compared to the other (e.g. IHL), imposes a greater number of obligations on a party to the conflict, such party is simply bound by those additional obligations as well.
I hope, therefore, that the ICC’s decision on the OTP’s request with regard to the situation in Afghanistan may be framed along the lines that I have traced in this post, in the interest that victims of armed conflicts may receive the widest possible protection, whilst at the same time refraining from creating incentives to the use of lethal force outside of areas of active hostilities.