[Just Security is publishing a series on the ICRC’s updated Commentaries to the Third Geneva Convention on Prisoners of War (2020). This GCIII Commentary series is published in collaboration with the ICRC’s Humanitarian Law & Policy blog and EJIL: Talk. The series editors for Just Security are Adil Haque and Ryan Goodman.]
Serving as a legal adviser in a multinational environment such as NATO has its challenges. I know having recently finished a six year term as Director of the Office of Legal Affairs at NATO HQ in Brussels. One of these challenges may seem elementary but is very real: how do you gain a quick understanding of Member State positions on various international law questions, especially when time is tight and evidence of State practice may be difficult to find, disparate or even non-existent? A second challenge comes with the multinational-specific legal difficulty that tends to arise in a different way than in a purely national context. As many commentators, including John B. Bellinger, III and Vijay M. Padmanabhan, Cordula Droege, Tristan Ferraro, Emanuela-Chiara Gillard, and Bruce “Ossie” Oswald, have pointed out, detention is usually high on the legal adviser’s list. And detention issues can vex the legal adviser to a multinational organization in special ways.
The ICRC’s new Commentary on the Third Geneva Convention (GCIII) can help with both of these challenges.
GCIII and Multinational Operations
Before proceeding, there’s a preliminary question: how exactly is GCIII relevant to contemporary multilateral operations? This is a valid question, especially since the overwhelming majority of detention-related issues since the end of the Cold War have occurred in situations typically characterized as non-international armed conflicts (NIACs). These include NATO’s operations in Bosnia and Herzegovina, Kosovo, and Afghanistan, where detention issues have had major strategic, operational, and public diplomacy consequences. For NATO, the situations that could be seen as international armed conflicts (IACs) in which the application of GCIII could arise have been limited to the Kosovo (1999) and Libya (2011) air campaigns, rather than land operations in which a large number of POWs could be anticipated. (Interestingly, the POW issues that did arise were with respect to allied forces that came under enemy control, such as the capture of three U.S. soldiers in what is now North Macedonia by Serbian forces.)
Planning for POW issues was, by contrast, an important part of NATO’s Cold War-era preparation for large-scale convention conflict in Europe. With the renewed focus on collective defence since Russia’s 2014 actions against Ukraine and the more general strategic shift toward great power competition in the global security community, POW issues – and thus GCIII – may well return to the agenda of NATO planners. Perhaps more significantly, GCIII is still relevant because recent NIAC detention practices and standards, such as the 2012 Copenhagen Process Principles and Guidelines, have ended up importing many requirements from IAC law. This has been NATO’s practice in many aspects of its detention policy since the 1990s. The UN Secretary-General’s 1999 bulletin on observance by UN forces of IHL goes even farther, applying mutatis mutandis a GCIII standard to all forms of detention related to UN operations.
We also should ask another background question: how much detention in a multinational operation is actually multinational? In NATO’s practice, the answer is none. Detention is carried out solely by individual allied forces and not by the international organization. (The allocation of responsibilities may work differently in other international organizations.) As a result, in NATO operations, each detaining State conducts itself in accordance with its own national and international obligations. There is some NATO doctrine in this area. This includes a 2007 Allied Joint Publication, which is a category of document agreed by allies as part of NATO’s extensive (and perhaps little known) standard-setting activities. (This specific document is not publicly releasable.) But it has proven difficult to get subsequent agreement among allies on detention standards. This makes GCIII and the information gathered in the 2020 Commentary on State practice on the convention’s implementation even more relevant for multinational legal advisers. The Commentary can be a useful source, among other things, for locating this information in one place, although it is of course not a substitute for consulting the relevant multinational and national operating documents.
Specific Challenges of Implementing GCIII in Multinational Operations
In terms of the substance of implementing GCIII in multinational operations, the Commentary pays special attention to two important issues. First, which State is to be considered the Detaining Power, which then bears responsibility for POW treatment and other issues under GCIII? Second, how does the principle of assimilation – according to which “the standard of treatment to which prisoners of war are entitled is in the first instance determined by reference to the domestic standards and laws applicable to members of the armed forces of the Detaining Power” (para. 1520) – work when multiple national legal frameworks are in play, or where an international organization is involved? While the answers to these questions will depend on the specifics of the multilateral operation, in particular the policies and practices of the “umbrella” international organization, the Commentary provides a useful way of approaching the questions conceptually.
The Identity of the Detaining Power and Transfers to Another Detaining Power
In its discussion of Article 12, the Commentary is right to point out that “[w]here States conduct multinational operations in the context of an international armed conflict, it may not always be apparent which State is the Detaining Power and thus responsible for the treatment of the prisoners” (para 1519). The Commentary gives the example of “joint patrols during which enemy troops are captured or surrender” (id), during which “it may not be clear which State is the Detaining Power in situations where combatants are captured by a patrol that consists of the national forces of different States.” (para. 1522).
As a practical matter, there may not be much doubt – at least in the way NATO operations are organized – as to which country’s personnel actually takes control of an enemy individual upon capture, surrender or whatever other means by which that person “falls into the power of the enemy.” This is a factual question where appropriate record-keeping requirements can (and should) be built into operational protocols. As the Commentary points out, there can only be one Detaining Power at a time.
The problem may be less an issue of confusion about who actually detains as a lack of willingness or ability of certain parts of a multinational force to actually detain. While this appetite may vary depending on the nature of the particular conflict, it was an acute issue in Afghanistan, especially with respect to the US-led coalition in Operation Enduring Freedom (OEF). This resulted in a situation where detainees were handed over to allies that did have the capability to detain, or perhaps calling in the ally that could do the detention to complete the operation. As a 2016 Norwegian government report recalled, “in 2002, the Foreign Minister stated in the Storting [Parliament] that if the Norwegian special forces in OEF should, contrary to expectation, take prisoners, they would in practice have no other option than to hand them over to US forces” (10.4.2). This points to the centrality of the question of transfers of POWs in the practice of multinational operations.
These situations also show that the legal obligations concerning protection of detainees apply to the initial detaining power even if its involvement was transitory and very short-lived. As the Commentary explains, once a POW is detained by one Detaining Power, “[f]or another Power to assume that control and responsibility, a transfer pursuant to Article 12(2) and (3) is required.” Articles 12(2) and 12(3) provide for well-known pre-transfer and post-transfer requirements. Before any transfer, the original Detaining Power “has satisfied itself of the willingness and ability” of the transferee Power, which must be a party to GCIII, to apply the Convention. After the transfer, if the original Detaining Powers is notified of failure to do so, it must “take effective measures to correct the situation or shall request the return of the prisoners of war.” These obligations sensibly apply not just to transfers across international boundaries but also to transfers of POWs between co-belligerents operating on the territory of the same State. The obligations would therefore apply in various situations common to multinational forces, such as quick transfers in the field of the sort described above or within a detention facility run by a multinational force.
An important technique in complying with pre- and post-transfer requirements has been the development of arrangements for detainee transfers. It is not clear from publicly available information how regularly such arrangements are put in place, but it appears that they are in wide use. For example, the US Department of Defense Law of War Manual states “[s]pecial agreements may be concluded with respect to POW transfers to the custody of another Detaining Power. U.S. policy has been to require a written agreement or arrangement before transferring POWs to a coalition partner” (Section 9.30.3). The footnote to this section refers to published agreements from the Persian Gulf War. Likewise, the UK Ministry of Defense’s 2020 Joint Doctrine Publication 1-10: Captured Persons mandates planning for coordination with coalition partners on agreements between troop contributing countries on the handover of captured persons.
The Principle of Assimilation and Multiple States
The principle of assimilation is a cross-cutting concept that “reflects an understanding that prisoners of war will be treated on the same terms as members of the armed forces of the Detaining Power” (para. 30). National legislation regarding armed forces is to be applied to POWs, unless the standards in the national legislation are less protective than a series of guarantees on a range of matters set forth in Article 82. In that case, the protections in Article 82 would apply. Even if national legislation “constitutes one, but not necessarily the governing, benchmark for determining the judicial and disciplinary treatment owed to prisoners of war,” the question arises as to which national legislation is relevant in multinational operations, particularly when a POW is transferred between allies. The Commentary’s introduction goes back to the importance of the Detaining Power that “[i]n the course of multinational/combined operations, it may not always be apparent which State is the Detaining Power and thus responsible for the treatment of prisoners. Clarity on the identity of the Detaining Power is required for the principle of assimilation to operate effectively” (para. 38). On the basis of this wording, it appears as if the benchmark legislation changes as the POW is transferred. This makes sense given the overall logic of the principle, but it may raise questions if the transferring Power offers conditions in its national legislation that are more beneficial in certain respects than the receiving Power. It is not clear what particular disparities might exist among allied national legislation and, if these are considered problematic, how this situation has been dealt with in practice such as through transfer agreements. On the one hand, disparities might not be a major cause for concern as long as the receiving Power meets the internationally-agreed GCIII baseline. On the other hand, there is a strong overarching argument that transfers should not disadvantage the POW. On balance, it would make strategic sense for allies to settle on a more protective regime for POWs since allied forces in enemy hands would be entitled to benefit from such an approach as well.
The Commentary’s treatment of questions involving POWs in multinational operations is necessarily brief. Although there has been much discussion of NIAC detention, there has been little practice in implementing GCIII in a strict sense. Moreover, while it appears from national military manuals and other doctrinal publications that NATO allies take their GCIII obligations seriously and recognize the need to plan appropriately, there is limited transparency on any policies or arrangements that may have been adopted. For example, while some national publications on captured persons have been released, NATO’s Allied Joint Publication is not publicly available. Transfer arrangements, if not published in historical studies as in the Gulf War or released via national litigation, are generally not available either. This state of affairs is a missed opportunity, especially given the universal adherence to GCIII. NATO allies in particular, with their ability to do important work on standards and their advanced (and ever-growing) corpus of national doctrinal documents, have an opportunity to be more transparent about their commitment to implement GCIII fully if the situation were to arise. Such transparency is a “good news story” that would highlight NATO’s commitment to IHL and its broader human security agenda, which includes a number of issues related to the implementation of IHL and other obligations in armed conflict.
There is also more space for multilateral work in this area. The 2020 UK Ministry of Defence doctrine on captured persons (CPERS) specifically highlights the need for multinational coordination on “drafting memoranda of understanding and military technical agreements between troop contributing nations in relation to the shared use of CPERS holding facilities, intelligence sharing and the handover of CPERS.” This is an area that appears ripe for planning work in multinational venues like NATO. Work could begin with a study of the typical provisions of such arrangements that have already been in use with a view toward identifying best practices. There may even be residual knowledge from NATO’s Cold War-era planning that could revitalized and updated to include modern issues like the potential use of biometric information to identify POWs. Such a project would be an ideal way for NATO to continue its long-standing cooperation with the ICRC in a practical way.
This article reflects the author’s personal views only and does not reflect the views of NATO or any NATO ally.
Elizabeth Stubbins Bates, Geneva Convention III Commentary: Unpacking the Potential of “Ensure Respect” in Common Article 1
Keiichiro Okimoto, The United Nations and the Third Geneva Convention
Catherine O’Rourke, Geneva Convention III Commentary: What Significance for Women’s Rights?
Jean-Marie Henckaerts, “Commentary: ICRC unveils first update in sixty years.”
Cordula Droege, GCIII Commentary: ten essential protections for prisoners of war