Developing the Law of Non-International Armed Conflict: A View of the Harmonization Project

Editor’s Note: This post is part of the joint series of posts hosted by EJIL:Talk!, Lawfare and Intercross (blog of the International Committee of the Red Cross) and arising out of the Transatlantic Workshop on International Law and Armed Conflict held in Oxford this summer.

This post is a response to Professor Sarah Cleveland’s post on the Columbia-based Harmonization Project that Professor Cleveland and Sir Daniel Bethlehem are leading. That project explores the potential for applying the law of international armed conflict (IAC) in non-international armed conflicts (NIACs), as a means of developing the law applicable in the latter. The conclusion of the project is that the large majority of the rules applicable in IACs can be transplanted into NIACs without amendment and that this should be done by States either multilaterally or via unilateral declarations.

A detailed, rule-by-rule consideration of the degree to which parity between the law of IAC and NIAC is practicable is a very useful endeavour. Indeed, historically this has been the method by which the law of NIAC has developed. It is clear why this should have been the case. When the first international humanitarian law (IHL) treaties were adopted in the mid-nineteenth century, international law was still, by and large, a law governing inter-State relations. Matters that did not directly engage such relations, including NIACs, were thus generally excluded. Customary rules did of course develop to govern certain NIACs, such as the doctrine of belligerency, but these often applied only where another, non-party State was affected by the conflict.

As international law expanded to include the regulation of purely intra-State matters (reflected in human rights instruments, as well as the Genocide Convention, adopted in the aftermath of the Second World War), this basis for marginalising NIACs began to fall away. Rules traditionally applicable only in IACs could now move over into NIACs. And indeed this is what has happened: common Article 3 to the 1949 Geneva Conventions and Additional Protocol II were based on the law of IAC. This is also true of the customary rules recognised by the ICTY and ICRC.

It is therefore only natural that we should look to the law of IAC in developing the law of NIAC. This post, however, will offer some words of caution in adopting this method of humanising NIACs. In particular, it will be argued that both general and specific arguments militate against this supposedly self-evident means by which to develop the law of NIAC.

General Concerns

The Harmonization Project declares its goal as being to build upon current obligations in NIACs—it is limited to IHL and does not seek to make a claim regarding the relationship between IHL and human rights law. However, it seems to me that one cannot avoid such questions when considering proposals for developing the law of NIAC. Indeed, if one’s goal is further to humanise NIACs (as the Harmonization Project’s seems to be) then one must tread carefully in proposing the extension of IHL in toto to NIACs. As David Kretzmer has shown, far from increasing protections, this method could in fact undermine existing protections.

It is rather trite now to note that a State’s international human rights obligations continue, prima facie, to apply in armed conflict, including (and perhaps especially) NIACs. Thus, whereas in 1949 a Lotus-like presumption arguably applied in NIACs such that, absent IHL, no other rules of international law applied, the situation now is very different. If we were to extend the entire corpus of IHL to NIACs, we would need to consider how those rules interact with existing human rights obligations. No doubt one would be faced, reasonably or not, with lex specialis-type arguments that those IHL rules now displace incompatible human rights obligations. The result could, with regard at least to certain matters (e.g. targeting and detention), be a lowering of protections, rather than a raising of the floor. Indeed, this would be a consequence of the fact that harmonization would involve not only protective rules but also permissive rules. Dapo Akande and I have written here and here on the absence of permissive rules from the current law of NIAC; harmonization would change this.

The emergence of international human rights law has therefore significantly altered the context of these debates on the distinction between IACs and NIACs. Far from representing the final stage in the process of raising the level of protection for victims of NIACs, harmonization could be the final nail in the coffin for this endeavour. This is especially so for it would apply a single legal regime to all NIACs—a category which, since 1949, has expanded considerably to include a vast number of permutations, from relatively low-intensity internal conflicts at one end, to large-scale NIACs involving multi-national forces at the other. Underpinning IHL is a basic presumption that it is necessary to revert to its more permissive rules whenever an armed conflict exists. One must question whether this general presumption of necessity shouldapply in all situations labelled a NIAC.

Specific Rules

None of the above is meant as an absolute rejection of harmonization. Rather it is a caution against assuming harmonization to be a progressive method. In certain areas, the law of IAC can function as a useful starting point, but simple harmonization would be problematic.

Let us take detention (internment) as an example. The law of IAC contains two principal internment regimes: that for prisoners of war (under the Third Geneva Convention, GCIII) and that for civilians (under the Fourth Geneva Convention, GCIV). Harmonization would involve applying both regimes in NIACs. The GCIII regime, however, poses significant problems for NIACs. That regime comprises two key elements: internment on the basis of status and internment for the duration of hostilities. Both elements are unsuitable for application in most NIACs. The first, status-based internment, is problematic by reason of the absence of any elaborated notion of status in NIAC. Even if one could develop an appropriate definition of ‘combatant’ for NIACs (whether it is the ICRC’s ‘continuous combat function’ test, the Israeli Supreme Court’s notion of contribution to the cycle of hostilities, or the various functional criteria developed by the DC courts), the second element of the GCIII regime is equally problematic. This is because the temporal contours of NIACs are notoriously indeterminate—if there is one thing more difficult than identifying precisely when a NIAC comes into existence, it is identifying when hostilities cease such that one could speak of an ‘end’ to the conflict. Applying the GCIII presumption of internment for the duration of hostilities could therefore result in extremely prolonged detention without any form of review.

The GCIV internment regime, covering civilians, is more appropriate for NIACs, given its individual, conduct-based approach. GCIV provides that a civilian may be interned where necessary for reasons of security, provided that they have periodic reviews of the actual necessity of their internment and are released as soon as the individual reasons justifying their internment cease. However, this should not simply be transposed verbatim into NIACs. Rather, we should use GCIV as a starting point and take this opportunity to develop these rules in line with subsequent practice. Thus, in line with the ICTY in Delalić(para 329), we might say that the review body must have the power to order release. Similarly, in keeping with State practice, we could include a requirement that the review body be independent from the chain of command that ordered the initial detention. The point here is simply to remind ourselves that the treaty rules applicable in IACs are very much a product of their time; one should not be too quick to assume that these skeletal rules constitute the benchmark for developing the law of NIAC.

Conclusions

This post has offered two lines of critique against the orthodoxy of harmonization, and one may now ask how these are to be reconciled—on the one hand, I have suggested that drawing on the law of IAC could undermine existing protections, whilst on the other I have suggested that certain rules might be useful starting points in developing the law of NIAC. I have offered a proposal along these lines in a recent article. Simply put, my view is that, at least in particular respects, the law of NIAC should be developed, using the law of IAC as a starting point (but not an end in itself). However, for States wishing to access these more permissive rules, they should do so via their human rights treaty obligations, e.g. through contextual interpretation of those human rights obligations, where possible, or by derogation therefrom, again where possible. This dual approach of developing NIAC law, but accessing it through human rights law, has a number of advantages. First, requiring a State to access the permissive aspects of IHL through their human rights obligations would ensure that it is not simply assumed necessary to revert to those permissive rules. Rather, necessity and proportionality in the prevailing circumstances would need to be demonstrated. Second, developing the law of NIAC would clarify the minimum standards that States must honour. It would constitute the floor below which no conduct may fall even in cases of permissible derogation. Finally, whereas there is great uncertainty surrounding whether and, if so, to what extent armed groups might be bound by human rights law, the law of NIAC binds States and non-State groups equally. Developing the law of NIAC would therefore add to the obligations that bind the non-State side.