The UK Court of Appeal will soon hear the appeal in Serdar Mohammed v. Ministry of Defense, a highly important case in which the UK High Court held that the long-term detention of a suspected Taliban commander by British forces was unlawful—and, by extension, so was the general long-term detention policy of British forces in Afghanistan. (See Ruchi Parekh’s terrific primer and preview of the upcoming Court of Appeal hearings).
The lower court’s judgment deservedly received accolades from international law experts for its extraordinarily informed and thoughtful analysis across a range of challenging international legal questions. I agree with a core part of the High Court’s holding, but I respectfully disagree with a significant part of the rationale the court used as a basis for reaching that result. My hope is that the Court of Appeal will not repeat those mistakes.
An important notation: The UK government relied on my 2009 article in the American Journal of International Law in its briefs submitted to the High Court. I accordingly take very seriously the task in this post to explain why I believe my article is consistent with the High Court’s holding, but contradicts a significant part of the court’s chosen rationale. A more extensive version of the arguments in this post will also be published as an article in the U.S. Naval War College’s International Law Studies (2015). I will alert readers when that article is available.
Not in this post, but in a subsequent one, I will discuss perverse implications of the Mohammed judgment. The High Court decision has been celebrated by human rights lawyers, including a guest post and letter to the editor on this blog. But I aim to show that a part of Mohammed’s conceptual framework threatens to undermine human rights and humanitarian protections in wartime.
1. The Mohammed framework
The High Court advanced two propositions:
Proposition 1: (No) Authorization
International humanitarian law does not provide authorization (the source of legal authority) for a power to detain in non-international armed conflicts
Proposition 2: (No) Regulation
International humanitarian law has nothing to say about the grounds on which states may preventively detain particular individuals in non-international armed conflicts (eg. whether states may lawfully detain individuals who pose an imperative security threat) or what procedures should apply to detention in non-international armed conflicts.
With respect to the the first proposition, the court concluded that because international humanitarian law (IHL) does not authorize detention in non-international armed conflict (NIAC), the government needed to secure authorization either from Afghan domestic law or a Security Council resolution.
Part of the basis for the first proposition, according to Mr. Justice Leggatt, was the second proposition. That is, proof that IHL does not authorize detention in NIAC is that neither treaty law nor customary international law contain any rules setting forth who may be detained (in US constitutional parlance, “substantive due process”) and according to what procedures (in US constitutional parlance, “procedural due process”). On this view, there is a complete gap in IHL when it comes to those substantive and procedural determinations. That is, we might as well be talking about WTO law. The WTO does not authorize or prohibit detention in NIAC, and has nothing to say about the substantive and procedural basis for detention.
I agree with the first proposition. I disagree with the second. And, indeed, the second proposition is unnecessary; the first proposition stands without it.
2. Authorization
I agree with Mr. Justice Leggatt’s holding that IHL does not provide authorization to detain in NIAC, and thus “the only potential sources of a power to detain are considered to be the host state’s own domestic law (i.e. in this case the domestic law of Afghanistan) and [UN Security Council Resolutions].”
[As an aside: if UN Security Council Resolutions suffice, one might posit that the UN Charter or the use of force regime more generally authorize detention as an incident of waging battle, but that is not a question I explore here.]
Despite the government’s reference to my AJIL article in its brief, that article is consistent with the Mohammed judgment’s first proposition. A major point in my article is that IHL does not prohibit (it allows for) detention of civilians who pose a security threat in NIAC. And the High Court held that IHL does not authorize (nor does it prohibit) detention in NIAC. As Lawrence Hill-Cawthorne and Dapo Akande helpfully explain, all those claims are compatible. Indeed, Hill-Cawthorne and Akande explain, correctly in my view, the relationship between the High Court’s first proposition and the claims in my AJIL article:
“[To] argue that IHL does not create a legal basis to intern in NIACs is not inconsistent with Professor Goodman’s view. IHL does not restrict states with regard to detention in NIACs anymore than it does restrict their ability to detain in IACs. States are not prohibited from detaining in NIACs, and, in that sense, are therefore permitted by IHL to detain. IHL simply does not itself provide a legal basis to do so. That legal basis must be found elsewhere.”
As Hill-Cawthorne and Akande explain more broadly in their excellent analysis, it is routine for areas of law to regulate a practice without providing a source of authority for that practice. By way of example, international human rights law restricts the grounds on which a state may detain an individual and imposes procedural safeguards when a state does so. But international human rights, as a body of law, does not provide the source of authority to detain. If anything, the reverse is true. A state must have a basis in domestic or other legal authority to exercise the power of detention.
It should be noted that Mr. Justice Leggatt goes further (in dicta) by asserting that IHL does authorize detention in international armed conflict. The High Court’s judgment invoked the wording of Article 21 of the POW Convention (and by implicit reference presumably similar language in Articles 41-43 and 78 of the Civilians Convention).
It is not clear, however, that IHL even provides the source of authority for states to detain in international armed conflicts. Instead, IHL may be better understood as a generally prohibitory legal regime—consisting of rules and prohibitions. Accordingly, it is not as though states needed the 1949 Geneva Conventions to have the authority to detain or kill in armed conflict. Consider, for example, Derek Jinks’ exposition of the general structure of the legal regime in an earlier post at Just Security:
“IHL should not be understood as conferring authority on states in the strong sense. IHL rules, in the main, are prohibitory. These rules establish a floor of humanitarian protection–crafted in light of the vulnerable circumstances common to organized hostilities.
An important feature of the IHL regime makes clear that it does not provide affirmative authorization to kill, capture, or detain. IHL is, in one important respect, a second-order legal regime–governing only the conduct of hostilities.”
Jinks wrote the above before the Mohammed judgment, and he applied that same insight in an exquisitely reasoned Just Security post evaluating the High Court’s opinion. Notably, Hill-Cawthorne and Akande also explain very well why IHL does not authorize the power to detain, though they restrict their claim to the power to detain in NIACs.
3. Regulation (of security-based detention)
Where did the Court derive the ideas for proposition 2?
The PIL Claimants told the court that in NIACs “IHL does not supply any answers” to substantive questions of who can be detained and procedural questions of what safeguards should apply. The Claimants also stated (quoting Els Debuf 2013):
“‘[F]or an implicit legal basis to detain or intern to exist… it should at least be possible to deduce from it the grounds and procedures in accordance with which a person can be deprived of his or her liberty. However, neither CA3 nor AP II defines who can be interned or detained in [NIAC], for what reasons, in accordance with what procedures, and for how long.’”(my emphasis added)
In other words, IHL is as good as WTO law. That is, IHL (like WTO law) includes no authority to detain as clearly evidenced by its purported lack of anything to say about the conditions under which detention in NIAC can be exercised.
The Court accepted this argument. As for treaty law, Mr. Justice Leggatt stated in language that closely tracks the above text from the Claimant:
“I do not see how CA3 or AP2 could possibly have been intended to provide a power to detain, nor how they could reasonably be interpreted as doing so, unless it was possible to identify the scope of the power. However, neither CA3 nor AP2 specifies who may be detained, on what grounds, in accordance with what procedures, or for how long.” (my emphasis added)
Mr. Justice Leggatt extended that reasoning to his analysis of customary international law (paras. 258 & 261), and ended with the two propositions:
“I have concluded that in its present stage of development IHL does not provide a legal power to detain nor does it specify grounds on which detention is permitted nor procedures governing detention.” (my emphasis added)
The second proposition—especially in its strongest form that IHL essentially provides no answers to grounds and procedures for detention—is not well supported.
First, the very structure of IHL provides an answer to the outer boundaries of permissible state actions in NIAC. That was a major claim in my AJIL article, namely, the proposition that state actions which are permitted by IHL in international armed conflict are permitted by IHL in NIACs. This is not reasoning by analogy. It is reasoning by structure. I wrote:
“More fundamentally, IHL in international armed conflict–and the Fourth Geneva Convention in particular–is directly relevant because it establishes an outer boundary of permissive action. States have accepted more exacting obligations under IHL in international than in noninternational armed conflicts. That is, IHL is uniformly less restrictive in internal armed conflicts than in international armed conflicts. Accordingly, if states have authority to engage in particular practices in an international armed conflict (e.g., targeting direct participants in hostilities), they a fortiori possess the authority to undertake those practices in noninternational conflict. Simply put, whatever is permitted in international armed conflict is permitted in noninternational armed conflict.”
In short, IHL rules regulating international armed conflict are with little or no exception more restrictive across the board than those regulating states in fighting an internal conflict (where states’ sovereignty is at its zenith). It is difficult to conceive of an action that a state would not be permitted to take in a NIAC that would suddenly become permissible if the conflict were internationalized into an IAC. Put another way: what victim of armed conflict would prefer Common Article 3 and Additional Protocol II rather than the full protections of the Geneva Conventions and Additional Protocol I? As an illustration, consider how the jurisprudence of the International Criminal Tribunal for the former Yugoslavia, dating back to its first judgment in Tadic, rests on this basic understanding. Prosecutors in The Hague have fought mightily for the rules of international armed conflict rather than NIAC to apply to the behavior of defendants, because the former body of rules imposes a greater set of prohibitions. There was never any thought that a more permissive regime would apply in international armed conflict.
So far we have discussed the permissive boundaries of detention in NIAC but what about limitations on states in these contexts? IHL also imposes a set of prohibitions on the grounds for detention in internal armed conflict. That is, multiple sources conclude that IHL prohibits arbitrary deprivation of liberty in NIAC (see footnote 12 of the AJIL article, for example). Subsequent to that law review article, several important states through the Copenhagen Process—including “specially affected” states which is a significant category for customary international law purposes—explicitly accepted such restrictions on detention in NIAC. Consider also the ICRC’s statement in a Background Paper on detention for the regional consultations 2012-2013: “In terms of grounds for internment, the ICRC, along with a growing international consensus of experts considers that ‘imperative reasons of security’ is an appropriate standard for internment in NIAC.” And a Report by a group of experts convened by the ICRC and Chatham House “quite easily” reached a consensus that in NIACs “parties to a conflict may capture persons deemed to pose a serious security threat and that such persons may be interned as long as they continue to pose a threat.” (See also the ICRC’s customary international humanitarian law Rule 99: Deprivation of Liberty; update: see also this assessment of Common Article 3 in a 2009 article I wrote for the Naval War College’s International Law Studies).
On what theory did the High Court base its notion that IHL includes no regulation whatsoever of grounds for detention?
The theory of the High Court (and the Claimants) is that states wanted to avoid recognizing the authority of nonstate actors such as insurgents to detain. But as Hill-Cawthorne and Akande describe so well: “There is good authority for the view that regulation of conduct by international law does not imply authorization of that conduct or acceptance of the legality of that conduct.” Indeed, the High Court (and the Claimants themselves) accepted this logic. That is the court (and Claimants) applied that logic in reasoning why IHL’s regulation of the conditions of confinement should not be equated with an authority to detain. In short, it is clear that regulation of an action (especially a practically inevitable action in war) does not translate into recognition of the right or authority to engage in that act.
By way of another example, consider trials. If the High Court were correct about bestowing legitimacy or legality on the actions of nonstate actors, IHL would not recognize the prospect of such parties holding a trial and imposing criminal sentences, yet Common Article 3 does just that. Like detention, the legal system recognizes the fact that nonstate actors will engage in such actions and IHL thus creates a set of legal rules to regulate such procedures. As Sandesh Sivakumaran, a leading authority on NIACs, wrote: “[I]t is not so much about a right to intern as it is about the need to regulate the existing practice of parties to non-international armed conflicts. … Indeed, a number of armed groups have enacted rules on offences, arrest, and trial, and these could and should be developed to include the modalities of internment.”
It is unclear, but the High Court’s theory for proposition 2 may also rest on a notion that the IHL detention scheme in international conflicts includes only status-based detention with corresponding privileges such as for POWs. The PIL Claimants stated: “Perhaps the most important reason for this is that … the internment powers in the Geneva Conventions are based on status (e.g. prisoners of war) and those categories (with their corresponding privileges) have no equivalence in NIAC-IHL” (see also para 5 of Annex A and para. 230 of the judgment). However, IHL in international armed conflict also includes detention without such status categories or privileges under the Civilians Convention (covering individuals who pose a security threat, saboteurs, and the like). And, it is now well recognized that IHL in NIAC does include status-based categories such as combatants and civilians which are potentially relevant for detention.
The explanation that states avoided the application of IHL to detention in NIACs because they wanted to avoid assigning legal authority to nonstate actors also suffers a problem with respect to contemporary practice. Here’s the flaw: If it were true that regulation of detention in NIAC implies recognition of nonstate actors’ authority to detain, what is the explanation for why so many states have since then explicitly endorsed such regulations? Have powerful military states suddenly agreed to recognize nonstate actors’ authority to detain? Have these governments’ decided to wipe out their ability to pass domestic legislation to prohibit forcible abduction and detention by insurgents? Hardly. It is because the connection between acceptance of restrictions on detention and recognition of the authority to detain does not exist in the way the High Court believed.
* * *
At bottom, the most important point is that these debates about proposition 2 are irrelevant for the decision at hand in Mohammed v. Minister of Defense. Proposition 1 is sufficient and well-established without it. I suppose there may also be a middle ground here that avoids some of these debates. One might argue that the rules regulating detention in NIAC, although they do exist, are relatively rudimentary. I suppose, courts might conclude that an authority to detain cannot exist with such a limited set of rules to define its scope. But the notion that no such rules exist is mistaken, and, as I aim to demonstrate in a subsequent post, that notion is also dangerous. It can produce unintended negative consequences for the scheme of humanitarian protections in wartime. The Court of Appeal should avoid repeating this mistake.
[Editor’s note: See also Just Security’s other coverage and analysis of Mohammed v. Minister of Defense.]