Cross-posted on Just Security.
Earlier today Charlie Savage of The New York Times broke the story that while serving as the Legal Adviser at the US State Department Harold Koh wrote two major opinions on the extraterritorial application of human rights treaties, urging the Obama Administration to abandon the previous categorical position that the International Covenant on Civil and Political Rights can never apply outside a state party’s territory. The first opinion is on the geographical scope of application of the ICCPR, is dated 19 October 2010, and is available here. The second, on the geographic scope of application of the Convention against Torture and its application in situations of armed conflict, is dated 21 January 2013, and is available here. The two opinions, probably obtained by Savage in yet another leak from within the Administration, are a fascinating read. Koh essentially adopts almost all of the critiques levied against the existing US position, which he sees as increasingly untenable, and provides his own (relatively moderate) model of how the two treaties should apply outside a state’s own territory.
Savage also reports that despite Koh’s opinions the Administration has decided not to abandon the previous US position, simply because it fears (or at least a sufficient number of its component parts do) that accepting that human rights treaties apply extraterritorially would make its collective life more difficult, as everything from extraterritorial drone strikes to NSA surveillance could fall within the purview of the ICCPR. We shall soon see if Savage’s reporting is correct – the US is up for periodic review before the Human Rights Committee next week, and this is bound to be one of the first questions asked. As I’ve explained before, the US Fourth Periodic Report and a follow-up communication to the Committee merely registered the US position and the criticism thereof, without reiterating it, thus leaving the door open for change. If Savage’s reporting does prove to be correct and the US now clearly reiterates before the Committee that the ICCPR cannot apply extraterritorially because its Article 2(1) is supposedly crystal clear and unambiguous when it says that ‘[e]ach State Party to the present Covenant undertakes to respect and to ensure to all individuals within its territory and subject to its jurisdiction the rights recognized in the present Covenant,’ an important opportunity will have been missed.
The leak of Koh’s two opinions was probably timed to coincide with the hearings before the Committee, and will provide the Committee’s members with ample fodder to criticize a government whose position was expertly dismantled by its own legal advisor. Even if the US decides to stick to its guns, its position has been undermined. I frankly cannot think of a similar situation in which the disclosure of internal legal advice and the timing of that disclosure have so fatally compromised a state’s public legal position on a matter of comparable importance (a good, but imperfect, comparison point might be Lord Goldsmith’s advice to the UK government with regard to the invasion of Iraq). The Committee will have a field day. If I was the US delegation in Geneva, I think it would be far better to cut one’s losses and stick to the deliberate constructive ambiguity in the US Fourth Report, without reasserting the previous position or engaging the Committee on this issue in any detail.
Of the two opinions, the ICCPR one is more interesting, and I’ll address it first, since the territorial scope of the CAT is an easier question. In many of its articles the CAT refers to state obligations extending to ‘any territory under its jurisdiction’, a formulation that is clearly not limited to sovereign title over territory, unlike the ICCPR ‘s seemingly dual requirement that individuals must be both within the state’s territory and subject to its jurisdiction. Koh’s ICCPR opinion is convincing in arguing that the text of Article 2(1) is not all that clear and unambiguous, since the ‘and’ can easily be read as an ‘or’ (as the Committee did), and since the ‘within its territory and subject to its jurisdiction’ qualifier could only refer to the positive obligation to ensure human rights, but not limit the negative obligation to respect them.
Koh is also convincing in arguing that the US position hasn’t been either as consistent or as longstanding as is sometimes claimed, really starting only from 1995 and the then-legal adviser’s oral response to a question put to him by a member of the Human Rights Committee, that the issue was never seriously considered within the US government before that, and that neither the ICCPR’s object and purpose nor its travaux support the strictly territorial US position. Koh’s view is thus (p. 4) that:
• In particular, as detailed below, it is my considered opinion that a better legal reading would distinguish between the territorial scope of the Covenant’s obligation to “respect” and to “ensure” Covenant rights.
• A state incurs obligations to respect Covenant rights – i.e., is itself obligated not to violate those rights through its own actions or the actions of its agents – in those circumstances where a state exercises authority or effective control over the person or context at issue.
• A state incurs obligations to ensure Covenant rights – either by legislating or otherwise affirmatively acting to protect individuals abroad from harm by other states or entities – only where such individuals are both within its territory and subject to its jurisdiction, since in such cases the exercise of such affirmative authority would not conflict with the jurisdiction of any other sovereign.
The model that Koh proposes, which relies on the distinction between positive and negative obligation, is similar to the one that I have argued for in my book. Where I would part ways with Koh’s analysis is first when he argues, relying on Bankovic, that the personal conception of jurisdiction as authority and effective control over individuals would not extend to the killing of an individual, at least not in active hostilities (p. 5). As I’ve argued before, there is no non-arbitrary way of limiting the personal conception of jurisdiction, which collapses into the proposition that a state exercises authority, power, or control over an individual whenever it is engaging in conduct capable of violating that individual’s rights, be it through detention, killing, or just by reading their email. Second, I think Koh’s model is overly limiting in saying that the positive obligation to ensure human rights would only accrue on US territory – that obligation should in my view extend whenever a state de facto has effective control over territory, as the European Court has found in Lozidou. In other words, when it was the occupying power in Iraq the US had the positive obligation to ensure human rights of the Iraqi population even from attacks by third parties. The extent of that obligation will of course vary depending on the state’s capacity and on the legal limits imposed by other relevant rules of international law, such as IHL.
Perhaps most notably, Koh is trying to sell his model to the rest of the Administration by arguing that it wouldn’t require dramatic changes in existing practices, which largely already comply with the relevant standards as a matter of policy (at 49 ff). But while I would fully agree that human rights law is sufficiently flexible so as to accommodate valid state concerns in the extraterritorial context, Koh is still overstating his case – the ICCPR’s impact is not as small as he tries to make it seem. In particular, his dismissal of the impact of human rights on targeting and detention practices based on the current US understanding of the lex specialis principle in my view rests on very shaky foundations.
As for the CAT opinion, its perhaps most interesting feature is how Koh addresses and criticizes the positions held by some of the legal officers of the previous administration, and I’m sure others will comment on this at greater length. As a purely legal matter, I found particularly striking his finding that the CAT presumes a general, territorially unlimited obligation of states not to commit torture which also applies in armed conflict – an obligation not spelled out in the Convention, but implied in it. I think this is exactly right, and indeed this is an approach similar to that of the ICJ in the Bosnian Genocide case when it found that the obligation to prevent genocide in Article 1 of the Genocide Convention implies a (territorially unlimited) obligation of states not to commit genocide themselves. He extends the same reasoning to the negative obligation to refrain from committing cruel, inhuman and degrading treatment (p. 46). He similarly finds the Article 3 CAT non-refoulement obligation to be territorially unlimited (pp. 4-5), while the positive obligations to prevent torture etc. would depend on de facto territorial control.
In sum, had the Administration adopted the views expressed in Koh’s two memoranda it would have gone a long way in appeasing its critics and towards having a principled position on the territorial scope of human rights treaties. Yet, on the other hand, the fear that the ICCPR will have a more significant impact than Koh has argued is also understandable, and it is no wonder that the Administration is finding it hard to budge from the status quo. But let’s see what the next week in Geneva will bring.