DOD or CIA in Yemen?

The ace Wall Street Journal national security reporting team, followed closely by the Washington Post and the AP, have been reporting in the last couple of days on the CIA being tasked to carry out an expanded Predator drone targeted killing program in Yemen. I’ve been meaning to blog on this, but meanwhile Robert Chesney poses the following question over at the Lawfare blog (where you can find links to these articles and an expanded discussion):

[W]hat really struck me about the stories was their common reference to a particular legal argument. Here is Siobhan Gorman and Adam Entous in the Wall Street Journal:

The U.S. military strikes have been conducted with the permission of the Yemeni government. The CIA operates under different legal restrictions, giving the administration a freer hand to carry out strikes even if Yemeni President Ali Abdullah Saleh, now receiving medical treatment in Saudi Arabia, reverses his past approval of military strikes or cedes power to a government opposed to them.

And here’s Greg Miller in the Washington Post:

Because it operates under different legal authorities than the military, the CIA may have greater latitude to carry out strikes if the political climate shifts in Yemen and cooperation with American forces is diminished or cut off.

And here’s Mark Mazzetti in the N.Y. Times:

By putting the operations under C.I.A. control, they could be carried out as a “covert action,” which can be undertaken without the support of the host government.

The common theme is that the CIA can conduct strikes contrary to or at least absent Yemen’s permission, whereas JSOC cannot. I can imagine any number of policy reasons why this might be so, why it might be the preferred course. But in each piece the claim is that this is a matter of legal obligation, not just policy preference.

And so here is my question for readers: What is the nature of the legal obligation supposedly at work here? Obviously covert action does not require host-nation consent. That’s not the curious part here. The curious part is the suggestion that DOD activity does require host nation consent as a matter of law. What law so requires?

The thing that jumps out as potentially relevant here is that, absent host nation consent, a use of force on that state’s territory presumptively violates its sovereignty under UN Charter Article 2(4). So perhaps the idea is that this is ok for the CIA to do via a covert action, but not ok for DOD to do. But is this really a dispositive concern here? Article 2(4) is not ironclad. An objection from Yemen under Article 2(4) would fail in the face of a U.S. Article 51 self-defense claim insofar as Yemen is unable or unwilling to use its authority to address the threat posed by AQAP itself. So long as one accepts that such unable/unwilling arguments can be made, the Yemen scenario seems like a very plausible place to invoke it. If that’s right, then DOD would not face an Article 2(4) hurdle in acting without Yemen’s consent.

Could it be that there are facts available internally that make the unable/unwilling argument implausible as to Yemen after all? I’m skeptical. Could it be instead that someone has concluded that DOD in particular simply may not avail itself of the unable/unwilling argument, as a matter of law? I’m not aware of such a law. Might it all merely reflect a high level of discomfort with the unwilling-and-unable test as a legitimate concept? Is it about something else entirely? Or is it all just a question of policy preference, not properly described as a legal distinction after all? Reader input is most welcome, whether you actually know or are just guessing!

Before turning to the legal question, let me say first that one of the most important features of the WSJ and WaPo reporting was the observation that apparently one of the reasons the CIA was being tasked with the mission was because of its experience in Pakistan not merely in running drones – which, after all, are often actually piloted by USAF – but rather in the utterly crucial intelligence-gathering operations on the ground that make possible what the drones do with missiles. The success of the drone program in counterterrorism operations in Pakistan has come about, so I have been told, on account of the CIA managing over the past several years to set up its own ground-level intelligence gathering operations in both Afghanistan and Pakistan – independent of Pakistan’s ISI. That independence has been crucial, for obvious reasons (to readers of The Onion, anyway) and apparently that ability to independently determine targets, not just independently strike at them, has greatly irritated Pakistan’s military.

This illustrates a crucial feature about targeted killing through drone warfare or, for that matter, using human teams. It is not solely a technology, the technology of drones, but instead equally or more dependent on an extraordinary intelligence effort at the ground level in order to identify targets. Drones, in their surveillance role, can be useful, but nothing substitutes for the ground level intelligence network. In that sense, the fear that critics sometimes have (that drones are a kind of weird mixture of globally ubiquitous surveillance-and-attack system that can strike anywhere around the world and at the same time a kind of flock of Predator ronin, restlessly seeking out targets anywhere, Paris France or Paris Texas) of the technology is misplaced, as I have written in this draft essay on SSRN. Drone warfare as a technology of targeted killing in counterterrorism is global in the sense that the drones can be piloted globally – but the most important intelligence is local and ground-level. If they are to be truly useful as targeted killing platforms, they must be tethered to local intelligence-gathering.

But now a brief thought on Bobby’s question. I wonder whether the journalists here actually do mean a truly legal distinction, as distinguished from an operational policy position that is being expressed as an internal legal policy position. I mean by that, an alterable and not necessarily absolutely held interpretation of an internal legal position, including interpretations of existing regulations and executive orders, within the intelligence community. That would not put it into “law” in the sense of deep interpretations of the UN Charter or international self-defense, etc. The legal positions might simply be existing secret executive branch legal opinions, whether of the Office of Legal Counsel or others, concerning existing executive branch legal policies.

That is, I find it hard to imagine that there is a difference of opinion between DOD and CIA on the scope of what the President can order in the way of action by one agency or the other with regards to whether Yemen is unwilling or unable to prevent the use of its territory by non-state actor terrorist groups. That has been US government policy stretching at least back to the Reagan administration. I would find it hard to imagine that there could be a legal view that it is lawful for the President to order the CIA to engage in armed covert action in Yemen without its consent – but that this would be a problem for the US armed forces. Among other things, why have a military if not to at times engage in military action on the territory of another sovereign without its consent? And as to attacking non-state actors, we’ve been doing that for a long time, in multiple places, including using DOD, so I can’t really believe that is at issue.

So my outsider’s guess is that this has nothing to do with international law jus ad bellum or even domestic law as such, and is more likely that there is a division of labor being established around operational capabilities. Those include, as the articles mention, the view that the CIA is better equipped to set up a program integrated with ground level intelligence gathering crucial for success. The legal aspect of this, I would guess, would be in some set of secret execute orders [ corrected, sorry typing late night] (the so-called “exords”) directed to DOD that are both “law” in the sense that these are legally binding and limiting instructions, but more in line with operational orders, revisable by the President and not “law” in the sense of international law on consent.

I would guess, as well, that part of the issue here does indeed lie with “covert” as distinguished from “clandestine.” Meaning, I would guess that a reason why the CIA is being tasked with this is partly in order to set up a genuinely covert intelligence gathering operation on the ground, in which the agents are covert and deniable should the US government wish to treat it that way. DOD personnel, by contrast, including special ops teams, might operate clandestinely – but if taken prisoner, will be acknowledged by the US government. I would guess, purely as an outsider, that the nature of the intelligence-gathering contemplates the former rather than the latter, and that this is a legal issue within the executive branch.

Finally, however, looking forward. I think this new operation, following on the Pakistan operations, raises an important question as to whether we need a conceptually new category in US domestic law – a new legal category in Title 50 of “deniable” alongside the existing category of (genuinely) “covert.” (I’ll be speaking tomorrow at a UVA National Security Law Program conference on this topic, and it’s one I plan to raise with experts there.)