International Law

Thoughts on Ken Anderson's Post on the CIA Drone Program

By Robert Chesney
Friday, October 21, 2011, 11:59 AM

Over at Opinio Juris, Lawfare Book Review Editor Ken Anderson raises a series of important questions about the CIA drone program.  In that post, Ken very kindly notes the relationship of these questions to my Title 10/Title 50 project, and asks me to weigh in.  My thoughts appear below, after short restatements of the 8 issues Ken highlighted:

(1) Does CIA involvement in the use of deadly force in itself violate IHL?

I don’t think so. For some time after 9/11, the US took the position in the context of military commissions that the use of lethal force by an unprivileged individual was itself a war crime (see, e.g., the Khadr prosecution).  My understanding is that this is no longer the U.S. position (note that in al Nashiri, for example, the charge of murder in violation of the law of war is hooked to an allegation of perfidy, not just the absence of combat immunity).  The latter view is the correct one, in my view, and it speaks directly to the question of whether CIA control over the drone program is in itself an IHL violation.  That is: CIA personnel may be exposed to (theoretical) criminal liability under the domestic law of certain countries (Pakistan), but that does not make the sheer fact of CIA involvement a war crime.

(2) Is the CIA sufficiently accountable in relation to IHL constraints on how force is used?

I agree with Ken that it would be smart policy for the executive branch to make public much more detail about the legal standards (and associated enforcement and training procedures) that notionally would control in relation to any program involving the use of lethal force.  I don’t agree with Philip Alston that such transparency is itself an actual requirement of international law (see Philip’s very interesting article here).  By the way, the link to Hank Crumpton’s speech, and a transcript of his exchange with Mike Lewis in regards to early CIA efforts to bring in JAG advisers in relation to drones, is here.

(3) Why continue to insist on having the CIA conduct such operations “covertly” when everyone knows they exist?

Ken points out that some take the view that Pakistan would be still-less cooperative if we were to remove the fig leaf of deniability that follows from our insistence on not formally acknowledging the drone program there.  That’s a question for the judgment of the diplomats and analysts, as Ken says.  All I really have to add on that point is that the decision that operations in Pakistan must be deniable does not require that the CIA conduct them. The President can direct other agencies to conduct covert action, and in any event a military-commanded and militarily-executed drone program in Pakistan would qualify as “traditional military activity” exempted from the definition of “covert action” in 50 USC 413b(e).  So why still insist on CIA control?  Again the real explanation is probably diplomatic above all; it’s not just that Pakistan’s leaders want the fig leaf of deniability for the operations in general, but that they want to be able to make specific claims about the limits on U.S. military involvement in Pakistani territory. 

(4) News accounts strongly suggest that the various orders and findings that provide the military and the CIA with their operational parameters very likely produce differences in where and how easily the military and the CIA can use force in various locations.

I’m in complete agreement with Ken on this point.  I comb the public record on this point in my paper, doing my best to flesh out what the differences appear to be.  Ken notes that these are revisable, but not easily revisable, constraints.  That seems exactly right to me.

(5)  Ken’s fifth point concerns the operational integration of the CIA and the military, and the complexities this raises in terms of arguments for accountability to a military chain of command. 

I don’t have anything to add on this point.

(6) Here Ken concludes that the CIA’s experience, capabilities, and relation to the underlying intelligence network does the bulk of the work in explaining the CIA’s lead role in some drone contexts, as opposed to variations in the relevant legal constraints.

That sounds about right, combined with (i) the diplomatic sensitivities noted above and (ii), to a very limited extent, the constraints associated with variations in the Exords and Findings.  Put another way, I see the various legal complexities associated with the Title 10 / Title 50 debate as being consequences rather than causes of particular operational arrangements.  They are important consequences to understand, to be sure, and may justify a number of reforms (I suggest several in my article).  But it is much less clear that the legal nuances drive decisions as to who performs which actions where.

(7 and 8 ) A very important point here about the indigenous ground forces that the CIA reportedly controls in Afghanistan:

Ken foresees these forces taking on still more significance in the aftermath of a U.S. withdrawal (such a withdrawal seeming over more likely in light of domestic political and economic conditions), which sounds correct to me.  One can readily imagine an environment that is quite similar to conditions from late 1998 through mid-2001, except this time with (i) far more clarity with respect to permission to use lethal force, (ii) far more intelligence about who is who in theater, and (iii) far more familiarity with the available proxy forces.  This will ensure more success than was possible in the 1998-2001 period, but whether it will be enough success is quite unclear.  The million dollar question is whether the intelligence collection system will continue to function well over time after the U.S. presence is stripped down.   It might well do so, if a multi-party peace agreement in Afghanistan (and perhaps Pakistan as well) creates conditions in which remaining targets are quite isolated and proxy forces are relatively free to operate.  If conditions are otherwise, the situation could slowly regress to 1998-2001 conditions. 

Ken also highlights a distinct point about the way drone strikes relate to HUMINT networks: The strike itself is but the tail end of a chain that primarily consists of collection activities, activities that are very much in the CIA’s control.   This reinforces point (6).