Targeted Killing

Questions About CIA v. DOD Drone Strikes

By Jack Goldsmith
Tuesday, May 13, 2014, 9:55 AM

Two reasons often given for the need to transfer targeting killing by drone from CIA exclusively to DOD are (i) collateral damage reduction, because DOD supposedly has stricter targeting criteria and better law-of-war compliance, and (ii) more robust oversight.  Neither reason is sound.

Taking the second point first, as we have often noted on this blog, the Intelligence Committees are much more informed about CIA drone strikes than the Armed Services Committees are about DOD drone strikes, and appear to exercise more rigorous oversight.  According to Diane Feinstein, SSCI “receives notifications with key details of each strike shortly after it occurs, and the committee holds regular briefings and hearings on these operations—reviewing the strikes, examining their effectiveness as a counterterrorism tool, verifying the care taken to avoid deaths to non-combatants and understanding the intelligence collection and analysis that underpins these operations.”  Such reporting is required by law, since CIA has a statutory duty to “keep the congressional intelligence committees fully and currently informed of all covert actions.”  By contrast, DOD has no such legal duty to the Armed Services Committees, and apparently does not report its drone targeting activities in nearly the same detail as CIA does to the Intelligence Committees.  The Armed Services Committees' inferior oversight is confirmed in the most recent draft Defense Authorization Act, which, as Bobby noted last week, would require “case-by-case notifications to SASC and HASC when the military carries out a targeted kill or capture operation outside of Afghanistan.”  The need for such a law implies that such notifications are not now the norm.

As for the first point, as we have also frequently noted, DOD appears to make mistakes in its drone targeting much more frequently than CIA.  According to Mark Mazzetti in his book The Way of the Knife, such errors (and related issues) convinced President Obama to transfer the task of finding and killing Anwar Al-Awlaki in Yemen from DOD to CIA, which did the job.  This is likely not a result of difference in legal compliance – it is more likely a result, as Mazzetti noted, of superior CIA intelligence.  Ken Dilanian’s LAT story yesterday, which Bobby already discussed, appears to confirm this point, at least with regard to the latest supposed DOD error in a drone attack in December that allegedly killed many civilians.  According to Dilanian, “the CIA informed the command before the attack that the spy agency did not have confidence in [DOD’s] underlying intelligence.”  If it turns out that CIA was right about the persons targeted and DOD was wrong – Dilanian reports that the “National Counterterrorism Center, which coordinates terrorism intelligence from multiple agencies, is somewhere in the middle, saying the evidence is inconclusive” – it will be further confirmation of CIA’s relative superiority, at least with respect to intelligence.

All of this might make one wonder why there is (halting) movement to transfer CIA drone operations to DOD.  But more fundamental questions puzzle me.  For starters, I am not sure I understand what it means to transfer drones from CIA to DOD, since Air Force personnel appear to be pulling the trigger in any event.  I think the answer is that CIA and DOD have (a) different bases from which they launch their drones (for Yemen strikes, Mazzetti reported, CIA operates from bases in Saudi Arabia and JSOC operates from bases in Ethiopia and Djibouti);  (b) different forms of intelligence support for drones; and (c) different aims and authorities in their drone operations.  As a result, CIA and DOD make different decisions about when, where, and at what to fire.

But then further questions arise.  Why aren’t DOD and CIA drone operations entirely integrated – using the best intelligence, the best command structure, the best law-of-war compliance mechanisms, and the like – into a single organizational structure where the DNI controls intelligence and the DOD executes military strikes in the chain of command based on this intelligence?  To what extent are (or should) drone targeting decisions be coordinated by the NSC?  What is the NCTC’s role?  Why does DOD keep going its own way on intelligence, apparently, over the years, with unfortunate results?  To what extent is the CIA-DOD split driven by the Title 10 v. Title 50 distinction, either historically or presently?  If Title 10 v. Title 50 is driving this, why does the President allow it to do so, since he possesses enormous discretion about which track to use, and since DOD can engage in covert (or otherwise non-acknowledged) actions and CIA can engage in non-covert (i.e. acknowledged) actions?  How much do budget and oversight and related jurisdictional and turf disputes between the intelligence and armed services committees drive the non-integration of drone strikes?  And finally, are there underappreciated benefits from allowing CIA and DOD to compete over drone strikes?

I am confident that some of these questions betray ignorance about what is really going on.  But that is the point.  At the level of elite public consumption, at home and abroad, the debate has been about “transferring drone strikes” from CIA to DOD.  This is a simplistic and misleading formulation.  The important questions, I think, include the ones in the previous paragraph.