In an exchange of posts over the past few weeks about the most recent casualty figures released by the Director of National Intelligence for theaters outside areas of active hostilities, Gabor Rona and Ryan Goodman have explored Rona’s argument that the data in the DNI report cannot be trusted. To recap:

  • In his February 7 post, Rona argues that the DNI report is “not credible” because U.S. casualty statistics reflect (1) the likely overapplication by the U.S. government of the law of armed conflict (and accordingly its rules permitting individuals to be targeted on the basis of “combatant” status) to situations where he believes higher human rights law standards ought to apply, (2) a definition of “combatant” that Rona believes is too broad and (3) a definition of “imminence” that he believes is too loose.
  • Goodman’s rejoinder of February 22 questions one of the core assumptions in Rona’s thesis–i.e., that the United States overstates the scope of the al Qaeda/associated forces/ISIL conflict and therefore is likely mislabelling significant numbers of civilian casualties as combatant deaths. Among other things, Goodman notes that there is widely-acknowledged ongoing armed conflict in Yemen, Somalia, and Libya–all countries where the U.S. has recently stated it has ongoing military operations and that the U.S. government has described as “outside areas of active hostilities” for purposes of the Presidential Policy Guidance.
  • Rona’s reply of February 23 recognizes that there is armed conflict in each of these countries, but argues that the U.S. government has not supplied sufficient evidence to prove that it is party to these conflicts. He also argues that his mistrust of DNI’s data is rooted in the U.S. government’s use of the terms “armed conflict,” “combatant,” and “imminent”–which he suggests are “at odds with the established understanding of those terms under international law.”

While it hardly seems likely that we will close the distance between the two sets of posts with this one, we wanted to weigh in on a few key points of disagreement before the exchange is retired to the digital archive, at least to make sure that they are on the record. A few statements that caught our collective eye in Rona’s last post are as follows:

1. Rona argues that, (w)e don’t know where the Administration considers itself to be at war, and therefore we don’t know if it using the more ‘liberal’ killing rules of the law of war in places and situations that do not, in fact, amount to war.” Assuming that this refers to the Obama Administration, we read the record very differently. Recognizing that there is a great deal of political sensitivity around the question of the geographic scope of the conflict, the Administration chose its words very carefully, but made three essential overlapping points: First, the United States is in an armed conflict–the domestic and international law underpinnings for which are most comprehensively summarized in the December framework report–with al Qaeda and the Taliban, their associated forces, and ISIL. Second, the existence of this conflict does not, in the words of President Obama, give the U.S. government the right to strike “wherever we choose,” given the barriers imposed by state sovereignty under international law, which can be overcome by, among other things, the consent of the territorial sovereign where strikes are to take place. Third, of the six countries identified in U.S. government war powers reporting and the December framework report as places where the United States conducts military operations, the three that the DNI report does not describe as areas of active hostilities (Libya, Somalia, Yemen) have all consented to U.S. operations on their territory. What, then, is so troubling about this picture? The U.S. government has made clear that it is in armed conflict with certain groups that are operating in countries that have consented to the use of U.S. military force, on their territory, to engage those groups. If the criticism is that the U.S. government can be a bit fuzzy about whether it is relying on a law of armed conflict justification or a self-defense/imminence justification when it takes action against those groups, then that may be a conversation worth having, but we’re not sure we see much cause for mistrust in this–since what it effectively means is that the United States might, in theory, occasionally rely on a self-defense/imminence justification instead of a less stringent law of armed conflict/status justification when acting in situations where both are available.

2. Rona argues that “Even in contexts that clearly are war … our government applies an overly broad definition of the term ‘combatant.’” As a threshold matter, we note that it goes too far to suggest that there is as clearly an “established” understanding as Rona posits of what “combatant” means in the context of a non-international armed conflict with an organized armed group like al Qaeda or its associated forces. Beyond a general sense that “combatant” connotes individuals who are not “civilians” and vice versa, there are important areas that still lack consensus. Indeed, from our perspective, one of the major challenges in international law over the past fifteen years has been to try to flesh out a common perspective on these terms, at least among the United States and its like-minded partners, in the absence of a black letter definition. The ICRC’s study on Direct Participation in Hostilities made an important contribution to this discussion but can’t be relied upon as expressing the view of States on this matter. What’s more, the U.S. government’s involvement in efforts to find common ground with some of its closest partners yielded a range of views. While there is still work to be done in narrowing that range and in identifying the areas of overlapping agreement, the U.S. government has offered, in a number of places, the contours of its definition–including in Brian Egan’s speech to the American Society of International Law and the December framework report. The Department of Defense has also included its views on the subject in its Law of War Manual. We do not doubt that Rona may take issue with some of the elements laid out in these sources, which is perfectly fair, but we do think it important that any conversation about the merits of the U.S. government’s approach to combatancy look at these sources, and also acknowledge where there is a lack of consensus about the metes and bounds of related terminology under international law.

3. Rona argues that, The initial U.S. position [concerning the recent Yemen raid] was that there were no civilian casualties. While the government swiftly retreated from that claim, the fact that it made the claim at all speaks volumes.” We see it very differently. It is important to bear in mind that U.S. and partner militaries often find it hard, in the initial period following a strike, to confirm civilian casualties, even in cases where there has been extensive pre-strike surveillance (which we note is standard,). Often as not, this is because the relevant command either lacks sufficient post-strike information, or has not been able responsibly to analyze it. Perhaps the early denials of civilian casualties in the Yemen raid should have been caveated more explicitly. What is more important, though, is that the work to confirm the existence (or not) of civilian casualties continued, and that findings were made promptly public. While there is much in what has been reported about the Yemen raid that could be cause for concern, the way in which CENTCOM handled the disclosure of civilian casualties–pivoting quickly to acknowledge the fact (if not the number) of such casualties–does not strike us as rising to the top of that list. Indeed, commentators who have criticized the United States for lacking transparency in its targeting policy have lauded the speed and manner in which the Pentagon acknowledged civilian casualties in the Yemen raid.

4. Rona argues that “As long as our government asserts definitions of ‘armed conflict’, ‘combatant’ and ‘imminent’ that are at odds with the established understanding of those terms under international law, its claims of civilian casualties are likely to be lowballed and simply cannot be trusted.” The point we made above about the term “combatant” applies to “armed conflict” and “imminence” as well. To suggest that there is such an “established understanding” of these terms that would undercut the DNI report strikes us as being unfounded. Take “armed conflict.” While we do not dispute that there is a fairly standard view that “armed conflict” connotes a certain level of violence between a state and a group at a certain level of organization, Rona doesn’t suggest that the United States has failed to satisfy this test when it comes to the various groups under discussion. What he does seem to question is whether a conflict with these groups can be said to extend to countries away from the hot battlefields of Afghanistan, Iraq, and Syria–i.e., to countries like Libya, Somalia, and Yemen. While the U.S. government’s position (and ours as well) is that it can, the question is certainly a fair one to pose. What is not persuasive, however, is to suggest that the answer is “no” as a settled matter of international law (if anything, as the writings of Anthony Dworkin show, European policies and legal perspectives on this issue have merged toward the United States in recent years).

Similarly, it is perfectly fair to probe the U.S. government’s approach to “imminence.” In so doing, however, it is important to bear in mind the point that we made above–i.e., that in each of the countries we are talking about (Libya, Somalia, and Yemen), there would be a LOAC justification for status-based targeting when dealing with the groups in question. Accordingly, whatever one’s views on the U.S. approach to imminence, any casualties involving group members could appropriately be tallied as “LOAC” combatants. Importantly, the imminence issue that Rona appears to be disputing is a policy overlay; to be sure, it has played a significant role in shaping U.S. counterterrorism activities in these three countries, but it is a legally unnecessary requirement, on top of what the law of armed conflict demands. We also note that any analysis of the U.S. government’s position on imminence should take into account the full discussion set forth in a speech by former Attorney General Eric Holder, a speech by former Homeland Security Adviser John Brennan and the Office of Legal Counsel’s Memo, in addition to the short quote that Rona offered.


In wrapping up this discussion, we do not mean to suggest that a U.S. government lawyer and an academic or civil society lawyer like Rona can’t arrive at different views about civilian casualty counts, particularly given different views of the relevant legal frameworks and different approaches to the relevant definitions. We also believe it is important for the U.S. government to continue to work on refining its definitions, developing international support for them, and putting out as much information as possible about them. But those are much more tailored claims than Rona is making in his recent posts .They go less to the issue of “trust” and more to recognizing that the law of armed conflict applies to these situations and that, uncomfortable as it may seem, some aspects of the law simply aren’t as settled as we all might hope.


Image: Shilo Thompson and Brandon Walker, both aircraft mechanics with the 361st Expeditionary Reconnaissance Squadron, load an AFM-114 Hellfire missile onto an MQ-1B Predator unmanned aerial vehicle at Ali Air Base, Iraq, on July 9, 2008 – Dept of Defense.