Gabor Rona of Human Rights First took some time out from today’s hurricane to send me the following thoughts on my post from Friday engaging Glenn Greenwald on the subject of counting civilian casualties in drone strikes. I offer some thoughts in response afterwards:
Ben says that he is ”not confident (that he knows) what the proper set of assumptions are to bring to the conversation about counting civilian casualties.” I’m glad to see he’s raising the issue.
The proper set of assumptions are extant and readily available for popular consumption. The reason it might appear otherwise is that the U.S. government does not acknowledge them and many American opinionators merely pick up from where the government leaves off.
The first assumption is that we look to law for the answer. There are some who dispute this, but I will presume that on this, at least, Ben and Lawfare readers are on board.
The second assumption is that international law constrains what we do. Domestic law doesn’t address who can be killed in armed conflict, or outside of armed conflict, except for provisions like the standard prohibitions against murder and the more specific prohibition against assassinations, all of which are somewhere on the continuum from irrelevant to undispositive. It is conventional and customary international humanitarian law (or law of war, or law of armed conflict, or jus in bello) that provides the most detailed guidance on who can be killed in armed conflict. It is conventional and customary human rights law that provides the most detailed guidance on who can be killed outside the context of armed conflict.
The third assumption is that in armed conflict not of an international character, namely between a State party to the conflict and a non-State armed group that is party to the conflict, or between two or more non-State armed groups that are party to the conflict, targetability is limited to members of the armed forces of a party and those who may not be members of the armed forces but are directly participating in hostilities.
The fourth assumption is that in the absence of armed conflict, targetability is limited to those who present an imminent threat of deadly harm that cannot be ameliorated through means other than deadly force.
Now, I know these statements leave room for uncertainty. And we can discuss how terms such as “combatant,” ”militant,” “terrorist,” “insurgent,” “associated forces,” “imminence,” and indeed, “civilian” either contribute to, or confuse, the issue. But my third and fourth assumptions are not mere assumptions. They are statements of conventional and customary international law and they provide much more guidance than what our government has admitted. The reason our government has been coy is simply that some of its more influential voices aren’t particularly interested in, and don’t want to be constrained by, the law.
But the law is worthy of respect not merely because it is the law. It also dictates sound policy. Killings that are outside the scope of international law jeopardize international counterterrorism cooperation and risk creating more enemies than they eliminate, not to mention that they terrorize civilian populations.
The fact that someone like Ben, who has been making a career out of these issues for years, just now comes around to asking what the rules are, is a testament to the power of official American denial. There are many of us who have been flogging the applicable law for a long time. For example, Daphne Eviatar and I did so in a Foreign Policy article in May. I’ve even detailed this stuff, including the impropriety of justifying killing with terms like “militant,” on Lawfare before. But our voices, and those of the rather considerable number of international legal authorities that question U.S. legal interpretations, are a mere drop in the ocean of contrary views that ignore, or denigrate, or misrepresent the applicable law and dominate the U.S. conversation.
I have some ideas about why this is so. Unlike Europe, which is the other part of the world where the modern laws of war developed, we in the U.S. have no living memory of the effect of war on our civilian population. We need to go back to the Civil War for that. Europe, on the other hand has a different perspective, from WW I through to the Balkans. Europe calls it humanitarian law. We call it the law of war or the law of armed conflict. Europe understands that body of law to heavily emphasize the protection of civilians. We interpret that body of law to heavily emphasize the protection of our armed forces.
These are the meta-conditions under which we operate. I’m glad to see Ben articulate his uncertainty and hope that he’ll continue to ask uncomfortable questions that cast doubt on the legality and legitimacy of U.S. practice.
Gabor here has laid out, lucidly and clearly, international law standards for targeting, and he has made an impassioned case for fidelity to these principles. I actually have little argument with them, particularly as Gabor explicitly acknowledges that they “leave room for uncertainty.” Though I am certain that differences would open up if we probed, as he has stated things here, I have little argument with him.
The trouble is that this broad agreement, felicitous as it is, doesn’t really solve the problem of counting civilian casualties. Because the problem in some ways is less one of the legal principles that should guide U.S. forces in targeting than the factual assumptions one makes after the fact in counting civilian deaths as an intellectual matter. This latter project is surely informed by the law, as Gabor suggests in his first assumption and as I noted in my post. But it is not entirely answered by the law.
The reason is factual uncertainty—and factual dispute—as to whom we have killed. Take as an example Gabor’s third assumption: “in armed conflict not of an international character . . . targetability is limited to members of the armed forces of a party and those who may not be members of the armed forces but are directly participating in hostilities.” This assumption, with which I fully concur, accounts—I would guess—for the overwhelming majority of the legal justifications by the United States for its drone strikes. That is to say, in nearly all cases in which the United States conducts a drone strike, it has concluded that its target is either part of the armed forces of the opposing side or a civilian directly participating in hostilities. Now let’s assume, as I hypothesized in my post, that the United States conducts a strike at a target it believes in good faith and with strong evidence, is a legitimate enemy target—a compound used by Al Qaeda and filled with guys with guns, say. And let’s imagine that our forces have identified some, but not all, of those guys with guys as Al Qaeda operatives. And let’s imagine even that they have identified one of the people in the compound as a major terrorist figure. On this basis, I would have no doubt that U.S. forces would—believing they are acting entirely consistently with Gabor’s principles—hit the compound. And let’s imagine that when they do so, they kill both the high-value leader and some of the guys with guns whose specific identity they do not know.
In this situation, the question of the legality of the strike and the question of how one counts the civilian deaths are related, to be sure, but they are not the same question. The United States would argue—quite correctly, in my view—that the strike was legal to the extent that U.S. forces had observed the principles of distinction and proportionality, and the United States might point out as well that there is no requirement in law that collateral damage be zero.
But that legal analysis of the propriety of the strike would not answer the question of how many civilians had been killed. What if the family of one of the guys with guns announced that he was actually a student or a journalist? What if another family claimed their relative had been a doctor? Because we are generally not able to get detailed bios of each person killed in a strike, we end up making certain factual assumptions—as well as legal assumptions of the type Gabor describes. These factual assumptions will affect our count a lot. If we assume as an empirical matter that people with guns in a compound that is a lawful target, we will count fewer civilian casualties than we will if we assume that people whose family claim they are civilians were not directly participating in hostilities. If we assume that folks whom Pakistani Army sources describe as “militants” to journalists are part of enemy forces or civilians directly participating in hostilities, our numbers of civilians will be far lower than if we assume that someone about whom we know little individually should not be tagged by association. These differences don’t reflect necessarily a difference in legal principle. They reflect, rather, a different set of background factual assumptions in the absence of perfect information.