War Stories

Reading Between the Lines of Obama’s Drone-Casualty Report

The death toll is less significant than the fact we are dropping bombs outside of war zones in the first place.

Supporters of Defense of Pakistan Council, a coalition of religious and political parties, chant anti-US slogans during a protest against the US drone strikes in the Pakistani tribal region, in Peshawar on November 10, 2013.
Supporters of Defense of Pakistan Council protest in Peshawar against U.S. drone strikes in the Pakistani tribal region on Nov. 10, 2013.

A. Majeed/Getty Images

Just before the Fourth of July holiday weekend, President Barack Obama released data on how many people—combatants and civilians—that his drone strikes have killed since he took office. Despite the tight deadlines, reporters noticed one salient fact: The numbers were considerably lower than those compiled and published by non-government organizations over the last seven years.

Who’s right: Obama’s director of national intelligence or the outside analysts? Or does anyone really know, and what difference does it make anyway?

Obama’s report—the first such public document ever—stated that the U.S. launched 473 strikes from Jan. 20, 2009 until Dec. 31, 2015, killing 2,372 to 2,581 combatants and 64 to 116 noncombatants.

New America, over the years, has counted a few more strikes (505), a wider though not inconsistent range of combatant fatalities (1,860 to 4,049), but more than double the number of civilian deaths (216 to 254, or possibly higher still. The group counts an additional 160 to 271 deaths as “unknown” as it’s unclear whether they’re combatants or civilians).

The Long War Journal is closer to the administration on total strikes (471) but closer to New America on civilian deaths (207). The Bureau of Investigative Journalism (442 civilian deaths) is close to New America’s estimate if all the uncertain deaths are counted as civilian.

A few things need to be said about all of these estimates. They include not just drone-fired weapons but other kinds of airstrikes, including cruise-missile strikes, against presumed terrorists. They also count only strikes in areas outside of official war zones—that is, they exclude strikes or deaths in Iraq or Afghanistan. They do include strikes in Pakistan, Yemen, and, to a much lesser extent, Somalia, which are conducted mainly by the CIA, not the U.S. military—and where, until now, the U.S. government has not admitted to launching attacks at all. (This is progress, of sorts.)

Why are the numbers so different, especially on casualties? The administration’s report—which acknowledges that the other groups’ estimates are “significantly higher”—puts forth a few reasons. First, and no doubt the biggest of all, is that there are “inherent limitations in determining the precise number” of deaths, especially in “non-permissive environments.” That is, in areas where we cannot easily send soldiers or spies to count bodies on the ground in the aftermath.

Imagery from drones, aircraft, and satellites can help intelligence analysts take such counts, but much of this is guesswork and, until very recently, government leaders—neither politicians nor military officers—took little interest in the question. In part to avoid comparisons to the Vietnam War–era’s body counts, precision in such matters hasn’t been a high priority. More than this, though, nobody knows the real numbers. Nor is it always obvious whether the people killed were combatants or innocent bystanders, especially in places and conflicts where the combatants aren’t wearing uniforms.

In the early years of drone strikes under both George W. Bush and Obama, U.S. intelligence officials tended to count all military-age males in the vicinity of an attack as “combatants.” However, this new report seems to be based on stricter criteria. It states that a dead body is counted as civilian if “there is insufficient basis” for tagging it as a combatant. The report also acknowledges, “Males of military age may be non-combatants.”

Another reason for the difference in numbers, according to the report, is that (ahem) the government knows more than private organizations do. New America, the Long War Journal, and the Bureau of Investigative Journalism base their numbers on news stories from a wide variety of media sources. The government draws on these open-media sources but also on what the report calls “sensitive intelligence,” including imagery, human agents, signals intelligence, and geospatial intelligence—all gathered before, during, and after the attack. Though the report doesn’t elaborate, these sources would include NSA intercepts of cell-phone conversations and GPS traces that pinpoint the location of the people having these conversations.

That said, the report’s estimate of civilian casualties is almost certainly too low. For instance, a U.S. cruise missile attack in Yemen, on Dec. 17, 2009, killed 41 civilians. This is widely known and was confirmed on background by U.S. officials not long after the fact. Those 41 deaths are presumably included in this report’s data. Yet, given the report’s estimate of a total 64 to 116 civilian deaths as the result of 473 strikes over a period of seven years, this would mean that—excluding the Yemen cruise-missile strike—as few as 23, and no more than 75 civilians, were killed in this entire timespan.

This is very difficult to believe and raises serious questions about how the intelligence agencies go about gathering and calculating the data to begin with. It, in short, raises questions about the entire report’s credibility.

In a larger sense, though, the specific numbers of civilian deaths—whether they total 60 or 200 or 400—are less significant than the fact that there are civilian deaths in these countries caused by U.S. airstrikes at all. This is an issue not just of morality, human rights, and international law, but also of national-security strategy: Some of these innocent victims’ friends or relatives might react to their sudden, seemingly senseless murders from the sky by sympathizing with, or even joining, the terrorist group that was the target.

Overall, it is a good thing that Obama is, however belatedly, releasing these numbers. It is a better thing still that along with the report, he issued an executive order that, beside calling for greater measures to protect civilians in these sorts of attacks, requires the director of national intelligence to release a public report each year estimating the resulting casualties. If the pilots, analysts, and officials involved in launching these attacks know that the president wants to see all the data on casualties, they might take more care in gathering the data.

Peter Bergen, director of the New America project that’s been tabulating drone strikes and their casualties, said in a phone conversation Tuesday that the report’s very existence suggests a “migration” of these operations away from the CIA’s secrecy to the military’s more open practices. If that is true, it may soon be possible—finally—to have a free, informed debate on whether hurling thousands of bombs and missiles from the sky, in a country with which we’re not at war, is a good idea in the first place.