Skip to main content
Menu
There’s Not Enough Data On Civilian Drone Casualties

President Obama offered a personal apology on Thursday to the families of Warren Weinstein and Giovanni Lo Porto, two hostages of al-Qaida who were accidentally killed by a U.S. drone strike in Pakistan in January. Weinstein, an American citizen, and Lo Porto, an Italian national, had been held captive since 2011 and 2012, respectively.

The apology was highly unusual, but the death of civilians in U.S.-led drone strikes is not. Drones killed between 421 and 960 civilians in Pakistan between 2004 and April 12, 2015, according to estimates from the Bureau of Investigative Journalism, which assembles its drone database using news reports from around the world, information published by WikiLeaks, and, in three cases, its own field investigations into ambiguous reports of deaths.

More Politics

Weinstein and Lo Porto are, per The Long War Journal, the first civilian casualties of drone strikes in Pakistan so far in 2015. As of April 13, the site, which is a project of the Foundation for Defense of Democracies, had not logged any civilian casualties as a result of the five known drone strikes carried out so far this year.

It’s hard to know how many casualties may be missing from the Long War Journal’s statistics. Weinstein and Lo Porto died in January, but their deaths were not acknowledged until April. The very existence of drone strikes is often classified, making it difficult for researchers to assess civilian casualties.

In his statement, Obama explained that, after he learned about the deaths of Weinstein and Lo Porto, “I directed that the existence of this operation be declassified and disclosed publicly. I did so because the Weinstein and Lo Porto families deserve to know the truth.”

Declassification and open discussion may be a comfort to these two families, but most families of civilians killed in the drone war have not received similar explanations or apologies. Because the government does not publish the names of civilians killed in these attacks, or even release estimates of the number of civilians who’ve died, several independent organizations have tried to fill the gap.

Like the Bureau of Investigative Journalism, the Long War Journal relies on a mix of aggregated press reports, particularly from the Pakistani press and original reporting. Between 2006 and April 13, 2015, the Long War Journal estimated that drones killed 156 civilians in Pakistan.

As part of its The Year of the Drone project, the New America Foundation studied civilian deaths using a smaller pool of news reports than the Bureau of Investigative Journalism or the Long War Journal, trading comprehensiveness for a higher degree of confidence in the reports it includes. Between 2004 and April 12, 2015, it estimated that 258-307 civilians had been killed in Pakistan.1

The Columbia Law School Human Rights Clinic conducted an audit of all three of these tracking organizations’ tallies, annotating their data sets where different organizations had varying estimates of civilian deaths for a given strike. In a report published in October 2012, Columbia estimated that drones killed 72-155 civilians in Pakistan in 2011 alone.

Estimates vary considerably, with most organizations preferring to acknowledge the ambiguity of their data by reporting a range of casualties rather than offering the false precision of a single number. Information on this topic is not generally available from the government directly, and when the United States does discuss deaths following drone strikes, its methods for distinguishing militants from civilians have drawn criticism. The Obama administration labels any man of military age killed in a drone strike as a militant by default.

Keeping statistics on deaths in a war zone is difficult, particularly when soldiers are carrying out strikes remotely and will never visit the places they bomb. It took the United States more than a month to learn that its forces had killed Weinstein and Lo Porto because of the difficulty in verifying who, exactly, was killed after the blast has gone off.

Because the government keeps its statistics under wraps and independent groups must work with data that’s been through an international game of telephone — news organizations work with local freelancers, who sometimes dispatch yet another layer of proxies to speak to witnesses — the only fact we can be certain of is that whatever the true tally of civilians killed in drone strikes is, it has now gone up by two. And while it’s likely that somewhere between 156 and 960 civilians have died in U.S.-led drone strikes, the number of personal apologies by the president of the United States to the families of the deceased is considerably lower.

Footnotes

  1. Although the Long War Journal starts with a larger pool of raw data, it uses different filtering criteria than the New America Foundation.

Leah Libresco is a former news writer for FiveThirtyEight.

Filed under

Comments Add Comment