Armed Conflict Cybersecurity & Tech Foreign Relations & International Law

Troubling Visualization of U.S. Drone Campaign in Pakistan

Ritika Singh
Tuesday, March 26, 2013, 3:34 PM
This is sure to satisfy your targeted-killing jonesing for the week: Out of Sight, Out of Mind, a website launched by a company called Pitch Interactive, documents---quite stunningly in visual terms, if not altogether fairly---the scope and scale of the U.S. drone program in Pakistan. The site description reads: “This project helps to bring light on the topic of drones.

Published by The Lawfare Institute
in Cooperation With
Brookings

This is sure to satisfy your targeted-killing jonesing for the week: Out of Sight, Out of Mind, a website launched by a company called Pitch Interactive, documents---quite stunningly in visual terms, if not altogether fairly---the scope and scale of the U.S. drone program in Pakistan. The site description reads: “This project helps to bring light on the topic of drones. Not to speak for or against, but to inform and to allow you to see for yourself whether you can support drone usage or not.” Actually, this is not quite honest; the data visualizations are designed to speak against drone usage. But they dramatically illustrate the number of drone strikes that have taken place between 2004 and 2013, and, in a separate view, they provide information about the victims of each of the strikes. The casualties fall into one of four categories: children (red), civilian (maroon), other (grey), and high-profile targets (white). The site contends that the numbers currently stand as follows:
  • Number of children killed: 175
  • Number of civilian deaths: 535
  • Number of other deaths: 2,348
  • Number of high-profile target deaths: 47
  • Total casualties: 3,105
The presentation of the data set is deeply flawed, which is a shame because the site does the best job I’ve ever seen of illustrating the acceleration of the drone campaign and the density of the strikes over the last decade. The data come principally from the Bureau of Investigative Journalism, save for data on high-profile targets, which come from the New America Foundation. These are two organizations that have played major roles in the accounting of casualties from drone strikes. The trouble with the presentation is two-fold: First, the BIJ’s estimates of civilian casualties are significantly higher than those of other counting organizations. So by building a site that takes these data at face value, Pitch Interactive is actually taking a position on a hotly-disputed question. That’s their right, of course, but they are not simply “informing” the public either. Pitch Interactive compounds this problem with its treatment of the “other” category, which is almost surely composed overwhelmingly of legitimate targets. These are, after all, people who BIJ is not contending are civilians; and they do tend to be people whom press reports have categorized as “militants.” The visualization, in fact, does not have any kind of category for folks short of high-value targets, such as “militant” or “belligerent” or “combatant.” It describes the “other” category, rather, as follows:
The category of victims we call “OTHER” is classified differently depending on the source. The Obama administration classifies any able-bodied male a military combatant unless evidence is brought forward to prove otherwise. This is a very grey area for us. These could be neighbors of a target killed. They may all be militants and a threat. What we do know for sure is that they are targeted without being given any representation or voice to defend themselves.
Put simply, the visualization is implying that of 3,105 drone-strike casualties, only 47 are known to be legitimate kills. And that is nonsense. Gathering and reporting data on the number of militant versus civilian casualties based on media accounts fraught with discrepancies and outright contradictions is inherently problematic. There are also critical disputes about definitions: Which people count as victims? Which are legitimate targets, even if not high-value personalities? There is simply no perfect way to handle these problems and, thus, the number of civilian casualties in general will necessarily remain a matter of varying estimates and ranges. What I’d like to see is a site as well executed as this one that estimates drone strike casualties based on assumptions and legal views selected by the user. To cite only one example, the visualization would be different if you clicked on a button that said “an able-bodied military-aged male on a site known to house an Al Qaeda leader is presumed to be a legitimate target” versus if you clicked “anyone not individually identified as an Al Qaeda figure should be presumed to be a civilian.” Now that would be an interesting juxtaposition.

Ritika Singh was a project coordinator at the Brookings Institution where she focused on national security law and policy. She graduated with majors in International Affairs and Government from Skidmore College in 2011, and wrote her thesis on Russia’s energy agenda in Europe and its strategic implications for America.

Subscribe to Lawfare