Skip to content

Are Drones More Likely to Inflict Civilian Casualties?

By Taylor Marvin

MQ-9. US Air Force photo by Kristi Machado.

MQ-9. US Air Force photo by Kristi Machado.

At Duck of Minerva, Charli Carpenter highlights a recent Guardian piece by Spencer Ackerman relaying the claim by a classified Department of Defense study suggesting that drone strikes over a yearlong period in Afghanistan “caused 10 times more civilian casualties than strikes by manned fighter aircraft.” The study suggest that this higher incidence of civilian casualties is due to insufficient training of drone pilots compared to their in-the-cockpit counterparts. Carpenter notes that if this is true, the argument that drones are a more humanitarian option than crewed aircraft would be substantially weakened.

(Though, importantly, drones strikes within Pakistan are often held up as the more humane option compared to Pakistani military action, not US airstrikes conducted by crewed aircraft, so the study’s findings would not necessarily affect this higher-profile controversy over the ethics of drone strikes.)

There are numerous reasons to doubt the validity of these findings: First, as Carpenter remarks, the study itself is classified — making its data and methods impossible to verify. Secondly, as Dan Nexon argues, there are many within the DoD, and particularly the US Air Force, that find the prospect of a future where fighter jocks take a backseat to drones deeply unattractive and have an incentive to discredit UCAVs. Finally, there’s an obvious potential bias here: as other commenters noted, in the last decade the United States has disproportionately used drones to conduct signature strikes in civilian areas, while crewed aircraft are more often used in more traditional combat strike roles. It’s entirely possible that the manner in which drones are used, rather than operator training or anything inherent to remotely piloted aircraft, is more likely to cause civilian casualties. Unless the study in question is made public we won’t know if this bias is accounted for.

But beyond these issues it’s important to remember that this study has little relevance to the debate over drones themselves. There is little reason to suspect drones are inherently more prone to collateral damage than crewed aircraft, at least in the manner which they have recently been utilized. After all, when conducting signature strikes in undefended airspace — drones’ signature mission in the last decade — there is little qualitative difference between remotely piloted aircraft and crewed strike aircraft. Instead, what makes drones different is their political baggage, which is typically understood as less restricting than crewed strike aircraft. In many situations — in particular, contested airspace — remotely piloted aircraft will likely suffer more from limited situational awareness and other operational limitations than crewed aircraft, limitations that may make them more prone to killing civilians. But there is again little reason to suspect that the limitations of contemporary unmanned platforms have affected drones’ propensity for inflicting civilian casualties today. Instead these civilian casualties are more likely due to the United States’ division of labor between crewed and uncrewed strike aircraft, rather than the inherent qualities of each platform.

Drone strikes will remain controversial, as they should. But it’s important to untangle debates over the consequences of UCAVs themselves, and of their applications. I’d argue that this study is more relevant to the latter than the former.

Advertisements
No comments yet

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: