The public part of President Obama's new war strategy is to put more soldiers on the ground in Afghanistan.
The not-so-public part is to expand our use of drones in Pakistan. In this morning's New York Times , Scott Shane reports that Obama "has authorized an expansion of the C.I.A.'s drone program in Pakistan's lawless tribal area" to parallel our Afghan surge.
Human Nature has been updating you on this trend for a while. Drones make it easier for us to fight in Pakistan and Afghanistan. They enable us to kill our enemies from a distance, without putting our troops in harm's way.
But does that distance deaden us to the significance of collateral damage? Do drones turn war into a video game ? Do they increase the risk of civilian casualties? An Amnesty International official makes that argument to the Times : "Anything that dehumanizes the process makes it easier to pull the trigger."
The argument makes sense. But by insulating their operators from danger, drones do more than dehumanize the killing process. They can hover over their targets, give their operators time to study the scene, plan, and think before firing. As Shane points out, "Operators at C.I.A. headquarters can use the drones' video feed to study a militant's identity and follow fighters to training areas or weapons caches, officials say. Targeters often can see where wives and children are located in a compound or wait until fighters drive away from a house or village before they are hit."
Which effect is greater: dehumanization or hover time? That's an empirical question. And the best way to answer it is by monitoring civilian casualties.
But that isn't an easy thing to do. Do you count only the dead you can verify as civilians from the drone's video feed? In areas controlled by the Taliban, do you wait for official investigations? There's no CSI Waziristan. Alternatively, do you trust whatever some local honcho tells the Pakistani press? Do you believe the eight guys who died in the targeted building were farmers? The former methods are likely to undercount civilian deaths; the latter is likely to exaggerate them.
Based on media reports, the New American Foundation credits the drones with an ugly 2-1 ratio of militant to civilian casualties. The Long War Journal reports a much more heartening ratio of 9-to-1 . And a government official offers Shane an estimate of 20-to-1, claiming the drones have killed only about 20 civilians.
I'm skeptical of the high count. If the 2-1 ratio were correct, I'd expect more outrage from the Pakistani countryside than we've seen. On the contrary, Shane notes that in a recent survey of Pakistani professionals in the targeted areas, half of them said the drone strikes were accurate, and most said they strikes didn't foment anger at the United States.
But I don't buy the low count, either. Never trust killers to report their death tolls. That's a good rule even when the killers work for your government and are fighting bad guys. Casualty counts, especially of civilians, must be independently checked.
I'd like to think that drones, because of their video technology and hover time, are making it easier to kill the enemy without killing civilians. If so, we should make them a bigger part of our strategy in Afghanistan and Pakistan. They could even police the area once the surge is over and our troops are coming home.
So let's figure out an internationally credible way to check their performance in sorting the good guys from the bad. If this is a better way to fight insurgencies, let's prove it.