An Algorithm For War

by Zoe Pollock

Ron Rosenbaum makes a convincing moral argument against drone-porn war crimes:

Drones mean you don't need to win hearts and minds if you're allowed to blow away the bodies of "the enemy" without risking U.S. lives. But at what cost? Few of us have wanted to scrutinize too carefully a program that holds out the tempting promise of "victory" and thus the withdrawal of large numbers of troops from Afghanistan sooner rather than later. Or to look at the downside: that drone slaughter—whether or not it's a war crime—is counterproductive, creating generations of potential terrorists from the families of the innocent victims of careless carnage. A 2009 Brookings Institution study estimated that for every "militant" killed, there were 10 civilian casualties. And critics have pointed out that each of them will have 10 grieving relatives who will become "militants" or supporters in all likelihood.

Of course, there's a lot of controversy over the percentage of noncombatants killed in the drone strikes. One study, not very convincingly, puts civilian casualties at slightly above 3 percent. Another says 10 percent, another a full one-third, Brookings far more. Do these different numbers yield different moral conclusions? Are the drone strikes defensible at 4 percent murdered innocents but indefensible at 33 percent? There's no algorithm that synchs up the degree of target importance, the certainty of intelligence that's based on, and potential civilian casualties from the attack. It's a question that's impossible to answer with precision. Which suggests that when murdering civilians is involved, you don't do it at all.