by Robert Wright

People continue to ponder whether the reported killing last week of the leader of the Pakistani Taliban will be a big plus for our side. More generally: Does killing terrorist leaders do much good? I’ve seen only one long-term study on the question, and, in the case of religiously-based terrorist groups, its findings weren’t encouraging: “The result that consistently stood out from this research was the propensity of decapitation strikes to cause religious organizations to become substantially more deadly.”

My own assessment is sunnier: Chances that the killing of Baitullah Mehsud will help our cause, hurt our cause, or have no effect are, respectively, 33 percent, 33 percent, and 33 percent. This is based on a rigorous multidecade study of organizations I’ve worked for. Here’s the data: When someone leaves the organizationwhether through resignation, getting fired, or even deaththe chances that their replacement will be more effective than them, less effective, or about the same are: 33, 33, 33.   

And why should we expect things to be any different when it’s a terrorist leader who is being replaced? 

We know that terrorist organizations aren’t exactly meritocracies. Their leaders don’t answer to a board of trustees or face elections, and few seem ethically opposed to ruthlessly hanging onto power. So their organizations may abound with capable but untested heirs who have been secretly chanting, “Put me in, coach.” Besides, most organizations could use a little shaking up now and then. Sure, the transition may be messy, but, hey, terrorism is a messy business.

An analogy: If you’re trying to put a candy company out of business, killing the occasional top-ranking executive probably won’t get the job done, or have much effect on the company’s fortunes at all. (Unless you kill with such frequency that the pool of aspiring replacements dries up, and in the terrorism business that would require, I think, a much higher rate of killing than we’ve generally mustered.) What you need to do is reduce demand for that brand of candy.

Now, reducing the “demand” for a brand of terrorism is hard to conceptualize. But it would include reducing the amount of grass roots support for the terrorists both in their own habitat and around the world. And one problem with trying to kill terrorist leaders with drone strikes is that the number of civilians you kill in the process can actually increase that support. 

Of course, civilian deaths can have the opposite effect, if enough of the public blames the terrorists for them. Still, I suspect that, on balance, collateral damage helps the terrorists and hurts us. One reason is that, even if civilian deaths cause the local populace to turn on the terrorist organization (a big if), they can still help terrorist recruiters in the wider world. So, come to think of it, my 33 percent estimate of success may be a bit high, all things considered.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.