Kilcullen: Send either lots of troops or none at all
November 12th, 2009
Some quick notes from Georgetown, where David Kilcullen has just addressed students and faculty at the Center for Peace and Security Studies. Highlights below:
- We’ve suffered from only incrementally increasing the number of troops over the years. The Taliban has proven itself capable of absorbing the impact from an additional 10-30 thousand troops. We need to either “overmatch” them with a substantially larger deployment or not send any at all (or possibly draw down).
- Whenever we send more troops, violence will spike almost by definition. This is for two reasons: a) the observer effect, more troops on the ground means more eyes on the ground, means more incidents get reported; b) more combatants means more combat. “It’s like opening the fridge door and the light goes on.”
- The oft-touted 1:50 (or 20:1,000) ratio is “flawed.” It was based on post-war reconstruction studies done by the Rand Corporation, not on actual insurgencies. Successful COIN campaigns have employed ratios that vary widely. It also refers to total security forces, not just — in our case — American troops. Finally, it’s better to think about the military presence functionally, rather than numerically.
- There is “not much point” to negotiating with the Taliban right now. This is because the Taliban believe they are winning and so have no reason to bargain. Our goal should be to fight first and hard, to convince them that they should talk.
- “Where local officials sleep” is a good indicator to track progress. In the film, I Am Legend, Will Smith must get home before the vampires come out to feast. Similarly, in Afghanistan today some 70% of provincial governors sleep in Kabul instead of the provinces they govern. This is bad.
- Successful counterinsurgencies take 15-20 years. Unsuccessful ones take 9-11 years. Since 1816, 80% of counterinsurgencies have been successful, but when you control for whether those campaigns are being waged on domestic or foreign soil and whether the governments in question were willing to negotiate with the insurgents, the number can vary widely. Counterinsurgents have won only about 20% of the time when the government has not been willing to negotiate and when the intevening force was of foreign extraction.
- There is “no universal silver bullet” for winning counterinsurgencies; “there are no templates.” Counterinsurgency itself can best be described as “a battle for adapation…against an enemy who is evolving.”
- COIN should be viewed as “a subset of stability operations” because it is not a strategy.