Metrics Schmetrics

Much discussion about metrics at the Center for a New American Security's (CNAS) annual conference today.  As I mentioned in a previous post, CNAS' report on Afghanistan put forward Afghan civilian casualties as the primary metric of success.  

Afterwards, a co-worker and I discussed metrics some more, and soon realized our focus on metrics is a little misplaced.  War is more complex and unpredictable than any other human endeavor.  We tried to come up with metrics that would have been appropriate in either of the World Wars, and came up empty-handed.  I'm not sure what metric could have told the story of Operation Iraqi Freedom.  In particular, the turning point, the Sunni Awakening, couldn't have been predicted by any metric. 

So metrics are not a magic bullet.  But metrics do serve two useful purposes.

1. They drive out bad metrics.  Bad metrics are a crutch.  In Vietnam, one of the official responses to failure was to start counting enemy casualties.  If an official metric is established, there's no room to lapse into bad metrics.   

2. They provide a concrete way to transmit commander's intent.  Most commanders will tell their troops to avoid killing civilians.  Such platitudes have more impact when a team's performance is measured in civilians protected, rather than bad guys killed.

Question for tomorrow: if we achieve success in Afghanistan, and create a secure, prosperous, well-governed state, will that make it easier to go after Al Qaeda Central in Pakistan?