The Limits of Retroactive War-Assessment

Kevin Drum is absolutely right that initial reactions to Obama's speech yesterday will have nothing to do with the political impact of the Libya operation compared with actual results:

In the end, Obama will be judged on whether his approach works. If U.S. involvement really stays limited; if Qaddafi finds himself out of a job within a few weeks; and if the aftermath of the war isn't too disastrous, then Obama will be vindicated, congressional approval or not. If any of these things doesn't happen — and I'd pay particular attention to the last of them — he'll be in trouble. As with all things, success justifies nearly anything.

Unfortunately, this isn't really a sound way of looking at things. Bad bets still sometimes pay off. And it seems to me that one big problem with initiating wars is that they have relatively unbounded downside risk. You could start five wars, have four of them "work" well enough, and still come out as a loser on net. In particular, I think this is a problem with moving beyond helping the rebels defend Benghazi and insisting on Gaddafi's departure. Gaddafi is bad, and if he goes a better government might come into place, but the Libyan state could also collapse and create a new humanitarian catastrophe. And whatever happens, the reality is that western officials are making this decision while possessing very little relevant information.