Wednesday, September 12, 2007

Lies, dammed lies, and Iraq statistics

Over the past couple of days, General David Petraeus has been presenting his version of the 'situation' in Iraq to various congressional committees. I sat and watched some of this testimony and was impressed to see the General use a wide array of charts and graphs to back up his claims of positive, but limited progress.

I have no particular political axe to grind, I'm all in favor of a solution over there, whatever that turns out to be. But one chart in particular did interest me. It was this one (sorry about the quality):

If you can't read it, this the number of attacks in Iraq (from top to bottom) against infrastructure, from IEDs (found and exploded), sniper and small arms attacks, and mortar/rocket attacks.

The General used this chart to support the notion that the surge (the increase in US forces since Jan, which peaked in Jul) is working to reduce attacks. I think it is, but how is it working? Have the increased offensive actions of US units destroyed the ability of insurgent groups to mount these attacks? Or has the increased number of US troops just made it more difficult to carry them out?

Either way it's a good outcome right? Well yes, but also no. If it's the former, you are destroying the capability for future attacks. If it's the latter, you are curbing attacks in the immediate future, but not really in the medium to long term - you're just putting your finger in the dike so to speak.

In the testimony I saw, the General didn't really elaborate on which mechanism was at work. There was no accompanying graph of captured or killed militants. And no indication of militant numbers who just went back home to sit out the surge because their chances of capture were higher.

So what can the graph itself tell us about which mechanism might be at work? One interesting observation is of the steepness of the decline post July. If militants were still going flat out to bomb and kill US troops and civilians in that post July period, even with the surge and very successful offensives by US troops, you wouldn't expect the decline to be so steep. Why? Because it's been shown in the past that these attacks can be carried out with relative ease and despite heavy US troop presence. If the surge was slowly eroding away the ability of militant groups to launch attacks, you would expect the decline to be less steep - more of a gradual fall.

More likely, the militants have just stopped attacking. They are having a bit of a rest. Re-grouping and waiting for the right opportunity to begin in earnest again. This is far more likely to be represented by a steep, rather than gradual decline in these numbers. Given the sheer amount of activity from Jan to May 07, you would think a rest would be in order!

If this is the case, the force reduction plan might just be 'taking the finger out of the dike' enough to cause a new flood of violence. Let's hope not

Digg this
Sphere: Related Content

No comments: