Show Us the Data and the Methodology, Petraeus

Via Steve Benen, here is hilzoy at Obsidian Wings on those U.S. military statistics supposedly showing decreased levels of violence in Iraq:

Here are some of the things we know about these statistics: they don't include Sunni-on-Sunni violence, or Shi'a-on-Shi'a violence. They don't include car bombings. There are unexplained changes in the figures from one report to the next. They don't seem to take seasonality into account.

More discussion below the fold.

The reason for not including car bombings is just ludicrous:

"Car bombs and other explosive devices have killed thousands of Iraqis in the past three years, but the administration doesn't include them in the casualty counts it has been citing as evidence that the surge of additional U.S. forces is beginning to defuse tensions between Shiite and Sunni Muslims.

President Bush explained why in a television interview on Tuesday. "If the standard of success is no car bombings or suicide bombings, we have just handed those who commit suicide bombings a huge victory," he told TV interviewer Charlie Rose."

Gosh. By the same logic, why include any deaths at all among the civilian casualty numbers? After all, people who kill people are often bad guys, and who wants to hand any bad guys a victory?

The reason for not including Sunni-on-Sunni or Shi'a-on-Shi'a violence involves a much deeper mistake.

Cast your minds back a few years. Back then, a lot of people thought that the only people we had to worry about were the insurgents. The growth of Shi'a militias -- well, that wasn't really a problem. (There are some pretty amazing quotes to this effect in this article by Spencer Ackerman, which I provide some excerpts from here.) We characterized the problem too narrowly: as I wrote back then, "The insurgency is the militia (or: group of militias) that's dangerous to our troops. All the militias are dangerous to Iraq." But we focussed only on the insurgency; as a result, we missed the chance to try to deal with the militias when they were much, much weaker.

Now, we're doing it again. The danger we should be worried about, it seems to me, is a civil war. One form that a civil war might take is: Sunnis fighting Shi'a. But that is not the only possibility. It's not even the only actuality. In Basra and large areas of the south, Shi'a are fighting Shi'a for control. The recent fighting in Karbala was Shi'a on Shi'a. If we do not focus on this, we'll be making the same mistake we made before: defining the problem too narrowly, and waking up one morning to realize that the part of the problem we were ignoring has become the main event. That being the case, to define the relevant set of casualty statistics to exclude Sunni-on-Sunni or Shi'a-on-Shi'a violence is to make a profound mistake.

There are various lists of articles trying to measure the actual casualty figures, or questioning the government figures. See, for instance, here, here, and here. But there is an easy way to resolve these issues. If the government were to release its figures, and explain the methodology behind them, then it would be clear whether they had been cherry-picked or not. ...

In Steve's words: "Does anyone seriously believe that if the internal data pointed to legitimate progress in Iraq, the administration wouldn't release it?"