Assessing the Significance of Attack Data in Iraq--Contrarian

Seems like I'm due to challenge conventional wisdom (warning--this is off the cuff and not based on research):

The media reports that the military reports the average number of attacks per month in Iraq has steadily increased over the years since the end of significant military operations announced by our leader.

But this may not be significant, at least not as significant as it sounds. Counting and reporting is one of the things bureaucracy does. You need a flow of information up the channels to allow decision makers to make good decisions. So the data sounds useful. But changes in the data can reflect a change in reality or a change in the reporting mechanism.

In the early days of the war, DOD did no body counts; that was too much like Nam. My impression is that Rummy and Myers ran away from statistics as fast as possible. But as time has gone on, everyone searches for data, so the attack figure is one that has come to the forefront. My guess is that the military has become more and more concerned about the accuracy of such figures. They're probably also responding to continuous pressure from the media for such information. As the military changes and learns, it inevitably changes the data reported.

Take an example: someone in the lead vehicle in a convoy hears a couple bullets fly over head. That's all, nothing more. I don't know whether that counts as an "attack" today, but I'll bet it didn't count as an attack in July 2003. As a bureaucracy, the military has probably issued instructions on what counts as an attack and probably has a form, hopefully on-line, for reporting details. I don't know when they did that, but my experience is that you refine the instructions. (Did the bullets injure anyone? Was it Iraqi or U.S? Did they cause any property damage?) As time passes, there's also less downside to reporting attacks. Everything gets easier with experience, including reporting. There could also be some perverse incentives at work--don't report attacks, it looks as if the unit is in a quiet area so they may get moved to a hotter area. Report attacks and you may attract more resources. (There was a review of a book over the weekend, possibly in NYTimes--some ex-soldier reporting that his captain ordered an artillery barrage because the captain of a nearby unit had ordered one. ) That's the sort of thinking that leads to inflation of medals.

All this is, I suspect, documented in assessments of our experience in Vietnam. Iraq may not be Vietnam, but some of the dynamics of a (military) bureaucracy never change.