A caveat

I realized that a death rate can be defined in two ways: it can be defined in an absolute sense as I did below, or it can be defined in a relative sense whereby it takes into account the number of troops on the ground. Interestingly, depending on what you do you will get different answers. USA today notes that the death rate for National Guards is going down [1], while the Spokesman-Review concurs with my findings [2, or google cache 3]. There are arguments for use of both of these deaths rates. I think the main question to ask is this: does more troops on the ground mean more military engagments? If the answer is yes then the relative death rate is probably more correct. If the answer is no then the absolute works fine. Regardless, the absolute numbers still tell an important story: more troops are being killed today than at the beginning of the occupation.

A last note, Iraq Body Count just published a set of statistics of civilian deaths in Iraq [4]. IBC makes all of its data available to the public, though these data require some cleaning. I would encourage people to do their own analysis with these data. Sometimes statistics new analyses reveal new information or better ways to present the old information.