June 2007: 3.6
May 2007: 4.26
Last year, July 2006: 1.48.
Overall daily average to date is 2.48. Total US dead as of today: 3659.
Iraqi monthly military and police fatalities: 232.
Previous military/police fatality rates
June 2007: 196
May 2007: 198
Last year, July 2006: 132.
Total Iraqi military dead: 7333.
Note that I've seen media reports suggesting the Iraqi military casualty figures are significant undercounts.
Iraqi monthly civilian fatalities: 1441
June 2007: 1146
May 2007: 1782
Last year, July 2006: 1063.
To-date civilian partial total (stats begin only in March 2005): 35,597.
Note that the civilian numbers are far less accurate than others (most likely to be greatly underestimated, or even ridiculously underestimated), but could still be useful in determining trends.
Comments: Now eleven months in a row with American casualties above average, no prior bad stretch lasted longer than three months. The overall average for American/non-Iraqi foreign fatality rate continues to move up, from a low of 2.29 deaths daily.
As before, civilian casualties remain terrible. The rate seems to hover around a level that is nearly twice as bad as a year ago, and three times worse than in 2005. Neither we nor the Iraqis realized how good we had it back in 2005.
Six months have passed since the troop escalation began, with no indication in these statistics that it has accomplished anything, except possibly as a contributor to higher US military casualties, with no trend yet for Iraqi military and police.
Three of the last four months have had especially high British casualty rates, so we'll have to see if that continues.
The NY Times screw up is here:
BAGHDAD, July 31 — The death of a marine in western Iraq brought the American military death toll to 74 so far in July, on course to be the lowest monthly figure this year....
It's a terrible reporting job. First, the death toll is significantly worse than the four-year average, so playing it as good news is ridiculous. Second, there's no point in trying to draw a trend based on a single month. I've seen right-wing bloggers do it countless times - take one month of slight improvement from an overall worsening trend, claim victory, and ignore the stats for months more until the process can repeat. The Times isn't keeping good company.