Some Ugly Numbers

Not so long ago, a comment drew my attention to the heavy losses of officers the US Army and Marines were suffering in Iraq. I've been putting off a full blog post on this for a while, because it involves maths and ghoulishness, but after I saw this at the Washington Monthly:

"The Army is currently having to eat its seed corn to maintain the rotations. The Army Times reported that 44 students were removed from the Command and General Staff Officers Course to bring the 4th ID up to strength in majors before its redeployment to Iraq. Other individual and unit training is being either canceled or shortened to maintain the rotations.

This week's Time reports that there is a major shortfall in the number of sergeants (E-5s or three stripers for you civilians) and so all E-4s will be considered eligible for promotion to sergeant rather than having their individual qualifications examined."

I thought I'd better get on with it. Now, the figures we need are obviously the breakdown of casualties by rank, with if possible some longitudinal series so we can look for a trend. (You can get the stats here).

Now, the official categorisation (as here - note the figures are not entirely comparable because they exclude losses before the 1st of May 2003 and are current at the 26th of February) in the US forces breaks down into three groups: officers, grades E5-E9 and grades E1-E4. Doing the sums, 57.6% of the dead were in the lowest band, 32.5% in the E5-E9 (sergeant) band, and 11.0% were officers. The figures for the wounded (for this we have to use the not fully comparable figures here) are as follows: officers made up 6.39% of the wounded, senior NCOs 31.78% and junior ranks 61.83%.


What does this mean? Was my commenter right when he compared this to the British army at the Third Battle of Ypres in 1917? For this we need to know what percentage of the forces each group make up, as well as some historic data for comparison. One point is already clear; the rank and file are more likely to become casualties and survive. Their leaders are more likely to die. (Note, this is not an stats error - when I did a draft calculation using the less-current figures only, a similar relationship was evident.) This is shown more strongly with regard to officers, but is still present for the E5-E9 band.

The officer corps is roughly 15% of the total force. Totalling up the (slightly late) figures, we get a figure for officers of 7.6% of the total casualties; so, a marked underrepresentation. However, this takes no account of the sizeable number of staff officers in various large headquarters in Baghdad who are not really exposed. No officer above lieutenant colonel has been killed (although six of that rank have been), so the real base of comparison would be the percentage of officers up to that rank. Leaving out 4,539 Army and Marine officers of colonel's rank and above gives us a figure of 14.0%, but this includes all units and organisations anywhere in the world. The second, senior-NCO band, are some 31.7% of total casualties, but make up only some 24% of the force worldwide - a marked overrepresentation. Plugging the figures back in, 38% of the force (assuming the MNF Iraq has a similar demographic to the US Army and Marine Corps worldwide) accounted for 42% of the dead.

For comparison, officers made up 12.1% of the US dead in Vietnam. It has been widely remarked on that Western armies have tended to suffer fewer deaths as a proportion of total casualties over the last century, a consequence of better medical treatement - which would suggest a similar or higher implied danger in Iraq. Another complicating factor is whether or not psychiatric cases are included in the casualty figures.

What's the point of all this? Well, it is not as bad as we may have thought; but it's still not good, and the death rate for the army's leadership core is significantly higher than that for the majority of its members. This is much more marked for the crucial layer of senior NCOs.