Stefan Sharkansky points out that, according to some statistics, a black man is 36% more likely to die under normal circumstances in Washington DC, than during combat operations in Iraq. (Link via Instapundit.)
Now, of course, you could achieve the same drop in mortality by shipping those black Washington DC inhabitants to somewhere else in the US. It seems slightly specious to single out the black 25% of the military, then narrow it down further to those who live in Washington DC (or, I imagine, similarly violent and/or dangerous areas of the US), and say that for them, well, they're better off in Iraq. What about the rest of the troops?
But even if this drop in mortality applied to all of the troops stationed in Iraq, it would still be disgracefully small. Consider the resources available to the US Army - prior investment in R&D and procurement, staffing levels, logistical support - compared to the similar resources applied to reducing violent crime in the streets of America. Consider the sheer imbalance in strength and equipment between the US soldiers and the insurgent forces in Iraq.
If, despite all that, the death toll in Iraq, after "major" combat operations have ended, is still only 36% lower than one of the deadliest places in peacetime USA, then we have to conclude that something has gone dreadfully wrong, either politically or militarily, in the way the occupation and rebuilding of Iraq is being carried out.