Not all deaths in the military come from combat, or during wars:
Interestingly, in the raw numbers, the wars in Iraq and Afghanistan had death counts that were actually matched or surpassed by those of the military in the early 1980s, when the US was mostly at peace. When the graph is by deaths as a percentage of the active duty military, that changes:
In this case, the percentage of deaths rises higher during the 2000s than it was during the 1980s.
Some of the differences come simply from size. The military of the 2000s was about two-thirds the size of the military of 1980s and thus, by demographics alone, that larger population would see more deaths. But what is most notable is the way that both accidents and homicides have gone down.
Accidents from 2005-2010 averaged approximately one-third the 1980-85 number, and murders from 2005-2010 averaged approximately two-fifths of the 1980-85 figures. Some of that reduction is from the smaller force, but the reduction is much more than the reduction in force size.
Given its population, it is interesting to think of the military as a quite large American city. Using 2010 figures, in 1980, the US military was Houston. In 2010, it was San Antonio. Just like those two cities, the military comes with all the ordinary everyday occurrences of life that have nothing to do with war or combat and some of those – accidents and homicides – lead to the death of its residents.
P.S. I didn’t talk about suicide in the military because that’s worthy of an entirely independent post, which I will get to if and when I can.