Austin,Tex. -- Which American communities pay the highest price for the war in Iraq? A look at the demographics of soldiers killed reveals that Iraq is not the war of any one race or region. Rather, it is rural America's war.

Altogether, a nearly equal percentage of Americans aged 18 to 54 live in counties with a million or more inhabitants as live in counties of 100,000 or fewer. And yet, of the soldiers who have died in Iraq, 342 came from densely populated counties while 536 came from smaller ones. Derived from Pentagon and census data, this chart shows the Iraqi war death rates for every 100,000 people ages 18 to 54 by the size of their county's population.

The difference is visible not just in the size of a soldier's county of origin, but also in its location. Counties disconnected from urban areas tend to have higher death rates, regardless of population size. Small rural counties have a death rate nearly twice that of counties that have the same population but happen to be part of metropolitan areas.

Why should this be? It's not that Iraqi insurgents are singling out rural soldiers, or that commanders are putting them at particular risk. Rather, the armed forces themselves must be disproportionately drawn from rural communities - a fact not immediately discernible from recruitment data, which report the race, age and education of recruits, but not their home counties.

This is above all an economics story. Military studies consistently find that a poor economy is a boon to recruiting. The higher rate of deaths from rural counties likely reflects sparse opportunities for young people in those places.

When the Iraq war memorials go up in years to come, these monuments to heroism and sacrifice will be found less often in thriving urban centers than in lagging rural communities.

First published in The New York Times, July 19, 2005