What is the best way to figure out death rates in major cities and in the armed forces?
January 11, 2007 9:22 PM   Subscribe

What is the best way to figure out death rates in major cities and in the armed forces?

A friend of mine recently told me that the death rate of American soldiers was 60 per 100,000 soldiers while the D.C. death rate was 80.6 per 100,000 persons for the same period, meaning you are about 25% more likely to be shot and killed in the U.S. Capital than you are in Iraq.

I just couldn't believe it. I was in such shock I had to see if she was right.

I went to the Census Beurau's website and got the populations of Washington D.C. and followed a link from them to get Iraq's population counts. I then went to the FBI's website where I got the murders and violent crimes counts for D.C. After that I went to the Department of Defense website to get the number of soldiers platooned in Iraq as well as their mortality rate.

Washington D.C.'s murder rate: 35.4 in 100,000 (198 out of 550,521)
Baghdad's murder rate: 95 in 100,000 (this I got from the Brookings Institute)
American soldiers in Iraq's murder rate: 568 in 100,000 (824 out of 145,000)

Now, I want to engage my friend in a thoughtful and well-researched debate but as I went to a school that had no math requirement, I just wondered if anyone out there knew of a better way to calculate the murder rate. I was also curious if I went about gathering my info in the best possible manner (especially the Baghdad murder rate).
posted by jaybeans to Law & Government (7 answers total)
Best answer: Your numbers are probably good (they pass the eyeball test, anyway,) but here are things to confirm if you haven't already done so:

1) Are all these rates calculated over comparable periods of time (e.g., 12 months?)

2) Does the DOD death count include homicides only or does it also include victims of accidents, disease, etc (my guess is no.) If so, you should be comparing to overall death rate in DC, not murder rate.

3) Does the DOD death count include mortalities attributable to injuries suffered in combat in Iraq but that occurred after the victims were relocated to medical facilities out of theater (my guess is no.) E.g., if a soldier dies of complications six months after getting hit by an IED but after he's been flown back to a hospital in the States, is that included in the DOD numbers? That could seriously impact the DOD rate but I have no idea how you'd get those numbers.

4) You should say which years each of these rates apply to. Hopefully they're the same 12-month period, or reasonably close (within, say a year or so of each other.)

I'm sure a stats maven could come up with more but that's the few that come to mind here.
posted by Opposite George at 9:36 PM on January 11, 2007

Math isn't the issue. Comparing apples to apples is.

100% of people die (and don't confuse "death rate" with "murder rate"). However people in the military are generally young and healthy and much less likely to die (yet) of something other than being shot.
posted by winston at 9:39 PM on January 11, 2007

this episode of this american life details the science, methodology and tragedy of a death count study in iraq.

TAL episode 320
posted by tigamonki at 9:50 PM on January 11, 2007

However people in the military are generally young and healthy and much less likely to die (yet) of something other than being shot.

This is true of homicide victims in DC, too.
posted by toxic at 10:09 PM on January 11, 2007

Or, what winston said.

There are a number of reasons why looking at death rates by age cohort in DC and Iraq (as opposed to murder rates) seems like the way to go - after all, to the soldier it really doesn't matter if it's a bullet that does the job or a jeep accident; the relevant question is "will service in Iraq or walking the streets of DC be more likely to result in my death?"

And keep in mind, as he said, that unhealthy people aren't admitted to the armed forces, so you'd expect fewer soldiers dying of, say, causes attributable to congenital problems than the folks back home. So it'd be best to get the death breakouts by general category (disease, accident, homicide, etc.) and compare the breakout and totals.

And just so you know, I've done this exercise a couple of times before -- though I didn't save my work so the data sources are lost -- and it always turned out that young Americans serving in Iraq were much, much more likely to die in a twelve-month period than their buddies back home, where much, much means by a factor of three or four or so. I am not a statistician so maybe there was some major flaw in my methodology but it confirmed to my satisfaction that the "more dangerous in DC" talking point is basically full of crap. Good luck with this.
posted by Opposite George at 10:14 PM on January 11, 2007

Kieran Healey at Crooked Timber covered this thoroughly a few months ago. There's a spirited debate in the comments, too.
posted by maudlin at 10:17 PM on January 11, 2007 [2 favorites]

I worked up the Wyoming Index back in 2003, which made a similar analysis and found Wyoming was more dangerous than Iraq as well. The thing to remember is this is statistics. Now that full 2002 stats are available I see 4,153 people died in Wyoming while the population was 498,703, for an overall rate of 0.83%, much worse than the military death rate of 0.27%. But let's just grab the right age bracket, 15-44: 336 deaths, so the rate is 0.07%, much BETTER than the military rate. For comparison, the death rate of those 75 years and older in Wyoming was 0.46%.

It's all in the way you split the data. My intention with my article was to demonstrate the media's image wasn't spot on so I didn't delve to deep or bother to update it. I suspect you all will find the same thing, but stats are so dicey there isn't much of a "truth" to be found that won't be challenged in some way, for instance, do natural deaths count for the armed forces figure, how about suicide or illness? Maybe the illness wouldn't have hit them had they been at home. Etc. This page [PDF] splits it out a bit for the armed forces but isn't concentrated on the Iraq personnel.
posted by jwells at 5:21 AM on January 12, 2007

« Older Driving with parking brake?   |   Wiping out the debt Newer »
This thread is closed to new comments.