December 14, 2007

Questionable Numbers

A USA Today article earlier this week noted the increasing confirmed or suspected suicides among members of the armed forces, but provided questionable figure for civilian suicides for comparison. The military suicide rate was pushed to USA Today by Senator Patty Murray, (D-WA). Murray was just one of 21 Democrats to vote against the resolution authorizing the invasion of Iraq.

According to the USA Today article:

A record number of soldiers 109 have killed themselves this year, according to Army statistics showing confirmed or suspected suicides.

The deaths occur as soldiers serve longer combat deployments and the Army spends $100 million on support programs.


Those numbers show 77 confirmed suicides Army-wide this year through Nov. 27 and 32 other deaths pending final determination as suicides.

The Army updated those statistics Wednesday, confirming 85 suicides, including 27 in Iraq and four in Afghanistan.

The highest number of Army suicides recorded since 1990 was 102 in 1992 a period when the service was 20% larger than today.

A total of 109 suicides this year would equal a rate of 18.4 per 100,000, the highest since the Army started counting in 1980. The civilian suicide rate was 11 per 100,000 in 2004, according to the Centers for Disease Control and Prevention.

The discrepancy between military and civilian suicide rates—18.4/100,000 for the military, and 11/100,000 for civilians—is certainly shocking.

But it isn't necessarily accurate in an oranges-to-oranges comparison.

For example, an Associated Press account published today states that the civilian suicide rate for one segment of the population, middle-aged Americans 45-54, has risen dramatically, and that it isn't as far from the military rate as the USA Today article states.

The rate rose by about 20 percent between 1999 and 2004 for U.S. residents ages 45 through 54 far outpacing increases among younger adults, the U.S. Centers for Disease Control and Prevention reported.

In 2004, there were 16.6 completed suicides per 100,000 people in that age group. That's the highest it's been since the CDC started tracking such rates, around 1980. The previous high was 16.5, in 1982.

Experts said they don't know why the suicide rates are rising so dramatically in that age group, but believe it is an unrecognized tragedy.

The general public and government prevention programs tend to focus on suicide among teenagers, and many suicide researchers concentrate on the elderly, said Mark Kaplan, a suicide researcher at Portland State University.

"The middle-aged are often overlooked. These statistics should serve as a wake-up call," Kaplan said.

For a like comparison to be made, one can—and perhaps should— try to compare the military suicide rate against the most demographically-comparable civilian group, and not the entire U.S. population.

When this is done, the CDC figures show that the 2004 age-adjusted suicide rate for civilian men—which would most closely correlate to the mostly male military population—is at 15.2 per 100,000, just 1.4/100,000 different than the military figure. This isn't an oranges-to-oranges comparison with military deaths, but at least we're closer to talking citrus in both instances.

The highest overall suicide rate among the groups studied was among males 65 or older, at 28.9 per 100,000.

For men, getting old seems to be a far greater risk factor for suicide than going to war, but then, I'm not a statistician.

Posted by Confederate Yankee at December 14, 2007 10:10 AM

I agree CY - it is an apples and oranges comparison. It would be better to contrast the Army suicide rate to that of other high stress, high danger jobs such as police and firemen.

I was an XO in a company that had a soldier suicide. It's not a pleasant experience and we lost a good troop that day.

RIP Spc. Herrick

NOTE: BTW, the article above mentions Herrick's room "indicators he dabbled in witchcraft or the dark arts". That it fails to mention is a) it was a shared room and b) Korn and Maralyn Manson posters do not constitute dabbling in the dark arts.

Posted by: Dan Irving at December 14, 2007 03:38 PM

And... the vast majority of military suicides did not occur in a combat zone.

Posted by: Chuck Simmins at December 14, 2007 05:12 PM

Re Senator Murray-


Posted by: rosignol at December 14, 2007 09:06 PM

"the deaths occur as soldiers serve longer combat deployments..."

That's journalist-ese for "I have no evidence of a connection so I'll slimily insinuate one."

Let me try: "suicides of Mac users occur as increasing numbers migrate to the latest release of the operating system." Which is just as dishonest.

Posted by: pst314 at December 14, 2007 09:06 PM

Hmmm... perhaps they could try comparing the suicide rates of US service members with members of Al Qaeda and its affiliates.

That's just as valid a comparison... in fact, it might be more valid because they're currently opponents in the War on Terror.

-taking tongue out of cheek-

For a truly valid comparison, perhaps someone could dig up the stats on suicides in the armed forces during WWII, and compare it to now? In other words, pre-Vietnam and post-Vietnam.

I would honestly be interested in seeing that sort of comparison done.

Posted by: C-C-G at December 15, 2007 10:12 PM
Let me try: "suicides of Mac users occur as increasing numbers migrate to the latest release of the operating system." Which is just as dishonest.

The problem with your suggestion is that the suicides have not, as far as I can tell, been broken down by OS. The statistics in the article specifically address military personnel, and it is a fact that deployments are longer than they once were. Correlation is not causation, but the article's stats have more correlation than your scenario.

Much as I hate to admit it, CCG has a point: the pre- versus post-Vietnam statistics would be pertinent data.

Another question: is the 1.4/100,000 adjusted rate even a statistically significant difference?

Posted by: novanom at December 16, 2007 01:07 PM

Yes you are correct in that among the older men ex-vets do have a much higher rate of suicide.

Posted by: John Ryan at December 16, 2007 05:48 PM

Thank you, novanom.

Upon further reflection, I'd actually like to see three data sets compared... WWII, Vietnam, and post-Vietnam (i.e. Gulf Wars I and II, along with the Afghanistan war). Comparing those three would give us the best possible picture.

Not that I expect the media to do it. They're still too invested in failure in Iraq.

Posted by: C-C-G at December 16, 2007 06:34 PM