COUNTERCOLUMN: All Your Bias Are Belong to Us

So, basically, the reporters went trolling on Lexis-Nexis and other databases to find "murder" within the same paragraph as "veteran" or "soldier," and built a front-page story around that research. They compared the pre-war numbers to the post-war numbers and found that, voila!, there's a difference. And then it looks like they cherry-picked the best anecdotes out of that research (including the ones where they could get interviews and photos) to craft a narrative which fit the data.

The article makes no attempt to produce a statistically valid comparison of homicide rates among vets to rates among the general population. Nor does it rely at all on Pentagon data about post-deployment incidents of violence among veterans. It basically just generalizes from this small sample (121 out of 1.7 million Iraq and Afghanistan vets, not including civilians and contractors) to conclude that today's generation of veterans are coming home full of rage and ready to kill.

I've got a one-word verdict on this article and its research: bullshit.

The New York Times found 121 cases in which veterans of Iraq and Afghanistan committed a killing in this country, or were charged with one, after their return from war. In many of those cases, combat trauma and the stress of deployment — along with alcohol abuse, family discord and other attendant problems — appear to have set the stage for a tragedy that was part destruction, part self-destruction.

Three-quarters of these veterans were still in the military at the time of the killing. More than half the killings involved guns, and the rest were stabbings, beatings, strangulations and bathtub drownings. Twenty-five offenders faced murder, manslaughter or homicide charges for fatal car crashes resulting from drunken, reckless or suicidal driving.

To compile and analyze its list, The Times conducted a search of local news reports, examined police, court and military records and interviewed the defendants, their lawyers and families, the victims’ families and military and law enforcement officials.

From the October 1, 2001 start of the Afghanistan war, that's about 26,000 troops/month. To date (Jan 2008) that would give about 1.99 million.

That means that the NY Times 121 murders represent about a 7.08/100,000 rate.

Now the numbers on deployed troops are probably high - fewer troops from 2001 - 2003; I'd love a better number if someone has it.

But for initial purposes, let's call the rate 10/100,000, about 40% higher than the calculated one.

Now, how does that compare with the population as a whole?

Turning to the DoJ statistics, we see that the US offender rate for homicide in the 18 - 24 yo range is 26.5/100,000.For 25 - 34, it's 13.5/100,000.

See the problem?

Damn, is it that hard for reporters and their editors to provide a little bit of context so we can make sense of the anecdotes? It's not in Part 1 of the article. And I'll bet it won't be in the future articles, either.

Because it's not part of the narrative of how our soldiers are either depraved or damaged.

Now put yourself in the place of a newspaper editor. Suppose you are asked to evaluate whether your paper should run a long article on a nationwide epidemic of murders committed by veterans of Iraq and Afghanistan--a crime wave that, your reporter suggests, constitutes a "cross-country trail of death and heartbreak." Suppose that the reporter who proposes to write the article says it will be a searing indictment of the U.S. military's inadequate attention to post-traumatic stress disorder. Suppose further that you are not a complete idiot.

Given that last assumption, I'm pretty sure your first question will be: "How does the murder rate among veterans of Iraq and Afghanistan compare to the murder rate for young American men generally?" Remarkably, this is a question the New York Times did not think to ask. Or, if the Times asked the question and figured out the answer, the paper preferred not to report it.

As of 2005, the homicide rate for Americans aged 18-24, the cohort into which most soldiers fall, was around 27 per 100,000. (The rate for men in that age range would be much higher, of course, since men commit around 88% of homicides. But since most soldiers are also men, I gave civilians the benefit of the doubt and considered gender a wash.)

Next we need to know how many servicemen have returned from Iraq or Afghanistan. A definitive number is no doubt available, but the only hard figure I've seen is that as of last October, moe than 500,000 U.S. Army personnel had served in either Iraq or Afghanistan. Other sources peg the total number of personnel from all branches of the military who have served in the two theaters much higher, e.g. 750,000, 650,000 as of February 2007, or 1,280,000. For the sake of argument, let's say that 700,000 soldiers, Marines, airmen and sailors have returned to the U.S. from service in Iraq or Afghanistan.

Do the math: the 121 alleged instances of homicide identified by the Times, out of a population of 700,000, works out to a rate of 17 per 100,000--quite a bit lower than the overall national rate of around 27.

But wait! The national rate of 27 homicides per 100,000 is an annual rate, whereas the Times' 121 alleged crimes were committed over a period of six years. Which means that, as far as the Times' research shows, the rate of homicides committed by military personnel who have returned from Iraq or Afghanistan is only a fraction of the homicide rate for other Americans aged 18 to 24. Somehow, the Times managed to publish nine pages of anecdotes about the violence wreaked by returning servicemen without ever mentioning this salient fact.

Labels: Crime, New York Times, Psychology, soldiers' issues, The media