A Relatively Scientific Experiment

I wrote here about the New York Times’ interminable–the first in a series, too!–but entirely bogus effort to portray veterans returning from Iraq and Afghanistan as traumatized killers. A team of Times reporters did extensive “research,” the methodology of which was never described, and found 121 cases in which a veteran of one of those theaters of war was accused of some form of homicide (including vehicular). The paper made no effort at any kind of statistical analysis, but merely reported that number as if it were somehow significant, and followed up with a few anecdotes and generalizations about how war is brutalizing–going all the way back, in fact, to Odysseus. How’s that for evidence?
As we and others easily showed, the Times’ total of 121 alleged homicides was a very low number, culled over a six-year period. If taken seriously as competent research, it supports a conclusion that veterans who have served in Afghanistan and Iraq are much less likely than other young men and women to commit homicide.
In Fayetteville, North Carolina, the local paper was intrigued by the Times’ claims and decided to check its own archives for evidence. Fayetteville, located near Fort Bragg, home to the 82nd Airborne and special operations units, is an excellent place to conduct the experiment; few localities, if any, have been home to as many soldiers returning from Afghanistan and Iraq.
The Fayetteville Observer checked its own archives, with predictable results:

Twelve Fort Bragg soldiers have been accused of killing 13 people in the six-plus years since Sept. 11, 2001, according to Observer records. In the six years before the terrorist attacks, 16 Fort Bragg soldiers were accused of killing 18 people.

There you have it: wartime and peacetime yield the same low homicide rates for soldiers. In reporting these findings, the Observer referred to the claim made by the New York Times that reported instances of alleged homicide involving a veteran increased by 89% in the period 2001-2007 compared with the six-year period preceding the war in Afghanistan. That claim, insofar as it related to Fort Bragg personnel, was refuted by the Observer’s research.
In my post, I didn’t even mention the 89% figure because I thought it was so silly. Obviously, the attention now being paid to soldiers and to combat veterans is far greater than it was before the Afghanistan and, especially, Iraq wars. Newspapers now are correspondingly more likely to mention the fact that a person accused of homicide is a combat veteran. The Times article itself illustrates the point perfectly: the paper’s reporters and editors are eager to sell the idea that veterans of Iraq or Afghanistan are likely to be combat-stress-crazed, and prone to violence, in a way that would not have occurred to them or any other newspaper when writing about the peacetime armed forces prior to 2001.
This factor could easily account for the 89% figure, but I’m pretty sure there is more going on. The Times was completely silent as to its methodology, which in itself refutes any pretension to scientific method. But does anyone think that the Times dispatched reporters to small towns all across America, hunting through old microfiche records for references to veterans? I don’t. I’d bet that all or substantially all of the Times’ research was done on the internet, by googling combinations of words like “veteran,” “soldier,” “Iraq,” “murder,” and so on. A reader with time on his hands could probably duplicate the Times’ research and come up with their 121 instances in a couple of hours.
There are obviously two biases at work here. The first is the general rule that it is easier to get information on more recent events than less recent events. If you set out to learn about one set of events that happened from 2001 to 2007, and a second set of events that happened from 1995 to 2001, other things being equal, it will be easier to find information on the later ones. This bias is greatly accentuated on the web. The farther back you go in time–back into the dim mists of the 1990s–the less information you will find available in newspaper archives. Some go way back, others don’t.
Taking all these factors into account, it is remarkable that the Times found only an 89% increase in stories mentioning that a person accused of homicide was a veteran. The Fayetteville Observer’s simple comparison illustrates how far afield the Times was led by its bias against the armed forces and against the Afghanistan and Iraq wars.


Books to read from Power Line