Statistics

Ralph Peters gets his stats right: the New York Times purposely misleads

I’m a veteran and haven’t killed anybody in years. But if you read the New York Times you’d be right to worry that I might.

The Sunday, 13 January 2008, edition of the Times spent four pages! detailing that, in the four and three-quarter years since the Iraq war began, returning soldiers, sailors, and airmen came home horribly scared—mentally, of course—and committed 121 murders. Which is a big number, no question; and probably some, or even most, of the people killed didn’t even have it coming to them.

Military writer Ralph Peters, in today’s column for the New York Post, shows that about 350,000 soldiers have come back from both the Iraqi and Afghanistani wars. That makes the per-year murder rate equal to about 7.3 per 100,000.

Time to seriously fret about the mental health of soldiers? Perhaps we should lock them down for a cooling off period until they loose their aggressiveness.

It was at this point that Peters did what any good statistician would have done: he refused to look at the statistic in isolation. He asked: is 7.3 a lot, or is it a little? How can you find out? It’s easy: by going to the Bureau of Justice web site and looking at the murder rates per 100,000 in a demographic most similar to that of GIs, which are 18-24 year-olds:

The civilian murder rate is 26.5 per 100,000

which is more than 3.5 times higher than for GIs! Incidentally, the murder rate for 14-17 year-olds is 9.3; and for those 25-34 it is 13.5, both higher rates than for GIs. It isn’t until you reach the the 35-49 year-olds do you find a lower rate at 5.1 per 100,000. As Peters says the Times

unwittingly makes the case that military service reduces the likelihood of a young man or woman committing a murder.

But his best work comes when he notes

In 2005 alone, 8,718 young Americans from the same age group [as GIs] were murdered in this country. That’s well over twice as many as the number of troops killed in all our foreign missions since 2001. Maybe military service not only prevents you from committing crimes, but also keeps you alive?

Peters has called on the Time’s “public editor” Clark Hoyt (who is in charge of correcting errors) to acknowledge the paper’s purposeful character assassination of our veterans. Add your voice to Peters’s: Hoyt’s email is public@nytimes.com.

Update: 16 January 2008.? Good thing I bought a bigger hat. NYPost.com

Categories: Statistics

7 replies »

  1. NY SLIMES at it again. Don’t hold your breath for a correction or acknowledgement of their obvious bias.

  2. “The civilian murder rate is 26.5 per 100,000”

    Whoa, a 200% increase over the veteran murder rate? Just incredible.

  3. “The civilian murder rate is 26.5 per 100,000”

    i’m sure the “paper of record” would then find a way to blame this figure on the Bush Administration or the second amendment…

  4. I’m new to your site. I love it. Keep up the good work. I quickly scanned your in-progress book and it looks like you are on a good track for making statistics interesting.

    As a Nam vet, I always thought we returnees got a bad rap in the statistics department, especially the way we were painted as being nothiing but a bunch of baby killers, murderers and rapists. It really torqued my jaw. I am so pleased to see folks like you fighting against this sort of tyranny.

    Regarding this piece of garbage by the Times: other web sites have done their homework, reviewed the cases and further reduced the incidence of violent behavior in Iraq/Paki vets. The Times’ way of counting violent behavior (another misleading statistical trick) is seriously awful journalism. For you and those battling against these lies and distortions, I thank you. Our vets deserve a lot better.

  5. 1. Read Burket’s book. Stolen Valor.

    2. Just to be fair, you ought to make the obvious gig on the second point of dividing deaths overseas in combat zone/number served. And then compare versus pop here versus deaths here. 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *