When is a high suicide rate not a high suicide rate?

When is a high suicide rate not a high suicide rate?

Right in line with the blog entry just below, here’s another story with statistics that don’t add up. James Taranto at Best of the Web Today looks at a USA Today article on the high suicide rate among police. It says that the suicide rate for police is 18 per 100,000 while among the general population it is 11.1 per 100,000. Why so high? Their experts say it’s because of the easy access to firearms (of course) as well a macho culture afraid to make mistakes or seek help.

Unfortunately, as Taranto points out, the newspaper’s statistics don’t add up. For one thing, they’re comparing apples to oranges. Since 88 percent of police are male, they should have compared the suicide rate to men and since men are far more likely to commit suicide, it’s radically shifted. For all men, the suicide rate in 2002 was 18.4 per 100,000, almost exactly the same as that for police. Take out the teens and for men 25 to 44 it’s 22.2 per 100,000 and for 45 to 64, it’s 23.5 per 100,000.

What we see is that in fact the suicide rate among police is not “alarmingly high,” but in fact may actually be lower than what you’d expect from a population that is 88% young and middle-aged men.

Yet another big expose that isn’t and another case of reporters and editors looking for a big splash instead of news.

Technorati Tags:, , , ,

  • The New England Journal of Medicine did the same thing in 1989, when they tried to say that gun ownership was conducive to suicide; they did this by measuring the number of suicides committed using a firearm as a subset of all suicides, instead of measuring the number of suicides by firearm as a subset of all gun owners.

    It’s known as value bias error, and I learned to recognize it in Sociology 101; to see it done by a peer-reviewed scientific journal was (at the time) shocking.  That’s why I have such a jaded view of all the “peer-reviewed” global warming hype we are bombarded with today.