Saturday, September 28, 2013

Suicides Four Times Higher in High Gun States

The Baltimore Sun

Several studies have linked gun ownership to the risk of suicide by firearm, according to Dr. Matthew Miller, the new study's lead author from the Harvard School of Public Health in Boston.

But some critics question whether people living in states where lots of residents own guns are inherently more suicidal than those who live in places where ownership is less common.

To address that critique, Miller and his colleagues gathered state-by-state data on gun ownership, suicide attempts and suicide deaths from a variety of sources, including the U.S. Centers for Disease Control and Prevention (CDC) and health surveys.

They compared the 16 states with the highest gun ownership rates to the six states with the lowest rates. Both of those groups included about 62 million people.

In the high gun ownership group, 51 percent of adults lived in a household with firearms, versus 15 percent of adults in the low gun ownership group.

There were about 7,300 firearm suicides in the states with the most guns - including Alabama, Montana and West Virginia - in 2008 to 2009, according to findings published in the American Journal of Epidemiology.

That compared to 1,700 suicides by gun in the low ownership states, such as Hawaii, Massachusetts and New York.

The number of non-gun suicides in the two sets of states was similar, at about 4,200 and 4,300, respectively. What's more, state-wide rates of suicide attempts did not differ based on levels of household gun ownership.

This reminds me of one of our favorite posts about guns and women. The universal truth prevails: where there are more guns there are more problems. In other words, guns do more harm than good.


  1. Guns do more harm than good? Not according to the National Academies of Science. And what this study fails to explain is why Canada and Ireland have suicide rates that are similar to ours.

  2. The toothless grin hillbilly says it again; any poll, or study you use is a lie. Only his studies and polls tell the truth. That's hillbilly sense YA HOO!

  3. Suicides BY FIREARM are four times higher.

    One of these days I'll get you to stop conflating "gun death" stats with overall rates. I think we're close.

    1. You didn't read it through. I know you look for those magic words and then stop. But, a bit further on it said the non-firearm suicides were about the same. That means the total suicides are 4 times higher in those high-gun states, as I said.

    2. I did read it through. Firearm suicides are 4x. Other suicides are equal. That makes total suicides 2x. To be correct you would have to either add "firearm" in front of suicide, or replace the "four" with "double".

    3. Gun suicides: 7300 / 1700 = 4.3
      All suicides: (7300 + 4200) / (1700 + 4300) = 1.9

    4. What kind of tricky shit is that supposed to be? Just add up the gun suicides and the non-gun suicides in both high-gun and low-gun states. It couldn't be simpler, but leave it to you to complicate it right into incomprehensibility.

    5. What "tricky shit"? What you asked is what I just did in the last line, and it comes out to 1.9 times higher- not four. 11,500 suicides in high gun states, vs. 6,000 in low gun states.

      Tricky shit? Don't challenge me on math. I've been doing basic arithmetic since I was 4 years old.

    6. So, now your calling my 4X remark wrong and correcting it with 2X, is that it?

      That means that high gun states have twice the suicides, not twice the gun suicides. Correct?

      I admit my mistake with the 4X. You were right. But, do I understand you correctly that you're admitting that high gun states have twice the suicides as low gun states?

    7. "Yes" to all three of your questions.

      This shouldn't come as a big surprise to you, though. I have said many times how there is a correlation of guns and suicide. In my various posts about correlation calculations I have showed that the Pearson coefficient is around 0.7, which is a pretty solid correlation. I'm not the type to hide from or manipulate numbers. I will say the methods these guys used are pretty poor. They only looked at less than half the states, while my computation looks at all of them. They did that because the common reader can related to sum totals better than a single coefficient number, but their method is far more susceptible to manipulation or being misleading. In this case, they looked at 16 states on one side, and 6 states on the other- both sets with roughly 62 million people. That 62 million line can be manipulated. They can crunch the numbers for different set sizes and pick the one that best matches the agenda to publish. I'm not saying with certainty that they did that, but I am trying to show you how easy it is to manipulate results when you only look at a subset of data instead of the entire body. In the case of suicides, most subsets will show what they want to see (because there is a correlation) but they can still pick the one with the biggest shock value: "it's double!"