Writing, publishing, geekdom, and errata.

Whoops - seems like another gun ownership meme really IS wrong, after all.

1 comment
I saw this a few days ago on imgur, so I wasn't surprised to see it show up on Facebook.

At which point I was surprised to see someone nit-picking (and calling "misleading") the Swiss person's description of their own military service, whether or not someone can have ammo at home, and so on.

Which really ignores the fact that the original meme is misleading as hell. All Mr. Shekler is pointing out (and validly) is that there are a lot of confounding variables, and that maybe... just maybe... gun ownership isn't the defining variable in the crime rate.

But this is where sociology comes in. Because we have a natural experiment that we can use to test this hypothesis:  The United States of America.



We have access to information about gun ownership rates and crime rates. If gun ownership is the defining characteristic, then we should be able to see a pretty straightforward correlation. (Remember, rates are per capita.)

I didn't look too hard - I got violent crime rates from census data and gun ownership rates from this (admittedly quite biased) article, but they'd sourced their data from USACarry which is kind of biased in the other direction... so I went ahead and used it.

For simplicity's sake, I took the states with the ten highest and ten lowest gun ownership rates (in 2007). Then I plugged in their ranking in violent crime, and made a nice little graph.


If there was a really strong correlation between violent crime rates and gun ownership rates, the orange dots should trend right along with the blue dots.  And yet they do not.

I'm sure someone with more time (or perhaps a sociology paper to write) could use this same data and plug it into SPSS or SAS and see if it comes close to hitting an alpha of 0.05 ... but I'm going to go out on a limb and guess that we're not going to see either a strong or significant correlation with this data.

Or in other words:  Yeah, this data suggests that the Swiss guy is right. If you want to talk about why Switzerland has one of the lowest crime rates in the world, you might want to talk about something other than guns.



The actual table of the data is screencapped below if you're interested.


1 comment :

Clark Mcconachy said...

Nice article, I tried to do a similar thing a few months back. Problem is tou can't really get solid data from the US relating to firearms ownership because of their lax registration laws.