Gun Ownership Neither Increases Nor Decreases the Crime Rate

[Recursion]

Last edited Tue Apr 23, 2013, 08:41 PM - Edit history (2)

http://www.psmag.com/culture/gun-ownership-neither-increases-nor-decreases-crime-rate-55473/

Recently, Alex Berezow analyzed the correlation between the number of privately-owned guns per capita in a country and the rate of homicide by firearms. This is a sensible first step to answer the question of whether reducing the number of guns in a society makes it safer. However, the greater purpose of gun control is not to reduce firearm homicide specifically, but to reduce the overall rates of murder and violent crime in a country.

...

A simple correlation analysis cannot tell us which of these ideas, if any, are correct, but it can shed light, to some degree, on whether gun control is associated with less violent crime. Let’s look first to see if there is a correlation between the number of guns per capita and the number of intentional homicides, of all types, per capita in 172 countries:



The answer, clearly, is no. In fact, the statistically insignificant trend is toward slightly (as inferred from the negative slope of the line) fewer homicides as gun ownership increases.

Further, we will follow Alex’s lead investigating this correlation by removing some countries. Let’s remove all countries with very poor development (e.g., some African and South and Central American states), states with extremely high murder rates (nearly all of which have very low gun ownership) and states currently involved in civil wars or major domestic unrest. This leaves 72 more developed countries to analyze. Are homicide rates higher in these countries with more guns?



The answer is still no. There is no reasonable way to cherry-pick any sample of countries to arrive at a significant correlation, or even a hint that reduced gun ownership lowers overall homicide rate.



...

First, there is no correlation between the number of guns per capita and the overall homicide rate. So people who believe fewer guns will reduce the homicide rate may be wrong. Second, there is no correlation between the number of guns per capita and the rate of assaults and robberies. So people who believe guns make society safer by reducing overall crime may be wrong, too.

This is presenting familiar data in a new way (it asks about all homicides whereas previously people have been posting gun deaths: you are more likely to be killed by a gun in a country with more guns, but you're not particularly more likely to be killed).

That aside, I'm interested in the US here. Guess which dot we are? Not very difficult, is it?

We're very far to the right, meaning we have more guns than most other countries.

Now check out that line, that's the regression. That's what tells us where our homicide rate "should" be if our homicide rate varied with gun ownership like it does for other countries. We're way above that. (We're also way above it on the version of this chart that gets posted more often on DU, looking at only gun deaths rather than all homicides.) That is, we're more violent than the model says we should be. If we lower our gun ownership rate to the median and maintain that variance we would still be among the most violent countries on earth (and, per this correlation, we would have a higher homicide rate than we do now, though it would be a lower gun death rate).

EDIT:

I did a regression below for the G-20 (which basically matched Hartsfield's results) and the G-8 (which showed an equally weak positive correlation), FWIW. And, while I'm editing, why not include the charts: