The Correlation Between Homicide And Firearms

Last Updated on November 14, 2020 by Oddmund Groette

Mainstream media and politicians always call for the limitation of firearms to avoid murders and violence. Because of this, I wanted to look at some numbers by myself. This article is superficial and does not in any way explain any cause and effect, which I believe are more complex than we imagine.

First up is simply a correlation of all countries: Estimated number of guns per capita by country and their respective intentional death rate (by using the column called “homicide rate”. The source is Wikipedia, but numbers are presumably from official sources. Thus, the source might be disputable, but as far as I can see this is the best I can get with my limited time.

Here is the scatter plot and a regression line for “best fit” among all countries:

As you can see there is a negative correlation: the more firearms among the population, the less murders. The one in the right corner is the USA: lots of guns but far away from being the country with the most homicides.

The above scatter plot includes all countries. What if we look at only OECD countries from the above sample? At least they should be more relevant as a group:

This time the “best fit” regression line turns slightly positive. There are two outliers: the USA in the right corner and Mexico in the upper left. Mexico has little registered firearms, but a very high murder rate (the war on drugs?). If we exclude these two countries we get this scatter plot:

It turns slightly negative.

I grew up in Norway. Norway has lots of firearms, 8th in the world per capita, but at the same time the 3rd lowest murder rate.

  •  
  •  
  •  
  •  
  •  
  •