After the recent tragic school shooting in Connecticut, USA, I saw a lot of traffic on my facebook site focused either on the need for stronger gun control, or the ineffectiveness of stronger gun control when it comes to reducing the likelihood of these kinds of tragedies. That last bit of the sentence is critical. Nobody argues that gun control can completely stop all gun violence, but there is a lot of mis-information out there on the link between gun violence and gun ownership, or gun violence and gun laws. At first I found myself responding to posts by my friends on facebook, making probabilistic arguments based on my own memory of the link between gun ownership and gun violence. But these arguments get tricky, and so I decided to take a quick look at the data myself. Country by country data is simplest to work with, so I have stuck to those data.
Gun Violence as a Function of Gun Ownership rates for all countries
The first graph is the relationship between gun ownership (guns per 100 residents) and gun violence (gun related deaths per 100,000 people), across all 75 countries where that data is easily obtainable (see end of post for data sources). If I were an advocate for gun rights, I might feel pretty good – there appears to be no correlation at all between gun ownership and gun violence among countries. Here in the United States, we look somewhat remarkable in our capacity to keep our guns in check. We have loads of guns, and a lot less gun related violence than many countries. Countries like… El Salvidor, Jamaica, Honduras, Gautamala, Swaziland, Columbia… Hmmm. These countries are not exactly comparable to the US. In a perfect world, we’d like to have data on a series of countries that are mirror images to the U.S. in all ways (size, GDP, urbanization, age of country …) except access to guns. We don’t have that of course, but we get a lot closer by restricting the data to “similar” countries. There are a number of metrics out there we could use to do this. Here I use the UNDP Human Development Index (HDI) as a way of creating a similar set of countries. The HDI combines data on life expectancy, education, and income indices to rank countries into four tiers of human development – very high, high, medium, and low.
In the second graph, I restricted the data to all countries in the top HDI category (Very High HDI [HDI > 0.79]). The picture changes dramatically. Now we see a clear positive relationship between gun access and gun violence. The US clearly has the most gun violence and the most guns. I have fit a linear regression through this data (Adj. R2 = 0.47), and if we think the linear relationship is reasonable to begin with, we could begin to think about why some countries are above the line and others are below the line (for example, the US is considerably above the line, and thus it appear that US citizens not only have more guns per capita, they also have a harder time keeping their guns pointed away from other people compared to many other countries). I would not walk too far down that path without adding a lot more nuance to the analysis. There are all sorts of things we might want to
Gun Violence as a Function of Gun Ownership rates for countries with a Very High Human Development Index Score