Is there a correlation between guns owned and guns used? I was asking myself that question, downloaded some data from Wikipedia and created these scatterplots.
Visualizations are realized in Mike Bostock´s d3.js.
These graphs are, in essence, an extended version of Mark Reid´sDeath vs guns plot.
General deaths by gun versus gun ownership
Plotting gun ownership versus gun deaths results in this graph:
OECD countries
Non-OECD countries
It´s interesting to check out other distributions by only looking at OECD-countries (use the checkboxes to the right of the graph).
Update: Plotting the Guardian data leads to this graph:
OECD countries
Non-OECD countries
Again, filtering between OECD- and Non-OECD-countries shows two very different distributions. It also becomes clear that data is everything if we compare it to the Wikipedia results.
Homicides, suicides, accidental deaths versus gun ownership
On Wikipedia, the gun-related deaths are additionally divided into homicides, suicides and accidental deaths (this data isn´t available for all countries). Distinguishing between these different types of deaths results in these graphs:
Homicides versus guns:
OECD countries
Non-OECD countries
The `homicides´ distribution looks very similar to the general gun-deaths plot, both for OECD and non-OECD countries.
Suicides versus guns:
OECD countries
Non-OECD countries
The `suicides´ is more interesting, as it looks like as if there was a much clearer correlation between people dying from self-inflicted gun wounds and the number of guns in a country. This might, however, also stem from the bias in the data towards OECD countries.
Accidental deaths versus guns:
OECD countries
Non-OECD countries
Finally, this is what accidents versus guns looks like. Again, there seems to be a correlation: more guns means more accidents (and the other way round!).
