Wednesday, December 19, 2012

The Geography of Gun Violence

The Newtown shooting has occasioned endless features and articles on gun violence and gun control.  While I'm not sure that anything can be done to stop a mass killer intent on carrying our his mission, our murder rate is off the charts -- as Jim has pointed out.  Using 2009 data, we have a gun homicide rate that is is some 15 times greater than the combined homicide rates of 22 other high income countries.  That's just crazy and, if a mass killing generates a discussion about that, I'm not not complaining.


The frenzy is such that even old articles are being dredged up.  I saw an article originally published in January 2011, after the Tucson, Arizona shootings:  The Geography of Gun Deaths.  It compared the firearm deaths -- accidental, homicide and suicide -- for the 50 states and the District of Columbia.   Here's a map showing the comparison:  




Then, came the interesting part.  As the author, Richard Florida states:


With these data in hand, I decided to look at the factors associated with gun deaths at the state level. With the help of my colleague Charlotta Mellander, we charted the statistical correlations between firearm deaths and a variety of psychological, economic, social, and political characteristics of states. As usual, I point out that correlation does not imply causation, but simply points to associations between variables.

First of all, let's look at the factors that do not correlate with gun death rates:  rates of unemployment, illegal drug use, stress,  neuroticism, mental illness and population.  

Now, what are the factors which can be assocaited with high gun death rates?  The factor with the highest correlation to gun deaths was which candidate carried your state in 2008.  If your state went for Obama, breathe a sigh of relief; there are not many gun deaths in your state.  Be concerned, however, if it went for McCain.   Also, if your looking for safety, pick a state with a high immigrant population.  A high immigrant population had a high correlation to low gun death rates.   Also, look for a state with a lot of college graduates.   Anyway, here's the complete graph.  You can see for yourself.  




 

4 comments:

James R said...

Fascinating! Like he says, "correlation does not imply causation", but it's great dinner conversation, nonetheless.

Big Myk said...

Whenever one of my blog posts generates a great dinner conversation, I consider that to be a good day.

James R said...

That is the ultimate goal. I'll try to use it in Christmas dinner conversation, which gives you a Nobel Prize of sorts.

Big Myk said...

Yes, I agree. A great dinner conversation is the ultimate goal. Either that, or a devestating power spike in volleyball.