Guns, Homicides and Data
I came across a fascinating post at Jon Udell’s blog, “Homicide rates in context ,” which starts out with this graph of 2007 data:
Jon’s post says more than I care to on this subject right now, and points out questions worth asking.
As I said in my post on “Thoughts on the Tragedies of December 14th,” “those who say that easy availability of guns drives murder rates must do better than simply cherry picking data.”
I’m not sure I believe that the “more guns, less crime” claim made by A.W.R. Hawkins claim is as causative as it sounds, but the map presents a real challenge to simplistic responses to tragic gun violence.
Hawkins appears to have based his map on this one, but collapsed the gun ownership data categories of 0-5 and 5-10 per 100,000 into a single 0-10 per 100,000 category. You might think that this blurring of the data was done to bolster his point; I couldn’t possibly comment.
It seems that the real takeaway here is that the effect of gun ownership on homicide rates, if there even is one in either direction, is swamped by confounding factors. I would guess that levels of poverty are a much stronger driver of homicide rates. It might be interesting to see what happens when you correct for these confounding factors (although I suspect some of them are very hard to quantify – things like the general level of social cohension).