Jeff Soyer has some good continuing discussion of the Harvard Study. One thing that strikes me is, if they are arguing that ease of obtaining guns because of their prevalence is a factor in homocides, wouldn’t the overall number of guns per capita be a more relevant number? For instance, Pennsylvania’s household owership rates is unremarkable, but we have one of the highest number of guns per capita of any state. Our murder rate, however, is about the same as New Jersey, despite the fact that we have Philadelphia, and New Jersey has no large cities, and very low rates of gun ownership and guns per capita.