>If the author’s thesis is correct, then there should be a noticeable difference once these groups are exposed to easily accessible guns. That doesn't seem to

>If the author’s thesis is correct, then there should be a noticeable difference once these groups are exposed to easily accessible guns. That doesn't seem to track logically, to me. You're assuming that the assertion is only useful if those are the only things that cause gun violence, rather than the primary and addressable things. Of course it's a myriad of things of varying impact and varying interaction with each other. This test can't realistically control for those things, nor can it even control for the basic effects of transitioning between two countries/cultures that would create noise in the data. This test could certainly produce useful information, but it's too naive to prove or disprove the notion that availability and laws contribute strongly to US gun homicide rates.