[potato3732842]

>This is why it’s misleading to talk about state-level stats without accounting for density: Idaho has a lower crime rate because it is mostly rural and has a single large city, which isn’t that big. Crime is a function of population, not land.

Don't you mean function of density or was that a slight of hand rather than a typo? Like compare Wyoming to 1/16 of NYC or 16x Wyoming and compare it to all of NYC. They're about equal in population but the rates per capita are per capita so they're unchanged whether you multiply one or divide the other.

[acdha]

Yes, density would have been a better choice - what I was trying to get at is that when you have a lot of people in close proximity you have more social interactions which can turn negative. For example, here in DC violent crime is largely limited to a few areas where drunk people get out of bars late at night and various crews are fighting over territory, so the numbers go up but most people in the neighborhood aren't affected. The crime rate always goes up in the summer because people are out on the street where they can get into arguments, and everyone's a bit touchy during a heat wave.

You certainly have things like rural gangs, too, but if things are spread out you just don't have that critical mass to ramp the numbers up. This also plays out in other types of crimes – cars get stolen anywhere there are cars, but thieves are playing the odds and it's easier not to attract in a dense population while they'd stick out if they started going up some stranger's driveway in a place where there's no other traffic. When that Kia lock exploit was in the news, there were bored teenagers basically treating street parking as a shopping mall because the supply was huge and until they actually touched a car there was no crime in walking down a sidewalk.