The Atlantic has a fun guess-the-homicide-rate-over-timegame!
Turns out I’m really good at this game.
But I shouldn’t boast; I have no excuse not to do well. I show this chart literally half a dozen times in each and every class I teach.
What I don’t like is how dismissive they are of the current increase in violence, the largest percentage increase in homicide in decades. They quote the Brennan Center, which has been bending over backwards to downplay the recent increase in killing. (Lest there be any evidence of an effect whose name shall not be spoke…. You know, the one that starts with F, son.) The Center wants us to see those dead bodies not as real lives who mattered, but statistical flukes.
I'm still not convinced on the cause(s) of increased homicides in 2015 (and 2016?). But, you are completely right about the Brennan Center downplaying the significance of the recent increase. It's literally one of the largest (the largest?) increases in homicide rate ever, yet they imply it's just the normal ups and downs of the rate.
Prof, slightly off-topic: Have you been following Heather MacDonald's series of guest posts at the Volokh Conspiracy?
JSM
I'm not sure the homicide rate tracked over time is a good measure of crime rate, because medical survival rates have constantly improved. A better measure, reflecting "citizen perceptions", would probably be a composite of violent personal crimes like aggravated assault, ADW, and rape, and property crimes like burglary and larceny.
Shootings would be better. But that data is harder to get. Homicides are fine, even with better medical care. It's a small and consistent bias over time. Just slant the whole line up a bit if you're looking at over a decade.