Bernard Harcourt is finding some very interesting and powerful correlations between crime rates and institutionalization rates. And you can’t just look at prisons. I’m not an expert in the field, but I suspect that this may be groundbreaking, with some interesting implications for public policy.