The newest Atlantic (not yet online) contains an article about Memphis's experience with shutting down its noxious projects and offering housing vouchers to their low-income inhabitants, who use the vouchers to move to other areas of the city. The concept has been applied across the country. In Memphis, though, it's had the unfortunate effect of spreading all over the metropolitan area what were once isolated concentrations of crime. And overall crime rates in the city are way up.
Motivating housing voucher programs is the idea that if high-poverty, high-minority, high-crime neighborhoods are dispersed--if the residents of those neighborhoods move to more economically and racially integrated settings--than deleterious activity will wane. It's an idea that's been extended to k-12 education, too: If poor or minority students are removed from all-minority, high-poverty neighborhoods (and their schools), they'll do better academically. But it's not that simple. Nor is it true that other forms of shuffling kids from school to school to improve classroom "diversity" does much for the educational prospects of the shuffled. Dangerous neighborhoods are dangerous for a variety of reasons, but at the core it's because they're inhabited by... criminals, who, when transplanted to better neighborhoods, are simply able to steal better merchandise. Bad schools are bad not because of who sits next to whom, but mostly because of??the... bad teachers and bad administrators??who work in them. And good schools are good largely because they're staffed by people who are good at what they do.