A long term (25 years!) longitudinal study has just released data that found that children who were assigned to "child and parent centers" that provided parenting assistance, education and other services were less likely to be arrested 25 years later. Other longitudinal studies have found no linkage with early education and crime but this study seems convincing. Questions about whether small programs can be scaled up remain.
I'm also interested in how we talk about early education - these sorts of studies lend themselves to calculating "return of investment" on education. Is it naive to think education might be valued independently of economic interests? Is investment language the way we must phrase topics to get them approved in tight budgetary times? I'm not sure.
That's why police chiefs are one of the most unlikely backers of proposals for universal pre-kindergarten. Still, most of the studies that show pre-K reduces crime are also studies that pre-K skeptics criticize the most, saying those programs are hard to replicate at a state or national level.