The evidence in the area of education policy is encouraging - we have identified several ways to prevent crime from happening, rather than engaging in after-the-fact punishment offenders that comes at great cost to us (and them). One problem is that the most effective educational strategies that have been identified to date tend to target young children, so that any crime-prevention benefits arise only with some delay. For that reason there is also great value in considering social policy interventions that might help prevent criminal offending among people who have already reached the highest-risk offending ages of adolescence and young adulthood.
Mentoring programs such as Big Brothers / Big Sisters (BB/BS) have the potential to reduce risky or criminal behavior among adolescents by connecting them with adults who may be useful sources of advice or guidance, or help youth maintain a pro-social orientation. Public / Private Ventures carried out a rigorous randomized experimental study of BB/BS and found that children assigned to mentors reported lower levels of risky behavior than other children. One caveat is that since the study relied entirely on self-reported behavior, we cannot be completely sure whether the program reduced involvement with risky behavior or instead just the willingness of youth to self-report risky behaviors.
The Quantum Opportunities Program (QOP) was funded by the Ford Foundation to provide educational help, work experiences and financial incentives for success to at-risk ninth grade students in five cities. Randomized experimental evaluations of QOP found children assigned to the experimental groups had higher test scores, graduation rates, and post-secondary school enrollment rates than youth who were not assigned into the program, although some of the program results seem to be driven by just a single site (Philadelphia).
Rather than working with individual youth to change their individual risk factors for involvement in crime or violence, an alternative approach is to try to address community-level factors that might contribute to delinquency. A large body of social science theory predicts that growing up or living in a disadvantaged, dangerous urban neighborhood may increase youth involvement with crime and violence by affecting access to quality schools or job prospects or positive role models, the quality of local policing, and the local peer environment. HUD has funded a randomized experiment to examine the effects of moving families out of highly distressed public housing projects, known as the Moving to Opportunity (MTO) demonstration. MTO was launched in 1994 in five cities (Baltimore, Boston, Chicago, LA, and New York), with eligibility limited to families with children living in some of the nation's most disadvantaged public housing projects. Via random lottery some families but not others are given the opportunity to relocate to less distressed and dangerous neighborhoods, which a large body of social science research hypothesizes should reduce youth involvement with delinquency and violence. Evidence from a 5 year follow up found a more complicated pattern of results, with MTO moves causing short-term reductions in violent crime arrests for both boys and girls, but after several years some increase in property-crime arrests for male youth. The 10 year evaluation is currently out in the field, for which Crime Lab member Jens Ludwig is serving as project director. Results should be available sometime in 2010.