A couple of years ago, the city of Chicago started a summer jobs program for teenagers attending high schools in some of the city's high-crime, low-income neighborhoods. The program was meant, of course, to connect students to work. But officials also hoped that it might curb the kinds of problems — like higher crime — that arise when there's no work to be found.
Research on the program conducted by the University of Chicago Crime Lab and just published in the journal Science suggests that these summer jobs have actually had such an effect: Students who were randomly assigned to participate in the program had 43 percent fewer violent-crime arrests over 16 months, compared to students in a control group.
That number is striking for a couple of reasons: It implies that a relatively short (and inexpensive) intervention like an eight-week summer jobs program can have a lasting effect on teenage behavior. And it lends empirical support to a popular refrain by advocates: "Nothing stops a bullet like a job."
Researcher Sara Heller conducted a randomized control trial with the program, in partnership with the city. The study included 1,634 teens at 13 high schools. They were, on average, C students, almost all of them eligible for free or reduced-price lunch. Twenty percent of the group had already been arrested, and 20 percent had already been victims of crime.
Some of the students were given part-time jobs through the program, working 25 hours a week at minimum wage ($8.25 in Illinois) with government or non-profit employers. They worked as camp counselors, office assistants, or in community gardens, among other places. Other students in the treatment group worked 15 hours a week at similar jobs, but also received 10 hours a week of "social-emotional learning" time, where they learned skills to manage their emotions or behavior that might get in the way of employment. All of the students in the program received mentors as well. The teenagers in the control group participated in neither part of the program.
Heller used Chicago Police Department data to follow what happened to all of the students in the 16 months after the program began. In the crime data, there was no difference between the students who got the counseling and those who did not, suggesting that the group working 25 hours a week may have acquired some of the same social-emotional skills on the job. There was a big difference, though, in the violent crime arrest data between the teenagers who got jobs and those who did not:
A lot of things could be going on here. Teenagers who might have committed crime to get money would no longer need to when they have a job. If their added income allowed parents to work less, they may also have gotten more adult supervision. It's also possible that students who were busy working simply didn't have idle time over the summer to commit crime — but that theory doesn't explain the long-term declines in violent arrests that occurred well after the summer program was over.
Heller, in fact, found that most of the decline came a few months later:
That long-term benefit suggests that students who had access to jobs may have then found crime a less attractive alternative to work. Or perhaps their time on the job taught them how the labor market values education. Or maybe the work experience may have given them skills that enabled them to be more successful — and less prone to getting in trouble — back in school.
This one study can't identify exactly why a summer jobs program might change the trajectory of teens at risk of becoming violent. It also raises the possibility that teenagers with summer jobs might have more money to spend on drugs (drug arrests for the treatment group were slightly higher than for the control). These results do suggest that cities could get a lot of payoff for the minimal cost of a summer-jobs program — particularly if it targets teens before they drop out of school. As Heller writes:
The results echo a common conclusion in education and health research: that public programs might do more with less by shifting from remediation to prevention. The findings make clear that such programs need not be hugely costly to improve outcomes for disadvantaged youth; well-targeted, low-cost employment policies can make a substantial difference, even for a problem as destructive and complex as youth violence.