A summer jobs program for teenagers appears to cut the rate of violent crime, according to a new study by a University of Pennsylvania researcher.
And not because the youths were too busy working to break the law. Those who were randomly chosen to get the eight-week positions were arrested for violent offenses 43 percent fewer times than their peers, and most of that difference occurred during the 13 months after the jobs were finished.
The findings by Sara B. Heller, an assistant professor of criminology at Penn, are reported in Friday's issue of the journal Science. Teens in the study were generally from lower-income families, and one-fifth of them had previously been arrested.
Heller, who conducted the study in Chicago in cooperation with the city government, said it was not entirely clear why the summer jobs seemed to have a lingering positive impact after they concluded, at the end of summer in 2012.
One factor may have been "soft skills" learned on the job, such as conflict resolution and self-control, said Heller, who came to Penn from the University of Chicago in 2013. Each youth was partnered with a mentor, who may have helped teach those skills, she said, calling the results "surprising and really exciting."
"We don't have a lot of success stories for reducing violence among disadvantaged youths," she said.
Indeed, there is scant evidence that teen jobs programs can have a lasting impact on crime, said Dan Bloom, a policy area director for MDRC, a New York-based nonprofit policy research group. A few programs have been found to lower crime rates but they had a residential component and were costly, he said.
The Chicago program, dubbed One Summer Plus, cost less than $3,000 per youth, said Evelyn Diaz, commissioner of the city's Department of Family and Support Services, which oversaw it. Of that total, about half went toward the teen's wages ($8.25 an hour - the minimum in Illinois - for 25 hours a week) and the rest went toward paying the mentors and administrative costs, she said.
"You just don't expect for a short-term, relatively low-cost program to have such enduring effects," Diaz said.
Bloom, who was not involved with the Chicago research, said he would want to see additional research to make sure that such programs reduce crime, but he said that Heller's study was well-designed and that its findings promising.
"It's always better to replicate a finding like this, but this is already a lot more than we knew before," Bloom said.
He also credited the administration of Chicago Mayor Rahm Emanuel for a willingness to see whether the program was effective. Heller, the study author, said the same. "There are some policymakers who are afraid of finding out if programs they like don't work," she said.
The 730 teens who were offered jobs were picked at random from among 1,634 applicants. Their number of arrests for violent crimes was slightly lower than that of the remaining 904 teens while the jobs lasted, but the difference did not become statistically significant until six months into the study - three months after the jobs were completed.
From there, the gap continued to widen. At 16 months, police data revealed 5.1 arrests for violent crimes per 100 youths who were offered jobs, compared with 9.1 violent-crime arrests per 100 teens who were not.
Heller said that the results may underestimate the impact of the jobs program, as a quarter of the teens who were offered jobs did not accept them, though some of those teens found other jobs on their own.
Some of the teens who were not offered jobs also found employment on their own, but Heller she did not yet know how many. She said the number was likely to be low, citing a past study that found just 9 percent of African American teens were employed. Of the teens in the study, 95 percent were black.
Heller said it was too soon to perform a full analysis of whether the program paid for itself, as she would need to know whether the program yielded any long-term effects on employment, among other factors. Still, she said, early evidence suggests that the program's price tag may already be outweighed by such benefits as reduced costs to the justice system and less suffering by victims.
The jobs program did not appear to have an impact on the rates of nonviolent crimes. Heller said the reason might be that although the jobs may have taught teens to resolve conflicts, those other types of offenses do not involve conflict.
The jobs included serving as a day-camp counselor, cleaning vacant lots, planting community gardens, and performing clerical work for nonprofit groups and government offices. Those in office jobs had some tedious tasks, such as paper-shredding.
"We explained that everybody's first job is a horrible job," Diaz said. "You've got to do some of that."
One mentor was assigned to every 10 youths. These adults were available to give advice around the clock, and some went beyond the call of duty, Diaz said.
When one young man got into trouble with the law and had to go to court, the mentor appeared before the judge to recommend that the teenager stay on probation rather than go to a detention center, she said.
Philadelphia also is known for having a strong teen-jobs program, Diaz said. She said her office had exchanged ideas with Philadelphia Youth Network, a local nonprofit that works to get teens employed.
A key to both cities' approaches is that the teenage years are not too late to make a difference, she said.
"Lots of people will write off teenagers, especially if they've already gotten in trouble with the law," Diaz said. "We don't give up on any child."