In fiscal year 2012, about 60,000 youth age 16-24 participated in Job Corps, the country’s largest youth job training and readiness initiative. According to the Department of Labor, the program is designed to steer kids “toward brighter futures filled with self-confidence, independence, and productive employment.” Job Corps houses participants in facilities leased by the government, and connects them to vocational specialists who teach them how to speak and act, as well as specific skills like auto repair.
This year, Job Corps costs the government $1.7 billion, or about $28,400 per individual served. By far the most expensive per-individual training program, it begs the question: is this spending worth it? According to Dr. Peter Schochet, author of a 2008 study of Job Corps, the answer is yes:
“no other program serving [disadvantaged youth] has been shown to be as effective as Job Corps…In order to make a difference with this population, you have to spend money.”
While Dr. Schochet is an ardent advocate of the program, the program’s success depends entirely on your choice of metrics. On one hand, Job Corp model shows that participants are less likely to engage in criminal activity and typically see an earnings bump in the years immediately following the program. However, the effects do not last. Four years after completing the JobCorps program, which last an average of nine months, Job Corps participants earned only 22 cents per hour more than their peers ($7.55 vs. $7.33/hour). Eight years after completion, participants earned a whopping $26 more per year than nonparticipants.
Perhaps even more distressing is the fact that participants and nonparticipants were getting jobs in virtually the same, relatively unskilled industries, with 21.3% of participants worked in the service industry after four years, vs. 20.8% of non-participants; 20.9% of participants worked in construction, vs. 20.3%; 13.9% of participants worked as mechanics, vs. 13.1%. Job Corps seems to work, but barely, and the effects are
It’s hard to argue that this population of disadvantaged youth should not be engaged in programs like this. After all, helping them get ahead in life really helps everyone: by reducing crime and the size of the welfare state. The question, then: is if there is a better way to allocate over $28,000 per person to help them get ahead?
Here’s an idea- rather than spending it on a bureaucratic infrastructure (JobCorps had a $104 Million budget authority for facilities alone last year), spend it on human capital. Last year, Teach for America had an estimated acceptance rate of 13.6%, about the same rate as NYU business school. TFA rejected about 40,000 applicants, and similar programs like City Year and NYC Teaching Fellows rejected scores more. The human capital is there. The average cost of sending 3 kids through JobCorps is $84,000, significantly more than the average TFA teacher’s salary. We’re not suggesting to expand TFA.
In fact, where new TFA teachers are overwhelmed by classrooms full of difficult students fail, mentors in a new initiative with a smaller set of students might succeed. By focusing all of their time on just three youth, some close to their own age, teachers might make a real impact that lasts for years. They might teach their students skills learned in internships or summer jobs. They might spark an interest in higher education, and might also help their students and their parents navigate the halls of admissions and financial aid offices.
Or they might team up with their students to pursue policy research. Schochet’s study is trenchant, but the data it relies on dates from 1994-2003. All dollar amounts it uses are in 1995 dollars. No major independent study on JobCorps has been done since. $1.7 billion could buy a lot of impact studies. If we’re to ensure the best results for disadvantaged youth, who lack deep pockets, we should perhaps spend more time measuring if what we are doing works, because what little data we have suggests that it doesn’t.