Absence of conflict of interest.
- The study's objective was to examine the impact of YouthBuild on employment, earnings, and education outcomes.
- The study was a randomized controlled trial that randomly assigned young adults to the YouthBuild program or a control group. The authors administered a survey 48-months after study enrollment and used statistical models to compare outcomes of participants in the treatment and control groups.
- The study found that YouthBuild participants were significantly more likely than control participants to be enrolled in an education/vocational program, earn a high school diploma, and be employed at the 48-month follow-up.
- This study receives a high evidence rating. This means we are confident that the estimated effects are attributable to YouthBuild, and not to other factors.
Features of the Intervention
YouthBuild began as the East Harlem Youth Action Program in the 1970s to provide education and vocational training to young adults. The program now serves over 10,000 young adults at over 250 organizations in the United States each year and provides several services to reengage disconnected youth in employment and education. These services include education designed to lead to a high school diploma or equivalent credential; vocational training for an in-demand industry like construction; services focused on building youth development (e.g., leadership training and community service); supportive services geared to help individuals in training and employment; case management, workforce preparation; life skills training; and counseling. The program typically lasts between 6-12 months and offers youth a stipend or living allowance while participating in the program. Programs also used additional screening measures including academic skills tests, program specific criteria, and Mental Toughness Orientation to determine participant readiness to participate in the program. All YouthBuild programs are locally operated by community-based organizations, educational institutions, or government agencies and the program model provides flexibility to adapt program components, as needed, to fit community contexts. The program serves young adults between the ages of 16 to 24 who are not currently enrolled in high school and are from one of the following groups: low-income family, migrant family, in foster care, are ex-offenders, have a disability, or have parents who are incarcerated.
Features of the Study
The study was a randomized controlled trial of 75 YouthBuild programs selected from those receiving funding from the U.S. Department of Labor or Corporation for National and Community Service in 29 states, the U.S. Virgin Islands and Washington DC. The random assignment and enrollment procedures differed across the locations, where each program determined the eligible pool of applicants based on their own standard selection processes (e.g., administering basic skills tests or conducting interviews). Young adults between the ages of 16-24 years old were recruited from August 2011 through January 2013. A total of 3,929 eligible participants were enrolled in the study. Of the eligible participants, 2,700 were randomly assigned to the treatment group and 1,229 were assigned to the control group. Those assigned to the treatment group received the YouthBuild program and those in the control group were given information about other education and employment services available in the community but were prohibited from enrolling in the YouthBuild program for two years after random assignment. The study sample was predominantly male (64 percent) and African American (63 percent), with an average age of 20 years. The majority of participants did not have a high school diploma or equivalent (91 percent) and 30 percent were parents. Data sources included a survey administered at 48-months after program enrollment and administrative records on employment, earnings, and educational attainment. The authors used statistical models with controls for demographic, education, and program characteristics to compare outcomes of participants in the treatment and control groups.
Education and skills gains
- The study found that treatment participants were significantly more likely than control participants to have a high school diploma or equivalency credential at the 48-month follow-up (55 percent vs. 46 percent).
- The study found that treatment participants were significantly more likely than control participants to be enrolled in vocational school at the 48-month follow-up (33 percent vs. 22 percent).
- The study also found that treatment participants were significantly more likely than control participants to have enrolled in postsecondary courses (27 percent vs. 22 percent) and have attended college (21 percent vs. 13 percent) at the 48-month follow-up.
- The study did not find any significant differences between the groups in receipt of a school-based trade license/training certificate or a postsecondary degree.
- The study found that treatment participants were significantly more likely than control participants to have ever been employed (89 percent vs. 85 percent) and currently employed (51 percent vs. 46 percent). However, based on administrative data, these differences were only significant two years after program enrollment (63 percent vs. 60 percent).
Earnings and wages
- The study found that treatment participants had significantly higher current average weekly earnings than control participants at the 48-month follow-up, earning $33 more per week). However, based on administrative, there were only significant differences in earnings one year after program enrollment and control participants were earning $314 more than treatment participants.
Considerations for Interpreting the Findings
The authors estimated many related impacts on employment and earnings outcomes. Performing multiple statistical tests on related outcomes makes it more likely that some impacts will be found statistically significant purely by chance and not because they reflect program effectiveness. The authors did not perform statistical adjustments to account for the multiple tests, so the number of statistically significant findings in these domains could be overstated. Also, the study reports a less stringent statistical significance level, considering p-values of less than 0.10 to be significant, though it is standard practice to consider statistical significance if the p-value is less than 0.05. Only results that demonstrate a p-value of less than 0.05 are considered statistically significant in this profile.
Causal Evidence Rating
The causal evidence presented in this report is high, because it was based on a well implemented randomized controlled trial. This means we are confident that the estimated effects are attributable to YouthBuild, and not to other factors.