United States General Accounting Office. (1996). Job Training Partnership Act: Long-term earnings and employment outcomes (GAO/HEHS-96-40). Washington, DC: U.S. Government Printing Office.
- The study’s objective was to estimate the impact of Job Training Partnership Act (JTPA) programs on the long-term earnings and employment of male and female out-of-school youth and adults. This review focuses on the findings for out-of-school youth.
- The programs’ effectiveness was evaluated by analyzing outcomes five years after program entry for a subset of participants in the National JTPA Study, a random assignment study that examined impacts of JTPA programs in 16 local service-delivery areas.
- The authors found no statistically significant impacts of JTPA on earnings or employment rates five years after program entry.
- The quality of the causal evidence presented in this study is high because it is a well-executed randomized controlled trial. This means we have confidence that estimated effects would be attributable to JTPA programs, and not other factors.
Job Training Partnership Act
Features of the Intervention
The JTPA, which was enacted in 1982, provided block grants to state and local governments to provide employment training to people facing difficulties obtaining employment. JTPA programs provided services such as remedial education and job-search assistance in local service-delivery areas. Typically, providers of such services included community colleges; community-based organizations; and vocational, technical, and proprietary schools. Program participants in JTPA who were involved in the study typically received three to four months of training, at an average cost of about $2,400 per participant.
JTPA replaced the Comprehensive Employment and Training Act, which had similar job training goals. JTPA was itself replaced by the Workforce Investment Act of 1998.
Features of the Study
The study examined the impacts of JTPA program participation on the long-term employment and earnings of male youth, female youth, adult men, and adult women; this review focuses on the analysis of male and female youth. To be eligible to participate, youth had to be ages 16 to 21 and not attending school at the time of application to JTPA.
The study drew on a subset of the sample used for the National JTPA Study (NJS), which randomly assigned more than 20,000 eligible out-of-school youth and adults in 16 local service delivery areas to treatment and control groups from November 1987 to September 1989. The current study examined long-term employment and earnings outcomes for that portion of the NJS sample for which five years of follow-up data were available in a Social Security Administration earnings records database (13,699 individuals; approximately 68 percent of the original sample). The analysis sample included about 1,700 male and 2,300 female youth.
The study reported no statistically significant differences between the earnings or employment rates of youth who had access to JTPA programs and those who did not in any of the five years after random assignment.
Considerations for Interpreting the Findings
The authors noted that about one-third of those assigned to the treatment group did not actually enroll in JTPA. In addition, a substantial minority of the control group received some JTPA-like services from other resources available in the community. Because all youth were included in the analysis in their originally assigned groups, this reduces the difference between the services received by the treatment group and those received by the control group, which could understate the programs’ impacts.
In addition, the authors used t-tests, not regression models, to determine the statistical significance of differences between treatment and control group outcomes. Although this is an acceptable method of testing differences in a well-executed randomized controlled trial, the precision of these tests could have been enhanced by estimating regression models that included controls for baseline earnings.
Causal Evidence Rating
The quality of the causal evidence presented in this study is high. This means we have confidence that estimated effects would be attributable to JTPA programs, and not other factors. However, the analysis found no statistically significant effects.