education and either vocational education or actual jobs. Again the results were equivocal. On one hand, the Youth Incentive Entitlement Pilot Projects was effective at reducing birth rates and increasing both teen employment and school enrollment. This program guaranteed part-time employment to youth during the school year and full-time employment during the summer if they remained in high school. On the other hand, four very rigorously evaluated employment and education programs were not effective in reducing pregnancies or birth rates. This evidence suggests that vocational and employment programs are not effective in decreasing pregnancy or birth rates.
In summary, the strongest evidence for program effectiveness using a youth development framework comes from the Teen Outreach Program. Why was this service-learning program effective while the others were not? We do not have an answer for this question because causal links in the interventions were not evaluated. Therefore, we do not know whether to attribute its success to the service-learning component or to other characteristics of the program, such as strong bonds with adults. Nevertheless, the program evaluations do offer promise that a youth development approach may provide opportunities that reduce the risk of teenage pregnancy.
Furthermore, Kirby summarized the strengths and limitations of each evaluation based on the following methodological principles: sampling, random assignment, sample size, long-term follow-up, measurement of behavior (sexual and contraceptive behaviors), statistical analyses, publication of results, replication, and independent evaluation. Kirby concluded that few studies abided by high standards for each principle, thus limiting generalizations about which specific programs or types of programs reduce sexual risk-taking and teen pregnancy.
The most comprehensive review of positive youth development program evaluations was funded by the Department of Health and Human Services and compiled by Catalano, Berglund, Ryan, Lonczak, and Hawkins (1999). This report examined evaluations of positive youth development programs solicited from a wide variety of sources. The authors reviewed 77 program evaluations and chose 25 for their report (see Table 6–1 for details). Only research designs using a control or