A national quasi-experimental evaluation of Learn-and-Serve America, service-learning programs funded by The Corporation for National Service (CNS), was conducted by The Center for Human Resources at Brandeis University (1999). CNS is a government entity created to fund programs that support school and community-based efforts to promote community service. For example, in 1994-95, the Corporation gave approximately $30 million in grants to support over 2,000 programs involving more than 750,000 youth. The Learn-and-Serve evaluation covered the years from 1994 to 1997 and consisted of an ethnicially and socioeconomically diverse sample. The programs were selected through a structured sampling process from a group of 210 randomly selected sites. Although there was variation among program characteristics, all of the programs chosen for the evaluation had been in operation for at least one year, reported higher than average service hours of participants, reported regular use of student written and oral reflection, and were school-based and linked to a formal curriculum.

Across the sites, students provided, on average, 70 hours of direct service, such as working as a tutor, in a nursing home or in a homeless shelter. Over 75% of participants had face-to-face contact with those that they were serving, such as meeting with students or senior citizens. This service also generally involved both individual and group projects. Very importantly, 80% of the program participants felt that they had made a contribution to their community and over 60% said that their service involved real responsibilities, a chance to do things themselves, and opportunities for discussion. Post-program results demonstrated short-term effectiveness. Aside from educational implications (e.g., school engagement, higher school grades) high school participants had significantly higher service leadership attitudes and civic attitudes than the comparison group. Program participants were also significantly more likely to have volunteered for and spent more time doing community service. Middle school participants, aside from having better social studies grades and fewer arrests and instances of pregnancy, were also significantly more likely than controls to have volunteered, but did not volunteer significantly more hours.

Participants were also evaluated one year after the program to determine longer-term associations with positive outcomes. Of the original 1,000 participants, 764 were involved in the follow-up (460 program participants and 304 comparison participants). The only effect that remained for middle school students was a marginally significant difference on number of arrests in the previous six months (i.e., participants having fewer arrests). High school participants provided almost twice as many hours of volunteer service as comparison students, but this effect was only marginally significant. Service leadership was significantly greater for participants, but the size of the effect decreased over the past year. Also, these longer-term associations may be confounded by individual characteristics. For instance, those who continued to be involved in community service after the end of the program were significantly more likely to volunteer and have college aspirations, and less likely to consume alcohol than those who did not continue to be involved in community service. It is quite possible that this exemplary group of youth were the reason for any statistically significant impacts of the program at the one-year follow-up.