In 2002, the New York City Department of Education (NYCDOE) announced a new district-wide admissions process that placed more emphasis on student choice and created over 100 new small schools to improve the high school educational experiences of students in the lowest income areas of the city. These new schools are commonly referred to as Small Schools of Choice (SSC) because they do not screen potential students based on prior academic achievement and are located in the neighborhoods they serve.
Prior research on SSCs has found large, positive impacts of enrollment on high school graduation rates and student outcomes. Building on these findings, MDRC researchers constructed a dataset that links administrative data from NYCDOE high school records, National Student Clearinghouse postsecondary enrollment data, and New York State employment and earnings records to assess the trajectories and outcomes of high school graduates in both postsecondary education and the labor market. NYC’s high school assignment lotteries create a naturally occurring experimental design, making it a prime example of how already-collected administrative data can be used for a longitudinal, retrospective evaluation. Within this study period, over 21,000 students entered the lottery for a seat in an SSC, providing a study and control group of students who did receive a seat and students who did not receive a seat or enroll, respectively.
Four cohorts of SSC students were followed after graduation from high school—three for four years, one for six years—to investigate “whether the positive effects of SSCs translate into impacts on students’ postsecondary degree attainment and performance in the labor market.” Since earnings data were not available for the full sample of students, researchers looked at a subset of each cohort for which they had earnings data, putting the proper controls in place to ensure the earnings sample for each cohort was representative of the overall group of students followed.
They found that, while the original positive effect SSCs had on college enrollment decreased in the four years after high school graduation, it was still evident among SSC graduates, as compared to their counterparts who attended a different style public high school. Additionally, researchers found that SSC graduates have higher rates of college enrollment than the control group and were also as likely as their counterparts in the control group to be employed. When both forms of “productive activity” are considered together, SSC enrollment positively effects the number of students who have a job and enroll in postsecondary education overall. Researchers note that future work will look at student outcomes for six years post-graduation with more detailed information on education behavior, such as course credits and enrollment from local institutions.
It’s worth underscoring that integrated administrative data has enabled MDRC to conduct multiple on-going evaluations of this kind, demonstrating the unique value of research that is able to look across agencies, outcomes, and lifecycles. Analysis of linked datasets can help demonstrate the complex ways in which policy changes affect the lives of specific populations, and in the case of the SSC research, offers promising evidence that this NYC program improves postsecondary and labor outcomes for young people.
Read the brief here
ICYMI: Mapping neighborhood network sites in Chicago
MDRC is also building the base of data-driven programs and projects that incorporate community expertise and knowledge. Specifically, MDRC’s Chicago Community Network (CCN) study uses community-informed surveys to expand local data about neighborhood organizations to include the interrelationships between the agencies themselves. The latest and final feature of the series features CCN data as customized network maps, among other graphics and visualizations, which are relevant and digestible to stakeholders with varying levels of data literacy. Overall, their work increases public access to actionable intelligence that supports growth and positive change at the neighborhood level.