Study
Valentine and colleagues (2017) used a randomized lottery system to examine the impact of summer employment on school attendance, academic achievement, and employment among youth who applied for the Summer Youth Employment Program (SYEP) administered by the New York City Department of Youth and Community Development (DYCD). Using results from the city’s randomized lottery system, the study authors compared the outcomes of youth who had won a placement in the SYEP through the lottery (treatment group) with those who had applied to the program but had not won a place (control group).
An initial set of lottery winners were randomly selected in May of each year, and youth who won places in the program were notified and given 5 days to respond to the offer. Then, during a secondary lottery, youth were randomly selected to fill the remaining spots of those who had not responded or had declined the initial offer. This process was repeated until all places were filled, or the program period began.
The total sample for this study included 264,075 youths between the ages of 14 and 24 who applied to SYEP for the first time between 2006 and 2010. There were 116,919 youths in the treatment group and 147,156 in the control group. Approximately 52.4 percent of the treatment group was between the ages of 14 and 15, 28.6 percent were between 16 and 17, and 18.9 percent between 18 and 24. The treatment group was 46 percent male. In regard to race/ethnicity, roughly 43.5 percent were Black, 30.5 percent Hispanic, 12.2 percent white, 8.4 percent Asian, and 5.5 percent other. Of the treatment group, 86.3 percent had an educational status of high school or below, 11.8 percent had been employed the year before SYEP and earned roughly $494, and 79.4 percent were eligible for free or reduced lunch.
Similarly, 54.2 percent of the control group was between the ages of 14 and 15, 28.4 percent between 16 and 17, and 19.4 percent between 18 and 24. The control group was 45.8 percent male. In regard to race/ethnicity, 43.2 percent were Black, 30.6 Hispanic, 12.1 percent white, 8.5 percent Asian, and 5.6. percent other. Of the total control group, 85.9 percent had an educational status of high school or below, 11.8 percent had been employed the year before SYEP and earned roughly $500, and 79.6 percent were eligible for free or reduced lunch. There were a few statistically significant demographic differences between the treatment and control groups, including the number of 18–24-year-olds and those in other age categories (14–15 and 16–17), the borough of residence, and whether they had been employed the year before SYEP. There were no other statistically significant differences between the treatment and control groups.
Intent-to-treat and regression analyses were conducted for several outcome measures, including total employment, total earnings, school attendance, high school graduation rates, and college degree completion. The NYC Department of Education provided data on youth’s high school attendance and graduation in NYC and enrollment in and graduation from 2- and 4-year colleges in the United States. This data was only available for youth in the sample who received funding from NYC schools. Quarterly records from the New York State Department of Labor were used to measure employment and earnings in jobs that were not part of SYEP. Such records did not include information on jobs located outside of the state or informal jobs (e.g., babysitting). Lastly, DYCD provided data on applicants’ characteristics, lottery results, and participation in SYEP. Follow-up analyses were conducted every year for up to 5 years after program participation. Five-year follow-up findings were not included in this review because over the course of the 5 years, youth in both the intervention and the control groups were able to apply for SYEP in any year in which they were eligible. Therefore, over the course of 5 years, youths in both groups may have participated in SYEP for multiple years, and the findings at the 5-year follow up would not represent the impact of SYEP on all lottery winners in a given year. The study authors conducted subgroup analyses.