Abstract
The Degree Project experiment was designed to test the efficacy of a promise scholarship (a type of financial aid and free college) in Milwaukee, Wisconsin. In 2011, half of the schools in Milwaukee Public Schools (MPS) were assigned to treatment (i.e., it is a cluster randomized trial). To increase statistical power, we used paired randomization based on pre-treatment values of school-level college-going rates. All high school ninth graders in the treated schools were offered up to $12,000 merit-based grant that could be used to attend the vast majority of two- and four-year colleges in the state of Wisconsin. To receive the funds, students had to achieve a cumulative 2.5 GPA, attend class 90 percent of the time, graduate from an MPS high school on-time, and fill out the FAFSA. The study team had access to rich (non-public) data including K-12 academic outcomes, surveys of students during high school, college attendance and graduation, employment, earnings,. incarceration, and teen pregnancy. The study tracked students for eight years from the point of treatment.