Performance-Based Scholarship Demonstration


For many low-income college students, one of the biggest barriers to attendance is cost. While federal and state financial aid is available to help with tuition, fees, books, and some living expenses, students still often have unmet need, particularly if they are from the poorest families or are independent from their parents. Working while going to school is one possible answer, but too many hours on the job can contribute to poor academic performance and dropping out. Loans are another possible answer, but many low-income students are reluctant to take on debt — especially if they have doubts about their ability to complete a college degree, or if they do not come from families where college attendance is the norm.

MDRC launched the Performance-Based Scholarship Demonstration in 2008 to test an innovative strategy for addressing two policy objectives: increasing the financial support available to low-income students, and creating an incentive for such students to complete their courses and make more timely progress toward degrees. The idea was to provide a supplement to existing federal and state financial aid that is contingent on enrolling in a minimum number of credit hours and making passing grades. The performance-based scholarships were paid directly to students (rather than to the colleges or universities they attend) in order to reward students for their progress and to allow them to make choices about how best to support their schooling. For some, this meant buying books or paying for transportation to their college campus; for others, it meant cutting back on work hours or hiring a babysitter for their children during finals week.

The Performance-Based Scholarship Demonstration was based on positive short-term findings that emerged from MDRC’s Opening Doors Demonstration in Louisiana, which had a number of positive effects for students, including boosting students’ credit accumulation, grades, and persistence. The Opening Doors Demonstration targeted low-income parents and, as a result, the study sample was composed of older, unmarried, and mostly female students. Unfortunately, the devastation inflicted by Hurricane Katrina disrupted students’ education, and plans for a longer-term follow-up to look for impacts on degree receipt were discontinued.

While the Louisiana short-term results were impressive, there was an open question about whether performance-based scholarships would be effective in other college settings and for a broader range of target groups. In addition, there were questions about the payoff of varying scholarship amounts and durations.

Agenda, Scope, and Goals

With anchor funding from the Bill & Melinda Gates Foundation and commitments from other funding partners, MDRC tested variations of performance-based scholarships at eight colleges and one intermediary across six states. Each program targeted low-income students with high unmet financial need, based on the cost of attendance and gaps in state financial aid. The hope was to inform potential changes in state policy and the design of scholarship programs.

The Performance-Based Scholarship Demonstration was designed to produce rigorous evidence of the programs’ impacts and answer policy-relevant questions:

  • Can performance-based scholarships be implemented widely, in diverse settings, for diverse groups of students?
  • Do performance-based scholarships affect important academic outcomes, as measured by enrollment, the number of credits students earn, and attainment of degrees and certificates?
  • Do the impacts of performance-based scholarships vary for different scholarship programs or for different types of students?
  • What is the cost to implement performance-based scholarship programs?
  • Are performance-based scholarship programs cost-effective?

Design, Sites, and Data Sources

More than 5,100 students received scholarships at two- and four-year institutions in six states: Borough of Manhattan and Hostos Community Colleges in New York City (with support from the Robin Hood Foundation); Lorain County, Owens, and Sinclair Community Colleges in Ohio (with support from both the Chancellor of the Ohio Board of Regents and the Ohio Department of Job and Family Services); the University of New Mexico (with support from the Open Society Foundations); Hillsborough Community College in Florida (with support from both the Helios Education Foundation and the Open Society Foundations); Pima Community College in Arizona (with support from both the Helios Education Foundation and The Kresge Foundation); and statewide in California through the California Cash for College program in partnership with the Los Angeles Area Chamber of Commerce Foundation and the California Student Aid Commission (with support from the College Access Foundation of California). We also worked with UNCF, the nation’s largest minority education organization, to provide technical assistance for implementation of performance-based scholarships at three UNCF member institutions, Benedict College and Claflin University in South Carolina and Florida Memorial University in Florida.

Because the scholarship dollars were limited, the demonstration used a random assignment research design to compare the outcomes of students who receive a performance-based scholarship (on top of their regular financial aid) with the outcomes of a control group who would not receive the performance-based scholarship but would have access to any other available scholarships. MDRC used a variety of quantitative data sources, including transcript data, other administrative data, and surveys, to determine the impacts of the scholarships. Qualitative data from focus groups and interviews were employed to help understand why the scholarship intervention had an effect on outcomes.

MDRC has results from the programs in all six states, in addition to the original results in Louisiana. Among the findings are the following:

  • The programs modestly increased degree completion and academic progress.

  • The programs did not have a substantial impact on persistence (measured each year by whether students enrolled in school).

  • The scholarships worked for a variety of low-income students with different characteristics, including at-risk groups that traditionally perform poorly.

  • Evidence from the study in California suggests that offering students a scholarship in their senior year of high school increases enrollment in the first year of college.

  • At sites where student services accompanied the performance-based scholarship, the use of these services increased markedly as a result of the program.

  • The performance-based scholarships increased students’ total financial aid, even when sometimes helping to reduce their dependency on loans.

  • The scholarships cost additional money but did not increase the average cost per degree.