Co-Principal Investigators: Bridget Terry Long (NBER/Harvard University) and Philip Oreopoulos (NBER/University of Toronto)

Purpose: The purpose of the proposed research is to test the effects of simplifying the financial aid application process and providing assistance at scale. Findings from previous efficacy studies found that these two supports led to a 20 percent increase in college attendance in our target areas in Ohio and North Carolina. The team will recruit participants in this scale-up evaluation from the 3.5 million individuals who use free tax-filing services, and will randomly assign families to different types of assistance with the financial aid process. The researchers will evaluate the effects of the simplification of the financial aid application process and assistance on college outcomes.

Project Activities: Researchers will use an experimental design to investigate the impact of the simplification of the application process on the likelihood of college attendance, access to financial aid, college choice, and college persistence. Participants will be assigned to the control condition or one of three levels of assistance. The project team will gather long-term outcome data for students who participate in both the experimental and control conditions of the study.

Products: The products of this project will be evidence of the effectiveness of the simplified financial aid application process when implemented at scale, and peer reviewed publications.

Structured Abstract

Setting: The intervention will take place at about 9,000 free tax-file sites across the United States (i.e., Volunteer Income Tax Assistance sites) run by the Internal Revenue Service. The researchers will collaborate with a group that provides tax software at each of these sites.

Sample: The proposed research will focus on families with college-age children and incomes below $45,000 per year. These individuals would likely qualify for financial aid if they attended college.

Intervention: The intervention involves several levels of assistance to families completing the Free Application for Federal Student Aid (FAFSA). As individuals complete their taxes, the researchers will use their tax data to pre-populate the FAFSA form. The researchers will then offer different levels of assistance to different individuals. The levels of assistance vary from assistance with filing the FAFSA form to interpretation of the likely aid eligibility.

Research Design and Methods: The research has an experimental design. In the initial year of the study, families who agree to participate in the research will be randomly assigned either to the control group or one of the three treatment groups. There are three treatment groups and each group will receive a different intervention. One group receives aid information only (i.e., receive a pamphlet which includes information on college costs and benefits). Among those receiving FAFSA help, some receive personalized assistance while others do not. The research team will gather data from administrative data sets (e.g., the National Student Clearinghouse) to determine if participation in the treatment groups resulted in improved long-term student outcomes relevant to access and persistence in college.

Control Condition: The control group will receive basic information about the FAFSA and the college application process.

Key Measures: The researchers use administrative data that describes individual's college attendance, access to financial aid, college choice, and college persistence. The data will be drawn from datasets maintained by the National Student Clearinghouse and the Department of Education. These outcomes, which include completing the Financial aid application, financial aid receipt, college attendance, and college persistence, will be investigated for the entire sample as the well as for subgroups (by income, race, gender). The use of administrative data allows the researchers to inexpensively track all of the study participants over time.

Data Analytic Strategy: Given the randomization involved, simple comparisons of the control and treatment groups will identify the treatment effects. Researchers will estimate both the effects of offering the service (intent-to-treat effects), and the effects of using the service among individuals for which a FAFSA is filed (treatment-on-the-treated effects). The former will also use fixed effects multivariate regression analysis with covariates to estimate the former, and regressions with instruments to identify the "treatment-on-the-treated" effect.

Related IES Projects: National Center for Postsecondary Research (R305A060010)