Panel Paper: A Randomized Experiment to Increase FAFSA Completion among Public Housing Residents

Saturday, November 4, 2017
Water Tower (Hyatt Regency Chicago)

*Names in bold indicate Presenter

Michael DiDomenico, U.S. General Services Administration and Calvin C Johnson, U.S. Department of Housing and Urban Development


Despite the fact that college is on average a smart investment, low-income students are less likely both to start and complete post-secondary education than their wealthier peers (Ma, Pender, & Welch, 2016). Recent experiments have shown that interventions that increase Free Application for Federal Student Aid (FAFSA) completion or renewal also can increase rates of college attendance and persistence (Bettinger et al., 2012; Castleman & Page 2015). The US Department of Housing and Urban Development, US Department of Education, and the General Service Administration’s Office of Evaluation Sciences fielded an experiment designed to increase FAFSA completion among HUD-assisted youth. In one large public housing authority, the study team randomly assigned public housing households with youth ages 17 to 24 to be sent several communications urging them to complete the FAFSA.

Those selected to be sent communications were sent a series of 4 physical mailers, 3 robo-calls, and if the household provided an email address, 4 email messages. Messages targeted common reasons for not filing, including incorrect assumptions about eligibility and missed deadlines (Davidson, 2013). While the messages were largely informational, they also incorporated behavioral science insights. For example, one message included a social norm, indicating that 4 out of 5 students in the housing authority who complete the FAFSA qualify for a Pell Grant (Gerber & Rogers, 2009), while another included an implementation intention prompt, asking students to write down the time and place they planned to complete the FAFSA (Gollwitzer, 1999).

The study is sufficiently powered to detect a policy-relevant changes, with a total of approximately 50,,000 individuals split evenly between the treatment and control groups. Communications were sent between February and April 2017. This paper reports on the effect of the messages on FAFSA completion rates, with possible later analyses including college enrollment. In addition to building on prior research and policy aimed at reducing barriers to FAFSA completion, this study provides evidence of how federal agencies can collaborate to improve efficient access to public benefits.