Do Emergency Micro-Grants Help Financially-Disadvantaged Students Succeed?
Over the last 10 years, tuition and fees at degree-granting institutions have risen by 27 percent, making it more difficult for students, especially those already struggling to cover basic needs like housing and food, to afford to remain in college in the face of unexpected financial trouble. In many cases, unpaid term balances prevent students from continuing in the current term or enrolling in the following one, and as a result, students dropout or are automatically dropped. Unpaid balances can be particularly devastating in a state like Georgia, which mandates that institutions drop students who don’t pay their full tuition balance by the first week of classes.
Some institutions have responded to this problem by implementing micro-grant or emergency aid programs that alleviate financial distress of students in high need. One of the pioneering examples, and one of the largest, is the Panther Retention Grants (PRG) program developed by Georgia State University (Georgia State), which automatically covers unpaid term bill balances of up to $2,500 for students in good academic standing, allowing them to remain enrolled in their current term and register for the following one. Since 2011, the program has awarded more than 14,000 grants totaling $18.9 million to over 11,000 students, with an estimated ROI for Georgia State of $4 to $7.8 million.
With support from the Bill & Melinda Gates Foundation, Ithaka S+R will assess the impact of the PRG program on students’ academic outcomes and estimate its broader institutional return on investment, culminating with the publication of a research report in 2020. We aim to employ a regression discontinuity (RD) design, a quasi-experimental approach that utilizes an artificial quantitative threshold for treatment eligibility (in this case, an unpaid balance cutoff of $2,500) to mimic random assignment to control and treatment groups. An additional key feature of the program is that, unlike most emergency aid programs, eligible students automatically receive the PRG covering their unpaid balance without having to identify, apply for, or accept such assistance. This eliminates selection bias – the possibility that eligible students who take up the treatment are different, on average, from those who don’t – allowing us to estimate the causal impact of the program on students’ outcomes.
In addition to assessing the program’s impacts on enrollment, credit accumulation, debt, and graduation, Ithaka S+R will analyze whether there are disproportionate benefits for low-income, first-generation, and underrepresented minority students, groups more likely to face financial challenges, which may contribute to the known equity gaps in higher education.
On the institutional side, programs such as PRG appear to be cost effective for an institution like Georgia State, whereby the tuition revenue generated by retaining PRG recipients who would have otherwise left the institution exceeds the program’s operational costs. Drawing on our findings pertaining to student retention through the PRG program, we will assess the impact on institutional finances by comparing the costs associated with the program to changes in revenues.
The PRG program, like other emergency aid programs, removes the most immediate financial obstacle to remaining enrolled at the institution. Nonetheless, eligible students’ financial challenges may persist, either forcing them to enroll in fewer credits, which derails their time to graduation, or dropping out a few semesters later than they otherwise would have. Understanding the impact of these grants on retention in immediate and future semesters, as well as on degree progress and completion, will provide the higher education community with important insights into whether such focused and targeted programs provide sufficient assistance for different groups of students.
 The exact methodology to assess the program’s impact will be finalized once data from Georgia State is collected, cleaned, and tested to ensure that a regression discontinuity is the appropriate model given the distribution of awards around the $2,500 threshold.