The impact of providing community college students information about large financial aid grants for living expenses

Students assigned to the treatment group will receive informational emails explaining how the amount of financial aid they are eligible to receive through the Student Success Completion Grant (SSCG) varies with the number of units they enroll in. The emails highlight the additional grant amounts available when enrolling in higher course loads, particularly the financial bonus for enrolling in 15 or more units per term. The intervention is designed to reduce information frictions by clearly presenting the relationship between course enrollment intensity and grant eligibility. Emails will be sent twice—once shortly before the start of the fall term and again prior to the add/drop deadline—to maximize the likelihood that students receive the information before finalizing their course schedules.

2025-08-01

2025-09-30

We will track students for at least three years from the initial intervention (2025-26 through 2027-28). Our two primary outcomes are enrolling in at least 15 units and completion (degree attainment or transfer) within 3 years. Specifically, we define the first outcome as a binary indicator equal to 1 if a student enrolls in 15+ units in the fall and spring semester of Year 1. Completion is defined as a binary indicator equal to 1 if a student earns either a certificate or associate degree, or transfers to a four-year institution within three years.

We will also examine the intervention’s impact on a variety of secondary outcomes, including overall community college enrollment, total units enrolled and accumulated and financial aid receipt. We will also explore longer-run outcomes of persistence and time-to-degree.

The experiment will be conducted on the full population of 177,154 California residents who applied for financial aid for the first time by submitting the Free Application for Federal Student Aid (FAFSA) or California Dream Act Application (CADAA) for the 2025-26 academic year, were eligible for Cal Grant B and listed a California Community College on their application. Notably, this sample includes some students who listed both a CCC and a four-year institution (e.g. California State University, University of California, etc.) on their application and is consistent with prior years . These new students who file and meet Cal Grant eligibility requirements (e.g., GPA and financial need cutoffs) for the first time (e.g., they are not returning Cal Grant recipients) and list a CCC on their FAFSA/CADAA

Once CSAC provides this list of all eligible students, we will randomly assign students to either the treatment (50%) or one of two control groups containing 25% of the sample each. The first control group can be thought of as a “pure” control where these students do not receive any information. The second control group receives a generic email regarding their Cal Grant B eligibility. As described below, we include this email control group to allow us to estimate treatment-on-the-treated (TOT) effects among those students who open the email. Importantly, assignment to treatment status will be made independent of whether a student actually enrolls at a CCC or receives a Cal Grant B.

The control email will provide generic information about Cal Grant B eligibility and include the basic text CSAC already uses in their communications to students. The treatment email will include the text of the control email plus additional details on the total aid they would receive with bonuses for enrolling in various unit thresholds (e.g., <12, 12, 15 units).

We will send the emails at least twice, with the timing chosen to maximize the chances that the information will influence how many units students take. Specifically, the first email will be sent just prior to the start of the fall semester (early August) and the second shortly before the drop/add deadline (end of August). The beginning of the CCC Fall semester ranges from the first week to third week of August, with the deadline to add or drop classes usually around early September.

Not available

Randomization was conducted at the individual student level using administrative data from the California Student Aid Commission. The analytic sample was restricted to new financial aid applicants eligible for Cal Grant B who listed a California Community College on their application. Prior to randomization, a small number of records with inconsistent identifying information within Social Security numbers were removed to ensure each student had a unique and reliable record.

Students were then assigned to experimental groups using a reproducible randomization procedure with a fixed random seed. Random assignment was stratified by award status (new applicant), institution listed on the financial aid application, and age group (under 20, 20–24, and 25 or older). Within each stratum, students were randomly ordered using a uniform random number and assigned to one of three groups: 50 percent to the treatment group receiving the informational email, 25 percent to a control group receiving a generic Cal Grant email, and 25 percent to a pure control group receiving no email.

Individual

No

177,154 individual students

177,154 individual students

88,577 in the treatment
44,288 in the control email
44,289 in the pure control (no email)

Based on data received from the California Student Aid Commission, there are 177,154 new students in the 2025–26 academic year who are eligible for Cal Grant B and list a California Community College on their FAFSA. Our intervention provides the information treatment at the start of the Fall 2025 semester to a randomly selected 50 percent of these students. We conducted power calculations to determine the minimum detectable effect sizes (MDES) for both intent-to-treat (ITT) and treatment-on-the-treated (TOT) estimates. We focus on two binary outcomes: enrolling in 15 or more units, which is the behavior directly targeted by the information intervention, and degree completion. Based on prior research, we assume a baseline control mean of 23 percent for enrolling in 15 or more units and 9 percent for degree completion. All calculations assume a significance level of 0.05 and statistical power of 0.8. Given the large sample size, the study is powered to detect very small effects. For the full sample, the experiment can detect ITT effects as small as 0.56 percentage points for enrolling in 15 or more units and 0.38 percentage points for degree completion. Even under conservative assumptions about treatment exposure—for example, if only 25 percent of students open the email—the study can still detect treatment-on-the-treated effects as small as 1.39 percentage points for enrolling in 15 or more units and 0.96 percentage points for degree completion. Overall, the experiment has ample statistical power to detect policy-relevant effects.