Emailing eligible college students resulted in more than 7,000 students applying for CalFresh benefits

CalFresh benefits can help college students make ends meet while attending college, but not all eligible students apply. One contributing factor may be that students are not aware they are eligible. Therefore, outreach efforts informing them of their eligibility could help increase take-up rates. To test this, we designed and conducted two experiments that leveraged an expansion in CalFresh eligibility for students that went into effect in early 2021. In response to the pandemic, Congress permitted a temporary expansion to college student eligibility for SNAP (Supplemental Nutrition Assistance Program), or CalFresh as it is known in California. The expansion went into effect in January 2021 and will last through the end of the federal public health emergency. The California Department of Social Services (CDSS) and the California Student Aid Commission (CSAC) partnered with The People Lab (TPL) and the California Policy Lab (CPL) on two randomized experiments to evaluate whether outreach about this policy change could increase the number of students who applied for and eventually enrolled in CalFresh. One experiment, conducted in February and March 2021, tested the impact of email outreach. The second, conducted in June 2021, tested the relative effectiveness of different messages and modes of communication on the same group of students who were newly eligible. This brief reports the results of the first experiment.1 In a randomized experiment with 285,731 California college students who were eligible under the temporary expansion and were not already receiving CalFresh in January 2021, we found that an email from CSAC led roughly 2–3% of recipients to apply for benefits through GetCalFresh.org within a week of receiving the email. The increase in applications translated into an increase in CalFresh enrollment of 1.5 to 2.5 percentage points. Effects were even larger for students who received a second follow-up email.This work has been supported, in part, by the University of California Multicampus Research Programs and Initiatives grants MRP-19-600774 and M21PR3278