@EdWorkingPaper{ai26-1485,
title = "Meeting People Where They Are: Experimental Evidence on Embedded Supports, Service Use, and Educational Outcomes",
author = "Kelli A. Bird, Benjamin L. Castleman",
institution = "Annenberg Institute at Brown University",
number = "1485",
year = "2026",
month = "May",
URL = "http://www.edworkingpapers.com/ai26-1485",
abstract = {Many public services suffer from persistently low take-up despite high potential returns. A growing body of evidence suggests that information alone does little to close this gap; instead, hassle costs and default access points may be binding constraints on utilization. We test whether reallocating existing services to where people already are – rather than requiring them to seek out centralized offices – can improve utilization and outcomes.

We study this question in community colleges, where even modest frictions can impede students from engaging with available support services. We conducted a multi-semester randomized controlled trial in which treatment sections had tutoring and advising staff embedded directly in the classroom. Faculty and staff targeted these supports toward students at risk of academically struggling. Control sections followed standard instructional practices. Critically, both treatment and control students had access to the same on-demand tutoring and advising through centralized campus offices.

Embedding targeted supports into the classroom generated large increases in utilization: within the pilot course, treated students were 31 percentage points more likely to meet with a tutor, 29 percentage points more likely to meet with an advisor, and 12 percentage points more likely to meet with their instructor. Treatment students were 5 percentage points more likely to pass their courses and 4.6 percentage points less likely to withdraw. Impacts spilled over beyond the focal course: treated students were more likely to meet with faculty and staff outside the pilot course, and showed gains in overall semester GPA and persistence into subsequent terms. Our findings demonstrate that resource reallocation can meaningfully improve outcomes for underserved students.},
}