Title: When and Why Does College Advising “Work”: Evidence from Advise in TN
Abstract: College advising programs increase the likelihood students apply to and enroll in higher education. However, few are proven effective at scale. We leverage the rollout of Advise TN across 33 communities to estimate causal impacts of a novel advising program on college enrollment, persistence, degree completion, and workforce participation. With complementary event-study and robust difference-in-differences strategies, we show this program raised college enrollment rates by 3 points (or 6%) at scale, especially among Hispanic, female, and rural students. We then interrogate mechanisms to explain this success with administrative records and unique student-advisor interaction data. We show increases to college-going are driven by larger improvements to early task completion, including filing the FAFSA and applying for state aid. We also descriptively show that program design matters, where college enrollment rates vary significantly by advising intensity, modality, and student-to-advisor ratios. We do not detect changes in students’ later college outcomes or employment and argue this is also due to variation in program design, including a focus on short-term information, medium-term task completion, and long-run skill development. Our study greatly expands knowledge on advising programs and is among the first to interrogate how programs come to “work” at scale.
