Association for Public Policy Analysis & Management

Panel: Experimental Results on the Benefits of Supporting Full Time College Enrollment

(Education)

Improving low-income students’ college completion rates is one critical means to increase economic mobility and reducing inequality.  This panel would present the findings from four randomized trials, three of which have sufficiently long follow-up to demonstrate that, contrary to most prior study findings, it is possible to achieve large gains in college completion rates.  The study findings also suggest that the right combination of requirements and services may be able to reliably replicate these gains.

Two of the studies evaluate a program, Accelerated Study in Associate Programs (ASAP), originally developed and implemented at the City University of New York (CUNY) and subsequently replicated in a separate study at three colleges in Ohio.  Previous findings for CUNY ASAP found unprecedented impacts on the program’s goal of three-year associate degree attainment.  One panel paper would present currently unpublished six-year impacts for CUNY ASAP on degree-receipt and related outcomes.  This paper would also describe early impacts from the Ohio ASAP study.

The other two papers would present findings on a program model originally developed in San Antonio, Texas, Project QUEST, and subsequently implemented in nine other sites, including at the Valley Initiative for Development and Advancement (VIDA) in the lower Rio Grande Valley.  One paper would present currently unpublished six-year impact findings from a study of Project Quest itself which show very large effects on earnings, as well as on other educational and economic outcomes.  The other paper would present currently unpublished two- and three-year findings from a study on VIDA which show educational gains that approach those of CUNY ASAP.

Although the four programs represent two different program models, and populations with varying levels of educational need and educational goals, they also have very similar program features.  All four require full-time enrollment and support it with frequent mandatory counseling and financial support, including at a minimum support for tuition and transportation conditioned on program compliance.  The success of these common features and the fact that both models are implemented in multiple sites suggests that these large impacts are not simply due to chance, and potentially could be replicated more broadly with large effects.

The papers also speak to the value of marrying multiple sources of high-quality data with strong research designs for causal analysis.  All four studies rely on multiple data sources, including national college records from the National Student Clearinghouse) and/or local college records, national employment and earnings records from the National Directory of New Hires and/or state Unemployment Insurance earnings records, supplemented by high-quality surveys.  The studies also demonstrate the high value of rigorous randomized studies that serve as replication trials in establishing strong evidence for decision making.  Without the combination of high-quality measurement, well-implemented strong research designs and multiple studies of the same model, it would be easy for skeptics to dismiss the large positive effects found as “too good to be true.”