The Cost-Effectiveness of Providing Early College Programs in Rhode Island. REL 2025-012

Abstract

This study examined the cost effectiveness of three early college programs in Rhode Island: dual enrollment, concurrent enrollment, and Advanced Placement (AP). Each program had a positive effect on participants' enrolling in college within one year of high school graduation, persisting to a second year of college, and starting at a four year college. For the most part, students experienced these positive effects regardless of whether they were academically proficient at the start of high school. The average annual estimated incremental cost per student for each program (the cost beyond that of providing business as usual programs) varied considerably by local education agency. AP had the lowest estimated cost per student. Concurrent enrollment had a somewhat higher estimated cost per student, though the cost reported in this study does not include the substantial expense of obtaining the graduate level training required for high school teachers who teach these courses. Dual enrollment had the highest estimated cost per student for three key reasons. First, dual enrollment typically served a very small number of participants over whom fixed costs could be spread. Second, the fees paid to postsecondary institutions represented a relatively large portion of the program cost. Finally, some local education agencies dedicated more staff time to dual enrollment to support students in registering, persisting, and ultimately succeeding. While the size of program effects was similar across the three programs, the difference in costs meant that the cost per outcome was lowest for the AP program. The study findings should be considered in the context of the population of students served by dual enrollment. Compared with concurrent enrollment and AP, dual enrollment reaches more students from populations that have been less likely to enroll in college, such as students eligible for the National School Lunch Program. These students may have different support needs. Future research could provide policymakers valuable contextual information through a more comprehensive analysis of broader societal benefits such as workforce participation that are attributable to these programs, which was beyond the scope of this study.