The Causal Impact of Dual Credit Programs on College Students' Outcomes: Evidence From Texas

• In Event: Marriott Roundtable Session 17
  In Roundtable Session: Two Birds With One Stone? An Examination of the Effectiveness and the Implementation of Dual-Credit Education Programs

Abstract

As dual-credit (DC) programs become more prevalent, so too have debates about their effectiveness and efficiency in promoting postsecondary access and success. The expansion of DC education has been predicated on several untested assumptions, namely that it exposes high school students to the academic and behavioral expectations of college, reduces college attendance costs, and encourages students to commit to college before they graduate from high school (Bailey, Hughes, & Karp, 2002). However, some worry that these claims may be unfounded, expressing concerns that proponents of DC education are overselling its benefits. While these concerns have historically centered on questioning the academic rigor of DC courses, they have more recently focused on the transferability of college credits earned through DC programs, and the ability of DC programs to equalize educational opportunities for underserved students (Bailey, Hughes, & Karp, 2002; Gewertz, 2016; McGee, 2016).

Undoubtedly, testing the extent that DC programs affect student outcomes is important, but like many interventions in education, it is difficult because of problems with self-selection. Typically, the types of students that meet requirements needed to enroll in DC courses are those who already demonstrate high academic performance and other indicators of college readiness. Because of this, it is challenging to separate the effects of DC programs from the effects of students’ pre-existing characteristics on student outcomes. To date, only two studies (Edmunds, Unlu, Glennie, Bernstein, Fesler, Furey, & Arshavsky, 2016; Berger, Turk-Bicakci, Garet, & Hoshen, 2014) have effectively addressed this problem but have done so within the context of studying early college high schools, a particular model of DC education. Because of this narrow focus, we know very little about whether students respond positively to DC education as a general education intervention in the multiple ways it is delivered in practice.

This paper provides the first causal evidence on the effectiveness of DC programs on measures of academic achievement and efficiency. The study draws on over two million merged student-level administrative records from the Texas Education Agency and the Texas Higher Education Coordinating Board on eight cohorts of high school students who started their junior year between the fall of 1999 and 2006. We use an instrumental variable estimation strategy embedded within a differences in differences framework to estimate the effects of enrolling in DC education on high school graduation; college enrollment, persistence, and completion; and credits earned and time to degree. We disaggregate these analyses by student subgroups to understand the extent that students of varying characteristics are equitably served by DC offerings. Indeed, findings from initial Ordinary Least Squares analyses suggest that while DC students outperformed their non-DC counterparts on a wide range of outcomes, they were also more likely to be White and identify as academically gifted, suggesting that more advantaged students select into DC programs. Causal findings will offer insights into the potential of DC education to boost student success and increase efficient degree completion, information which is useful for policymakers to have as they consider options to expand or reform DC programs.

Authors

• Scott Delhommer, The University of Texas at Austin

• Holly Kosiewicz, Texas Higher Education Coordinating Board

• Trey Miller, American Institutes for Research

• Richard Murphy, The University of Texas - Austin