Association for Public Policy Analysis & Management

This study seeks to fill multiple gaps in the performance-based funding (PBF) literature by evaluating the impact on performance and efficiency goals as well as unintended consequences of the policy. While most PBF 2.0 policies allocate less than 10 percent of funds via metrics, and a couple allocate nearly all of funding through these metrics (dubbed PBF 3.0), I examine a moderate version of the policy that falls between these two extremes. I call this PBF 2.5, and use Arkansas as a case. I use a difference-in-differences approach to examine the impact of PBF 2.5 on degree production, cost efficiencies, and the potential exclusion of historically underserved students. I examine the potential for heterogeneous effect sizes across institutions in relation to each institution's level of dependence on the state for funding.

Findings reflect previous research. In response to a PBF 2.5 policy, Arkansas public colleges appear to limit the enrollment of Black and Hispanic students as well as Pell recipients. Similarly, the number of White students enrolled has increased suggesting a shift in recruitment and admissions processes to change the enrollment mix. Colleges appear to be limiting the number of students who statistically have a lower proclivity of persisting and completing their degree in order maximize institutional revenue under the PBF policy.

PBF appears to have little impact on the efficiency of degree production from an institutional or state perspective. Despite increased efficiency being an explicit goal of the Arkansas policy, the cost per credential does not appear to vary in response to the implementation of the policy. Similarly, the Arkansas PBF policy does not appear to increase degree production. Although the policy is intended to increase colleges’ focus on increasing degree attainment, there does not appear to be any effects on overall or bachelor’s degree production. There is minimal evidence that associate degree production may decrease, however the findings are inconclusive on this point.

Finally, the theoretical link between fiscal dependence and policy impacts is partially supported by my findings. Increased dependence does appear to be related to exclusionary policies and changes in enrollment mixes based on racial and ethnic lines. However, dependence does not appear to be a predictor of changes in the number of Pell recipients enrolled in a given institution. The average treatment effect of PBF 2.5 on efficiency and degree production appeared to be null, and the incorporation of fiscal dependency to test the potential for heterogeneous policy effects also drew null findings. However, the heterogeneous effects on exclusionary practices is an important consideration for policymakers. Given the evidence that fiscal dependence may impact certain behaviors of colleges, policymakers should be mindful of the potential harm changes in policy may have on highly dependent schools. Moreover, schools that are more highly dependent upon the state also tend to serve lower income students and more students of color. The disproportionate impact raises important equity concerns in moving to a performance-based funding model.