Association for Public Policy Analysis & Management

States provide a substantial amount of financial support to public colleges, and recently the method states use to allocate appropriations has changed. Since 2007, 34 states have adopted outcome-based funding (OBF). OBF replaced funding models which allocated appropriations solely using enrollment levels, and instead uses outcome measures like graduation rates. States adopted OBF to create an incentive for colleges to increase the academic performance of their students. However, OBF could lead colleges to reduce admissions to student groups who are less likely to graduate.

To prevent colleges from decreasing admissions to underrepresented student groups, many OBF models included “equity measures” -- an incentive to enroll specific student groups. For example, the equity measure in Tennessee provides four-year colleges $3,600 (40%) more appropriations for graduating a low-income students. While this type of incentive is widespread and substantial, we do not know if colleges responded to them by changing their admissions practices.

Using a regression discontinuity design, I examine whether the equity measure in Tennessee’s OBF model affected the enrollment decisions of low-income students. In Tennessee, the equity measure applies to low-income students, as determined by Pell Grant eligibility. As a result of the eligibility threshold, colleges receive a substantially different amount of appropriations for very similar students. Using individual-level data from four cohorts of high school graduates, I exploit the low-income status threshold and compare the enrollment decisions for students just above and below the eligibility threshold.

Overall, the regression discontinuity estimates show no difference in whether students enroll or in their college sector choice. One explanation for the null effects is that colleges cannot respond to the premium because they already admit most students. Since colleges did not respond to the equity measure, OBF could lead to a reduction in the enrollment of low-income students. I show descriptively that this is not the case, but I cannot causally estimate the combined effect of OBF and equity measure on the enrollment of low-income students. While the equity measure did not have the intended effect on enrollment, I show that it did redistribute appropriations to colleges that serve a higher share of low-income students. Taken together, my findings show that states considering adopting OBF should consider both whether colleges are able to respond to the new incentives, and how these incentives affect the distribution of appropriations between colleges.