Association for Public Policy Analysis & Management

There are over 200 promise programs (e.g., place-based financial aid programs) today, with 16 states operating at least one program statewide (Mishory, 2018). Their proliferation has spurred examination of design (Perna & Leigh, 2018; Taylor & Lepper, 2018), impact on student outcomes (Carruthers & Fox, 2016), and role in economic development (Miller-Adams, 2018). However, none have explored the effect of these programs on the ultimate distribution of aid (e.g., federal, state, institutional) within a state. We address a gap by exploring changes in aid at Tennessee’s community colleges after the introduction of Tennessee Promise. Prior research has cited the “fiscal appeal” of a last-dollar design to state governments because other sources of aid are leveraged first before scholarship expenditures (Wardrop, Divringi, & DeMaria, 2018, p. 2). Therefore, while additional state expenditures may occur to support the new grant program, we expect to observe large increases in federal aid (e.g., Pell) following the introduction of Promise.

We construct a panel dataset from IPEDS spanning academic years 2010-11 through 2016-17. We employ descriptive and difference-in-differences methods to estimate changes in total financial aid to institutions and the average financial aid award for first-time, full-time (FTFT) students. We also explore changes in the composition of aid across federal, state, and institutional sources. Our full sample encompasses 768 institutions across the 7-year time horizon. Counterfactual institutional comparison groups to the community colleges in Tennessee consist of all other community colleges in the U.S., the SREB region, and in contiguous states.

Descriptively, overall financial aid per FTFT student rose after the introduction of Promise from $4,374 in 2014-15 to $4,964 in 2015-16, an average increase of $590 compared to a national average decline of $119. Importantly, this increase was driven by increases in state aid alone ($844). In fact, average Pell and institutional aid per FTFT student declined by $218 and $38, respectively. This suggests the composition of aid changed substantially, with the state now carrying a greater share. To further examine these changes, we apply a DD model across the entire time horizon, including select covariates and college- and year-fixed effects, for our three outcomes of interest. Findings support a statistically-significant increase in average state aid per FTFT student after the introduction of Promise ($1,074, p < .001) and suggestive evidence (p < 0.10) that total Pell per student increased modestly ($246) and institutional aid fell ($82). These findings are consistent across counterfactual groups and robustness checks.

Our preliminary evidence suggests that Tennessee Promise did not lead to a meaningful increase of federal or institutional financial support after accounting for enrollment and demographic changes, but did significantly (p < .001) increase the state’s financial aid contribution. These findings suggest the last-dollar program may not have leveraged additional federal Pell funding per student as typically hypothesized (Perna, Leigh, & Carroll, 2018; Wardrop, Divringi, & DeMaria, 2018). These preliminary results beg additional research examining the fiscal sustainability of statewide promise programs and important tradeoffs between first- and last-dollar program designs.