Date of Award

12-2011

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Legacy Department

Economics

Advisor

Mroz, Thomas

Committee Member

Maloney , Michael

Committee Member

Simon , Curtis

Committee Member

Warner , John

Abstract

With a focus on the impact of postsecondary merit aid programs, this body of research uses a variety of econometric methods to evaluate treatment effects. Since the introduction of Georgia's HOPE (Helping Outstanding Pupils Educationally) scholarship in 1993, a large number of U.S. states have introduced merit-based financial aid programs. As a growing emphasis is placed on higher education and an increased level of scrutiny is directed toward postsecondary charges, it is important to formally investigate the impact of such programs on the behavior of higher education institutions and students.

Using multiple state merit aid programs introduced over the period 1993-2000, the reaction of postsecondary institutions to the increased ability of students to pay college costs due to newly available merit-based financial aid is investigated. Empirical evidence from difference-in-difference regressions suggests that public, 4-year colleges and universities have benefited from the implementation of merit aid programs. After the introduction of widespread merit aid in a state, public institutions increase the cost of tuition relative to similar institutions in nearby, non-merit aid states. As predicted, the change in pricing is largest in states where the value of the merit aid is tied to the level of tuition and the merit eligible population is large. The impact, however, is not immediate; sizable tuition increases are typically observed to begin in the third year of a merit program.
Next, the Georgia HOPE program is used to examine the reaction of private, 4- year institutions to the introduction of a merit-based financial aid program. In contrast to an earlier study (Long 2004), this study reveals smaller relative increases (1%-2.7%) in tuition and fees and smaller relative decreases (4.3%-5.4%) in institutional aid at private, 4-year Georgia institutions after the introduction of HOPE through the use of difference-in-difference regressions. While unable to implement significant increases in tuition and fees, Georgia private institutions may have been successful at reducing internal merit awards to HOPE recipients, partially offsetting the state-provided scholarship.
While the prior two chapters focus on merit aid's impact on postsecondary institutions, merit aid surely impacts the education-related decisions of college-age individuals. Here, the impact of merit availability on high quality military enlistments is examined. Many factors influence the decision of youth to enlist in the United States Army. Among these factors are the educational benefits provided by the Montgomery GI Bill program, which offers participants substantial funds for postsecondary study. Given that publicly provided merit aid is a substitute for military educational benefits, it is expected that the introduction of a state merit aid program will reduce the number of high quality Army enlistees from that state. Using both aggregate-level and micro-level data, an OLS and probit model confirm that the presence of a state merit aid program reduces the number of black, high quality enlistees.
The final chapter of this work seeks to replicate and summarize findings presented by LaLonde (1986), Dehejia and Wahba (1999, 2002), and Smith and Todd (2005) that evaluate the ability of non- experimental data and propensity score matching to reliably estimate treatment effects in social programs. Additionally, an assessment of their results is provided and general concerns regarding the use of propensity scores to facilitate the evaluation of treatment effects are discussed. As in the previous research, propensity score estimators are applied to data from the National Supported Work (NSW) Demonstration to estimate the treatment effect on post-training earnings of participants. A number of factors are identified that may weaken the performance of the estimator.

Included in

Economics Commons

Share

COinS