Economics ETDs

Publication Date

Spring 5-12-2018


Human capital production is central to economic wellbeing from a national perspective: it improves productivity, spurs technological innovation, and promotes sustainable economic growth over time. Equally important, investments in human capital are central to economic wellbeing at the individual level. College graduates tend to earn more money, are more employable, are better able to manage economic downturns, and have even been shown to have better health. Moreover, for many students, the college experience is an important lesson in living independently, developing social and professional networks, and generally taking on more responsibility in one’s life. It has been said that pursuing a degree in higher education is the largest investment one can make for the future. This work employs a variety of empirical strategies to better understand how students make choices regarding college: where to attend, what to study, and whether to work during college or not, for example. Chapter 1 provides background on two contentious issues in American higher education: the changing structure of financial aid and the longstanding trend of lengthening time to baccalaureate degree in the United States. Traditional financial aid has included subsidized and non-subsidized loans, need-based financial aid (e.g., Federal Pell Grant Program), and targeted merit-based financial aid, such as academic or athletic scholarships. However, the early 1990s saw the advent of a new type of financial aid: broad-based, state merit-based aid scholarships. I provide background information on how such scholarships are generally structured, paying special attention to New Mexico’s program, the New Mexico Legislative Lottery Scholarship (NMLLS). Chapter 1 then discusses longstanding trends in the time it takes undergraduate students to earn bachelor’s degrees in the United States. Focus is targeted at the potential costs of this phenomenon to state budgets, institutions, and the students themselves. Chapter 1 concludes by discussing responses to this trend by lawmakers and higher education officials, and whether such responses are warranted. Using a rich administrative data set from New Mexico’s flagship university, Chapter 2 examines whether the NMLLS resulted in any meaningful change in 4-, 5-, and 6-year completion rates comparing qualified resident students to nonqualified nonresident students before and after the program became effective. Propensity score matching is performed to mitigate any observable differences between resident and nonresidents. We find no overall completion effects in the aggregate, but do find economically meaningful divergent completion effects by academic preparation. It appears that results may be driven by students from families whose financial constraints are binding. Chapter three investigates whether college graduates suffer a wage penalty in the labor market for taking longer than traditional standards to complete an undergraduate degree. We view this as a test of whether employers view time to degree as a productivity signal in the labor market. According to human capital theory, there should be no wage penalty if students are accumulating the same amount of credits at a slower rate. However, if employers view lengthened time to degree as a negative productivity signal, then one would expect a wage penalty. Previous literature has found large and statistically significant wage penalties associated with lengthened time to degree, but we are not convinced they have adequately addressed the endogeneity of time to degree in the student wage equation. When we address endogeneity by controlling for institutional quality, student ability, and instrumenting for the student’s own time to degree with the average time to degree at their institution, we find no evidence that employers view lengthened time to degree as a negative productivity signal. Chapter 4 uses the same data set and methods as Chapter 2, but examines students’ choice of majors. I argue that one potential unintended consequence of broad-based merit scholarships is to discourage students from attempting more difficult majors, such as STEM, in order to maximize the likelihood of scholarship retaining. This may be particularly true for marginally academically prepared students. I find no evidence that either the likelihood of first majoring in a STEM field or earning a degree in STEM is affected in the aggregate by the enactment of the NMLLS. Statistically significant effects emerge when disaggregating by academic preparation, however—academically less prepared students are less likely to pursue a STEM major as a result of the NMLLS, while more academically prepared students are more likely to first major in a STEM field. This is in accordance with the previous literature as well as the theoretical model offered in the paper. Chapter 5 concludes, with a summary of main results from Chapters 2 through 4, with special attention being paid to the policy implication of these findings. Chapter 5 also offers suggestions for future work.

Degree Name


Level of Degree


Department Name

Department of Economics

First Committee Member (Chair)

Melissa Binder

Second Committee Member

Alok Bohara

Third Committee Member

Xiaoxue Li

Fourth Committee Member

Nancy Lopez




time to degree, merit aid, college completion, major choice, signaling

Document Type