Text

theses

Expanding access to quality education is important to both policymakers and the nation as a whole. Although many policies have been designed to address inequality in educational opportunities, research does not always support their ability to intervene effectively. Two such policies intended to increase educational opportunities are introducing competition in public school districts and providing access to postsecondary schooling via community colleges. My dissertation examines the effects of these interventions with respect to a recent cohort of students in a period of economic instability. In the first chapter of my dissertation, I examine the effect competition has on public schools in Massachusetts using publicly available data from the Massachusetts Department of Education. With eight years of data, from school year 2007-2008 to school year 2014-2015, I am able to construct a panel data set of public high schools. I define test scores as an outcome, using both statewide graduation requirements and the SAT. During this time frame, Massachusetts used the Massachusetts Comprehensive Assessment System (MCAS), which is comprised of three separate exams: Mathematics, English/Language Arts, and, starting in 2008, Science/Technology. As a measure of competitive influence, I construct the ratio of charter and private school enrollments to public school enrollments by district. Using a model which allows for school district fixed effects, I find little evidence of positive effects of competition from charter schools on public school student test performance. Increased private school penetration has negative effects on Math MCAS scores and positive impacts on Math SAT scores, both of which are driven by the suburban districts. Because this model heavily relies on the within-district variation to identify the impact, as an alternative, I employ a control function approach. This method allows for more flexibility in the model and identifies the effect using heteroskedasticity in the error terms. Because alternative school locations are not random, I first estimate a Tobit model to predict the alternative enrollment ratio. Using this in the second stage together with the control variable generated from the error terms, I estimate the effect of alternative enrollment on public school exam scores. Results from this approach indicate a positive impact of alternative schools on public school test performance. As these two methods are quite different, it is not surprising that they produce different results. The main concern moving forward is identifying the appropriate model. At this point, all I can conclude is that the effect of competition is highly dependent upon model choice, which in itself is fruitful for the literature. In my second and third chapters, I make use of the Education Longitudinal Study of 2002 (ELS: 2002) data set, produced by the National Center for Education Statistics, to analyze the effect of attending two-year colleges on eventual educational attainment and labor market outcomes. The NCES sampled 750 high schools, and within each school, about 30 students were randomly chosen to participate. Students were surveyed beginning in 2002, as high school sophomores, through 2012, eight years after most graduated high school. Compared to students who began at four-year colleges, students who started at two-year colleges earned about thirty fewer postsecondary credits, one fewer year of education, and were twenty-five percentage points less likely to earn a bachelor's degree. Further, I examine whether the gap varies by racial, socioeconomic, or academic differences. I find mixed results by racial groups, although Hispanic students seem to be most negatively impacted by starting at a two-year college, whereas Asian students do not seem to be impacted at all. Additionally, low income students who start at two-year colleges are less likely to earn a baccalaureate degree if they begin their postsecondary career at a two-year college instead of a four-year college. Finally, students with high school GPAs over 3.0 are disproportionately hurt by attending community colleges as they are less likely to earn a bachelor's degree, relative to their peers with high school GPAs above 3.0 who began at a four-year college. Finally, I consider the effects of educational choices in the labor market. By 2012, almost a third of the ELS respondents had completed some college but had not earned a degree. I find no significant difference in wages or employment status in 2012 between high school graduates and students with some postsecondary attendance but no degree. Further, there are benefits to earning a certificate or associate degree over some two or four-year college credits. I find that males see negative labor market returns in terms of income when earning some college credits without earning a degree, relative to students with a high school degree or GED. Women see no impact from earning some credits in two-year colleges, but have positive returns to some four-year college credits relative to high school graduates. For both genders, the wage benefits of earning a bachelor's degree range from twenty-two to sixty-four percent by age twenty-six. The range of returns is larger than earlier studies, but still suggests that earning a bachelor's degree provides the greatest wage benefits, even in an unstable economy and among those early in their career.

ETD_7422

Ph.D.

Includes bibliographical references

by Jessica Munz Scheld

doi:10.7282/T33N25PC

ETD doctoral

The author owns the copyright to this work.