Benefit Spillovers and Higher Education Financing: An Empirical Analysis of Brain Drain and State-Level Investment in Public Universities
By Chinmany G. Pandit
This paper analyzes the impact of out-migration of college graduates on state higher education investment. A three-stage least squares regression model with state and year fixed effects is developed and estimated, addressing the relationship between state legislative appropriations, tuition, and educated out-migration across 49 U.S. states from 2006-2015. The results support the notion that states respond negatively to benefit spillovers in higher education: for every one percent increase in the rate of educated out-migration, state appropriations decrease by 1.92 percent (roughly $140 per student). These findings suggest that an education subsidy
provided to states may be necessary to prevent underinvestment in higher education.
Advisor: Thomas Nechyba | JEL Codes: H7, H75, I22, I28, R23
Early Identification of Students at Risk for High School Dropout
By Derek Lindsey
For years, many have hoped to identify why high school students drop out. Typically, studies focus on factors identified in high school or middle school. By tracking a cohort of North Carolina students from third grade onward, we attempt to identify areas for intervention even earlier in order to prevent dropouts. Indeed, we find that variables that can be viewed as indicators of high risk for drop out in middle school are already measurably present as early as third grade. This suggests interventions can begin when students are still very young and when treatment is likely to be more effective.
Advisor: Thomas Nechyba | JEL Codes: I2, I20 | Tagged: Dropout, Education, Elementary School, Graduation, High School, Middle School
An Assessment of Teach for America Effectiveness and Spillover Effects in North Carolina
By Thomas Burr
Teach for America, while a relatively small cog in the grand scheme of education reform in America, has become something of a flashpoint for debate between the educational establishment and a new generation of reformers. In the first part of this research, I add to a growing number of studies on the effectiveness of TFA teachers by preforming regression analysis of student outcomes in grades 3-5 in North Carolina from 1995-2009 and find that, as measured by end of grade (EOG) math and reading test scores, first-year TFA teachers produce gains that are statistically indistinguishable from experienced teachers and approximately .09 standard deviations higher than other first-year teachers in math and .05 standard deviations higher in reading. In the second part of this research, I build off of Jackson and Bruegmann (2009), who for the first time showed evidence of peer effects between teachers, meaning that the outcomes of your own students can be affected by the quality of the other teachers in your grade. After confirming the results of Jackson/Bruegmann with three additional years of data, I add TFA status as an additional observable characteristic into the equation and find a statistically significant and positive effect to having a peer TFA teacher in your grade across several models.
Advisor: Thomas Nechyba | JEL Codes: I2, J24 | Tagged: Education, Peer Effects, Spillover, Teach for America
Tracking Decisions in North Carolina’s Public High Schools
By Michael Harris
This paper analyzes the criteria employed to assign students into tracked English and Mathematics classes across public high schools in North Carolina. Specifically, I examine the probability of high track placement moving from eighth grade to ninth grade classrooms based upon both achievement and demographic factors. Analysis is performed at both the school and district level. Although student performance does affect placement at both levels, there are other personal characteristics that are significant factors in determining track assignment. The main finding is that being black has a positive effect on high track placement at the district level, but a negative effect at the school level. The former appears to be linked to residential segregation, while the latter suggests a within-school bias that has important policy implications.
Advisor: Thomas Nechyba