Three essays on economics of education

Abstract

This dissertation includes three chapters. In Chapter 1, I estimate the effects of gender and race/ethnicity matching between students and high school math and science teachers on students' STEM enrollment and completion in colleges. My sample includes over 100,000 Missouri high school graduates entering a four-year public university in Missouri between 2001 and 2010. Potential bias from non-random teacher-student sorting within and across high schools is mitigated by examining student exposure to teaching staff broadly, and the use of within-high school variation for identification. I find no evidence of matching effects on students' postsecondary STEM outcomes. My findings for gender matching are more precise than for race/ethnicity and rule out modestly positive impacts. In Chapter 2, I estimate the effects of nursing articulation agreements between the University of Missouri-Columbia (MU) and numerous public community colleges on the likelihood that community college students transfer to the MU four-year nursing degree program. I use difference-in-differences specifications that leverage variation in the availability of nursing articulation agreements with MU across community colleges and over time to identify the effects of these agreements. I supplement these models with triple-difference specifications that additionally leverage variation in transfer rates across initial majors in community colleges. I find no statistical evidence that nursing articulation agreements affect student transfer rates. Supplementary analyses indicate that my null findings are not driven by supply constraints in the MU nursing program and instead reflect the failure of the articulation agreements to spur demand for nursing education among community college students. In Chapter 3, we examine how teachers from two alternative preparation programs—Teach for America (TFA) and Kansas City Teacher Residency (KCTR)—contribute to the teacher labor market in and around Kansas City, Missouri. We show that TFA and KCTR teachers are more likely than other teachers to work in charter schools, and more broadly, in schools with high concentrations of low-income, low-performing, and underrepresented minority (Black and Hispanic) students. TFA and KCTR teachers are themselves more racially/ethnically diverse than the larger local-area teaching workforce, but only KCTR teachers are more diverse than teachers in the same districts in which they work. In math in grades 4-8, we find sizeable, positive impacts of TFA and KCTR teachers on test-score growth relative to non-program teachers. We also estimate positive impacts on test-score growth in English Language Arts (ELA) for teachers from both programs, but our ELA estimates are smaller in magnitude.