Essays in Economics of Education

Loading...
Thumbnail Image
Degree type
PhD
Graduate group
Economics
Discipline
Economics
Education
Economics
Subject
Funder
Grant number
License
Copyright date
01/01/2025
Distributor
Related resources
Author
Shure, Kristen, Beamer
Contributor
Abstract

This dissertation focuses on questions related to economics of education in three chapters. The first chapter studies the effects of competition between schools on student academic achievement and school quality. I make use of a nationwide standardized test for all students in Mexico to study this question. The other two chapters utilize U.S. data to study questions related to labor markets and schools. In the second chapter, I study the role of information frictions in early career teacher turnover through a discrete choice dynamic programming model. My third paper, in Chapter 3, examines the effects of four-day school week on parental labor supply in the western and rural U.S. using a triple difference-in-differences approach. I summarize all three papers in my dissertation in more detail below. The first paper, titled "Public vs. Private: School Quality and Competition in Mexico”, explores how increased school competition affects educational quality and student academic achievement. I characterize increased competition as a broader choice set of public and private schools in a local area. My research aims to understand whether and to what extent greater school competition leads to improved outcomes for all students or to cream-skimming to the detriment of public students. I also use my estimated model to analyze the effects of a range of actual and hypothetical educational policies on academic achievement and on inequality. To answer this question, I develop an equilibrium model where parents select schools from locally available options, while schools choose their quality attributes. This framework includes a value-added test score model that captures the influence of school quality on academic achievement and incorporates peer effects. I utilize this value-added test score model to measure school quality but also as an input in my structural estimation for test score production. I estimate the model using data from Mexico, where students are assigned to a default public school based on residence but can opt for a nearby public or private school. The analysis draws on three data sources: a national administrative database on Mexican schools, school quality metrics from a school census, and teacher wage data from the decennial Census. The study focuses on school choice at the critical transition between sixth and seventh grade, when students transition from primary school to middle school (grades 7-9). School quality estimates reveal that quality is positively correlated with school size and negatively correlated with student-teacher-ratio, aligning with much of the class-size literature. Further, I find that households have a strong distaste for far away schools. On the school side, I find that the premium for a one unit increase in teacher quality is equivalent to about half of the wage coefficient on experience. In counterfactual simulations, I evaluate the effects of competition on student performance and school quality choices by exploring policy changes that may affect local competition. I consider an alternate school choice environment where students do not have freedom to choose alternative public schools beyond their closest school. This allows for measurement of the effects of local public school competition. I also consider a private tuition voucher that discounts the price of private schools and consider a public school subsidy for teacher quality that makes hiring high-quality teachers cheaper for public schools. In all three counterfactual simulations I measure the effects on student performance and school quality. My interest in education markets extends beyond my job market paper. I study decentralized matching markets through the lens of the labor market for early career teachers in "Information Frictions and Teacher Turnover”, co-authored with fellow Penn graduate student Zach Weingarten. The job market for early career teachers exhibits high rates of instability, with 44% of new teachers leaving the profession during their first five years, and we analyze whether this can be partly explained by information frictions. We develop and estimate a dynamic, partial equilibrium model of labor mobility that incorporates non-pecuniary information frictions for school climate and teacher workload. We embed information distortions into a discrete choice dynamic programming model of early career teacher labor mobility. We estimate the model using a novel combination of U.S. data on early career teachers, including the School and Staffing Survey (2007-2008), the Beginning Teacher Longitudinal Study (2007-2010) and the Baccalaureate and Beyond Longitudinal Study (1993-2003). These data provide us with a holistic snapshot of teachers’ experiences in the profession. We reconcile our findings through the policy lens of reducing teacher turnover. We find that a counterfactual policy which improves information outperforms each alternative policy considered in terms of improving teacher retention. Alternative policies considered include targeted wage premiums at hard-to-staff schools, large retention bonuses, and relaxed tenure requirements. Replicating the gains made through information revelation requires retention bonuses valued at 45% of teachers' current salaries. My third paper titled "Who takes care of the kids? Examining parental labor under four-day school weeks” further explores the relationship between schools and labor markets. I analyze the roll-out of the four-day school week in school districts across the western and rural U.S. (2000-2019). Districts transitioned to four-day school weeks to save money in terms of busing and building maintenance, but this transition presents unique challenges to households with school-age children in terms of childcare. I examine the effect of the four-day school week transition on family labor metrics using a triple difference-in-differences framework. Triple differences allows me to control for the age of children, which may affect the intensity of households childcare demands. By emphasizing family effects, I aim to better describe how households value childcare for school-age children. Previous research on the effects of four-day school weeks examines effects on students’ academic performance but has only slightly touched upon the effects on families at large. This paper uses data from the ACS as well as new data that I collected on the implementation of four-day school weeks by school district. Using a triple difference-in-differences approach, I find that the four-day school week implementation negatively affects the total labor income for both single and married women. I find no measurable effects for single or married men. I find that the four-day school week schedule has negative effects for female labor supply.

Advisor
Todd, Petra
Date of degree
2025
Date Range for Data Collection (Start Date)
Date Range for Data Collection (End Date)
Digital Object Identifier
Series name and number
Volume number
Issue number
Publisher
Publisher DOI
Journal Issue
Comments
Recommended citation