Essays on the Economics of Education

Abstract

This dissertation consists of three papers. In each paper, I leverage a different quantitative method to study a downstream effect of technology on educational outcomes. In the first paper, I study the impact of an online math learning program on 3rd through 5th grade math achievement in Louisiana. Employing Zearn Math usage metrics and administrative data from the Louisiana Department of Education, I find that grade-levels that programmatically used Zearn Math scored, on average, about 0.03 standard deviation units higher on Louisiana’s statewide math assessment. A placebo test using English language arts scores and several robustness checks suggest this may be an underestimate of the true effect. The findings offer timely information on curricular decision-making as pandemic-related school closures have led to a twentyfold increase in public school spending on Zearn-related materials between July 2019 and July 2021 and its presence continues to expand across the nation. The second paper, co-authored with Mary Laski, synthesizes previous theories explaining declines in teacher quality and expands on the current literature by incorporating a useful comparison group: the nursing workforce. We document historical trends in skill level, average and relative wages, wage dispersion, unionization rates, and quantity, and find important divergences in the teaching and nursing professions that cannot be explained by previous theories. We posit two new theories that align with our documented trends: technological innovation and occupational differentiation in nursing. We argue that trends in the nursing profession indicate that declines in teaching quality were (and are) not inevitable. The third paper examines the impact of instructional format and tutors on student outcomes in the context of an early literacy tutoring initiative. It presents the first within-program experimental evidence comparing the impact of in-person versus remote tutoring. Overall, we do not find statistically significant differences in students’ literacy outcomes between in-person and remote tutoring. However, substantial variation in student learning outcomes was attributable to differences across tutors, with the overall tutor effect estimated to be larger than that of instructional format. Additionally, students receiving in-person tutoring exhibited higher attendance rates and tutors reported closer relationships with their in-person students. These results suggest that while differences between in-person and remote tutoring may exist, the benefits of having a proficient tutor far outweigh these disparities.