Research

Working Paper

This paper studies the effect of state merit aid for in-state college attendance on student outcomes using national survey data on college enrollment and U.S. Census data on labor market outcomes. The econometric analysis exploits the staggered rollout of merit aid programs across states and yields evidence that the introduction of these programs led to a decrease in enrollment in relatively more selective out-of-state institutions and an increase in enrollment in relatively less selective in-state institutions. There is little evidence of consistent impacts on college degree attainment and no evidence of effects on post-college earnings for men. However, the data suggest post-college effects for women, including lower labor supply, lower earnings, and greater likelihood of having children. These results highlight the tension between states’ interest in retaining high-achieving students in the state and the distortionary effects of state merit aid on student college enrollment decisions. 

with Jing Liu and David Blazar, Working Paper

Media: Brookings, Axios, Fortune, code.org, EdSurge 

This study evaluates the impact of expanding high school Computer Science (CS) course offerings on students' college major choices and early-career labor market outcomes. Using longitudinal data from Maryland, we exploit variation from the staggered adoption of CS course offerings across schools. We find that taking a high school CS course increases students' likelihood of declaring a CS major by 8 percentage points and receiving a CS bachelor's degree by 5 percentage points. Access to CS coursework also raises students’ employment rates and early-career earnings. Notably, female, socioeconomically disadvantaged, and Black students experience similar or larger gains in CS degree attainment and labor market outcomes. However, low take-up rates of CS coursework, especially among underrepresented groups, underscore the need to expand participation in order to increase the supply of CS graduates and promote more equitable access to CS careers. 

with Nolan Pope and George Zuo, Working Paper

This paper studies how students’ K-12 academic performance relates to their long-term educational attainment and earnings. Using linked administrative data from Maryland, we estimate both descriptive correlations and causal effects via a teacher value-added framework. Both approaches show that math and English state standardized test scores are similarly predictive of broad educational attainment, but math scores are substantially more predictive of STEM degrees and earnings. Mediation analysis reveals that most of the English-earnings relationship is explained by educational attainment, while nearly half of the math-earnings relationship remains unexplained. Heterogeneity analysis confirms that math scores are more predictive of earnings across all student subgroups. However, the strength of this relationship is weaker for historically disadvantaged and lower-achieving students, while English scores show a stronger association with earnings for these same groups. These findings suggest that policies aimed at fostering economic mobility should consider differences in the strength of relationships between subject-specific skills and long-term outcomes across student groups.

with Thomas Dee, Elizabeth Huffaker, Maranna Yoder, and George Zuo, Work in Progress

The decision of whether to group (or track) students into classrooms by ability level has important consequences for student achievement, college readiness, and labor market outcomes. While research suggests benefits for tracked students, the practice may exacerbate inequality. We study automatic enrollment, or “opt-out”, policies that automatically enroll students in advanced coursework based on prior year test scores. These policies aim to broaden access to advanced coursework, particularly among under-represented groups. In this project, we study the adoption of automatic enrollment in middle school honors math classes in Dallas ISD starting in the 2019-2020 school year. We use regression discontinuity and differences-in-differences methods to study the impact of the policy on test scores, future course-taking patterns, and higher education outcomes for both tracked and untracked students in Texas. 

Work in Progress

This study investigates how teachers serve as role models who influence students' postsecondary and labor market outcomes. Using longitudinal student-level data from Maryland, I use value-added methods to assess teacher quality based on their impacts on college outcomes. The analysis explores teacher effects by subject area, whether teachers encourage students to pursue studies in fields related to their subject area, and how these factors influence employment and earnings. Evaluation of mechanisms explores whether teachers' long-term impacts operate through their impacts on students' test scores and behavior, or through shared demographic characteristics with their students. This study will inform policy discussions on college and career readiness as well as teacher quality and diversity.