Thesis work


The effect of higher-stakes grades on student achievement (job market paper)

When students lack motivation or effort, educational interventions that increase the quality or quantity of school resources may have no impact on student achievement. A more effective way of raising achievement could be incentivizing students to perform well in school. In this paper, I study whether students respond to non-financial incentives for higher grades, exploiting a reform in Stockholm that made compulsory school grades the sole criteria for admission to high school. Using a difference-in-differences design, I find that the reform increased students' grade rank in compulsory school by 3.5 percentile points on average. Estimates of the unconditional quantile treatment show that the largest shifts occurred in the middle of the grade distribution, where the performance incentives are strongest. I perform a variety of checks to support the hypothesis that these effects were driven by changes in student effort rather than changes in school grading practices. My findings suggest that behavioral responses from students drive the results. Thus, strengthening the performance incentives implicit in the design of the education system can have a positive effect on student achievement.


The effect of degree selectivity on teachers' initial job placements

Teachers with stronger academic credentials tend to work in schools with students from more advantaged backgrounds. This paper contributes to an emerging literature on the mechanisms that drive these sorting patterns. With register data covering all college graduates and teachers in Chile between 2007 and 2017, I examine whether earning a more selective teaching degree has a causal effect on where graduates teach at the start of their career. For identification, I exploit a college placement mechanism that generates hundreds of admission cutoffs around which access to more selective teaching programs is essentially random. Using the variation near these cutoffs in a regression discontinuity design,  I find no evidence that degree selectivity affects the type of schools where graduates find their first teaching job. This holds for several school attributes on which teachers typically sort: location; public or private ownership; and socioeconomic composition of the student body.


Teacher credentials, per-pupil spending, and student performance in adult education

This paper studies a large-scale educational expansion to evaluate whether shocks to school inputs have an impact on student performance. I analyze the spillover effects of a Swedish policy that temporarily doubled enrollment in adult education, thus putting considerable strain on school resources. Because the intervention targeted individuals age 25 and over, my analysis focuses on individuals under age 25 to mitigate concerns that changes in student composition drive my findings. First, I establish that students in regions subject to larger enrollment shocks also experienced stronger negative shocks to school inputs like teacher credentials and per-pupil expenditure. Then, I show that the stronger negative shocks to school inputs coincided with steeper declines in course completion. Taken together, the two sets of results suggest a causal link between school inputs and course dropout.


Other works in progress

  1. Peer gender composition and student outcomes in high school
    (with Aino-Maija Aalto)

  2. From epidemic to pandemic: Did the Corona outbreak affect study choices in Sweden?
    (with Aino-Maija Aalto and Dagmar Müller)

  3. Talking about my genderation: The economics of gender transitions in Sweden
    (with Ian BurnEmma von Essen, and Ylva Moberg)

©2020 by Lucas Tilley