Research

Despite an increase in women's educational attainment, they remain largely underrepresented in science, technology, engineering, and mathematics (STEM), especially in elite institutions. This paper examines the gender gap among highly selected STEM students in France. Using detailed administrative data from various sources, we show that although female students self-select more than male students to enter selective undergraduate STEM programs, they underperform in the competitive entrance examinations for elite STEM graduate schools, even when conditional on their academic achievement. We also observe an increasing gender gap over the time spent in undergraduate STEM programs, which is the main driving force behind the gender gap in access to elite STEM graduate schools. We thus explore the role of the learning environment during exam preparation. Specifically, we leverage variations in learning environments created by tracking within these selective undergraduate STEM programs to examine how heightened competitiveness during exam preparation affects gender performance gaps. Our findings reveal that increased competitiveness during exam preparation--measured either on average through a double-difference analysis within programs, through an event-study design before and after tracking, or at the margin of admission to the top track using a regression discontinuity design--exacerbates the gender gap in admissions to the most selective elite STEM graduate schools. These insights deepen our understanding of gender disparities in access to selective STEM higher education and shed light on the persistent gender pay gap observed at the top of the income distribution.

Notice relative au traitement des données collectées (RGPD CNRS) / Information about the processing of collected data (GDPR CNRS)

This study examines the impact of students’ relative rank in their 9th-grade class on their subsequent educational trajectory in France. In an education system where students are frequently ranked based on academic performance, these rankings can shape self-perception, motivation, and long-term outcomes. Leveraging exhaustive administrative data from 2014 to 2019 and naturally occurring variations in performance distributions across middle school classes, I analyze how class rank influences students’ high school track choices and academic performance. The findings indicate that higher relative rank in 9th grade increases the likelihood of enrolling in an academic high school track and completing the academic high school graduation exam (HSGE) on time. Additionally, at the intensive margin, students who were relatively highly ranked in mathematics (resp. in French) in middle school achieve significantly higher national percentile scores in mathematics (resp. in French) at the HSGE. However, this effect comes with negative spillovers, as stronger ranking in one subject is associated with lower performance in the other. These rank effects are highly heterogeneous: they are almost entirely driven by male and low socio-economic status (SES) students, suggesting that these groups are particularly sensitive to peer comparisons.

This project examines the critical issue of reintegrating formerly incarcerated minors into education and/or the labor market. In an unprecedented effort in France, we use and match administrative datasets from the Ministries of Justice, Education, and Labor, facilitating a robust longitudinal analysis. Our approach includes an econometric investigation aimed at isolating the causal effects of incarceration on educational and employment outcomes, controlling for individual fixed effects and socio-demographic factors. By doing so, we aim to study the underlying mechanisms that determine whether these young individuals can successfully reintegrate into society, shedding light on broader socio-economic implications and policy interventions.