Gustave Kenedi

High-achieving, low-income students persist in higher education at lower rates than their high-income peers, raising questions about what drives these post-enrollment gaps. Most financial aid research examines the U.S. setting, despite many countries combining modest tuition with living cost support. I study a generous French merit-and-need grant using a regression discontinuity design. Because eligibility is communicated too late to affect enrollment, I can cleanly identify effects on already-enrolled students. I find precisely estimated null effects on persistence, degree completion, and graduate degree enrollment. These results suggest that high-achieving students may be less sensitive to financial aid than lower performing peers.

We provide new estimates of intergenerational income mobility in France for children born in the 1970s using rich administrative data. Since parents’ incomes are not observed, we employ a two-sample two-stage least squares estimation. We show, using the Panel Study of Income Dynamics, that this method yields reasonably conservative estimates of (rank-based) intergenerational persistence measures relative to what would be obtained if parent income were observed. Our results suggest France is characterized by a strong persistence relative to other developed countries. 9.7% of children born to parents in the bottom 20% reach the top 20% in adulthood, four times less than children from the top 20%. We uncover substantial spatial variations in intergenerational mobility across departments, comparable to those observed across countries. We find that the upward mobility gains from geographic mobility are slightly decreasing in parent income and increasing in the income level of the destination department. The expected income rank of individuals from the bottom of the parent income distribution who moved towards high-income departments is around the same as the expected income rank of individuals from the 75th percentile who stayed in their childhood department.

How do students choose where to apply to higher education? Answering this question is important because the returns to higher education differ greatly across majors and institutions. This paper examines a previously overlooked factor shaping these choices: the high school environment. Specifically, we analyze how students’ applications and enrollments are influenced by the higher education trajectory of recent graduates from their high school. Exploiting random admission cutoffs in France’s centralized admission system, we compare high schools where older schoolmates were marginally admitted versus rejected from specific degrees (subject-institution combination). Our findings reveal significant older schoolmate spillovers: students are 6 percentage points (+19%) more likely to apply to and 2 percentage points (+45%) more likely to enroll in the same degree as a marginally enrolled older schoolmate. These effects are large, corresponding to roughly 45% of spillovers across siblings estimated in other countries. We find both teacher influence and homophily/role model effects mediate these cross-cohort spillovers. Lastly, we quantify the extent to which exposure to high-achieving older schoolmates affects the socioeconomic gap in applications to very selective degrees and find that equalizing this exposure would narrow the gap by around 10%. These results demonstrate how inequalities in higher education can be perpetuated or mitigated through high school peer networks.