Research
Working Papers
Late Bloomers? the Aggregate Implications of Getting Education Later in Life With Zsófia Bárány and Moshe Buchinsky CEPR Discussion Paper, DP18579 NBER Working Paper 31874
It is generally agreed upon that most individuals who acquire a college degree do so in their early 20s. Despite this consensus, we show that in the US from the 1930 birth cohort onwards a large fraction – around 20% – of college graduates obtained their degree after age 30. We explore the implications of this phenomenon. First, we show that these so called late bloomers have significantly contributed to the narrowing of gender and racial gaps in the college share, despite the general widening of the racial gap. Second, late bloomers are responsible for more than half of the increase in the aggregate college share from 1960 onwards. Finally, we show that the returns to having a college degree vary depending on the age at graduation. Ignoring the existence of late bloomers therefore leads to a significant underestimation of the returns to college education for those finishing college in their early 20s.Education Expansion, Sorting, and the Decreasing Wage Premium
This paper has received the ‘Honorable Mention’ Award at the 2022 IAAE Conference.
This paper studies the interplay between education expansion and workers and firms sorting in Portugal between 1987 and 2017. The Portuguese labor market is characterized by three facts: a decreasing high school wage premium, a dramatic increase in supply of high school graduates, and an increasingly unbalanced distribution of high school graduates across industries. To quantify the impact of the latter two on the former, I build a model of one-to-many matching where workers sort with firms based on their own preferences, their relative productivity within the firm, and substitution patterns with other workers. Using tool from the optimal transport literature, I solve the model and structurally estimate it on matched employer-employee data. Estimates suggest changes in sorting are mainly driven by heterogeneous increase in relative productivity of high school graduates relative to non graduates across industries. It acts as a mitigating force on the decreasing high school wage premium, but does not fully compensate for high school graduates' rise in relative supply.The data was kindly provided by the Instituto Nacional de Estatistica with the help of Francesco Franco at Nova SBE in Lisbon.
The Decreasing Returns to Experience for Higher Education Graduates: an Occupational Analysis New Draft
This article evidences the flattening of wage dynamics for higher education graduates in France between the cohorts entering the labor market in 1998 versus 2010. Postgraduates in the latter cohorts enjoy 10% less labor income over their first seven years on the labor market than postgraduates in the former cohort. The difference in average yearly wage growth is 2.2 percentage points between the two cohorts. I decompose variations in wage growth between cohorts by occupation, and find a strong relation between the influx of new graduates in an occupation and the flattening of wage progression in that same occupation, suggesting congestion is causing the flattening. I show two mechanisms are at play behind congestion: access to managerial positions, which reduced over time, and field of study-occupation mismatch, which bears an increasing weight on young graduates' wages over time.This project in conducted as part of a CEREQ workgroup (in French).
Work in Progress
Repeated Matching Games: an Empirical Framework, with Jeremy Fox and Alfred Galichon
On the Estimation of Scaling parameters in Separable Matching Models, with Arnaud Dupuy