[Online Appendix] [Replication Package]
[Online Appendix] [WP]
[WP]
Abstract: Women account for only 35% of global STEM graduates, a share unchanged for a decade. We use administrative microdata from centralized university admissions in ten systems to deliver the first cross-national decomposition of the STEM gender gap into a pipeline gap (academic preparedness) and a choice gap (first-choice field conditional on eligibility). In deferred-acceptance platforms where eligibility is score-based, we isolate preferences from access. The pipeline gap varies widely, from -19 to +31 percentage points across education systems. By contrast, the choice gap is remarkably stable: high-scoring women are 25 percentage points less likely than men to rank STEM first.
Abstract: This paper examines an early childhood, income-based affirmative action policy that allocates school admission points to children from low-income households based on two income thresholds. These thresholds create sharp discontinuities that we exploit using a regression discontinuity design. Our analysis draws on the universe of school applications submitted at age three, linked to academic and well-being outcomes nine years later. We find that the affirmative action policy increases both the number and quality of feasible school options available to low-income families and raises the likelihood of admission to their top-choice schools. However, families do not systematically rank higher–value-added schools or those with better average test scores as their top choices. Instead, the policy shifts admissions toward charter schools with slightly higher socioeconomic composition. Nine years after application, we find no significant effects on standardized test scores or student well-being. These findings suggest that early childhood affirmative action broaden educational options for low-income families but has limited long-term impacts.
Abstract: Across many settings, quasi-experimental studies find negligible effects of smaller classes on student learning. This is puzzling: teachers and parents favor smaller classes, and economists and education researchers have proposed mechanisms through which they should improve student outcomes - reduced classroom disruption (Lazear, 2001), greater use of individualized instruction (e.g., one-on-one help with math), and greater effectiveness of this instruction due to teachers’ increased capacity to provide it (i.e., fewer students to help). We address this puzzle using administrative and survey data from Madrid, a large school district in which class size varies quasi-experimentally. We confirm that class size has negligible effects on student learning and then ask: (i) do smaller classes activate the proposed mechanisms? and (ii) do offsetting behavioral responses - such as teachers focusing on student enjoyment, or reductions in student effort or parent involvement - help explain the results? We find that class size has only modest effects on classroom disruption and the use of individualized instruction, and negligible effects on the effectiveness of individualized instruction. While we detect some evidence of offsetting behavioral responses, we calculate that even in their absence, smaller classes would have limited effects on student learning.
(Draft coming soon)
Abstract: Gender occupational segregation remains a central barrier to closing the earnings gap, with women underrepresented in high-paying, math-intensive careers. We study whether evaluation environments shape early gender gaps using population-level administrative records covering all 6th-grade students in the Region of Madrid (Spain). We exploit multiple rounds of centralized school-level randomization that assigned identical, blind-graded standardized tests to be administered either internally by school staff or externally by invigilators, under otherwise identical testing conditions. External administration widens the mathematics gender gap by 0.06 standard deviations, with no corresponding effect in verbal performance. To examine potential mechanisms, we link administrative records to contemporaneous student-level attitudes and beliefs from a post-test survey. External testing increases girls’ reported mathematics stress relative to boys by approximately 18%, with no corresponding differential effects on self-confidence or effort.
The effect of university field of study on civic behaviours ( with Adam Altmejd and Luis Cornago). [Pre-Analysis Plan]
2) "El efecto de los compañeros de clase" Economía de la Educación. Editorial Fundación Ramón Areces. 2024
NOTE: Policy Reports are available upon request