Working Papers

"Reparations as Development? Evidence from Victims of the Colombian Armed Conflict," with Juliana Londoño-Vélez and Christian Posso. Revised and resubmitted at the American Economic Review.

Abstract: We study the effects of reparations for victims of armed conflict in Colombia. More than one million victims of forced displacement, homicide, and other conflict-related harms have received large lump-sum payments over the past decade. Using linked administrative microdata and difference-in-differences designs, we find that reparations improve the well-being of victims and their children. While modestly reducing formal employment, they increase entrepreneurship and investment in land, housing, and other durable assets, improve health and reduce acute-care utilization, and increase children’s postsecondary enrollment and test scores. Projected fiscal offsets reduce program costs, implying that benefits can exceed net fiscal costs.

"The Early-Career Effects of Mental Illnesst," with Kaveh Danesh and Christian Posso.

Abstract: Mental illness tends to arise when young workers are establishing their positions in the labor market. Using nationwide employer-employee records and medical claims data from Colombia, we use a matched difference-in-differences design to estimate the long-run effects of common mental health disorders on earnings. Among a cohort of young, formally-employed workers, newly diagnosed anxiety and depression reduce earnings for at least five years following diagnosis. In the short run, earnings losses mainly reflect fewer days worked; in the longer run, they reflect lower employment and wages. Part of the earnings effect is driven by reduced employment at firms with high wage premia, consistent with impaired mobility up the job ladder. We find no evidence of worsening mental health following plausibly exogenous job loss, suggesting that reverse causality is unlikely to bias our results. To examine whether mental health treatment mitigates the earnings penalty induced by mental illness, we use an instrumental variables design that leverages variation across areas and insurers in the probability of referral for specialty care. Preliminary findings suggest that treatment offsets most of the earnings penalty.

Published and accepted work

"Luck of the Draw: The Causal Effect of Physicians on Birth Outcomes," with Christian Posso, Estefania Saravia and Jorge Tamayo.  Accepted at the American Economic Journal: Applied Economics.

"Conflict, Social Protection, and Reparations," with Juliana Londoño-Vélez. The Handbook of Social Protection: Evidence and New Directions for Low- and Middle-Income Countries, edited by Rema Hanna and Ben Olken, March 2026, 381-396. Cambridge, MA: MIT Press. [Open Access]

"Reparations for Victims: Lessons from Colombia," with Juliana Londoño-Vélez and Christian Posso. AEA Papers and Proceedings, May 2023, 113: 342-46.

,"Vocational Training for Disadvantaged Youth in Colombia: A Long-Term Follow-Up," with Orazio Attanasio, Carlos Medina, and Costas Meghir. American Economic Journal: Applied Economics, April 2017, Vol. 9, No. 2. [Online Appendix]

Selected work in progress

"Healing the Gender Gap: The Impacts of Randomized first-job on Female physicians," with Christian Posso, Estefania Saravia and Jorge Tamayo.

“Can Reparations Help Victims Rebuild Their Lives? Evidence from Colombia using Survey Data,” with Juliana Londoño-Vélez.

"Measuring the Incidence and Reporting of Violence Against Women in Public Spaces," with Girija Borker, Samuel Likindikoki, and Ester Steven.

"Gender Targeting and the Psychosocial and Economic Effects of Livelihood Programs in a Fragile Setting," with Lasse Brune, Dean Karlan, and Florence Kondylis.

"Social Protection Effectiveness and Conflict: Evidence from a Randomized Experiment in Yemen," with Nausheen Khan, and Sikandra Kurdi.