Opportunity Bound: Transport and Access to College in a Megacity

This paper examines the causal effects of new transportation infrastructure on college enrollment, choice, completion, and early labor market outcomes. To estimate these effects, I use novel geolocated administrative data and a difference-in-differences strategy that exploits the rollout of two new public transportation lines in Lima, a megacity of 12 million people. My findings indicate that a 17% reduction in commuting time to college increases enrollment rates by 6%, primarily driven by private college enrollment. Moreover, female students influenced by this policy tend to enroll in low-quality private colleges, which are also connected to the new lines. The results are exacerbated when a female student resides in an unsafe district, as measured by the feminicide rates in her locality. In contrast, male students are more likely to enroll in public colleges, which are more dispersed throughout the city. Using a model of college choice, I find that for one standard deviation increase in wage returns, male students are willing to commute up to 55% more minutes than female students. In the medium and long run, access to transport increases an individual's likelihood of graduating from college by 12% and access to white-collar jobs by 6%. These results suggest that while improved transportation can increase human capital accumulation, the increase in opportunities is limited by gender differences in willingness to commute.

New* Paper Draft (Click Here)

Presented at:  AEFP, Seminario MAP, WEAI Graduate Student Workshop, UEA North America (University of Toronto), Columbia,  PUCP's Development Conference, NEUDC (Harvard University), SEA, LAUrban LACEA Rio de Janeiro, PUC Chile, Bucknell, Oberlin, Rutgers, Tecnologico de Monterrey, Wake Forest, ifo Institut, 8th Urbanization and Poverty Reduction Research Conference - World Bank, 2nd Annual World Bank Conference on Transport Economics, SOLE, PEA Annual Conference (U de Lima), CREED (UNC Charlotte), LACEA-LAMES Uruguay, Barnard College, Montana State.

Media and Policy: VoxDev, IGC Blog, CGD Blog, REDES Policy Brief.

College Licensing and Reputation Effects on the Labor Market

With Jose Flor-Toro and Matteo Magnaricotte

Working Paper (click here)

Presented at: NEUDC (2022), 9th Conference for Young Economists (UBC-UDEP), LACEA-LAMES (2022),  AEFP (2023), LAC-PaL Virtual Seminar (2025)

We study how better information on the quality of college education affects graduates' labor market outcomes. Between 2015 and 2021, Peru evaluated its existing colleges and awarded operating licenses to the 94 institutions meeting minimum quality standards. Using administrative labor market data and a staggered difference-in-differences approach, we find positive effects of positive news about graduates' human capital: within one year of the licensing announcement, wages increase by 8%, employment by 7%, hours worked by 8%, and the likelihood of being employed in a large firm and the public sector by 6% and 5%, respectively.  Most effects are concentrated among graduates with shorter or no tenure at their current job, while we don't find signicant effects for workers with longer tenure. This suggests that uncertainty about the productivity of workers is reduced over time, with public signals affecting workers' welfare.

Empowering the future of Science: Experimental evidence onSTEM skills and attitudes in Peru

* A previous version was titled: "Creating future female scientists: Experimental evidence on improving STEM skills and attitudes in Peru"

Presented at:  AEFP (2021),  Viernes Economico PUCP, UNMSM.

I study a program designed to reduce the gender gap in STEM fields by providing a bundled intervention targeting the abilities and aspirations of girls in primary school in a large city in Latin America. The program provides weekly science workshops which include lectures with top scientists, after-school activities and mentorship. I find that the program has no detectable effects on girls’ academic performance in school, but sizable effects on several other outcomes, including confidence in their science abilities, perceptions of non-STEM majors, and time use. I also document a trade-off between after-school study time and time spent on personal projects, suggesting a relative gain  in productivity given the null effects on academic performance and negative effects on effort.

Working Paper (click here)

Citizenship Policy and the Spread of Communicable Diseases: Evidence from the Dominican Republic

With Eduardo Campillo Betancourt and Jose Flor-Toro

Presented at:  NEUDC (Northeastern), SEA (2024)

Media: Vodou Economics

We study a controversial policy in the Dominican Republic in 2013 that targeted as much as 10% of the country's population based on their foreign ancestry and limited their safe access to services such as health. Beyond the direct negative effects such policies may have on the targeted group, we argue that there may be important indirect effects from such policies through the contagion of communicable diseases. We exploit the timing and differential exposure to this policy across the country, as well as highly disaggregated epidemiological data on diseases. Our difference-in-difference strategy provides evidence of a notable increase in the caseload of Dengue, a highly contagious disease. In contrast, there are no effects for noncommunicable diseases. We provide additional evidence across different specifications and when excluding outliers such as large urban centers and border regions. We also find that the effects are not driven by dengue suitability or in dengue case reporting by authorities. We argue that these results are due to a restriction in access to health services.

Working Paper (click here)