Working papers

When Organized Crime Moves In: Economic and Human Capital Disruption  (with Andrea Bocchino) [Working paper available on SSRN] 

This paper studies how organized crime affect local communities and human capital formation. Identifying these effects is challenging, as crime is endogenous to local conditions. We address this by leveraging the recent case of Ecuador, where criminal organizations from neighboring countries have rapidly established a new cocaine export route. This externally driven shock generated sharp increases in violent crime, allowing us to estimate causal effects using a difference-in-differences design based on proximity to areas prone to cocaine smuggling. Crime-affected areas experienced higher dropout rates among children at grades characterized by weak school attachment, the end of primary education and the first years of secondary school. While we do not find evidence of increased dropout among older students aged 15-18, individuals in this age group already out of education at the time of the crime surge exhibited a marked rise in risky behaviors, reflected in higher homicide victimization and earlier pregnancies. We also document severe economic disruption: household income fell by nearly 30%, driven mainly by a decline in informal employment. Declining earnings are a key mechanism linking crime exposure to school dropout. These findings show that the externalities of organized crime impose persistent social costs, deepening inequality and undermining human capital development. 

Exam Luck and Human Capital Accumulation (with Catalina Franco) [Working paper available on SSRN] 

Reject & Resubmit, Journal of Political Economy Microeconomics

We show that subtle standardized test design features shape long-term educational trajectories. Exploiting random variation in correct answer placement in a Colombian college entrance exam, we find that students are 5% less likely to answer correctly when the correct option appears last. This seemingly minor pattern is economically consequential: “unlucky” booklets with more Ds in math lower overall scores by 0.011 SD and first-choice admissions by 3%, diverting marginal students from a selective public university into lower-return vocational programs. The mechanism is consistent with students overlooking options at the bottom of the list when scanning answers sequentially under time pressure.

Work in progress