Job Market Paper:
Refugee Urban Shelters and Locals' Electoral Outcomes: Evidence from the Venezuelan Refugee Crisis in Northern Brazil (Draft Soon)
Abstract: Refugee camps and shelters in rural and secluded areas (often paired with constraints on refugees’ rights) remain the predominant form of aid provided by developing host countries, even though 78\% of refugees worldwide reside in urban areas. Since 2014, one million Venezuelans have entered Brazil, and the border between the two countries in Roraima has become the main entry point of an unprecedented migration flow. Diverging from this "standard" reception strategy, the Brazilian government granted comprehensive rights to Venezuelans and opened 11 urban refugee shelters in different neighborhoods of Roraima's capital. Leveraging the quasi-random placement of these shelters, I investigate how this "refined" reception policy affected locals' political choices. According to the results, Brazilian voters (particularly, male, low-educated, or young) living closer to shelters increased their support for far-right presidential and gubernatorial candidates at the expense of the incumbent governor involved in shelter policy efforts. Therefore, urban shelters triggered an accountability effect combined with a shift towards far-right populist candidates. The results are mainly driven by shelters hosting Venezuelan Indigenous people (an especially vulnerable, low-integrated, and culturally distinct subgroup of the refugee population). This reveals that the establishment of urban refugee shelters itself might be less important than their hosted population's cultural and integration features for explaining locals' political backlash.
Funding from UC Davis College of Letters & Sciences Dean's Graduate Summer Support Award and UC Davis Global Affairs Grad Grants
Working papers:
Refugees and Locals' Educational Outcomes: Evidence from the Venezuelan Refugee Crisis in Northern Brazil (with Gabriel Koraicho - Stanford)
Abstract: This paper studies how the influx of school-age Venezuelan refugees affected locals' educational outcomes in Boa Vista (Brazil) after 2017. The city is the capital of a northern Brazilian state (Roraima) and the main entry point for Venezuelan migrants into the country. To answer this question, we constructed a panel of schools and students using administrative data from 2010 to 2023. First, we found that Venezuelan students were placed in schools with enrollment capacity (lower student-to-teacher and student-to-classroom ratios), which was possible because of the destruction of amenities such as reading rooms, libraries, and computer and science labs. Our analysis also revealed that migrants were disproportionately placed in the first grade, were consistently older than their Brazilian peers in grades 1 to 5, and performed worse, particularly in portuguese exams. Peer effects were, however, limited and didn't seem to have affected the retention, dropout, or exam performance of Brazilians. Further analysis using not yet explored administrative data on grade placement exams will assess how the disproportional refugee first-grade enrollment affects Venezuelan educational outcomes. It will shed light on how to integrate into the host country's public schools migrant kids who are not fluent in the local language.
Funding from UC Davis Global Affairs Grad Grants and UC Davis Economic Dept. Research Award 2023 and 2024
Impact of a Multi-level Intervention on Social Cohesion among Refugee and Host Communities in Uganda (with Susan Awor, CEGA fellow)
Abstract: Using data from 4,561 households in refugee-hosting districts of Uganda, we examine the relationship between a livelihood support intervention and social cohesion. The intervention was non-randomized and targeted toward vulnerable households through a participatory wealth ranking approach. The “poorest of the poor” category received cash for labor, while the “active poor” category received loans or grants. We hypothesize that livelihood support reduces competition over scarce resources and improves social relations within communities. The results show that participation in the intervention is associated with reduced land-related conflict. Moreover, conflict dynamics differ by household characteristics, with female-headed households and those with more school-age children more likely to report conflict. Residing in refugee settlements is associated with higher levels of land and general conflict. These findings contribute to the growing literature on interventions for peaceful coexistence among refugees and hosts, highlighting the role of targeted livelihood support.
Work in Progress:
Benefits on the Bench: Workfare, Mental Health, and the Role of the Team (with Remy Beauregard - UC Davis)
Abstract: Recent studies have found large positive effects of working compared to pure cash transfers on various measures of mental health for labor demand-constrained populations. We hypothesize that some such benefits could stem from belonging to and training with a team even if one is not ultimately selected to work, which we term as “being on the bench”. This study proposes a novel field experiment to investigate the effects of being placed on the bench by a randomized employment lottery for forcibly displaced Venezuelan migrants in Roraima, a Brazilian border state. This population is understood to have both high unemployment and job-seeking as well as poor mental health, making them both appropriate and necessary to study. As we expect these effects to operate through alignment with the meaningfulness or mission of work being trained for, we introduce variation in the meaningfulness of tasks offered. Similar to previous studies, we also include a cash transfer treatment arm and a pure control for comparison.
Bordering on Effectiveness: The Uneven Impact of Visa Policies (with Renan Chicarelli – Duke)
Abstract: This paper examines the effects of Mexico’s recent imposition of visa requirements on travelers from Brazil on both tourism inflows and irregular border crossings into the United States. The policy was introduced in response to rising numbers of migrants transiting through Mexico to reach the U.S.-Mexico border, and it aimed to better regulate entries from countries associated with increased migration flows. Using variation in the timing and rollout of the policy, we estimate its effects on two outcomes: (i) short-term tourism to Mexico and (ii) U.S. Border Patrol encounters at the southern border. Preliminary results show that the policy significantly reduced travel from Brazil, including tourism. Correspondingly, encounters of Brazilian nationals at the U.S.-Mexico border declined.
Funding from the Institute for Humane Studies at George Mason University (grant no. IHS019318)