[World Bank Blog Post] [Newspaper Expresso (in Portuguese)] [Voice of America TV interview (in Portuguese)]
Abstract: As insurgencies rise in fragile states and international aid diminishes, finding effective non-material strategies to reduce civilian support for armed groups has become a critical policy challenge. This paper evaluates a novel field experiment in Pemba, the capital of Cabo Delgado, Mozambique, where an ongoing Islamist insurgency has displaced over one million people and raised concerns about social fragmentation and the need to win the hearts and minds of both locals and the population internally displaced by the conflict (IDPs). Participants were randomly assigned to small cohorts that engaged in four sequenced community meetings, facilitated by religious leaders and grounded in Islamic moral framing. Sessions addressed displacement, prejudice, radicalization, and reconciliation through structured dialogue. Outcomes were measured across five survey waves using direct questions, a novel graphical list experiment, and an Implicit Association Test. Effects were strongest among locals -- especially Muslims and those without prior contact with IDPs -- who showed large reductions in support for insurgents and increased religious tolerance. Among IDPs, positive shifts emerged among those exposed to violence or with low social integration.
Abstract: We examine the long-term impact of forced labor on individual risk preferences and economic decisions. For that, we focus on a policy of coercive cotton cultivation enforced in colonial Mozambique between 1926 and 1961. We combine archival information about the boundaries of historical cotton concessions with survey data collected specifically for this study. By employing a regression discontinuity design to compare individuals living in areas inside and outside the historical cotton concessions, we document significant disparities in risk aversion and agricultural patterns between communities. Our findings indicate that individuals from regions unsuitable for cotton production, yet subjected to the colonial cotton regime, exhibit higher risk aversion, are more likely to engage in farming, sell their agricultural produce, adopt technology, and save. These findings are predominantly driven by individuals who recall the coercive cotton cultivation in their communities, particularly women who bore the brunt of the colonial cotton regime. This paper underscores the enduring impact of colonial agricultural policies on risk and economic behavior, providing insights into the challenges post-colonial societies face in overcoming historical legacies.
Abstract: We analyse the historical influence of Christian missions on civil society in Mozambique, focusing specifically on trust. To do so, we use a georeferenced dataset on the location of Catholic missions in colonial Mozambique, and geocoded individual respondent data from the Afrobarometer Survey. We show that historical proximity to Catholic missions is linked to reduced levels of generalized trust. Our results contribute to the understanding of trust and social cohesion in current-day Africa, engaging with the ample literature on the long-term effects and path dependencies created by colonial rule in the continent. This paper highlights the importance of taking into account the specific context of each former colony and the circumstances under which missions operated, as well as their specific characteristics, to understand their long-term effects on society.
This paper explores how campaign rallies affect voters' preferences over candidates and public policies. It uses data from the 2016 U.S. Presidential Election and shows that in this context rallies held by Hillary Clinton significantly improved voters' preferences for her. Rallies held by Donald Trump generated two opposing effects: a decrease in the preference for Trump in urban counties, driven by republican and democratic voters; and a positive effect in suburban counties, driven by republican voters. Trump and Clinton rallies also shifted the policy preferences of suburban voters towards supporting more right and left wing policies, respectively. An event-study analysis shows the effects of Trump and Clinton rallies were short-lived.
Can Public Service Provision Enhance Social Cohesion and Prevent Conflict? Evidence from a Large-Scale Radio-Based Educational Intervention amid an Active Insurgency in Mozambique
Joint with Helpo (implementing partner) and UNU-WIDER.
Current stage: field activities and baseline data collection starting in Spring 2025.
Funding: UNU-WIDER and ISEG - University of Lisbon.
Current stage: field activities and baseline data collection starting in Spring 2025.
Funding: J-PAL / IPA Displaced Livelihoods Initiative; International Growth Centre; UNU-WIDER; ISEG - University of Lisbon.
The Legacy of Conflict on Socioeconomic Dynamics and Human Behavior: Evidence from Portugal’s Veterans of War
With João Pereira dos Santos, João Cancela, José Tavares, Luciano Amaral.
Implementing partner: Liga dos Combatentes (Portugal's veterans association).
Current stage: nationwide survey rolling out in Spring 2025.
Funding: Portuguese Science Foundation (FCT).
Current stage: draft written.
Funding: Brown University; Luso-American Development Foundation (FLAD).
Current stage: archival work completed and dataset assembled.
Funding: Brown University; Luso-American Development Foundation (FLAD).