This page gives a selection of my field projects.

Most of the projects below are experiments to learn about causal relationships, and make use of carefully designed measures to learn about behavior.

I register my research designs before data collection to avoid fishing. I also aim to make collected data publicly available as quickly as possible to aid other researchers.

DEMOCRATIC REPUBLIC OF CONGO

• The effects of emergency response. What is the effect of emergency relief to recent vulnerable populations displaced by rebel violence? We study the impacts of distributing vouchers for non-food items among over 1,000 displaced individuals. We collect data around six weeks after the distribution, and focus on four outcomes: physical health of children, adult mental health, social cohesion, and resilience. With John Quattrochi, Ghislain Bisimwa and Maarten Voors. Implemented by UNICEF and funded by 3ie.
  • Registered design
  • We also conduct a long run follow-up. We return to the communities one year later to learn about the same outcomes, plus women empowerment and migration: Registered design

 

• Improving service provision. How to improve the provision of basic services in the health and education sector? We study the impact of a development project in 339 villages that aims to strengthen the ties between service users, providers and local governments. With Chloe Fernandez, Ann Laudati and Eric Mvukiyehe. Implemented by the IRC and funded by DFID.
  • Registered design
  • Data
    
  • Video

 

• Long run impact of community-driven development. What are the long run impact of devolving financial and operational control over public goods to ordinary people? We collect data in 781 villages to learn about the impact of a large community driven development program ten years after program start. With Ann Laudati and Eric Mvukiyehe. Implemented by the IRC and funded by DFID.
  • Registered design
  • Data

 

• Resource distribution across networks. How does the position of initial recipients influence distribution of resources across a networks? We collect detailed network data on all households in 40 Congolese villages. We then distribute fertilizer and knowledge about fertilizer using different entry points based individuals’ network positions. With Jenn Larson and Maarten Voors.
  • Registered design

 

• Migration and migrant-native interactions. We collect detailed migration history and GPS information from 8,199 adults to understand migration patterns and hosting arrangements in Eastern Congo. Next, we conducted a set of novel trust and dictator games with 416 individuals to learn about the social dynamics between migrants and natives. With Neelanjan Sircar. Funded by Earth Institute’s AC4, CSDS and ASC. 
  • Registered design
  • Data

 

• Collecting conflict data using mobile phones. The Voix des Kivus project used cellphones to map local level (conflict) events in real-time from hard-to-reach areas. The project made use of “crowdseeding”, and received thousands of messages from Congo’s war torn South Kivu province. With Macartan Humphreys. Funded by USAID.
  • Project's website
  • ​Data

 

• Introducing democratic practices. The Financial Times called this “one of the world’s largest ever randomized trials”. We conduct a field experiment in 1,250 villages to learn whether a community driven development program can induce local level democratic practices. With Macartan Humphreys and Raul Sanchez. Implemented by the IRC and CARE and funded by 3IE and DFID.
  • Registered design
  • Data

SIERRA LEONE

• Community engagement and disease surveillance. In the wake of the Ebola outbreak, we work together with the Ministries of Agriculture and Health to study the impact of a One Health program. Across 350 villages, we explore whether community engagement and the introduction of community animal health workers can improve human and animal health outcomes.  With Macartan Humphreys, Niccolo Meriggi and Maarten Voors. Funded by the World Bank and the Netherlands.
  • Registered design

 

• Social status and discrimination. We conduct a set of original lab-in-the-field experiments with 750 participants in 46 villages in Sierra Leone. The goal is to learn about the importance of an individual’s position within different social networks for discriminatory behavior. With Neelanjan Sircar, Ty Turley and Maarten Voors. Funded by Cambridge University and NWO.
  • Registered design

KENYA

• Voter Registration. What are the barriers to voter registration in Africa? In 1,674 polling stations across seven counties in Kenya we implement a 2x3 factorial design that crosses two interventions. First, an informational intervention: providing citizens with basic information about voter registration via canvassing or SMS. Second, an operational intervention designed to reduce the cost of voter registration: bringing voter registration officers and technology to the village. The project is implemented together with the Independent Elections and Boundaries Commission. With Andy Harris and funded by NYU-AD.
  • Registered design