Decision Making and Governance within Churches: Evidence from Kenya (with Lorenzo Casaburi and Amma Panin)
Religious institutions in Sub-Saharan Africa are central to community life, yet little is known about how they decide on resource allocation. In Kenya, where 97.5% of the population identifies with a religion and 80% attend weekly services, we study decision-making in Evangelical and Pentecostal churches in Busia County. We implemented a pilot choice experiment in 37 churches, comparing the preferences of committees and regular members over four potential large-scale donations: recruitment events, a sound system, school fees, and theological training. Five committee members and ten regular members per church made incentivized choices individually, in groups, and jointly. Regular members favored school fees, while the committees preferred the other items. Differences arose from distinct goals, beliefs about returns, misperceptions, and limited delegation. Our results shed light on governance dynamics in decentralized religious organizations, with implications for church-led development.
Advancing Social Sensing to Enable Network Interventions in the Field (with René Algesheimer, Federico Cammelli, Mingmin Feng, Norina Furrer, Luca Lazzaro, Manuel Sebastian Mariani, and Radu Tanase)
Many policy-relevant behaviors – ranging from technology adoption to norm change and collective action – spread through social networks. A key question is how to design seeding interventions that trigger large-scale diffusion through social networks. While research shows that network structure and centrality shape diffusion effectiveness, applying these insights in practice remains difficult because collecting detailed network data is costly, time-intensive, and often infeasible. This study develops new techniques to estimate key individual and network-level variables for informing the design of seeding interventions. We test these methods in the field and evaluate their accuracy. Our approach will enable the design of network-based interventions in data-scarce settings and broaden the scope for real-world applications of social network theory.