This doctoral project uses field experiments and original survey data across three settings in Sub-Saharan Africa to study how insititutions, social networks and organizational allocation influence cooperation, investment, and collective decision-making. The studies show that formalizing verbal land-rental agreements into written contracts sharply reduces contracting frictions and increases tenants' perceived land access security, that the committees governing grassroots churches allocate resources in ways that diverge systematically from ordinary congregants' preferences, and that farmers' choice between a marketing cooperative and an independent trader is driven mainly by liquidity needs rather than social connections.
The project examines how contract design, organizational rules, and social ties shape economic decision-making in low-income settings, specifically, how formalizing informal contracts, allocating resources within collective institutions, and positioning within social networks affect investment, cooperation, and welfare among poorer and more vulnerable households. Across three studies, I study land access as a core household asset and how written contracts facilitate land transactions in rural areas, how grassroots churches allocate socially relevant resources and thereby shape informal redistribution, and why cocoa farmers choose to sell through formal cooperatives or private middlemen. A common mechanism across the three settings is that when contracts remain informal, or when organizational rules and enforcement are tilted toward better-off members, poorer and more vulnerable households face greater risk, invest less, and struggle to accumulate assets. Most of this work relies on original data that I collect through surveys, field experiments, and lab-in-the-field experiments.
The study "From Handshake to Contract: Evidence from Land-Use Agreements in Uganda" (with Arthur Laroche) asks whether formalizing verbal land-rental agreements into written contracts reduces contracting frictions and strengthens tenure security, in a setting where refugee households in Uganda's West Nile region rent farmland from Ugandan landlords almost exclusively through undocumented verbal agreements. In a randomized controlled trial with 514 landlord-tenant pairs, written contracts endorsed by local authorities reduce disagreement over key terms by roughly 80 percent, make agreements substantially more likely to specify essential provisions such as duration, and raise tenants' perceived land access security by 0.76 standard deviations, with suggestive evidence of longer investment horizons.
The study "How Do Churches Maximize? Study Design and Pilot Evidence from Kenya" (with Lorenzo Casaburi and Amma Panin) studies decision-making within decentralized Pentecostal churches, where a small elected committee typically allocates resources on behalf of a much larger congregation. Using an incentivized choice experiment run with committees and regular members separately and jointly across 37 churches, the study finds that committee and congregant preferences over donation options align in only half of churches: regular members favor school fees while committees favor a sound system, and committees show a markedly stronger preference for commitment devices. The pilot documents several underlying mechanisms, including divergent church priorities, differing beliefs about returns, and committees' misperception of congregants' preferences.
The study "Marketing-Channel Choice and Cooperative Preferences Among Cocoa Farmers in Côte d'Ivoire" (with Federico Cammelli, Mingmin Feng, Norina Furrer, Luca Lazzaro, and Radu Tanase) asks why cocoa farmers sell through a marketing cooperative rather than an independent itinerant trader, in a setting where the state-regulated farm-gate price leaves buyers to compete mainly on non-price attributes. Combining a survey of 377 farmers, a mapping of five village social networks, and a choice-based conjoint experiment completed by 321 of them, the study finds that farmers selling outside the cooperative system are distinguished mainly by individual financial circumstances, such as lacking a bank account or greater risk tolerance, rather than by social connections, and that stated cooperative preferences are dominated by access to credit and certification. Social-network position matters in narrow rather than general ways: farmers more central in the village borrowing network place less value on cooperative credit, consistent with informal networks substituting for formal credit access.
