Research
Publications
-
Beyond Targeting Women: Intrahousehold Contexts and the Effectiveness of Food Vouchers
-
The Effects of Nutrition Support on Behavioral Outcomes and Labor Productivity
-
The Effect of Conservation Policies on Wildlife Hunting and Consumption in North-eastern Madagascar
-
Accounting for Nutritional Changes in Six Success Stories: A Regression-Decomposition Approach
-
Drivers of Nutritional Change in Four South Asian Countries: A Dynamic Observational Analysis
Working papers
-
Maternal psychological distress, even below clinical thresholds, may constrain caregiving and child health. We randomize a five-session psychotherapy program integrated into community-based acute malnutrition treatment in Gombe, Nigeria, among 792 distressed but largely subclinical caregivers of severely malnourished children. Psychotherapy increases children’s weight-for-height by 0.21 standard deviations and mid-upper arm circumference by 0.17 cm, and suggestively reduces relapse readmissions by 39 percent. It modestly improves maternal mental health by 0.14 standard deviations, with the largest effect on anxiety. Mothers spend nearly an hour more per day on childcare, and intimate partner violence declines. Our results suggest that integrating psychological support into child health programs can improve both maternal and child outcomes.
-
Organic Farming, Intrahousehold Labor, and Productivity: Experimental Evidence from the Philippines
Agricultural technology interventions can reshape the gendered division of labor within farm households, yet whether these shifts improve or worsen productivity remains poorly understood. We conduct a randomized controlled trial among 600 vegetable farming households in the Philippines to evaluate the causal impact of organic vegetable technology adoption on crop productivity and intrahousehold labor allocation. Training and starter kits significantly increased organic technology adoption in both the short and medium run. In particular, the treatment increased adoption of labor-intensive homemade fertilizers and pesticides among poor households. We do not find any significant treatment effect on the aggregate number of workers or person-days. However, aggregate effects mask wealth-dependent heterogeneity in female labor. Pre-registered analyses reveal that among poor households, organic farming increased overall female labor, driven by time dedicated to fertilizer and pesticide application. In contrast, wealthier households maintained aggregate female labor but reallocated it away from these stages, suggesting they purchased market-ready organic inputs. While we find no overall effects on crop revenue or farm profitability, production function estimates indicate that treatment reduced the marginal product of female labor for profit—with suggestive evidence that this decline is concentrated among poorer households. These findings highlight how agricultural technology adoption can widen the existing gender productivity gap among poorer households by changing production processes.
-
The Effects of a Community-driven Nutrition Program and Block Grants on Household and Child Nutrition: Evidence from a Field Experiment in Madagascar
Chronic child undernutrition remains widespread in developing countries, and many of its determinants—from social norms around child feeding to the local disease environment and shared water and sanitation—are community-level conditions that household-focused interventions leave largely unaddressed. We conducted a cluster randomized controlled trial of a one-year child nutrition-focused community mobilization program, with and without block grants, in Madagascar. We find that pairing community mobilization with block grants improved household food security, water and sanitation, and women’s empowerment, along with children’s feeding practices and, suggestively, their growth. Community mobilization alone had limited effects. A community-driven approach can thus improve both the household and community environment and children’s own nutrition, underscoring the value of engaging the whole community rather than caregivers alone.
-
Time and Risk Preferences and Preventive Health Behaviors: Panel Evidence from Rural Malawi
Low take-up of preventive health products in low-income countries is often attributed to impatience. We ask whether time and risk preferences are consistently associated with preventive behavior, and whether long-run discounting and present bias are linked to it in the same direction. Using a three-wave panel of women of reproductive age in rural Malawi, we relate discount rates, present bias, and risk aversion, elicited through unincentivized multiple price list and investment tasks, to malaria and HIV/AIDS prevention. Individual fixed effects identify within-person associations, which we probe with lagged specifications and post-double-selection LASSO. Greater impatience is associated with more bed-net use, condom availability, and HIV education. Present bias is correlated with the opposite pattern for treated nets and HIV education, with impatient reversers 10.4 percentage points less likely to use an insecticide-treated net and 11.3 points less likely to report HIV education. Risk aversion is linked to greater condom availability and higher perceived HIV risk from a husband. Impatience is therefore not a uniform barrier to prevention, and distinguishing long-run discounting from present bias matters for how preventive tools are designed and delivered.
Selected work in progress
-
Complementing Peer Nutrition Groups with Community Mobilization: Evidence from a Field Experiment in Niger
-
Labor-saving Technology and Farm Household Welfare: Experimental Evidence from Direct Seeded Rice in India