Research
Work in Progress
Neighborly Business: Labor Exchange Networks and Wage Markets in Rural Thailand Job Market Paper
Why does unpaid reciprocal labor persist after rural wage markets develop? To answer this question, we study the allocation of household labor following surprise increases in crop returns. Using monthly bilateral paid and unpaid labor transactions from 16 Thai villages, we build a Bartik-type shift-share measure of crop-return innovations using unexpected changes in world prices and pre-set village crop specialization, interacted with each crop’s harvest timing. An unexpected increase in crop prices near harvest time reallocates both labor exchange and hiring toward established partners relative to previously untied potential providers. The findings show that established relationships help mobilize labor when timely agricultural work becomes more valuable and provide new evidence on the role of relational infrastructure in organizing paid work in rural labor markets.
The download link below always points to the latest version of the paper.
Harvesting Hope: Weather Shocks and Religious Expenditure in Sub-Saharan Africa Work in Progress
With Eva Raiber
We examine how weather shocks affect religious contributions in Nigeria using household panel data from the Nigeria General Household Survey (2010–2019), monthly SPEI, and Climate Hazards Group InfraRed Precipitation with Station data (CHIRPS). Severe drought and pluvial shocks have opposite effects on religious giving. Among households for whom these shocks are unusual, at least two severe drought months raise the probability of giving by 9.3 to 10.1 percentage points, about one quarter of the mean, while a severe pluvial month lowers it by 5.0 to 6.9 percentage points. The drought response is larger among Christian and urban households. The findings suggest wide nonlinearity in household response, which depends on the type of weather anomaly and exposure window, in addition to possible institutional differences in mobilization.
Under the Radar: Measuring Temporary Mobility in Household Surveys — Evidence from Malawi Work in Progress
Temporary mobility is an important margin of adjustment to income and environmental risk, but what can standard surveys tell us about household mobility and migration, and what do they miss? We study short-term and short-distance mobility in five cross-sections of the Malawi Integrated Household Survey (IHS) between 2004/05 and 2024/25. Exploiting the temporal representativeness of the survey’s fieldwork, we document variation in household size and composition across interview months in rural and urban areas, indicating that the population covered by individual absence statistics varies with interview timing. The patterns are driven by both adult and child counts and variation in child mobility is at least as large as that of adults’ in every rural round and in three urban rounds. We compare annual absence with casual work locations and reported migration after shocks and find that only 12.4 percent of rural adults and 14.4 percent of urban adults who report casual work outside the village or town are recorded as absent, and of 37 households reporting shock-related migration, 29 have no listed member recorded as absent. The analysis points to important gaps in what standard household surveys can infer about mobility in developing settings.
Prevention of Violence Against Women and Girls Through A School-Based Curriculum: Evidence from Kenya Work in Progress
With Ahmed Mushfiq Mobarak, Sreyashi Sharmin, and Junxing Gu (Yale University)
In Kenya, one-third of girls aged 13–17 experience sexual violence. This human rights violation creates substantial barriers to educational attainment and economic mobility. We evaluate the Empowerment Transformation Training (ETT) program, a school-based intervention implemented by Ujamaa Africa, using a large-scale randomized controlled trial. The program delivers gender-specific curricula to Grade 4–12 students to prevent violence against women and girls and promote gender equity.
We estimate causal effects on five primary outcomes: sexual assault incidence, school attendance, academic performance, teen pregnancy/marriage, and personal initiative. Our analysis examines economic mechanisms including human capital formation, social norms, and community externalities. The findings aim to provide evidence for cost-effective prevention strategies and to inform policy decisions on national program scale-up.