Research

This paper examines how social envy influences workplace productivity through a field experiment with sugarcane mill workers in Kakamega, Kenya. In settings where conspicuous productivity can trigger resentment, high-effort workers may intentionally reduce effort to avoid social punishment. I randomize payment conditions across 600 mill workers to vary the visibility of earnings and whether top performers are publicly highlighted. By comparing private versus public payment, and varying types of public recognition, I test whether conspicuous production reduces output due to envy, distinct from other peer effects such as social taxation.

This paper investigates how income inequality drives jealousy and sanctioning behavior in collectivist communities, focusing on the role of informal redistribution networks. In rural Kenya, where envy-related punishment is common, I use a randomized controlled trial with sugarcane mill workers to test whether workplace bonuses provoke sabotage, and whether sharing norms mitigate such effects. Participants are paired based on income equality and social ties, and outcomes include jealousy, punitive behavior, and life satisfaction. The study offers causal evidence on how envy and redistribution norms shape economic behavior and social cohesion amid inequality. 

Adverse childhood experiences are a widespread but under-recognized barrier to learning, especially in low-resource settings where access to mental health care is limited and barriers to education are high. Because trauma undermines emotional regulation, cognition, and school participation, there is a pressing need for low-cost, scalable interventions such as group-based programs that leverage peer support. This paper evaluates a randomized controlled trial in Zimbabwe, testing two components: (1) teacher training in trauma-sensitive pedagogy and (2) a peer-based student intervention, Freedom Clubs. In these clubs, students meet weekly to hear structured audio dramas depicting common traumatic experiences, followed by guided reflection and socio-emotional skill-building led by trained facilitators. Schools present a complex setting for such programs, as peer dynamics can amplify or suppress outcomes, particularly under stigma. I develop a theoretical model in which social connectivity, trauma exposure, and stigma-driven revelation costs interact to produce non-linear treatment effects by group composition. Consistent with the model, program impacts are strongest when peer closeness is either high or low, and weakest at moderate levels. Additionally, Freedom Clubs are most impactful for students with higher trauma exposure, despite higher reputational risk. Across groups, Freedom Clubs produce meaningful improvements in psychosocial outcomes on the intensive margin: students who attend more sessions experience reductions in PTSD and complex PTSD symptoms and more progressive gender attitudes. By contrast, teacher training alone yields few consistent gains, with some evidence of longer-run negative effects. Freedom Clubs are also associated with improved school attendance, though no measurable gains in exam performance are observed. Overall, the findings underscore the role of social context, stigma, and group structure in shaping trauma-informed interventions in schools.