research

We study the effects of a health insurance subsidy in Ghana, where mandates are not enforceable. We randomly provide different levels of subsidy (1/3, 2/3, and full) and evaluate the impact at seven months and three years after the intervention. We find that a one-time subsidy increased insurance enrollment for all groups in both the short and long runs, but health care utilization in the long run increased only for the partial subsidy group. We find supportive evidence that ex-post behavioral responses rather than ex-ante selective enrollment explain the long-run health care utilization results.

In settings where resistance and rampant misinformation against vaccines exist, the prospect of containing infectious diseases remains a challenge. Can delivery of information regarding the benefits of vaccination through personal home visits by local ambassadors increase vaccine uptake? We conduct a door-to-door randomized information campaign targeted towards COVID-19 unvaccinated individuals in rural Indonesia. We recruited ambassadors from local villages tasked to deliver information about COVID-19 vaccines and promote vaccination through one-on-one meetings, using an interpersonal behavioral change communication approach. To investigate which type of ambassador—health cadres, influential individuals, and laypersons—is the most effective, we randomly vary the type of ambassador that delivers the information at the village level. We find that the overall vaccination take-up is quite moderate and that there are no differences in vaccination outcomes across the treatment groups. These results highlight the challenge of boosting vaccine uptake in late stages of a pandemic.

The paper measures the effect of child market work on the long-term growth of human capital, focusing on the output of the human capital production: mathematics skills, cognitive skills, pulmonary function, and educational attainment. Our full sample is drawn from a rich longitudinal dataset Indonesia Family Life Survey (IFLS). We address endogeneity of child market work using provincial legislated minimum wage as the instrument. Our instrumental variable estimation shows that child labor negatively affects mathematics skills and pulmonary function, but not cognitive skills and educational attainment. We find heterogeneities in type of work. Those who work outside of family business have lower educational attainment than those working for family business.

Shifting beliefs and encouraging pro-sustainability behaviors to mitigate climate and environmental issues can be challenging due to their polarizing nature. This challenge is particularly concerning in Jakarta, the world's fastest sinking city, where many residents remain uninformed about land subsidence—its causes, severity, and implications. We conduct a large-scale online experiment to understand how to effectively communicate this environmental threat. We vary the perceived identity of the messenger (as a religious leader or scientist) and the narrative style of the video message (religious or scientific). Our results show that any version of message, compared to a placebo, shifts beliefs about causes and consequences of subsidence, increases pro-sustainability behaviors, self-efficacy, and institutional trust in addressing the issue. The messenger's perceived identity as a scientist---rated as more persuasive and trustworthy than an Imam---generates larger impacts on beliefs regardless of narrative style. A scientist delivering a religious narrative is effective in encouraging participants to spread awareness, while an Imam is better at building trust across stakeholders. Effects on beliefs are more pronounced among those with low prior knowledge, high institutional trust, and less reliance on groundwater. However, heterogeneous treatment effects on actions are limited even among the least informed. Overall, our findings demonstrate how perceived identity and narrative framing shapes public understanding of and action on environmental challenges.

Encouraging COVID-19 vaccination among hesitant individuals is vital for achieving herd immunity and curbing the spread of the virus. We conducted a large-scale randomized experiment in India to evaluate five interventions targeting individuals that remained unvaccinated after the initial mass vaccination campaign: (i) information only; (ii) information plus a gift worth $5 upon vaccination; (iii) information plus a 5% chance to win a gift worth $100 upon vaccination; (iv) information combined with improved vaccine accessibility; (v) information disseminated through community leaders. Our findings indicate that all interventions significantly increased first-dose vaccine uptake, with the guaranteed in-kind gift and accessibility interventions showing the largest effects. The ambassador intervention was particularly effective in promoting second-dose completion. Selection analysis reveals that economically disadvantaged individuals and those with limited access to transportation benefited most from these interventions. We find suggestive evidence of spillover effects on social networks, including friends and neighbors, highlighting the broader community benefits of targeted vaccination campaigns.