Publications
Does Conflict Reshape the Military–FDI Nexus? Evidence From a Dynamic Panel Analysis, with Narendra Regmi and Krishna Sharma, Review of International Economics (2026).
COVID-19 Vaccine Hesitancy by Smoking Status Among Ohio Adults, with Teferra, A. A., Nau, M., Tosun, L., Sahr, T. R., Freedner, N., & Ferketich, A. K., Ohio Journal of Public Health (2023).
Working Paper
Why give up the ability to choose? Evidence from a Nepalese land allocation program, with Abraham Holland and Bhisma K Bhusal.
Abstract: In theory, individuals should prefer a choice-based lottery such as random serial dictatorship (RSD) to a uniform random lottery (URL) when allocating private goods. However, household surveys, an information intervention, and a lab-in-the-field experiment from Nepal’s post-earthquake land allocation program show that almost half of participants prefer URL despite its weaker match quality. We find no evidence that comprehension, risk aversion, social pressure, or altruism explain this pattern. Instead, ambiguity about others’ choices and concerns about procedural integrity appear more plausible explanations. Our findings suggest that in contexts where state capacity is weak, allocation mechanism design must consider not only efficiency but transparency.
Beyond Generations: Old Age Allowance and Girls’ Educational Investment in Nepal, with Narendra Regmi, R&R at World Development.
Abstract: We study the intergenerational effects of Nepal’s Old Age Allowance (OAA), a universal non-contributory pension, on children’s educational expenditure. Using a fuzzy regression discontinuity design based on the program’s age-70 eligibility rule, we find a strong gender asymmetry: OAA receipt increases annual educational expenditure for female students by more than Rs 4,000, while the corresponding effect for male students is small and statistically insignificant. The increase for girls is concentrated in tuition and uniforms, indicating that additional household resources are used to cover formal schooling costs. Results by recipient gender do not support a simple female-control bargaining channel; instead, the evidence is more consistent with a financial-constraints interpretation in which daughters’ schooling is the margin most affected by limited household resources. Heterogeneity analysis further supports this view, with stronger effects in mixed-gender households, among relatively better-off rural households, and among households located farther from primary schools. The results suggest that social pensions can generate meaningful intergenerational spillovers and reduce gender disparities in educational investment even when they are not explicitly targeted to children.