Fueling Frictions: Nigeria Fuel Subsidy Removal and Early-Life Health

We study how removing a universal, untargeted subsidy affects healthcare utilization across the full sequence of early-life contact with the health system. Using Nigeria’s May 2023 petrol subsidy removal—a sharp, nationally unanticipated shock—we examine three outcome domains: antenatal care, birth outcomes, and childhood vaccination. Each domain uses a distinct identification design matched to data structure. For ANC and birth outcomes, we exploit variation in conception-cohort exposure: children born at different dates have different shares of their nine-month gestation window falling in the post-reform period, generating continuous within-cluster treatment variation conditional on birth-month seasonality. Post-reform exposure reduced ANC quantity and content quality, with proportionally larger effects in rural areas. Greater in-utero exposure widened urban-rural disparities in categorical birth size: rural births shifted toward smaller reported categories, while urban births shifted toward larger ones, consistent with an offsetting reduction in combustion-related PM2.5 that disproportionately benefited urban areas. For childhood vaccination, we exploit a within-child design: children born at different times face the same national immunization schedule, so comparing vaccine windows exposed before and after the reform—within the same child—controls for all child-level time-invariant unobservables. Post-reform coverage fell for vaccines requiring return clinic visits, with shortfalls expanding monotonically across the dosing schedule. Three pieces of evidence converge on disrupted physical access—not demand withdrawal—as the dominant mechanism. Rotavirus—newly added to the free national immunization schedule in 2022—and IPV, sustained by externally funded polio-eradication outreach, both maintain coverage despite the access shock. Rural households with older siblings show smaller vaccination shortfalls, consistent with transportation costs being shared across children. These results illustrate a broader pattern: the households capturing the largest implicit transfer under an untargeted subsidy need not be those bearing the largest adjustment cost when it is removed.