The politics and challenges of measuring mortality for humanitarian decision-making in Somalia, the Central African Republic and Bangladesh

Excess mortality is a fundamental metric of humanitarian crisis severity and should guide aid prioritisation. Despite intense political interest in mortality measurements, few studies have examined the political and institutional contexts in which mortality data are produced and used. To explore the challenges and opportunities of using evidence for humanitarian decision-making, we examined the political and organisational circumstances surrounding how mortality data are measured and used in three crisis contexts: Somalia, Central African Republic (CAR), and Coxs Bazar in Bangladesh. We used a political economy approach to analyse information from 23 national level interviews, document review, and observations at global mortality estimation meetings. Insecurity and bureaucratic obstacles to measuring mortality in areas outside government control created data coverage gaps in CAR and Somalia. Combined with the effects of censorship in CAR, these gaps likely limit humanitarian actors understanding of interventions that might be needed. Across settings, new crises brought additional financing and innovations in measurement, though they could not always be sustained. United Nations coordination systems and programmes provided the backbone for regular measurement work and new initiatives, but data sharing challenges created inefficiencies. In Somalia and CAR, mortality measurement needs to be scaled and better coordinated to influence decisions. In the context of widespread aid cuts, crisis-affected governments may take increasing ownership of certain areas of mortality measurement work, but the political nature of mortality data means that independent actors and funding are needed to ensure data are collected, counter censorship, and build capacities of diverse stakeholders.