Mozambique

Abstract Up‐to‐date subnational population estimates—disaggregated by age, sex, and to the lowest possible geographical level —are a key component of evidence‐based humanitarian action. In many humanitarian crises, however, large‐scale population mobility, combined with disruptions to national statistical systems, makes updating subnational population estimates challenging. Addressing these estimation challenges requires incorporating nontraditional population data sources, including mobility data generated for operational humanitarian response efforts, and addressing a range of data interoperability issues. In this paper, we present recent work by the United Nations Population Fund to update subnational population projections for Mozambique, Haiti, and Sudan by adapting applied demographic methods to a mix of traditional and nontraditional population data sources. These projections are used by UN Humanitarian Country Teams as a common population baseline for UN system‐wide decision‐making and operational response. We also highlight how cohort‐component Bayesian probabilistic projection techniques can be leveraged, alongside modeling of subnational mobility flows, to better measure and communicate uncertainty to humanitarian decision‐makers in limited‐data settings. Our work offers insights on key issues related to population data and estimation in humanitarian settings, including accounting for the impact of large‐scale crisis‐related mobility in subnational population projections.