National Identity-Seeded Verified Relationship Graphs: A Graph Neural Network Framework for Population Estimation in Data-Scarce Sub-Saharan Africa

Abstract Reliable population data are essential to equitable governance and development planning. Nigeria, Africa's most populous nation, has operated without a nationally accepted census for nearly two decades, producing a substantial demographic knowledge deficit affecting an estimated 220 to 235 million persons. Existing methodological responses, including satellite-based population modelling and microcensus extrapolation, estimate aggregate distributions rather than verifying individual existences and remain vulnerable to systematic over- and undercounting incentives. This paper introduces the National Identification Number-Seeded Verified Relationship Graph (NS-VRG), a novel computational demography framework designed to estimate national population through biometrically verified kinship-network traversal. The framework transforms Nigeria's National Identification Number infrastructure, administered by the National Identity Management Commission, into the anchor population of a directed relationship graph. Graph nodes are classified into three tiers: biometrically confirmed Anchor Nodes, relationally declared Semi-Verified Nodes, and Graph Neural Network-inferred Ghost Nodes. A population estimation formula integrates all three tiers with a satellite-derived coverage adjustment factor. The framework addresses seven structural limitations of the existing National Identification Number database: absent relational structure, underage under-enrolment, lack of continuous updating, duplicate records, and absence of mortality tracking. A federated privacy-preserving storage model, citizen-facing application interface, and required API extension endpoints are fully specified. NS-VRG positions Nigeria as the first country globally to estimate national population through a verified biometric kinship graph. This paper advances computational demography by demonstrating that biometric identity infrastructure can serve as the foundation of a living national population register, with assessed replication potential across African Union member states operating biometric identity systems.