Developing an Original Probabilistic Fertility Projection Model for Nigeria’s Sub-National Regions

Standard demographic projection methods produce deterministic fertility scenarios that understate forecast uncertainty, and developing an original probabilistic fertility projection model calibrated to the heterogeneous demographic contexts of Nigeria's six geopolitical zones provides both a methodological and empirical contribution to sub-national demography. This study developed an original Probabilistic Fertility Projection Model for Nigeria's Sub-National Regions, drawing on Bayesian demographic forecasting theory, time-series analysis of Nigerian TFR data, and expert opinion elicitation. A model development methodology was employed, combining systematic review of probabilistic demographic projection methods (52 publications from 2018 to 2024), analysis of NDHS TFR estimates from 1990 to 2021 for each geopolitical zone, Bayesian hierarchical model specification, and expert elicitation of probabilistic fertility trajectory assessments from 18 Nigerian demographers. The Bayesian hierarchical model treated each geopolitical zone as a partially exchangeable unit sharing prior information on fertility trend shape while allowing zone-specific deviations. Projection posterior distributions produced 80 percent credible intervals for zone-level TFR to 2040, with North West and North East zones showing the widest uncertainty ranges due to slower and more variable fertility transition. The model confirmed that divergent zone trajectories would likely persist to 2040, with North West TFR 2.8 points higher than South West TFR under the central scenario. Calibration validation confirmed well-specified uncertainty intervals. Expert review confirmed the model's original methodological contribution to Nigerian sub-national demographic projection practice.

Keywords: probabilistic fertility projection, Bayesian model, Nigeria sub-national, geopolitical zones, demographic forecasting