Probabilistic population projections with migration uncertainty.

Area of Science:

• Demography
• Population Studies
• Statistical Modeling

Background:

• Accurate population projections are crucial for global planning and resource allocation.
• Existing probabilistic population projections often rely on deterministic migration assumptions.
• Uncertainty quantification in demographic forecasting remains a key challenge.

Purpose of the Study:

• To develop probabilistic population projections incorporating uncertainty in fertility, mortality, and migration.
• To compare these projections with United Nations' probabilistic population projections.
• To assess the impact of migration uncertainty on global and regional population forecasts.

Main Methods:

• Probabilistic modeling of fertility, mortality, and migration for country-level population projections.
• Comparative analysis against the United Nations' Probabilistic Population Projections methodology.
• Out-of-sample validation to assess model calibration and predictive accuracy.

Main Results:

• Migration uncertainty is a substantial driver of population projection uncertainty for numerous countries.
• Prediction intervals for Northern America and Europe are over 70% wider due to migration uncertainty.
• Population projection uncertainty for Africa, Asia, and the global population remains largely unchanged.

Conclusions:

• Probabilistic migration projections are essential for refining population forecasts.
• The developed model demonstrates reasonable calibration through out-of-sample validation.
• Future demographic research should prioritize probabilistic approaches to migration forecasting.