Uncertain population forecasting.

Area of Science:

• Demography
• Population Studies
• Statistical Modeling

Background:

• Population forecasts are crucial for planning but are subject to inherent errors.
• Errors originate from initial population data and inaccuracies in predicting future vital rates.
• The linear Leslie growth model is a standard tool for population projection.

Purpose of the Study:

• To analyze error propagation in population forecasts using the Leslie growth model.
• To develop statistically sound prediction intervals for future population sizes.
• To compare model-derived intervals with existing forecast uncertainty measures, such as U.S. Census Bureau intervals.

Main Methods:

• Utilizing the linear Leslie growth model to simulate error propagation.
• Deriving vital rate predictions from a parametric statistical model.
• Employing mixed estimation to incorporate expert opinion and robust regression to assess model misspecification.

Main Results:

• Quantified the impact of initial population and vital rate errors on forecast accuracy.
• Developed prediction intervals that provide a range for future population estimates.
• Demonstrated the utility of the developed intervals in assessing forecast reliability.

Conclusions:

• Errors in initial population and vital rates significantly influence forecast accuracy.
• The developed prediction intervals offer a robust method for quantifying population forecast uncertainty.
• The methodology provides a framework for improving the reliability of demographic projections.