Probabilistic population forecasting: Short to very long-term

Area of Science:

• Demography
• Climate Change Economics
• Bayesian Statistics

Background:

• Population forecasts are crucial for governmental and private sector planning, traditionally using deterministic scenarios.
• Probabilistic forecasts are increasingly desired for accuracy assessment and risk-based decision-making.
• The United Nations has provided probabilistic population forecasts since 2015 using a Bayesian methodology.

Purpose of the Study:

• To review the United Nations' Bayesian methodology for probabilistic population forecasts.
• To extend existing methods for very-long range population forecasts up to 2300.
• To inform long-term projections of carbon emissions and the social cost of carbon.

Main Methods:

• Review of the United Nations' Bayesian probabilistic population forecasting methodology.
• Extension of the UN method by integrating expert review and elicitation with statistical approaches.
• Application to generate population forecasts beyond 2100 to 2300.

Main Results:

• World population is projected to grow through the 21st century.
• Population is expected to stabilize during the 22nd century.
• A decline in global population is projected for the 23rd century.

Conclusions:

• The enhanced methodology provides crucial long-term population projections necessary for climate change assessments.
• Extending probabilistic forecasting to 2300 improves the accuracy of social cost of carbon calculations.
• Probabilistic population forecasting is essential for informed long-range planning and risk assessment.