Sex-specific mortality forecasting for UK countries: a coherent approach

Area of Science:

• Demography
• Statistical Modeling
• Public Health

Background:

• Accurate mortality projection is crucial for actuarial and public health planning.
• Existing models often struggle with joint projections across genders and countries, leading to divergence.

Purpose of the Study:

• To introduce a novel gender-specific model for joint mortality projection in the UK.
• To improve the accuracy and coherence of mortality forecasts compared to existing methods.

Main Methods:

• Development of the 2-tier Augmented Common Factor (ACF) model.
• Utilizing a Poisson framework for direct modeling of death counts.
• Extension of the model to incorporate a cohort term for specific generational effects.

Main Results:

• The 2-tier ACF model demonstrates superior in-sample fitting compared to independent Lee-Carter or Li-Lee models.
• The model generates coherent mortality projections for genders within countries and countries within genders.
• The extended ACF model accounts for cohort effects, such as those seen in the UK's 'golden generation'.

Conclusions:

• The 2-tier ACF model offers a robust framework for joint gender- and country-specific mortality projections in the UK.
• This approach mitigates projection divergence issues inherent in independent modeling.
• The model's flexibility allows for the inclusion of cohort effects, enhancing forecast accuracy for specific populations.