Methods for comparing the mortality experience of heterogeneous populations

Area of Science:

• Demography
• Biostatistics
• Actuarial Science

Background:

• Understanding population heterogeneity in mortality is crucial for accurate risk assessment.
• Existing methods may not fully account for varying age-specific mortality patterns.

Purpose of the Study:

• To develop and demonstrate methods for estimating heterogeneity in individual mortality risks.
• To adjust mortality comparisons across populations for heterogeneity and selection effects.

Main Methods:

• Utilized Maximum Likelihood Estimates (MLE) to quantify mortality risk heterogeneity.
• Applied various assumptions on age trajectories of mortality.
• Analyzed US and Swedish cohort mortality data.

Main Results:

• Consistently found considerable heterogeneity in mortality risks across all tested models.
• Demonstrated the ability to adjust mortality comparisons using estimated heterogeneity.

Conclusions:

• The developed methods provide robust estimates of mortality heterogeneity.
• Accounting for heterogeneity is essential for precise comparisons of mortality risks and selection effects.