The (Im)precision of Life Expectancy Numbers

Area of Science:

• Biostatistics
• Public Health
• Demography

Background:

• Life expectancy is a key indicator of national health and social systems.
• Current methods for quantifying statistical imprecision in life expectancy are model-based.
• Published life expectancy figures often lack reporting of their statistical uncertainty.

Purpose of the Study:

• To introduce a more intuitive, data-based method for calculating the standard error (SE) of life expectancy.
• To extend the jackknife method for analyzing event rates more broadly.
• To describe the relationship between SE magnitude and the underlying data (person-years and deaths).

Main Methods:

• Development of a data-based standard error (SE) calculation for life expectancy.
• Application of the jackknife method to event rate analysis.
• Analysis of the correlation between SE and the number of person-years and deaths.

Main Results:

• A novel, intuitive, data-based SE method for life expectancy is presented.
• The study quantifies the statistical noise in life expectancy differences (year-to-year and between/within countries).
• Relationships between SE and data denominators (person-years, deaths) are described.

Conclusions:

• Agencies and researchers should report the imprecision of life expectancy figures.
• The number of decimal places reported should correspond to the quantified statistical uncertainty.
• This approach enhances the reliability and interpretability of life expectancy statistics.