Measuring the efficiency of mortality changes: an evaluation of past and forecast levels in advanced economies

James E Oeppen, Max Planck Institute for Demographic Research

The impact of an age-specific mortality change on life expectancy depends both on its quantity, and its “efficiency”. A large proportional change in a mortality rate may be inefficient if it occurs in an age-group that has little effect on life expectancy at birth. Since 1840, some countries have stayed close to the advancing linear frontier for life expectancy. This implies that they have efficiently shifted the focus of mortality change from the young, through adults, to the elderly. Mathematical demographers have defined many measures of the quantity of mortality change, but have given little attention to the concept of efficiency. This paper decomposes annual changes in life expectancy at birth for advanced economies into two components: quantity, and efficiency as defined for a social planner. The method can also be used to measure the efficiency of mortality changes for healthy life expectancy, and to measure efficiency with respect to particular causes of death. The all-cause mortality results show that in the past 50 years France, Japan and Switzerland have been demographically efficient, suggesting a positive association between the quantity and efficiency of mortality change. The United States has made both smaller and less efficient changes. The observed long term trend in efficiency is ignored by forecasting methods. Applying the decomposition to a European Union set of mortality projections reveals an assumption of almost total collapse of efficiency for all countries by 2050.