Mortality Rates, Mortality Conditions, and the Tracking of Progress in Life Expectancy

A population's level of period life expectancy is a biased indicator of period mortality conditions, due to the existence of cohort effects and mortality selection. It is also an indicator that has little relevance for the experience of actual individuals. In this paper, we discuss Lagged Cohort Life Expectancy (LCLE), or the life expectancy for the cohort currently reaching its life expectancy. We argue that LCLE is a useful mortality measure that provides information about levels of longevity currently being reached by actual cohorts of individuals. LCLE also makes a useful distinction in a population between individuals who have not yet reached the life expectancy for their own cohort, and those who have surpassed it. However, LCLE cannot be observed for the current year, because the cohort currently reaching its life expectancy is not yet extinct. Therefore the estimation of current LCLE must rely on assumptions about future mortality. We examine various forecasting approaches and show that current LCLE can be estimated with precision using the most simple mortality forecast ¿ one that applies current mortality to the remaining lifetime of cohorts currently alive. We estimate LCLE trajectories for four countries (France, Sweden, Japan, and the US) and discuss the usefulness of using LCLE as a summary measure of population health.