The Time Lag Between Period and Cohort Distributions of Adult Deaths

The risk of dying for humans has a characteristic W-shaped age-pattern, no matter if data are aggregated over periods or cohorts. We study the time lag between the best-fitting pairs of a period and a cohort distribution of deaths, from birth or from some later age. For four countries from the Human Mortality Database, we find remarkable regularity in the progression of the lag over time. This suggests that, knowing the lag, cohort distributions of deaths may be approximated by period ones, and vice versa.

• Register here