On period, cohort and population life expectancy

Abstract

Period Life Expectancy (PLE) is a measure of longevity valued for its sensitivity to short and long-term changes. However, it refers to a hypothetical cohort, not to a real population, thereby undervaluing longevity under declining mortality conditions. Other measures such as the Average Cohort Life Expectancy (ACLE) only partially overcome this limitation, still underestimating population longevity. This article introduces a new indicator, the Population Life Expectancy (PoLE), defined as the mean age at death of active cohorts in the studied population. Using a log-linear Poisson model with age-period interaction to project mortality of non-extinct cohorts, we estimated PoLE in Switzerland and Norway over 1876-2024, and compared it to PLE, Cohort Life Expectancy (CLE), and ACLE. PoLE clearly exceeded PLE, increasing from 63.3 to 89.7 for Swiss men (PLE from 37.7 to 82.4), and from 65.4 to 91.3 for Swiss women (PLE from 41.4 to 85.9), revealing a gain of about +50% over 150 years, rather than +100% suggested by PLE. Comparable results were obtained in Norway. PoLE was also higher than CLE until the mid-20th century, when the relation reversed, indicating that life expectancy is now higher for newborns than for those already alive, a tangible sign of human progress. Keywords: period life expectancy, cohort life expectancy, population life expectancy, real population, mortality projections.