A new approach to length of life inequality measurement

Abstract

The uncertainty around the limits of human lifespan has led researchers to implicitly treat that variable as unbounded. In this paper we categorize that variable as 'uncertainly bounded' and study the corresponding methodological and substantive implications for our understanding of lifespan variation. Being a bounded variable, the measurement of lifespan variability is affected by two problems: (i) the 'boundary effects' (the inequality of the distribution is related to the mean when the latter approaches the upper bound); and (ii) the '(in)consistency' problems (achievement and shortfall distributions can be inconsistently ranked by standard inequality measures). In this paper we propose new inequality indices overcoming such limitations. Applying the new measures to the Human Mortality Database we observe that the decline in overall lifespan variability typically associated with increases in longevity seems to stop and even reverse when life expectancy reaches a certain threshold. While our findings are influenced by the choice of the lifespan upper bound, they consistently point to the emergence of length of life inequality plateaus and trend reversals at higher longevity levels. This suggests the emergence of a worrying equality-efficiency dilemma, in which higher achievements in longevity would only be possible at the expense of higher lifespan variability.