Variation and Vulnerability: Unobserved Heterogeneity during Mortality Transitions
Michal Engelman, Johns Hopkins University
Emily M. Agree, Johns Hopkins University
The dramatic increase in life expectancy in most industrialized countries has been accompanied by a similarly striking compression in the variance of ages at death. We show that mortality variation in later life has nonetheless followed a contrasting pattern, with survivors to older ages becoming increasingly heterogeneous in their mortality risk. We argue that delayed mortality selection may be a result of ongoing improvements in survival at younger ages, and investigate the extent to which frailty models that account for changes in population composition over time capture both the temporal trend and age-pattern of mortality variability. We incorporate gamma-distributed frailty into a Siler trajectory representing the mortality hazard across the lifespan, and use maximum likelihood methods to simultaneously estimate the parameters of the resulting model. Our findings indicate that the distribution of frailty at older ages has been growing as survivorship increases, potentially accounting for the observed variability patterns.
Presented in Poster Session 4