The Impact of Disparities on the Age Trajectory of Mortality

James W. Vaupel, Max Planck Institute for Demographic Research
Michal Engelman, Johns Hopkins University

This paper proposes a new model for mortality disparities in heterogeneous populations. We define a mortality trajectory for a heterogeneous group with the lowest (or “best-practice”) levels of mortality at every age and model the observed mortality trajectories of two other heterogeneous populations with respect to the best-practice trajectory using a proportional hazard model. Starting with simulations, we assign parameters of frailty to populations with different levels of internal heterogeneity and use them to model disparities in their age trajectories of mortality relative to the best-practice trajectory. We show that this disparity model can produce realistic mortality trajectories for contemporary populations (e.g. the very heterogeneous U.S. and more homogeneous Japan) with relatively high life expectancies and mortality hazards that cross at older ages. Maximum likelihood methods are then used to estimate frailty parameters from empirical mortality data and test the model’s general applicability.