Using the Lee-Carter method to forecast mortality for a group of populations non-divergently

Nan Li, Duke University

Mortality differences among a group of closely-related populations are unlikely to continue to increase in the long run. Using a common factor to model the group’s age- specific mortality trend and a specific factor to describe an individual population’s special mortality dynamics, this paper extends the Lee-Carter method to forecast mortality for a group of populations non-divergently. Using the Human Mortality Database, non-divergent forecasts are provided for two-sex and low-mortality populations.