Forecasting of Singapore's future mortality rates : the Heligman-Pollard model and its comparison to the Lee-Carter model.

Abstract

In this project, we study mortality rates and life expectancies of the Singapore population using the Heligman-Pollard model – both initial and extended versions. The evolution of the parameters of both variants of the model are obtained, analyzed and discussed to form the basis upon which the future mortality rates and life expectancies are forecasted. The forecast results are compared against actual mortality data to investigate the accuracy of the estimates. The initial variant of the Heligman-Pollard model proved to be the better of the two, providing a better fit of data to the actual mortality rates. Projected mortality rates under the initial Heligman-Pollard model for Years 2011 and 2021 is then compared against those projected using the Lee-Carter model. Evaluation of the two models found the Heligman-Pollard model more superior in generating reasonable rates of mortality. Concluding remarks are made at the end of the paper with suggestions of possible future research.