Abhatoo - Centre Nationale de Documentation

The paper presents an adjusted version of the method of direct extrapolation of mortality by age and sex. The method is supplemented by additional procedures in order to improve its efficiency in the short-run and preclude implausible mortality patterns in the long-run. The short-run efficiency is improved by building the forecast on data from the most recent periods of age-sex-specific duration, when mortality dynamics exhibits steady trend. In the long-run, the rates of mortality decline are assumed to converge to a plausible function of age and sex, which is derived from the data based on the assumption that it is a monotonic function of age. Efficiency of such improvements is supported by data and forecasting results. The framework proposed provides a natural basis for introducing uncertainty into the projection. Based on preliminary structure of the probabilistic model, simulations were run to study estimation errors of model parameters and autocorrelations involved. They indicate that estimates of some of the parameters are ineffective and need further research.