The Most Frequent Methods of Smoothing and Extrapolating Mortality Curves and Applying them to the Czech Population
Abstract
The article introduced some most frequent methods of smoothing and extrapolating the mortality curve. All those methods could be divided into three groups.
The first group (represented by the traditional Gompertz-Makeham curve, Coale-Kisker or Beard model) leads to the probabilities of death close to one in relative young ages. Also the models in the second group (for example Weibull, Denuit-Goderniaux or modified Gompertz-Makeham model) limitary lead to probabilities of death approaching one, but it occurs in higher ages than for the models in the first group.
And the third group of models (Thatcher, Kannisto, Himes-Preston-Condran) is the most optimistic one, even in the highest ages the probabilities of death converge to values close to 0,6. It was shown that the selection of whichever of those models could influence the results obtained from the mortality table.
The differences in life expectancy at birth are around almost one year according to the various models.