Measuring the morphology of the life table

Jon Anson, Ben Gurion University of the Negev

As mortality declines to ever low levels, there has been growing interest in measures of the shape, or internal structure of the life table, as a basis for comparison. However, the usefulness of these measures in comparing life table morphologies is compromised by their close correlation with the mean level of mortality. The maximal human life span is, in all essences, fixed. The decline in mortality is thus a movement of deaths from the younger to the older years of life, and their concentration in an ever narrowing age band. The shape of the life table functions (mortality and survivorship curves) is thus largely determined by the average level of mortality, and any meaningful attempt to measure the shape needs to separate out the necessary effects of declining mortality. Otherwise, we are merely comparing mortality levels by another name. Yet, as mortality declines to ever lower levels, the difference between life tables is going to be more and more in the details of the distribution of mortality, and these measures are going to play an ever more important role. In the present paper we consider how the various morphological measures may be used to compare the internal distribution of mortality within the life table, while yet allowing for the changes which necessarily occur as mortality declines. This will enable us to identify differences in the shapes of mortality and survivorship curves, that is, in the distribution of deaths over the life span, which are not a necessary corollary of the differences in the overall level of mortality. Such differences may then be related to differences in the conditions (social, economic, physical) in which the population lives its life, as will be illustrated using examples from various contemporary and historical populations, at the national and sub-national level.