If fertility and mortality persist at a given level for several decades, they generate a population with a predictable age structure and rate of increase. Demographic measures have an essential inter-relatedness that derives from this phenomenon, commonly referred to as stable population theory. In general, if you know two of these measures, and have reason to believe that fertility and mortality have been approximately stable, you can predict the other measures. I want to exploit this opportunity for triangulation to see how firmly we should believe our estimates of Hadza fertility and mortality, and how well we have represented average Hadza demography during the second half of the twentieth century. I will show whether the age structure predicted by our estimates of fertility and mortality is matched by the age structure we observed in the field. In addition to age structure, several other measures can be predicted for a stable population and can be compared to field observations.

A stable population, one in which fertility and mortality remain constant for some decades, and in which migration is negligible, is something of an abstract ideal. In reality, fertility and mortality vary year to year, age structures are jagged (Figure 10.2), migration likewise varies and may occur in quite irregular bouts responding to some climatic or social opportunity or hardship. In small populations such as the Hadza, these variations become especially influential. Nonetheless, demographers of modern states, anthropological demographers, and historical demographers continue to find stable population models to be extremely useful. At the very least they can show when apparent results are inconsistent and require re-examination. Sometimes demographers have used stable population theory as a solution to the practical problems of measurement in less accessible populations. Especially useful have been the tables developed by Coale and Demeny (1983), and the Weiss (1973) models for anthropological populations have been widely cited. If one has data that allow one to select one of the models, then many other demographic parameters can be simply read off the model, or compared to further field observations to verify the choice of model. In the example best known to anthropologists, Howell (1979) used the Coale and Demeny (C&D) tables to obtain initial estimates of !Kung mortality and age structure.