The strategy of the demographic study

My aim in the first part of this book is to describe the basic demography of the eastern Hadza, and then debate some issues of hunter-gatherer population dynamics. Because hunter-gatherer populations such as the Hadza are small, mobile, difficult to study, have no written records or registered ages, and are the subjects of a number of uninformed and stubborn assumptions, I put a rather heavy emphasis on ways to check our results. This has shaped my central strategy: (1) estimate fertility, mortality, and migration; (2) look for evidence of stability in fertility and mortality; (3) if it is found, and migration is infrequent, use our fertility and mortality estimates as input to a stable population simulation that predicts age structure and structure-related measures; (4) match the predicted values to observed values. Ideally, these tests use data obtained by means that are different from those used to gather the data we used to estimate fertility and mortality in the first place.

If the checks fail badly, then we will know something is wrong with our methods or our estimates of fertility and mortality, or with the stable population assumption. If the checks match the predicted values, we may have more confidence in our results. I try to indicate the meaningfulness of the checks by using resampling methods, by comparing the observations to predictions from simulations that use higher or lower fertility or mortality as their input, and by comparison to other populations, especially the !Kung and Ache. I do not expect to confirm results to the great precision that one might expect in a large systematically recorded population such as an industrial nation. I do expect to be able to say whether we can confidently describe the Hadza as less or more fertile than the !Kung or Ache, or some other interesting population, and with higher or lower mortality than those populations.

One prerequisite for this strategy is that the Hadza are sufficiently near to a “closed” population.