We can now look at some estimates of the overall direction of transfer flows for preindustrial societies. I have examined data on the following populations:
- Kaplan (1994) presents consumption and production data for three different groups of Amazon Basin hunter-gatherer-horticulturalists, including the Aché, and generously made these available to me for further analysis. The age profiles themselves show that people continue to be net producers throughout their adult lives, including in old age. The elderly clearly continue to transfer food to their children and grandchildren. The Aché practice gerontocide, particularly of females, once the elderly are no longer able to produce.
- Dodds et al. (1996) provide estimates for five other Amazon Basin hunter-gatherer-horticulturalist societies, based on time-use data combined with assumed age profiles of consumption and hourly productivity. In these data, the net direction of transfers is still clearly downward, but there is apparently a stage in which the elderly produce less than they consume. For Figure 11-2, I have pooled these data for all five societies.
- Stecklov (1995) presents data for rural Cote D'Ivoire and analyzes them using the framework outlined above. Here also the net direction of transfers is strongly downward from older to younger, but at the same time the elderly consume more than they produce, by virtue of transfers from younger members of the population.
- Mueller (1976) develops age profiles of consumption and production from a variety of broad-based estimates representative of Third World agricultural populations. These data clearly indicate a net downward flow of transfers in the population, but again the elders receive transfers that enable them to consume more than they produce.
Figure 11-2 presents arrows for these societies. They all point distinctly downward, indicating that the direction of transfer flows is, on net, from older to younger. This contradicts the statement from Caldwell quoted above, if his statement is taken to refer strictly to transfers of foods and other easily measured physical items. The horizontal position and length of these arrows reflect the average ages of making and receiving transfers, and the figures indicate the difference in these average ages.
The answer to the second question, more specifically about the elderly, is that in some societies average transfers are always downward, even in old age (Kaplan, 1994), whereas in other societies, perhaps the majority, the older population consumes more than it produces. In this connection, however, note that the elderly provide many useful functions besides directly producing food. As Rosenzweig's (1994) study indicated, and as suggested by the quote from Simmons (1945), the knowledge and experience of the elderly can contribute substantially to food production, even when their physical contribution is limited. Furthermore, they may contribute in other intangible ways through leadership, maintaining order, etc. Finally, they may contribute childrearing services that