of the transfer flow, transfers within the family, and transfers through the public sector.
Figure 11-3 shows arrow diagrams for transfers within the family in the United States, grouped into four kinds: (1) bequests to children at death of the parent: (2) interhousehold transfer flows between no-longer coresident parents and children; (3) parental costs of higher education; and (4) parental costs of child rearing, net of children's earnings as teenagers.
For each kind, the arrow is positioned as described earlier, and the thickness of the arrow represents the average size of the flow per person in the population (for details of estimation, see Lee, 1994b). All four kinds of familial transfers are downward. The area of each arrow indicates the expected value of net future receipts of transfers over the life cycle of the individual or the household. The average household expects a receipt of -$44,000 in bequests (data not shown); that is, the average household has received $44,000 in bequests already and anticipates leaving a similar amount to its heirs in the future. Bequests flow downward across age. Other interhousehold gifts and transfers are poorly measured in the data used for these calculations, but better data confirm the net direction of such flows. The average child receives about $81,000 in childrearing expenditures (not counting parental time costs) and can expect to allocate a corresponding amount to each of its own children in the future, with the help of a spouse. There was an additional expenditure of about $6,000 per child for costs of higher education, which the child can also be expected to "repay" in the future. The figure shows massive downward transfer flows from parents to children within the family.
This pattern of downward familial transfers at all ages is consistent with Kaplan's (1994) measurements for the Aché and with the facts concerning postreproductive toothed whales reported in Carey and Gruenfelder and in Austad (in this volume). However, the pattern is not consistent with the other agricultural and preagricultural societies for which data were earlier described; these commonly showed that the elderly received transfers from younger members of the population, enabling them to consume more than they produced. However, the net flows up to surviving elders are overwhelmed in value by the value of net flows downward from parents to children earlier in the life cycle, so the net direction of transfers flows overall is definitely downward, as shown earlier in Figure 11-2.
There is an obvious reason why the elderly in the United States continue to make net transfers to their adult children, in contrast to most preindustrial societies so far examined (except for those studied by Kaplan, 1994). The United States, like other industrial nations, has a well-developed system of transfers to the elderly through the public sector. These public transfers substitute for transfers to the elderly from their own children. Indeed, many elderly apparently believe that these public sector transfers go too far and offset the flows by private transfers in the opposite direction. Figure 11-4 shows arrows for public sector net