No abstract

T he lively debate between Laurence J. Kotlikoff and Franco Modigliani presented in the Spring 1988 issue of this journal concerns an old question:what is the main motivation for saving and therefore for the accumulation of wealth? More specifically, what are the respective contributions to aggregate wealth of (1) saving for retirement (also known as "hump" saving); (2) precautionary savings (and "unintended" bequests) due to uncertainty about the length of life; and (3) planned bequests? Of course, other wealth holding motives are possible, but let us follow Kotlikoff and Modigliani in setting them aside for now. If Modigliani's life cycle hypothesis is to be viewed as a close to approximation of reality, then the bulk of existing wealth should have resulted from some combination of hump and precautionary saving.Our comment on this dispute attempts to advance two issues. First, the controversy involves an enormous gap between empirical estimates of the share of "inherited wealth" in total accumulation, even though the estimates are often based on the same data. We hope to clarify why the estimates vary so widely. Second, the Kotlikoff/Modigliani dispute is presented as an American issue, with little extension abroad. We will present some results from other countries that bear on the controversy. (Some of the estimates from Kotlikoff's paper originated in a 1981 paper he wrote with Lawrence Summers, which is why we sometimes refer to his position as the "Kotlikoff-Summers" argument in this comment.)The problem of conceptualizing the contribution of bequest to aggregate savings can be summarized with a thought experiment. Assume all bequests were confiscated.

Most models of family transfers consider only two generations and focus on two motives: altruism and exchange. They also assume perfect substitution between inter-vivos downward transfers and bequests. Based on French evidence, we show that parent-to-child transfers belong to three distinct categories (investment in child's education, ®nancial assistance, wealth transmission), and advocate a three-generation framework. Thus, transfer behavior of parents toward their children is strongly in¯uenced by the behavior of their own parents. There is also some evidence of the Cox and Stark demonstration effect: parents help their own parents, expecting to receive comparable support from their children. Such behavior can be regarded as indirect reciprocity: the bene®ciary does not give back to the initial giver but to a third person of another generation.

No abstract

No abstract