Private savings which are transferred intergenerationally through bequests make up a substantial portion of the U.S. capital stock. Most of these bequests do not get spent down by heirs and in fact usually increase in size as they are transferred to successive generations. In addition, bequests are overwhelmingly concentrated among the very wealthy. With this backdrop, Professor Barbara Fried engages in two related inquiries: First, who derives utility from bequests (donors, heirs, successive generations)? Second, what implications does this have for projecting the winners and losers from changing the taxation of savings, specifically in a shift from the income tax to a consumption tax? In order to determine who gains utility from bequests, Professor Fried first analyzes the various motives for bequest savings. These motives ,include the altruistic model, in which the donor gives out of generosity towards family members; the precautionary savings model, in which consumers save as a hedge against unforeseen contingencies and then bequeath whatever they do not consume; and the exchange model, in which a parent bequeaths money to a child in exchange for the child's care and services as the parent grows older. Professor Fried concludes that, under most theories of bequest motives, the level of bequests will increase under a consumption tax. For bequests that are disguised exchanges, the resulting increase in bequests will be used to finance a higher level of preclusive consumption for the donor generation, thereby generating utility for the donor generation only. For bequests motivated by altruistic and precautionary concerns, in contrast, the donor realizes at least a portion of the utility from bequest savings in a nonpreclusive form. As a result, any increase in bequests will generate increased utility for both donor and donee. A model that takes into account the double-level utility gains from increased bequest savings is likely to magnify the aggregate welfare gains from shifting to a consumption tax, as compared to the results reported by traditional models. It is also likely to magnify the distributional benefits of such a shift to the wealthiest five percent of Americans, who account for the majority of bequest savings.
Private savings ultimately destined for intergenerational transfer, through inter vivos gifts or bequests, account for a very substantial percentage of the U.S. capital stock at any one time. Estimates of the percentage range from 15 percent to as high as 70 percent, and average about 50 percent. In addition, bequests appear to be a luxury good--that is, they are heavily concentrated (as a percentage of lifetime income) among the very wealthy (the top quintile in lifetime income, and even more dramatically the top 5 percent).(2) Consistent with that finding, the data show small positive elasticities for bequests with respect to changes in lifetime earnings throughout most of the income range, and very high positive elasticities for the top quintile in lifetime earnings.(3) The data indicate that most bequests never get consumed by heirs--indeed, most people who inherit money leave a significantly larger estate to their own heirs than they were left (measured in real dollars).(4) All of this raises the question: What is all this money doing out there, if it is never spent? Who, across generations of a family linked by gifts and bequests, actually derives utility from these transfers and how?
The immediate impetus for the question derives from how we account for the distributional and welfare effects of changing the tax treatment of savings, such as by adopting a consumption tax in place of our existing income tax. A consumption tax lowers the effective tax rate on savings, as compared to our existing income tax. It is generally agreed that a lower tax rate on savings will generate aggregate welfare gains to society, by reducing the tax-induced intertemporal distortions in people's consumption choices and by increasing private savings available for productive investment, although the extent of these gains is hotly contested. It is also generally agreed that a lower tax rate on savings will disproportionately benefit the wealthy because they hold a disproportionate share of private savings. Again, the extent of the disproportionality of the benefit is hotly contested, and depends significantly on what version of a consumption tax is adopted.(5) Given the magnitude of bequest savings, the fact that bequests are heavily skewed toward the top income classes, and the fact that bequests are perpetuated across generations, the magnitude (if not sign) of both efficiency and distributional effects may well depend on how we account for the utility generated by bequests. This article attempts to sketch out those implications.
To illustrate the complexities entailed in accounting for bequest savings, take the following simple, but not implausible, example. A fellow named Lear starts off young adulthood as a dirt-poor tenant farmer. After working very hard for many years, Lear accumulates vast agricultural holdings, becoming the biggest landlord in England. When Lear dies at age eighty, he leaves his lands, valued at $10 million, to his beloved daughter Cordelia. Cordelia, who has a good job of her own, spends only a small portion of the annual rental income from her father's estate, leaving to her own daughter at her death an estate worth more (in real dollars) than what she got from her father. Cordelia's daughter leaves a similarly augmented estate to her daughter, and so on, until the lands are sold off and the proceeds finally consumed by Lear's impecunious great-granddaughter, Kate, ninety-five years after Lear's death.
Under our income tax, the land rents are taxed to each generation as received. Under most versions of the consumption tax, which do not tax unspent resources at death, tax on the rents will be deferred for four generations, until the savings are ultimately consumed. The value of that deferral across four generations is enormous. But to whom? Suppose that England shifts to a consumption tax when Lear is age fifty, thirty years before his death. Who gets credited with those thirty years of increased return to savings under a consumption tax? Lear? The great-granddaughter Kate? Everyone from Lear through Kate?
The question of who gains from bequest savings, while obviously relevant to ascertaining the winners and losers in shifting to a consumption tax, has received little attention in the legal tax literature. In conventional distributional analyses done by government agencies (the Joint Committee on Taxation, the Treasury Department, etc.) the answer is clear and uninteresting. Such analyses typically group individuals by income class based on annual income in the year of a fully phased-in tax change, and then estimate how each such income class's tax liability for that year will change under the new tax regime, assuming taxpayers do not alter their behavior in response to the tax change.(6) This measure of static revenue change as a percentage of annual income is a very simple view of the distributional winners and losers from tax changes--most economists would say hopelessly simplistic, although in the end it may be the most reliable and politically salable measure we can come by. For starters, it omits consideration of the excess burden (deadweight loss) of taxation.(7) It also ignores the fact that the distribution of lifetime incomes is much more compressed than annual incomes, with the result that a regressive tax appears more regressive, and a progressive one more progressive, when measured as a percentage of annual rather than lifetime incomes.(8) Be that as it may, under the conventional methodology for doing distributional analyses, accounting for the increased utility from all savings, including bequest savings, is straightforward in the shift to a consumption tax. Each annual income class benefits in proportion to its holdings of capital that would otherwise have generated taxable income in the year of transition. In the hypothetical above, Lear is the only one in his extended family who will gain in the transitional year from the shift to a consumption tax, since it is he who will avoid taxation on the land rents (and taxation on income from prior, invested land rents) accruing that year.(9) The tax liability thereby avoided will thus show up as an increase in Lear's wealth for the year--or more precisely, as an increase in the wealth of the annual income class of which Lear is a member in that transition year. It is irrelevant for these purposes that Lear's accumulating land rents are destined for bequest rather than lifetime consumption. It is irrelevant as well why they are destined for bequest: By assuming static taxpayer behavior, the model precludes any need to estimate how Lear might change his bequest savings behavior in response to changes in the tax rate on savings under different hypothesized bequest motives.
A number of economists have tried to model the winners and losers under tax changes in a more complicated fashion, by: (1) grouping income classes by lifetime rather than annual income; (2) measuring the effects of tax changes over a longer time horizon, in some cases including generations born after the change; and (3) adopting a broader measure of economic welfare that accounts for both the utility of leisure and the excess burden from taxation. The last of these changes requires economists to estimate a number of parameters built into taxpayers' utility functions, including the rate of pure time preference, intratemporal elasticity of substitution between labor and leisure, and intertemporal elasticity of substitution between present and future bundles of consumption and leisure. The last two changes in particular--shifting to a utility-based measure of welfare and measuring changes in utility across generations--raises the...