Transfer of Wealth
center on wealth and philanthropy
Media coverage, specifically the April 6, 2006 article in the Chronicle of Philanthropy, “Much-Anticipated Transfer of Wealth Has Yet to Materialize, Nonprofit Experts Say,”by Holly Hall, examines the estimate of the amount of household wealth to be transferred in the 55-year period from 1998 through 2052 in the report, “Millionaires and the Millennium”. The report, released in 1998 by the Center on Wealth and Philanthropy, was based on estimates from our Wealth Transfer Microsimulation Model (WTMM) which, in turn, was based on data on the amount and distribution of wealth from the Survey of Consumer Finances, sponsored by the Board of Governors of the Federal Reserve System, mortality rates based on data from the Center on Disease Control and the Center on Vital Statistics, life cycle savings rates estimated from the Survey of Consumer Finances, and historical patterns of the distributions of estates based on IRS estate tax data.
A major purpose of our analysis was to estimate the total amount of wealth that would be transferred in the 55-year period from 1998 through 2052. At the time most planners and developers were using a figure of $10 trillion to $12 trillion developed for another purpose by Robert Avery and Michael Rendell. Our estimate was $41 trillion (1998 dollars) during the 55-year period if household wealth grew at a secular rate of 2% in real terms. The model that produced this estimate incorporated life cycle savings rates, the 1998 distribution of wealth, and age and race specific mortality rates, and of course the assumed 2% rate of growth in household wealth.
In researching her April 6th article, Holly Hall contacted John Havens, Associate Director of the Center on Wealth and Philanthropy, to discuss the estimate. At the time that John Havens drafted his comments for Hall, contained in the links below, he identified three factors that could contribute to the discrepancy between the 20 year estimate predicted in “Millionaires and the Millennium” and the Giving USA estimates of charitable bequests:
1. The slide in the stock market and concurrent recession at the turn of the century and subsequent years was reflected in a decline in household wealth (after adjusting for inflation). This decline lasted three years (1999 through 2002) during which the household sector lost approximately 15% of its wealth. This was the deepest and longest decline in household wealth since the great depression. By 2004, household wealth had rebounded to its 1999 level. The Wealth Transfer Microsimulation Model assumes a low growth scenario of 2% positive real growth in wealth and this was not achieved in the early years of the scenario, leading to levels of charitable bequests below the levels implied by the 20 year estimates of the model.
2. There has been an increase the rate of inter vivos giving among wealthy households in recent years. There has been large growth in family foundations, donor advised funds, charitable remainder trusts, and other forms of planned giving and this growth is likely at the expense of charitable bequests. In our writings on wealth transfer we usually state that “if the transfer of wealth follows historical patterns, the transfer will be distributed as . . .” In fact the growth in the aforementioned vehicles of giving does not follow historical patterns so that the estimates of inter vivos giving produced by the model may be too low and the estimates of charitable bequests too high.
3. Until recently Giving USA based its estimate on data from federal estate tax returns plus an estimate for charitable bequests made by decedents not required to file a federal tax return. In any year the most recent estate tax return data published by the IRS is based on all returns filed in the year. Several years after they are released, the data for every third year is restructured, reanalyzed and again released for the same year but reported in terms of the year of death of the decedent. The charitable deductions classified by year of death vary, sometimes by more than 25%, from the charitable deductions classified by year of filing – sometimes higher sometimes lower. The estate data is still the best data available at the high end for charitable bequests but it is a noisy indicator of total charitable bequests in any given year. The part of the Giving USA estimate for charitable bequests made from estates not required to file federal estate tax forms is based on more assumptions and somewhat less reliable data than its estimate for decedents required to file estate forms. Given the lack of data concerning charitable bequests, the Giving USA estimates are reasonable ballpark estimates based on the data available, but they may contain considerable noise.
In addition to the above points, financial planners, development officers, fund raisers, and officers of nonprofit organizations continue to indicate that in their daily activities they see confirmation of a major increase in wealth transfer activity. This constitutes qualitative support that increased wealth transfer is taking place, often at substantially increased levels from the level of activity in 1999.
As Havens indicated in his correspondence with Hall, the Center on Wealth and Philanthropy had planned and has now started a project to review wealth transfer in recent years with a focus on the discrepancy between implications of the Wealth Transfer Microsimulation Model and Giving USA estimates. In its initial stages, the project may eventually develop new estimates of wealth transfer. We have not yet completed enough work to reach any conclusions. However, our initial work suggests the following:
1. There is substantial wealth transfer that has taken place since 1999, somewhat dampened by the downturn in household wealth from 1999 through 2002 but still at least as much as indicated in the 20 year low growth estimates of “Millionaires and the Millennium.”
2. Some of this wealth is distributed to charities by mechanisms triggered by the donor’s death but not completely captured by the estate tax data. Trusts, including charitable remainder trusts, are one such vehicle. Another is a life insurance policy with nonprofit beneficiaries that are transferred to the ownership of the beneficiary during the life of the donor – as part of financial and estate planning. A third consists of IRA and/or other retirement accounts in which a nonprofit is a beneficiary.
3. The recent dramatic rise in inter vivos giving reflects, in part, the increase in forms of giving triggered by the donor’s death but with spiritual and tax benefits to the donor during their life. For example, the donor gains a spiritual reward during his lifetime for donating a life insurance policy to his alma mater. Moreover, he gains an income tax deduction in subsequent years for his premium payments on the policy, which are recorded as inter vivos giving in income tax data. At his death, his alma mater obtains the death benefit that is not recorded in the estate tax data and consequently is not included in the Giving USA estimates of charitable bequests, although it is a gift to charity triggered by the donor’s death.
It will be a few months before we have arrayed and assayed enough data to reach conclusions relative to the validity of the short term (20 year) “Millionaires and the Millennium” estimates. In the narrow sense of a bequest in a will the short term charitable bequest estimates are probably too high; in the broader sense of transfer of assets to nonprofit organizations triggered by the death of the donor, it is unclear whether or not the 20 year estimates are too high and there is strong evidence that the 55 year estimates remain valid.
In the meantime, in order to provide background and context for the discussion of the wealth transfer estimate, we have included additional information in the links below. The correspondence presented in these links predates our current effort to revisit, review, and develop renewed estimates of wealth transfer in the near term. The conversation that ensued between Havens and Hall in preparation for Hall's Chronicle article took place largely via email and is included below.
Follow the first link to read Hall’s April 6th Chronicle article. Follow the second link to read the original research note that Havens prepared for Hall upon her request. Follow the third link to read the email conversation between Havens and Hall regarding Hall’s follow up questions.