Our suggestion is the same as a part of a recommendation made by the Tax Section of the American Bar Association in its 1965 Annual Report except in one respect. The recommendation would change section 678 as to the lapse of all 5 and 5 powers whereas our proposal would limit the change to cases where the holder of the power is the sole income beneficiary of the trust. We believe this liberalization should not be available in cases where there may be opportunities for minimizing income taxes on the donee from the exercise of a 5 and 5 power by having accounting income paid to other members of the holder's family or accumulated and taxed to the trust. Sections 8 and 25. Interest in a Trust or its Equivalent The current federal estate and gift tax laws permit property to be held in trust for the maximum period authorized by state law with the imposition of only one transfer tax upon the creation of the trust. During this period beneficial interests in the property may pass through several generations. Thus, through the use of trusts of long duration, several generations may presently share in the beneficial enjoyment of property without the payment of additional transfer taxes. Some people believe that present law is structurally unsound because it creates a tax preference (or tax incentive) for the use of trusts. The issue involved has become commonly known as the "generation-skipping" problem. Solutions have been suggested, but to date none has received anything approaching broad support. It is useful to consider these solutions as a prelude to discussing the approach taken in the Draft to this problem. This approach is contained in Sections 8 and 25 which for convenience will hereafter be referred to as "the Statute". The most extreme and complex solution that has been proposed is the one suggested in the Studies. It has been put in statutory form in section 505 of S.3378 introduced in the Senate by Senator Gaylord A. Nelson during 1972. The complexity of the Studies proposal has been criticized even by advocates of significant change in this area. See, e.g., Westfall, Revitalizing the Federal Estate and Gift Taxes, 83 Harv. L. Rev. 986, at 1006-1013 (1970). This solution is, as pointed out below, not consistent with its underlying premise, which is that property should be subject to a transfer tax in each generation. In order to accomplish this objective, a substitute (or additional) tax would be imposed upon an outright transfer, or a transfer in trust, of property to a person who is more than one degree below the transferor. The rate of tax would be 60% of the marginal transfer tax rate applicable to the transfers during the taxable period if made during life or to the transfers at death if made at death. Alternatively, the family would have the right to treat the transfer as if the property was first transferred to the skipped generation and, if this were done, the substitute tax would be the amount owed by the skipped generation. The application of the proposal to transfers in trust is summarized as follows: "When the generation-skipping gift or bequest is by trust, there would be generally the same options as to when the tax must be paid as would be available to the skipped generation had he elected to pay the tax. Thus, the transferor or his representative (i.e., executor or trustee) may elect to treat the taxable event as occurring at the time of the original transfer or as of the first day of any calendar quarter thereafter. In no event, however, may the tax be postponed beyond the date of the death of the last survivor among the group consisting of the transferor, his children, and any beneficiaries under the trust who are not within the category of individuals to whom a gift would be considered a generationskipping gift. At this time, it becomes certain that there is a generation-skipping transfer involved and no reason to further defer the tax. "The substitute tax would be computed on the value of the trust corpus as of the effective date of the election or the date of the taxable event. Thus, if an individual put $10,000 in trust, with the income payable to his son and the remainder to his grandchildren, and elected to pay the tax at the time he established the trust, he would be subject to a substitute tax on a transfer of $10,000. If, however, no election to pay the tax was made prior to the death of the transferor and his son, then at that time, a tax would be due computed upon the value of the trust corpus at that time. Since the transferor would be deceased at this time, the substitute tax in this case would be paid by the trustee out of the trust property. Any trust distributions prior to this time which skip a generation would also be subject to the substitute tax as applied to the amount of the distribution. "Once the substitute tax has been paid on the If The proposal in the Studies fails to apply the every generation theory of taxation to all transfers. the tax is to be imposed upon such a theory, transfers that do not move "downstream" (to a generation one or more levels below the transferor) but rather "upstream", usually to parents, or "sideways", usually to brothers and sisters, should be exempt from the tax. Yet the Studies' proposal does not contain such an exemption. In fact, it moves in the opposite direction by proposing the elimination of the previously taxed property credit allowed by section 2013, which has its primary significance in connection with transfers between brothers and sisters. We believe the proposal is not appropriate as to either outright transfers or transfers in trust. With respect to outright transfers, we do not understand why a substitute tax is appropriate in connection with a gift to a grandchild. Why should such a gift be penalized? There has been no splitting of benefits between generations as there is in a trust transfer. If there is an abuse in the generation-skipping area, in our opinion it exists only where such a splitting is present. Further, we know of no country or state that has imposed an additional transfer tax on outright generation-skipping transfers. While the fact that an idea has not been tried before does not automatically justify its rejection, it does suggest that an additional or substitute tax on such transfers is alien to the basic concept of fairness in transfer taxation. We agree with the resolution adopted by the ALI Project that "Under either a dual tax system or a unified tax, an additional tax should not be imposed on an outright transfer, or its equivalent." (page 7) We also disagree with the approach of the Studies proposal to transfers in trust. The proposal imposes a tax at the "wrong" time on the "wrong" person and is objectionable in this regard for three specific reasons. First, the additional tax is computed by multiplying the value of the taxable transfer by a percentage of the decedent's marginal tax rate applicable to his transfers at death or, in the case of a lifetime transfer, by a percentage of his marginal rate applicable to all transfers during the taxable period. This tax is inconsistent with the theory of the generation-skipping proposal because it is computed with reference to the rates applicable to the transferor. The tax should be computed with reference to the rates applicable to the estate of the deceased trust beneficiary, as is usually the case under the Statute. Second, the amount of the additional tax is dependent upon the rate applicable to the transfer. Thus, if the first inter vivos transfer is subject to the additional tax, the amount thereof would be lower than it would be if the same transfer is made at death or later during life after other gifts had been made. It is inappropriate to create an incentive for making transfers subject to the additional tax at any particular time to the extent this result may be avoided. The Statute avoids this problem by, in general, computing the tax with reference to the estate of a deceased trust beneficiary. Third, the election device is ill-advised and injects aspects of a lottery into the computation of tax because it permits the amount of the tax to be determined based upon the value at the time of transfer, or based upon the value at a subsequent time no later than the death of the last beneficiary who is no more than one generation below the transferor. This election, which is different from other elections now available to fiduciaries where it is possible to tell precisely at the time of the election what the tax effect of the election will be, places an undesired responsibility on the fiduciary with respect to the time of its exercise. No election is permitted under the Statute. We have other objections to the Studies' proposal. The substitute tax would be applied to a distribution of current income as well as to a distribution of principal (including accumulated income). Such an approach produces complexity and is uncertain in its application. Το illustrate, assume that a trust were to last until the death of the survivor of the transferor's children, when the property is to be distributed to his then living issue, per stirpes, and that during the trust term the trustee is to distribute income currently to the transferor's issue living from time to time, per stirpes. Every payment of income to a grandchild after his or her ancestor's death and prior to the death of the last surviving child would be subject to an additional tax and transfer tax returns would have to be filed each quarter. What is the source of funds to be used for payment of the substitute tax? Is it the income actually payable to the grandchild or is it the trust principal? If it is income, the substitute tax and the income tax on the gross income payable to the grandchild may exceed this income. If it is principal, the shares of other beneficiaries are adversely and unfairly reduced. The Statute avoids these problems by not being applicable to distributions of current income. Our estate tax law has always recognized a distinction between the limited interest of a trust beneficiary and the person who has outright ownership of property. If the income beneficiary of a trust is treated for estate tax purposes as if he owned the trust property outright, a determination is being made that the ownership rights are equivalent. They are not the same and should not be treated for transfer tax purposes as being the same, as they are under the Studies' proposal, merely because interests in a trust are given to successive generations below the transferor. Professor David Westfall has proposed a much simpler way of handling generation-skipping transfers in trust. Westfall, Revitalizing the Federal Estate and Gift Taxes, 83 Harv. L. Rev. 986, at 1006-1013 (1970). It is to grant a "parental deduction" of forty percent for property transferred outright to a child or in suc.. form that the property will be included in the child's estate at his death. The rationale behind this approach is that the incentive of the immediate tax deduction will be sufficient to prevent the use of generation-skipping transfers in trust. We believe there is considerable merit in this approach because it avoids the complexity inherent in any other approach, including that present in the Statute. Unfortunately, Professor Westfall does not have sufficient confidence in his solution to ignore the multiple generation-skipping trust. For such trusts, his February 27, 1973 statement on estate and gift tax reform to the House Committee on Ways and Means says that the Statute is preferable to the proposal of the Studies, if coupled with an uncomplicated solution, such as in his view the parental deduction, for a single generation-skipping trust.* The multiple generation-skipping trust should be ignored if Professor Westfall's approach should be followed. Subject to this qualification and to the qualification that the deduction would be available for * Professor Westfall's comments were based on a preliminary version of the Statute dated February, 1971. He raised four technical points which in his view required correction. of these points is met in the Statute. Each |