The primary purpose of using an irrevocable life insurance trust is to exclude the policy proceeds from taxation in the estate of the insured and the insured’s spouse. In the usual case, an unfunded irrevocable life insurance trust will rely on gifts from the trust grantor to provide the funds necessary to pay future premiums. Such gifts are subject to the gift tax. I.R.C §2503(b) provides for a gift tax annual exclusion of up to $11,000 (as indexed) per donee per year for gifts of present interests. Gifts of a "future interest" do not qualify for the annual exclusion. Consequently, a gift of a policy in trust or subsequent gratuitous transfers in trust of cash to pay future premiums would ordinarily be future interest gifts for which no annually excludable amount is available.
But if there were a way to convert these gifts of future interests to present interests, the gift tax exclusion would be available. Ingenuity found a way. The grantor makes a gift of cash to the trust. Upon receipt by the trust, each of the beneficiaries is given a right to withdraw a proportionate share of the cash gift and pocket the money. If the power to withdraw is not exercised within a set period of time, the right lapses. In that case, the trustee would then be permitted by the trust instrument to use the money for the premium payment. (Clearly, the principle works with gifts that are not intended to pay life insurance premiums as well.) The power to withdraw creates a present interest that brings the gift within the gift tax exclusion (to the extent of the $11,000 annual limit). This power has become known as a “Crummey power” after the case of Crummey v. Commissioner, 397 F.2d 82 (9th Cir. 1968), in which the Ninth Circuit held that an unrestricted right to the immediate use, possession and enjoyment of an addition to a trust, whether or not exercised (and whether or not likely to be exercised), makes the transfer a present interest for annual exclusion purposes.
The IRS’s Hostile Reaction
Of course, if Crummey powers are to have their intended planning effect, it’s essential that they not be exercised. A rogue beneficiary who actually took the distribution could ruin the arrangement. Unfortunately, in their zeal to limit this risk, planners sometimes structure the power so that it is difficult or impossible to exercise, for example:
· Withdrawal power that expires in too short a time to be effectively exercisable
· Withdrawal power granted to a beneficiary who is not notified prior to its expiration
· A Crummey power holder who is a minor signs off on a notice of right to withdraw
· A power to withdraw an amount from a trust which does not have sufficient liquidity during the exercise period to pay the withdrawal
· The grantee of the power is made to understand that if the power is exercised he/she will be punished in some way (e.g., disinherited)
Clearly, some of these conditions can cross the line into abuse, and over the years since the Crummey decision the IRS has encountered all of the foregoing situations, and has ruled on numerous factual variations, often holding that the purported withdrawal rights were not substantive, but merely illusory.
The IRS has also challenged Crummey arrangements in the Tax Court, but been rebuffed in significant decisions. In Cristofani v. Commissioner, 97 TC 5 (1991), the Service argued that the Crummey powers granted to the grantor’s five minor grandchildren were illusory, because unlikely ever to be exercised (they had only contingent interests). The Tax Court held that it was sufficient if the arrangement created a legal present interest:
As discussed in Crummey, the likelihood that the beneficiary will actually receive present enjoyment of the property is not the test for determining whether a present interest was received. Rather, we must examine the ability of the beneficiaries, in a legal sense, to exercise their right to withdraw trust corpus, and the trustee’s right to legally resist a beneficiary’s demand for payment . . . Based upon the language of the trust instrument and stipulations of the parties, we believe that each grandchild possessed the legal right to withdraw trust corpus and that the trustees would be unable to legally resist a grandchild’s withdrawal demand.
Since Cristofani, the IRS and the Tax Court have drawn lines in the sand. The IRS responded to Cristofani with AOD 1996-10, in which it rejected the Tax Court’s reasoning and announced that it would continue to challenge Crummey powers which do not represent bona fide gifts of present interests, in the form of rights to immediate access to trust assets; the withdrawal rights must be in substance what they appear to be in form. For its part, the Tax Court rejected the substance-over-form theory in cases such as Estate of Kohlsaat, T.C. Memo 1997-212 (1997), and Estate of Holland, T.C. Memo 1997-307 (1997). Whereas the IRS is willing to draw unfavorable inferences (such as an agreement not to exercise Crummey powers) from the fact that a beneficiary refrains from exercising a power without any economic justification (for example, if the Crummey powerholder is merely a contingent remainderman in the trust), the Tax Court’s position remains that expressed in the quote from Cristofani: what matters is whether the could legally exercise the right to withdraw and whether the trustee could legally resist a demand from the beneficiary.
This standoff makes the IRS’s most recent salvo, in TAM
The trust in question was an irrevocable trust whose assets were two whole life policies and some cash. The settler`s oldest child was the trustee. That oldest child, her spouse, and the settler`s younger child were the individual beneficiaries. Four §501(c)(3) charities were charitable beneficiaries. During the settler`s life the trustee had absolute discretion to distribute all or part of the corpus to the beneficiaries for their education, health, maintenance, and support. The beneficiaries had Crummey powers to withdraw proportionate shares of any transfer made to the trust, up to $10,000 per calendar year. (The charities’ proportions were 25%, 5%, 5%, and 5%.) When a transfer was made to the trust, the trustee was required to notify the beneficiaries, who then had 30 days to exercise their power to withdraw. In short, this was a typical Crummey arrangement.
The trustee never made any distributions to the beneficiaries during the settler`s life, and when the settlor died the assets were distributed proportionately to the beneficiaries. The settlor made some 11 distributions to the trust, and the trustee sent 44 notices of same to the charities. Several of the notices were defective in various ways. Some were undated. Others were sent so far in advance of the transfer to the trust that the Crummey power expired before the transfer was actually made. Some specified greater withdrawal amounts than the terms of the trust permitted, so that the trustee could not have satisfied withdrawal requests if all four charities had made them at once. Nonetheless, the taxpayer claimed that the transfers qualified for the gift tax charitable deduction under I.R.C. §2522, or, more to the point, for the §2503(b) gift tax exclusion.
The IRS’s Analysis
The IRS’s argument that the transfers do not qualify for the §2503(b) exclusion is based, a little curiously, on the common law of charitable organizations. “Under State law, a gift to a charity is considered as held in trust by the charity for its charitable uses. . . . Thus, a charitable gift must be applied by the charity to its charitable activities. The gift is inalienable by the charity for any other purpose. . . . “ Since “a charity’s officer or director is a trustee and fiduciary with respect to the charity and the charity’s property. . . . the officer or director is duty-bound to use care to preserve the charity’s property, protect the property against loss or dissipation, and hold the property exclusively for the charity’s charitable purposes.”
How does this black-letter law apply to the present case? The trustee had absolute discretion to distribute the trust assets, and the individual beneficiaries had ascertainable and therefore enforceable rights to distributions. From these facts the Service concludes: “Thus, any amounts the charities failed to withdraw could have been distributed to the individual beneficiaries during Decedent’s life. . . . if a withdrawal right was not exercised, the forfeited amount would become Trust property and be thereby exposed to dissipation for the private interests of Decedent’s family.” Thus, the failure to exercise the right to withdraw was a breach of fiduciary duty, and “in view of the strict prohibition on the use of a charity’s property for private purposes and the fiduciary obligations imposed on a charity and its directors, it is doubtful that any officer or director of a charity could properly participate in this kind of gamble, where funds charity purportedly controls are to be set aside for private utilization until some future date.”
The IRS’s argument is clear enough, and it seems to lead to a clear and persuasive conclusion: a charitable beneficiary with Crummey powers must exercise those powers. Not to do so would put the charity’s property rights at risk, constituting a breach of fiduciary duty. But this is not the conclusion the Service draws. Instead, it states: “we believe that there was a legal impediment prohibiting these withdrawal powers from ever becoming effective.” There is no further argument for the existence of this “impediment” except for the insinuation that the charities’ failure to withdraw “evidence that at least the charities’ understood that they were legally precluded from actively participating in this withdrawal arrangement that allowed funds to enure for private purposes.” Of course, this does nothing to establish the IRS’s position. A Crummey arrangement is a “gamble” for a charitable beneficiary only if it fails to exercise its withdrawal rights. But this shows the very opposite of what the IRS contends: so far from being legally unable to exercise its Crummey powers, a charity is compelled to exercise them. Since the IRS does not even offer any further characterization of the alleged “legal impediment,” it is hard to resist an uncharitable conclusion: the IRS conjures this impediment solely in reaction to the Tax Court’s position that to show that a Crummey power is illusory, the Service must show that it is legally unenforceable. Seen in this light, TAM 200341002 is another salvo in the Crummey war between the IRA and the Tax Court.
Conclusion: the End of Charitable Crummey Trusts?
The collateral damage of this skirmish could be severe. If the IRS is right, no charitable beneficiary of a trust can have Crummey powers. Clearly, this would be a change in tax planning theory and would require the revision of many existing trust arrangements.
There is little chance that the IRS’s position would be sustained if challenged in the Tax Court. Not only is the “legal impediment” unexplained and unargued, as explained above: it does not even fully meet the Tax Court’s emphasis on legality. In Cristofani and its progeny, the Tax Court stresses not only the beneficiary’s legal ability to exercise the power of withdrawal, but also the trustee’s legal ability to refuse that power. The IRS addresses only the first of these in the TAM. The Tax Court could easily note that the trustee in this case had no legal power to resist a demand from the charities. (In fact, the trustee was required to maintain a reserve to meet withdrawal demands while withdrawal powers were active.) And it could conclude, as above, that the charities were legally required to exercise their Crummey powers, not legally precluded from doing so.
Of course, if that happened the IRS might lose the battle but win the war. For if it became an established principle that charitable beneficiaries are required to exercise their Crummey powers, tax planners will stop drafting Crummey trusts with charitable beneficiaries.