Will Moore Be the End of Estate Tax Planning?

Austin Bramwell is co-leader of the trusts and estates group of Milbank LLP in New York and an adjunct professor of law at New York University School of Law, and Raquel Begleiter is an associate in the trusts and estates group of Milbank.

In this article, Bramwell and Begleiter argue that a Supreme Court ruling in Moore that would limit income attribution could threaten long-standing assumptions that underlie wealth transfer planning.

The Supreme Court may not yet have handed down its decision in Moore,¹ but the verdict among observers is already unanimous: The Moores failed to persuade the Court that they were unconstitutionally taxed on unrealized income.² Just minutes into oral argument, the Moores’ attorney, Andrew Grossman, conceded that the income at issue in Moore was indeed realized — by the corporation of which the Moores were shareholders.³ The argument thereafter focused not on whether income was realized (it was) but on whether that income could be fairly attributed to the Moores.

With the controversy framed that way, the Moores’ campaign to have the mandatory repatriation tax under section 965 struck down as unconstitutional appears all but lost. After all, as the justices themselves were quick to note, the tax system has for decades, in a variety of circumstances, attributed corporate income to shareholders. The Moores declined to argue that (for example) the subchapter S and subpart F attribution regimes were unconstitutional. That strategic decision, though it enabled them to disavow the more radical potential implications of their argument, put the Moores in the uncomfortable position of having to explain why the mandatory repatriation tax is unconstitutional while other attribution regimes are not. Few, if any, believe the Moores managed to do so successfully.

While Moore may not upend 100 years of tax law, it may yet have dramatic, if unanticipated consequences. For estate tax planners in particular, the Court’s interest in attribution may end up destroying assumptions that for decades have undergirded wealth transfer planning for wealthy individuals. In particular, it has long been assumed that a grantor’s payment of tax on the income of a grantor trust is not a gift for gift tax purposes. As discussed below, however, a Supreme Court holding that places limits on income attribution could undermine that assumption. If that happens, it would be the end of grantor trust planning in its current form.

II. Moore’s Pivot to Attribution

Moore was supposed to be a case about realization. After all, it reached the Supreme Court only after a Ninth Circuit panel had expressly, if improvidently, held that Congress has the power under the 16th Amendment to tax unrealized income without apportionment among the states.⁴ That holding occasioned a vigorous dissent (from the circuit’s denial of a motion for en banc review) from Judge Patrick J. Bumatay, who, joined by three of his colleagues, argued that unrealized income is not “income” within the meaning of the 16th Amendment.⁵ To the surprise of many, the Supreme Court granted the Moores’ petition for certiorari to decide whether “the Sixteenth Amendment authorizes Congress to tax unrealized sums without apportionment among the states.”⁶

By the time of oral argument, the Court seemed less than eager to decide that very question. The justices repeatedly asked about the consequences of potential rulings for the existing tax system, which suggests a reluctance to throw vast swaths of tax law into constitutional doubt.⁷ Moreover, as Grossman unequivocally conceded in response to Chief Justice John G. Roberts Jr.’s first question, there was, in fact, realization of income by the corporation. Justices Amy Coney Barrett, Sonia Sotomayor, and Brett M. Kavanaugh then pressed Grossman on whether the case presented a realization issue at all.⁸ Barrett even wondered if attribution was really a Fifth Amendment due process clause issue. If the answer is yes, then the Moores cannot prevail, for they did not even raise a due process clause argument before the Court.⁹ The Moores appeared to be in trouble.

As the government was arguing its case, Justice Neil M. Gorsuch took the discussion in a surprising direction. Likening the distinction between realization (including constructive realization, in the Moores’ formulation) and attribution to a difference in pronunciation (“po-tay-to/po-tah-to,” as he put it), Gorsuch suggested it was the same phenomenon by different names. He also did not think that attribution was a Fifth Amendment due process issue. The Court’s earlier cases, he suggested (apparently alluding to Heiner v. Mellon¹⁰), had not analyzed attribution under the due process clause.¹¹ In response, Solicitor General Elizabeth Prelogar, without agreeing with Gorsuch’s interpretation of the Court’s prior cases, conceded that the Court could analyze constitutional limits on attribution under the 16th Amendment.

Finally, Gorsuch asked Prelogar to articulate a list of factors that would be relevant to whether income could be constitutionally attributed to a taxpayer. In the ensuing dialogue, Prelogar agreed that any of the following factors would suffice to justify attribution of income:

• whether the taxpayer controls the entity;

• evidence of or potential for fraud in the use of the entity;

• the taxpayer’s overall relationship to the entity and its income;

• whether the attribution of income is unrelated to any privilege or benefit associated with doing business through the particular kind of entity; and

• the difficulty of otherwise collecting taxes on the income.

Gorsuch was pleased. “You’ve given me a very helpful list of factors from this Court’s history and practice, consistent with our precedent,” he said. In response, Prelogar merely cautioned against adopting a test for attribution that is based on whether the taxpayer had sufficient control. She did not reject the premise that there may be some 16th Amendment limit on attribution.

Needless to say, nobody can predict exactly how the Court will rule. But based on the oral argument, it seems safe to predict that the Court will not decide whether realization is a necessary feature of income within the meaning of the 16th Amendment. More speculatively, though still plausibly, the Court may address in some fashion the extent to which Congress can attribute income from one individual or entity to another. Perhaps the Court will incorporate the factors that Gorsuch and Prelogar discussed. Given the Court’s apparent reluctance to unsettle deeply entrenched tax regimes, any holding on attribution is likely to continue to give Congress wide latitude. Yet as discussed below, even a modest holding could have radical implications for estate tax planners.

III. An Inadvertent Gift Tax Exclusion

For decades, estate planners have exploited — to their clients’ advantage — a particular set of income attribution rules. Those rules, known as the “grantor trust rules,” are set forth in sections 671-679. When they apply, they cause a grantor or other person to be treated as the owner of a portion or all of the trust property. If ownership is attributed under section 671, all income, deductions, and credits of the deemed owned portion of the trust must be reported by the grantor or other person. Thus, it is possible for the grantor, rather than the trust, to be liable for the tax on a trust’s income.

When the grantor trust rules were enacted, Congress apparently assumed that most grantors would attempt to avoid deemed ownership.¹² After all, deemed ownership would cause more income to be taxed to the grantor at the grantor’s (generally higher) tax bracket and thereby deprive the grantor of the ability to shift taxable income to another taxpayer. Decades later, however, that assumption proved incorrect. Grantors instead began to welcome grantor trust status because it turned out to be a remarkably efficient vehicle for wealth transfer.

Two features of grantor trust treatment have combined to make grantor trusts popular planning vehicles. First, there is a mismatch between the grantor trust rules on the one hand and estate and gift tax rules on the other. That is, it is possible for gifts to an irrevocable trust to be considered complete for gift tax purposes¹³ and for the trust to pass outside the grantor’s estate for estate tax purposes,¹⁴ yet at the same time for the trust to be treated as owned by the grantor for income tax purposes.¹⁵ In those circumstances, the grantor remains liable for tax on the income and gains from the trust property, even though the trust property will pass free of estate tax at the grantor’s death.

Second, the payment of taxes by the grantor is not considered a gift to the trust or its beneficiaries for gift tax purposes. Rather, by paying income taxes on grantor trust income, the grantor is satisfying a legal obligation to the federal government. As the gift tax only applies to voluntary transfers, not to transfers that are compelled by a legal constraint, the grantor’s payment of tax cannot be treated as a gift.¹⁶ Conceding that principle, Treasury and the IRS held in Rev. Rul. 2004-64, 2004-2 C.B. 7, that the grantor’s payment of income taxes for a grantor trust is not a gift for gift tax purposes.

The result is a vast unwritten exclusion from the taxation of wealth transfers. By creating an irrevocable grantor trust for the grantor’s descendants, the grantor can effectively give his or her descendants tax-free returns during the grantor’s lifetime. Meanwhile, the grantor’s own estate is depleted by the payment of income taxes; economically, the depletion is equivalent to a deduction from the grantor’s taxable estate. Finally, at the grantor’s death, if the trust is properly drafted, the trust will pass free of estate tax so long as the grantor has not retained any taxable interests or powers that can pull the trust property back into the grantor’s gross estate. In summary, as the Senate Finance Committee’s Democratic staff correctly observed in 2017, the irrevocable grantor trust “allows the transfer to avoid estate and gift taxes on any appreciation that occurs after the initial transfer to the trust, while shifting income tax liability from the trust to the grantor, preventing income taxes from reducing the value of the property in the hands of the trust or beneficiary.”¹⁷

Not surprisingly, grantor trusts have become as infamous to policymakers as they are dear to the hearts of estate planners. Legislation introduced in the previous Congress would have generally subjected grantor trusts to estate or gift tax when grantor trust status ends.¹⁸ Treasury’s 2025 green book includes a proposal to treat the payment of a grantor’s payment of tax on grantor trust income as a gift. Shortly thereafter, Senate Finance Committee Chair Ron Wyden, D-Ore., introduced a bill to implement the green book proposal.¹⁹ So far, however, no reforms to eliminate the mismatch between the grantor trust rules and the estate and gift tax completion rules have gained momentum. And at least until recently, it seemed clear that a statutory change would be needed.

IV. The Extreme Reach of the Grantor Trust Rules

But the need for statutory reform may end up being largely superseded by Moore. The potential flaw in grantor trust planning is that it takes advantage of rules that go to extreme and dubious lengths to attribute income from one taxpayer, the trust, to another, the grantor. As noted, for gift and estate tax purposes, with a properly drafted irrevocable grantor trust, the grantor is considered to have completely divested himself or herself of ownership and control of trust property. Yet the grantor trust rules may still attribute ownership to the grantor, based on remarkably attenuated if not illusory connections to the grantor. Indeed, grantor trust status can be — and often is — achieved under the grantor trust rules even if the grantor literally has no control or beneficial interest whatsoever.

Examples abound. To start with a common one, a trust instrument can give any person, including a person other than the grantor, the power to substitute trust property with property of an equivalent value. According to the IRS’s own guidance, conferring that power, so long as it is exercisable in a nonfiduciary capacity, is enough to cause the grantor to be treated as the owner of the entire trust.²⁰ This result is obtained even though the grantor cannot benefit from the exercise of the substitution power, cannot direct how the power will be exercised, and cannot cause the power to be renounced.

Similarly, grantor trust status will generally arise when the trustees have discretion to distribute income and principal among a class of beneficiaries, such as the grantor’s descendants.²¹ There is an exception in the code if at least half of the trustees are not related or subordinate to the grantor (and certain other requirements are met). But that exception is negated if a person has the power to add to the class of beneficiaries.²² Thus, if a person has that power, the grantor will be treated as the owner of the trust, even if the grantor has no power to direct how the power will be exercised.²³

In response, it may be noted that the grantor at least chooses what the terms of the trust will be and who will hold powers over the trust. But over time the grantor’s initial choice may have little bearing on how trust property is administered. Trusts can last for longer than a lifetime, even as relationships change. Clearly, the grantor cannot be assured of always having influence over the same individuals who are initially designated in the trust instrument to have the powers causing grantor trust status. Moreover, it is not even true that the grantor’s initial designations will last. If a vacancy in a position that triggers grantor trust status arises, deemed ownership may continue even if the person designated to fill the vacancy is appointed by someone other than the grantor. Indeed, cautious planners, mindful that Rev. Rul. 2004-64 could be revoked, take care to put grantor trust triggers beyond the power of the grantor to terminate, so that the grantor trust status can never be treated as voluntary for gift tax purposes, even if Treasury concludes that Rev. Rul. 2004-64 went too far.

Perhaps the most extreme example of attribution in the grantor trust rules is attribution through the powers and interests of a former spouse. Under section 672(e), a grantor will be treated as holding any power or interest held by any individual to whom the grantor was married when that power or interest was created. Spousal attribution continues to apply even after the marriage ends, since the attribution is based on the spouse’s marital status at the time of creation, regardless of marital status thereafter. Thus, if a grantor includes a spouse as a beneficiary or designates the spouse as trustee, grantor trust status may continue even after the spouses divorce, no matter how acrimoniously. In other words, under the grantor trust rules, the grantor may be treated as the owner of a trust thanks to the powers of an individual who, as a practical matter, not only may not defer to the grantor’s wishes but may be actively spiteful.

The grantor trust rules overreach to such an absurd degree that grantors, finding themselves no longer willing or able to pay taxes on income that in no way belongs to them, have even litigated — unsuccessfully — to terminate grantor trust triggers. In Millstein v. Millstein,²⁴ for example, a grantor was unable to renounce the powers that created grantor trust status. According to the court, because only a trustee or a beneficiary could seek a court modification of a trust, and both in the case opposed the grantor’s efforts to terminate grantor trust status, the grantor trust triggers could not be eliminated. Still, for income tax purposes, despite the grantor’s total lack of control, beneficial enjoyment, and even informal influence, grantor trust status in Millstein presumably continued. Unlike the case of an ownership interest in a business entity, the grantor cannot give away or abandon ownership, for the grantor literally has no property interest that can be transferred or abandoned. Grantor trust status is a tax prison from which there is apparently no escape — except, perhaps, on constitutional grounds.

V. The Effect of a Moore Holding on Attribution

As noted, a decision in Moore may suggest that there is at least some 16th Amendment limit on Congress’s ability to attribute income from one taxpayer to another.²⁵ If the Court holds that the 16th Amendment imposes a limit on attribution, then if any code provisions violate that limit at all it would be the grantor trust rules. Consider only the list of factors that Prelogar offered to Gorsuch. None of them would justify all the forms of income attribution that are common in estate planning.

A. Control

Perhaps the most straightforward factor justifying attribution is a taxpayer’s control over the entity that actually realizes or owns the income. So central is the concept of control that Prelogar felt obliged to caution the Court not to treat it as the only test of whether income may be constitutionally attributed. In any event, as discussed, control by the grantor may be completely absent, yet the grantor trust rules may still cause attribution. Indeed, grantor trust status may continue to apply even though the gift tax test for surrendering dominion and control has been satisfied.²⁶ Control, therefore, does not justify or explain the broad reach of the grantor trust rules.

B. Fraud

Unscrupulous promoters have for decades promised willing audiences that income taxes can be avoided simply by holding assets in trust. These schemes are often shut down on sham transaction grounds.²⁷ Undoubtedly, the grantor trust rules, which automatically attribute trust income to the grantor, aid in enforcement by attributing income even in the absence of substance-over-form principles.²⁸ But the highly arbitrary and formalistic grantor trust rules were hardly designed to target fraud. Their purpose, rather, was to prevent the shifting of income when the grantor did not truly relinquish dominion and control.²⁹ Certainly, there is no connection between the rules that enable estate planners to achieve grantor trust status and any fraudulent device for hiding income. The goal of grantor trust planning is to pay taxes on income, not avoid them.³⁰

C. Relationship to Entity and Income

Perhaps the most nebulous of the Gorsuch-Prelogar factors is the overall relationship of the taxpayer to the entity and its income. Application of this factor would be highly fact dependent. In any event, the grantor trust rules can cause attribution even when the grantor has no relationship to the trust and its income, other than having created the trust many years before. Further, unlike in the case of a business entity such as a corporation or partnership, the grantor of an irrevocable grantor trust typically has no ability, even in theory, to have the income returned to the grantor. Indeed, as noted, unlike shares in a corporation or other equity interest in an entity, the grantor does not even have a property interest in an irrevocable grantor trust that may be renounced, transferred, or abandoned. In other words, there is often not even a relationship to the grantor trust and its income that can be severed.

D. Privilege or Benefit to Use of Entity

The Gorsuch-Prelogar factor that potentially goes the furthest in justifying any scheme of attribution is whether a particular form of ownership confers a “privilege or benefit,” for which attribution is the price. Understood that way, taxpayers submit (or potentially submit) to attribution whenever they choose one arrangement over another. Thus, in Burnet v. Wells,³¹ the Court held that the income of a trust that paid premiums on a life insurance policy on the grantor’s life could be taxed to the grantor, without violating the Fifth Amendment due process clause, because the grantor obtained a “privilege or benefit” from the use of the trust to provide life insurance protection to the grantor’s dependents and relatives. Incorporation of the privilege or benefit factor into 16th Amendment analysis would by itself give Congress broad powers to attribute income, including from trusts.

Still, the privilege or benefit factor would not necessarily justify all applications of the grantor trust rules. In Burnet v. Wells, the Court assumed that the grantor was personally benefiting from trust income, insofar as it was used to pay premiums on a life insurance policy held for the benefit of the grantor’s dependents. In many other circumstances, however, there is no conceivable privilege or benefit to the grantor of transferring assets to the irrevocable trust in question. And according even to the government’s lawyer in Moore, there must be a privilege or benefit to the selection of a particular form of entity to justify attribution. The mere exercise of freedom to divest oneself of assets is apparently insufficient to justify the attribution of income from the transferee back to the transferor. Yet under grantor trust rules, individuals can completely surrender all control and beneficial ownership of property and still have income from the transferred property attributed to them. That form of attribution likely goes beyond what even the privilege or benefit factor would permit.

E. Administrability

The final Gorsuch-Prelogar factor is whether attribution would aid in the administration of the tax system. Simply put, the grantor trust rules do not do so. On the contrary, they if anything introduce needless complexity in the income tax system. Anecdotally, for example, it is generally believed by practitioners that hardly any grantor trust actually satisfies its information reporting obligations correctly. In any event, the IRS is perfectly capable of taxing non-grantor trusts as separate taxpayers, as it has done since the income tax’s origins. It does not need attribution to collect tax on trust income.

In short, it is difficult to see how the Gorsuch-Prelogar factors can justify the extreme reach of the grantor trust rules. This is not to say that all attribution under the grantor trust rules is vulnerable to 16th Amendment attack. On the contrary, most applications of grantor trusts will continue to pass constitutional muster, just as they do today. A power to distribute or accumulate income to the grantor³² or to revest property in the grantor,³³ for example, would continue to be enough to attribute income to the grantor, even if Moore announces some limits on attribution. Similarly, income from foreign trusts, like income of foreign corporations, may continue to be subject to the broad attribution regime of section 679.³⁴

But wealth transfer planning with grantor trusts often does not rely on those straightforward instances of grantor trust attribution. Rather, grantor trust status is more often achieved through artificial provisions that deliberately exploit the formalistic grantor trust rules. Examples include third-party powers to substitute assets, to direct loans without adequate security, or to expand the class of beneficiaries. These powers are widely assumed to trigger grantor trust status, even though none of them is a means of retained grantor control, a fraudulent device, an indication of any relationship of the grantor to the trust income, or a privilege or benefit to the grantor. They may not survive, post-Moore, as constitutionally valid mechanisms for achieving grantor trust status.³⁵

Grantors, in other words, might wake up the morning after Moore is decided and discover that they have a constitutional right not to be taxed on grantor trust income. Under current law, the grantor trust rules enable them to take the position that they are paying taxes on trust income involuntarily and therefore that they cannot be considered to make a gift just because the federal government chooses to impose tax liability on them. But if grantors have a constitutional right not to be taxed on grantor trust income, but pay the tax anyway, then they are voluntarily satisfying an obligation that should, constitutionally, be considered an obligation of someone else. And if that is the case, then the grantor would potentially be treated for gift tax purposes as making a taxable gift every time the grantor pays tax on trust income. To bring about that result, the IRS would merely need to announce, in the wake of Moore, that it is reconsidering the extent to which the grantor trust rules can be constitutionally enforced.

Thus, if Moore explores the limits of attribution, even gingerly, estate tax planning as we know it may radically change. The use of grantor trusts to transfer wealth may come under scrutiny, even without an act of Congress. It would be ironic indeed if Moore, which was coordinated by conservative and libertarian groups generally hostile to wealth transfer taxation, ended up shoring up the estate and gift tax system. But history’s muse may have a sense of humor, even when it comes to taxes.