Capital Letters

Still Another Assault on Crummey Powers

Capital Letter No. 35
March 31, 2014

Treasury’s Fiscal 2015 “Greenbook” proposes to “Simplify Gift Tax Exclusion for Annual Gifts.”  But does this proposal to “simplify” simplify?  The Capital Letters answer is yes.

Dear Readers Who Follow Washington Developments:

The “General Explanations of the Administration’s Fiscal Year 2015 Revenue Proposals” (Treasury’s “Greenbook”) was released on March 4, 2014.  Nine proposals under the heading “Modify Estate and Gift Tax Provisions” (pages 158-72) included without significant change the seven proposals from 2013 (discussed in Capital Letter Number 33), plus two new proposals dealing with the annual gift tax exclusion and the definition of “executor.”  Because of the interest the annual exclusion proposal has generated and the subtle thoughtfulness the proposal reflects, this Capital Letter will feature that proposal.  A future Capital Letter will review the status of the other eight proposals.


This new proposal is on pages 170-71 of the 2014 Greenbook.  Its title is “Simplify Gift Tax Exclusion for Annual Gifts.”

The Greenbook cites Crummey v. Commissioner, 397 F.2d 82 (9th Cir. 1968), rev’g T.C. Memo 1966-144.  Of course “everyone” regards that case as the authority for relying on beneficiaries’ temporary, or lapsing, withdrawal rights with respect to gifts in trust to treat the gifts as if they were made outright to those beneficiaries in the amounts that could be withdrawn, thereby satisfying the “present interest” requirement of the gift tax annual exclusion under section 2503(b)(1) and Reg. §25.2503-3.

But the Crummey case was not really about that.  In both the Tax Court and the court of appeals, the IRS conceded the effectiveness of a lapsing withdrawal power held by a competent adult, including a powerholder who became an adult in the year in question.  The only issue was whether a gift subject to such a power in the hands of a minor who could not legally exercise it alone and for whom no guardian had been appointed was still a gift of a present interest.  The Tax Court said no; the court of appeals said yes.  But, indeed, neither court, nor the IRS, displayed any concern about lapsing withdrawal powers as such.  Hence “Crummey powers.”


The limitation or regulation of Crummey powers has been a popular theme of tax reform proposals.

The Clinton Administration (1998 and 2000)

In the Fiscal Year 1999 Greenbook, published in February 1998 (when Greenbook proposals were just starting to become more wide-ranging), the Clinton Administration declared (page 130) that “[t]he use of the Crummey power is essentially a legal fiction” and proposed to overrule Crummey and limit the annual exclusion to outright gifts.  In the Fiscal Year 2001 Greenbook, published in February 2000 (pages 186-87):

The proposal would conform the gift tax annual exclusion rule to the generation-skipping transfer tax rule [of section 2642(c)].  That is, the gift tax annual exclusion would not apply to any transfer to a trust for the benefit of an individual unless (i) during the life of such individual, no portion of the corpus or income of the trust may be distributed to (or for the benefit of) any person other than such individual, and (ii) if the trust does not terminate before the individual dies, the assets of such trust will be includible in the gross estate of such individual.

Both Clinton Administration Greenbooks targeted the proliferation of Crummey powers among non-beneficiaries or remote beneficiaries, stating:

In the Crummey case, the holder of the withdrawal power was the ultimate beneficiary of the trust.  In more recent cases, such as Estate of Cristofani v. Commissioner, 97 TC 74 (1991), and Estate of Kohlsaat v. Commissioner, 73 TCM 2732 (1997) [T.C. Memo 1997-212], the trust agreement has been drafted to give withdrawal rights to individuals who do not have substantial economic interests in the trust.  Typically, by pre-arrangement or understanding, none of these withdrawal rights will be exercised.

Joint Task Force of Several Professional Organizations (2004)

After the Economic Growth and Tax Relief Reconciliation Act of 2001 had created what turned out to be nearly 12 years of uncertainty and instability in the transfer tax law, ACTEC, the American Bankers Association, the American Bar Association’s Section of Real Property, Probate and Trust Law and Section of Taxation, the American College of Tax Counsel, and the American Institute of Certified Public Accountants formed a joint Task Force on Federal Wealth Transfer Taxes.  Most of the Task Force members were ACTEC Fellows, and Dennis Belcher was the Chair.  The Task Force’s 200-page Report in 2004 analyzed issues arising directly from the 2001 Act (long phase-out, one-year repeal, sunset, carryover basis, the retention of the gift tax, and the like), described possible alternatives to the current transfer tax system, and addressed specific problem areas and potential improvements to the current transfer tax system.  Section 16 (pages 95-99) cited the “complexity and uncertainty” created by the annual gift tax exclusion and described seven alternative simplifications (in four categories) that Congress could consider:

1.   Revise the present interest requirement:

a.   Deny the annual exclusion for gifts made in trust, except trusts for minors that qualify under section 2503(c) (often called “2503(c) trusts”).

b.   Deny the annual exclusion for gifts made in trust, except trusts that meet the requirements of section 2642(c), requiring a single beneficiary of the trust during that beneficiary’s life and requiring inclusion of the value of the trust assets in the gross estate of that beneficiary if the beneficiary dies before the termination of the trust (often called “tax-vested trusts”) (as proposed by the Clinton Administration in 2000).

c.   Provide that a withdrawal power does not meet the present interest requirement.

d.   “Codify” Crummey powers, but expand the lapse-not-treated-as-a-release rules of sections 2041(b)(2) and 2514(e) to apply to the greater of 5 percent of the value of the trust assets or the annual exclusion, thereby obviating partial lapses that produce “hanging powers.”

2.   Eliminate the present interest requirement and allow an annual gift tax exclusion for any gift in trust for a beneficiary if that beneficiary’s interest in the trust is susceptible of valuation.

3.   Limit the number of annual exclusions or the total amount qualifying for annual exclusions in any given year (which the Report admitted could penalize donors with large families).

4.   Expand the availability of section 2503(e) to permit, for example, expenses for assisted living or nursing homes for parents and expenses for room and board and other non-tuition expenses of children in college.

Joint Committee Staff Options (2005)

On January 27, 2005, the Staff of the Joint Committee on Taxation published a 430-page Report entitled “Options To Improve Tax Compliance and Reform Tax Expenditures” (JCS-02-05), as requested in February 2004 by Chairman Grassley and Ranking Member Baucus of the Senate Finance Committee.  Under the heading “Curtail the Use of Lapsing Trust Powers to Inflate the Gift Tax Annual Exclusion Amount” (pages 405-08), the Report, citing Cristofani, described the proliferation of Crummey powers among non-beneficiaries or remote beneficiaries as follows:

In recent years, taxpayers have used Crummey powers to achieve benefits extending beyond the conversion of future interests into present interests. Specifically, taxpayers have taken the position that the holder of the Crummey power need not even be a vested beneficiary of the trust, which creates the possibility of using multiple annual exclusions (one for each Crummey power holder) for what ultimately will be a gift to a single donee, as a practical matter.

The Report set forth three options that Congress might consider for improving the tax treatment of Crummey powers:

1.   Limit Crummey powers to “direct, noncontingent beneficiar[ies] of the trust.”  This would repudiate the broad use of Crummey powers sustained in Cristofani.

2.   Limit Crummey powers to powers that never lapse.  As the proposal acknowledges, “[t]his option effectively eliminates Crummey powers as a tax planning tool.”

3.   Limit Crummey powers to cases where “(1) there is no arrangement or understanding to the effect that the powers will not be exercised; and (2) there exists at the time of the creation of such powers a meaningful possibility that they will be exercised.  This option requires a facts-and-circumstances analysis of every Crummey power.”

These three options were estimated to raise revenue by $200 million, $700 million, and $200 million, respectively, over ten years.


The Obama Administration’s Concerns

The 2014 Greenbook points out that the use of Crummey powers has resulted in significant compliance costs, including the costs of giving notices, keeping records, and making retroactive changes to the donor’s gift tax profile if an annual exclusion is disallowed.  The Greenbook adds that the cost to the IRS of enforcing the rules is significant too.

Like previous proposals, the Greenbook also acknowledges the IRS concern with the proliferation of Crummey powers, especially in the hands of persons not likely to ever receive a distribution from the trust, and laments the IRS’s lack of success in combating such proliferation, again citing Cristofani and Kohlsaat.

The Description of the Proposal

The Greenbook offers the following explanation of the proposal:

The proposal would eliminate the present interest requirement for gifts that qualify for the gift tax annual exclusion.  Instead, the proposal would define a new category of transfers (without regard to the existence of any withdrawal or put rights), and would impose an annual limit of $50,000 per donor on the donor’s transfers of property within this new category that will qualify for the gift tax annual exclusion.  Thus, a donor’s transfers in the new category in a single year in excess of a total amount of $50,000 would be taxable, even if the total gifts to each individual donee did not exceed $14,000.  The new category would include transfers in trust (other than to a trust described in section 2642(c)(2)), transfers of interests in passthrough entities, transfers of interests subject to a prohibition on sale, and other transfers of property that, without regard to withdrawal, put, or other such rights in the donee, cannot immediately be liquidated by the donee.

To “eliminate the present interest requirement” is one of the catchiest introductions of a revenue raiser Capital Letters has ever analyzed.  Let’s see, the proposal would reverse the Service’s losses in Cristofani and Kohlsaat and even take back the Service’s concessions in Crummey by denying present interest treatment of lapsing withdrawal powers held even by current trust beneficiaries.  And it would codify IRS successes in denying present interest treatment for transfers of “interests in passthrough entities” in Hackl v. Commissioner, 118 T.C. 279 (2002), aff’d, 335 F.3d 664 (7th Cir. 2003) (interests in an LLC engaged in tree farming); Price v. Commissioner, T.C. Memo 2010-2 (interests in a limited partnership holding marketable stock and commercial real estate); and Fisher v. United States, 105 AFTR 2d 2010-1347 (D. Ind. 2010) (interests in an LLC owning undeveloped land on Lake Michigan), controversial successes achieved despite the fact that those donors transferred only interests in the entities and never claimed to be transferring “present interests” in the underlying assets.  And for the loss of the ability to make “Crummey gifts” it would substitute a gift-making budget of $50,000 that represents only about three and one-half annual exclusions, implying that a family with two sons, one daughter-in-law, and one grandchild – that’s my family, by the way – is too large.  Thus would the present interest requirement be tightened and made more harsh and arbitrary, confirmed by the fact that the proposal is estimated to raise revenues by $2.924 billion over 10 years.  The present interest requirement eliminated?


What Would Be Left?

Here, as Capital Letters analyzes it, is what apparently would be left as excludable gifts:

1.   Unlimited gifts directly for tuition or medical expenses under section 2503(e).

2.   Gifts up to $14,000 (currently) per donee per year, or $28,000 if split by spouses, consisting of

a.   simple outright gifts (other than gifts of interests in passthrough entities, property subject to a prohibition on sale, assets that, without regard to withdrawal, put, or other such rights in the donee, cannot immediately be liquidated, etc.) and

b.   gifts to trusts described in section 2642(c)(2) – that is, “tax-vested” trusts exempt from GST tax.

3.   Up to $50,000 annually of “mad money” for anything that’s otherwise impermissible or at least suspect.  There wouldn’t even have to be an arguable basis for the annual exclusion under current law; the Greenbook provides the simple example of “transfers in trust.”  If the first sentence of the Greenbook explanation is intriguing for offering to “eliminate the present interest requirement,” the second sentence maintains the theme of suspense by offering to “define a new category of transfers.”  In fact, what the Greenbook calls “a new category” is really everything else – everything that doesn’t meet the tightened present interest requirement.  That is the sense in which the present interest requirement turns out to be eliminated, at least to the extent of $50,000 per year.

Some may be inclined to interpret the proposal only as allowing up to $50,000 of annual exclusion gifts (also subject to the $14,000-per-donee limit) to avoid the tightened outright-or-tax-vested requirement.  That would help explain the comment that “a donor’s transfers in the new category in a single year in excess of a total amount of $50,000 would be taxable, even if the total gifts to each individual donee did not exceed $14,000.”  But, if the proposal were interpreted that way, it is hard to see how it would serve its stated objective of simplification – for example, in many cases it would still be necessary to sort through Crummey powers to identify the donees of additions to a life insurance trust – or how it logically could be described as “defin[ing] a new category of transfers.”


Inevitably, clients and their advisors will judge the proposal with reference to how their gift programs are affected.  Consider a life insurance trust, which the 2005 Joint Committee Staff “Options” (page 405) viewed as a “particularly common” context for using Crummey powers.   Consider a grantor-insured who has been giving a trust $50,000 each year to permit the payment of the premium, completely sheltered from gift tax by Crummey powers in five beneficiaries.  If this Treasury proposal were enacted, that use of Crummey powers would not work, but the $50,000 premium payment would be sheltered by the unrestricted “new category.”  But – and this is critical – the grantor would still be free that year to give the full $14,000 (not just $4,000 as when the Crummey powers absorbed the first $10,000) to each of those five trust beneficiaries.  Today, many clients would balk at the need to make such gifts outright, or only in trust until the donee is 21 under section 2503(c).  But under this Treasury proposal, that trust could continue to an older age, or for the donee’s life, just so long as the value of its assets was included in the donee’s gross estate, which could be achieved with a general power of appointment exercisable only by will.  That would be an upside for many donors, allowing more flexible trust planning while curtailing only the long-term tax planning elements.

Suppose, however, that the annual premium is about twice as much, say $98,000, currently sheltered from gift tax by seven Crummey powers.  The result is not as good, but it is not a disaster.  The grantor-insured would have to use $48,000 of lifetime exemption, or applicable exclusion amount.  But even a grantor who has maxed out the applicable exclusion amount with major gifts would probably find that the annual inflation adjustments to the basic exclusion amount under section 2010(c)(3)(B) would absorb that $48,000 of additional premium.  (Even this year’s inflation increase, the smallest since indexing began in 2012, is $90,000.  It was $120,000 in 2012 and $130,000 in 2013).  Not a welcome option in isolation, but remember that the grantor could still make $14,000 gifts to each of those seven donees, in trust for life if desired.

All these numbers are essentially doubled if married donors gift-split.  It is not clear if the $50,000 amount would also be indexed for inflation, although it would fit better in the current transfer tax landscape if it were.

Meanwhile, addressing the Greenbook objective of reducing compliance and enforcement costs, the proposal would prospectively eliminate the documentation and recordkeeping Crummey powers entail.  Notices to powerholders would be unnecessary and keeping track of unlapsed powers would be required only for past balances, which in many cases would begin to decline more sharply when the annual infusion of Crummey gifts is cut off.  Who among us has escaped the headache of trying to account for years of lapsing and hanging variations, especially when those variations are dependent on the value of an asset like a life insurance policy, which must be determined “at the time of the lapse” under Reg. §25.2514-3(c)(4)?  Add the end of wondering how temporary withdrawal powers may have affected past or current grantor trust status and the relief from audit or litigation hassles, and a really tough and scary looking proposal begins to look quite appealing after all.

Finally, while the proposal would codify the IRS passthrough entity successes in HacklPrice, and Fisher – successes, as mentioned above that many find controversial – those cases do reflect an unbroken chain of precedents, including Tax Court precedents.  Moreover, the proposal actually would waive those revenue-friendly precedents, again to the extent of $50,000 per year, removing at least one of the rocks and shoals of navigating estate planning with closely-held entities.

And Losers

But someone has to lose.  After all, the proposal is estimated to raise revenue by $2.924 billion over 10 years, averaging $292.4 million per year, which, at a 40 percent rate, translates into $731 million of lost annual exclusions per year.  That’s at least 52,000 Crummey powers, even after considering the unrestricted $50,000 annual exclusion and the use of remaining applicable exclusion amounts.  (The revenue estimates, which are in Table S-9 on page 190 of the President’s “Fiscal Year 2015 Budget of the U.S. Government,” are undoubtedly much more sophisticated than that and take into account behavioral changes such as forgoing gifts and the estate tax effect of making more trusts “tax-vested.”  The estimated revenue increases start at $70 million in fiscal 2016 and increase to $605 million by fiscal 2024.)

Certainly, the “proliferators,” the (shall we say?) less timid users of Crummey powers, and maybe people with very large families, will pay more tax or see their clients pay more tax.  The same is true, perhaps to a lesser extent, with entity planning.  Some taxpayers will not like the result.

There will also be some frustration with the tough effective date – simply “effective for gifts made after the year of enactment.”  While that would permit the balance of that year to make Crummey gifts as usual, the current natural limits on Crummey powers would moderate the floodgate effect.  And there is no proposed transition rule to maintain existing law for existing arrangements, as there was in a limited way in the 2000 Clinton Administration proposal.  But it is hard to deny that the maintenance of dual systems for a generation or two, which such transition rules often encourage, is not a pleasant prospect either, and relief from that burden, even at some tax cost, can be welcome.

On balance, a proposal that seemed at first blush to be a harsh and tired recapitulation of past efforts – cryptically, maybe even misleadingly, articulated – turns out to have much to commend it.  It will raise more revenue, which is what revenue raisers do, but the administrative relief this time seems worth it to this observer.  So, over 3,000 words later, is the proposal a simplification?  Yes.


But we cannot leave the discussion of this proposal without at least a passing comment about the history of the ideas it projects.  The limitation of the annual gift tax exclusion for transfers in trust to “tax-vested” transfers patterned after the GST tax exemption in section 2642(c), which is simultaneously an expansion of the use of “2503(c) trusts” by removing the age limit, has been discussed at least since section 2642(c) took on its current form in 1988.  It was the entire Clinton Administration proposal on the subject in 2000. 

But it was the 2004 Report of the Joint Task Force that set forth alternative approaches, reflecting the purpose of the Report to suggest options for congressional consideration and not to be an advocacy piece for particular approaches.  In fact the core of the Greenbook proposal is the blend of the section 2642(c) model (the Task Force’s alternative 1b) and the gross dollar-amount cap (the Task Force’s alternative 3).  Proposals like this reflect inputs from a vast array of sources, including IRS audit and ruling experience, examination of policy objectives, microeconomic analysis, scholarly commentary and other public input, and political balancing.  But any increment of contribution that might be traced to the Task Force Report should be gratifying.  It should keep us thinking, and writing.


Technical proposals like this typically are not immediately and eagerly embraced by Congress, particularly with the White House and the House of Representative controlled by different political parties.  Capital Letters has no prediction.  But, like any ideas, proposals like this do not ever completely go away.  Like recent tax reform proposals that seem so quixotic in the current political climate, these proposals and the thinking behind them go into the library of ideas for future use.  If others agree that this particular proposal reflects thoughtful balancing and could achieve real simplification, it could be checked out of the library sooner and more often than most ideas.

Ronald D. Aucutt

© 2014 by Ronald D. Aucutt. All rights reserved