Tax Notes logo

Taxing Gains at Death

Posted on Jan. 11, 2021
[Editor's Note:

This article originally appeared in the January 11, 2021, issue of Tax Notes Federal.

Harry L. Gutman
Harry L. Gutman

Harry L. Gutman is of counsel at Ivins, Phillips & Barker Chtd. He served as attorney-adviser and deputy tax legislative counsel in the Treasury Office of Tax Policy from 1977 to 1980 and as chief of staff of the Joint Committee on Taxation from 1991 to 1993. He thanks Richard Kinyon, Martin A. Sullivan, and Jon Talisman for their invaluable comments and technical review. Any errors are the author’s.

In this article, Gutman argues that the tax-free basis step-up at death should be repealed and replaced by a regime in which death and lifetime transfers are income tax realization events.

This proposal is the author’s own and should not be attributed to Ivins, Phillips & Barker.

Copyright 2021 Harry L. Gutman.
All rights reserved.

President-elect Joe Biden has proposed addressing income inequality in part by eliminating the tax-free step-up in basis for property passing at death.1 Although details are lacking, the proposal would address a glaring omission from a comprehensive income tax base — an omission that exacerbates vertical inequity and produces horizontal inequity, inefficient resource allocation, and significant revenue loss.2

In this article, I argue that tax-free step-up should be repealed and replaced by a regime in which death and lifetime transfers are income tax realization events.3 That would mean that when an individual dies owning property that has appreciated in value, tax would be payable on that gain rather than forgiven as it is under current law. Conforming rules would treat lifetime transfers similarly.

Section II of this article details the structural design elements of the solution — a task that has not been undertaken in the recent past.4 It also addresses the effective date and transition. There, the issue is the scope of the proposal — that is, whether it applies to all assets held on the effective date, appreciation occurring after the effective date, or assets acquired after the effective date. It is my view that specification of an acceptable transition regime is the key to the political acceptability of this proposal.

I. Background

A. Current Law and Policy Concerns

The Joint Committee on Taxation lists more than 230 income tax provisions as tax expenditures.5 An economic or social policy objective can be cited for virtually all of them. However, try as one might, no one can create a plausible tax, social, or economic policy justification for tax-free step-up. Indeed, in a world in which the reach of the estate tax is limited to less than 0.1 percent of all decedents,6 a provision that forgives the income tax on unrealized appreciation is particularly indefensible.

In the past, a principal tax policy justification for the estate tax was to ensure that those with capital income that has been exempted from current taxation bear their fair share of the tax burden.7 In other words, by subjecting the full value of included property to tax, the estate tax acted to protect the progressivity of the income tax system by capturing unrealized appreciation in the tax base. But that backstop role has diminished as the estate tax exemption has climbed. The estate tax at current exemption levels no longer achieves that role. Moreover, the estate tax is overinclusive in any event, because all property — including property with previously taxed appreciation — is subjected to the tax. Indeed, even if the estate tax exemption were returned from its current level of $11.58 million per taxpayer to its 2009 level of $3.5 million per taxpayer, it would not discharge its historic policy role in any meaningful way.

Here is an illustration of what is at stake:

Example 1: Twenty years ago, Ann Smith dropped out of college and started a tech company. Her ownership shares in that company are now worth $1 billion. For 20 years, she paid income tax on her $1 million salary until her retirement this year. But on the bulk of her income — the $1 billion of appreciation on her shares of stock — she has been allowed the tax benefit of deferring tax until sale. Even though she would like to diversify, it is safe to assume she will not sell the bulk of her shares during her lifetime because the tax-free step-up in basis at death creates a significant inducement to hold her shares. That is the lock-in effect, described later. If Smith needs cash, she can borrow, using her shares as collateral. Under current law, upon her death, the benefit of tax deferral on the increase in value of her shares — in this case, many decades of deferral — is transformed into tax forgiveness.

Smith’s situation is rare among the general public. Most income for low- and middle-income individuals, and even mildly wealthy people, is generated through wages, salaries, and business income on which they pay tax without deferral at ordinary income tax rates. Smith’s situation, however, is commonplace among the superrich. The bulk of unrealized appreciation in property other than homes and retirement accounts is held by a small fraction of the wealthiest people. Nobody becomes a billionaire through savings from paychecks. And the amount of that billionaire income is hidden from view because it is unrealized. Most economic analyses of the tax burden across income classes do not include this important component of economic income, purely for lack of data.

The proposal described later would tax Smith’s unrealized capital gain at death. She would still enjoy the benefits of decades of deferral of tax on her unrealized gain and the preferential rates provided to capital gains under current law. Twenty years of deferral, assuming a discount rate of 2 percent, cuts the effective tax rate by approximately one-third. Forty years of deferral cuts the rate by more than half. Assuming a 30 percent preferential capital gains tax rate, the effective rates of tax on Smith’s income would be approximately 20 percent and 14 percent, respectively.

The horizontal inequity that can result from tax-free step-up in basis at death is illustrated by a simple example.

Example 2: A and B are siblings. Each bought Stock X for $100,000. It is now worth $3 million, and each has decided to sell. A meets B in the street outside their broker’s office just after A has executed her trade and before B is going to do the same thing. A car hits them and both die. Assuming a 20 percent income tax rate and no estate tax, B’s heirs receive $2.42 million. A’s heirs get the stock with a new basis of $3 million and can sell it the next day and pocket the entire $3 million.8 That’s indefensible.9

As illustrated in Example 1, tax-free step-up induces individuals to hold property until death, which is an artificial impediment to sales that would normally occur. It therefore distorts capital flows. This phenomenon is called “lock-in.”

Finally, as is obvious, the failure to tax unrealized appreciation results in a major revenue loss. The JCT has estimated the tax expenditure revenue loss to be $218 billion for 2020-2024, plus $19 billion for carryover basis for gifts.10

B. A Bit of History

Recognition of a problem is one thing; solving it in a political world is a very different matter. Consider the following efforts:

  • In 1942 Treasury proposed that beneficiaries carry over the decedent’s basis in property transferred at death.11 (In a carryover basis regime, gains are not taxed at death; rather, the beneficiaries retain the transferor’s original basis and pay tax on all the gain when the property is sold.)

  • In 1963 President Kennedy proposed taxing the unrealized appreciation in property held at death.12

  • A 1969 Treasury study contained an extensive discussion of the reasons to change current law, as well as a comprehensive legislative proposal.13

  • In 1972 the American Bankers Association proposed an additional estate tax to apply to unrealized appreciation in property held at death.14

  • The Tax Reform Act of 1976 provided a carryover basis regime, ensuring the taxation of unrealized gain when the property is sold by the heirs. (Although the unrealized gain in property transferred at death will ultimately be subject to tax, carryover basis is a distinctly second-best solution because, as illustrated in Example 1, it continues the significant benefit of tax deferral until the property is sold by the transferee.) The 1976 provision proved controversial and was ultimately repealed before going into effect.15

  • In January 1977 the outgoing Republican Treasury proposed treating transfers by gift or at death as realization events.16

  • In connection with phased-in estate tax repeal under the 2001 tax act, a version of carryover basis actually went into effect for property acquired from a decedent in 2010.17

  • President Obama proposed treating transfers of appreciated property as a sale in his fiscal 2016 and 2017 budgets.18

It is fair to ask why, if both Democratic and Republican Treasury Departments have identified tax-free step-up as a problem and proposed essentially identical solutions, a permanent solution has not been enacted. I suspect the answer lies in the nature and distribution of the tax benefit. As noted earlier, most taxpayers do not have significant amounts of unrealized appreciation in securities. To the extent they have interests in appreciated property other than their residence, the bulk of that property is held in tax-favored retirement accounts and is subject to ordinary income tax when distributed. Those taxpayers are unaware that a problem exists, and consequently, they do not exert political pressure to resolve it. However, politicians are keenly aware of the issue (and its potential effect on their wealthy contributors). The combination of popular ignorance and political “sensitivity” produces a powerful inducement to dismiss the problem.

C. Anticipating the Response

My proposal will undoubtedly resurrect the arguments that have emerged in the past when an immediate recognition regime has been suggested, as well as the arguments used to support the repeal in 1980 of the carryover basis regime that Congress enacted as part of TRA 1976. As a Treasury official at the time, I was intimately engaged in the three-year legislative battle over carryover basis.

There were basically three reasons why carryover basis was repealed after heavy lobbying by the American Bankers Association and the tax and estate planning bar.19 The first was the alleged difficulty of ascertaining the basis of property when the owner dies. The second was the complexity of the rules intended to ensure that when property is subject both to estate and income tax, the total tax burden is the same whether the property is sold before or after death. The third was perceived inadequate transition relief. Each is discussed below.

1. Forced sales of assets.

As detailed later, under my proposal, deathtime recognition would be limited to marketable assets (for example, stocks, bonds and mutual funds in which most Americans hold their wealth) which, by definition, are liquid. While it is true that funding the payment of that tax may require the sale of some assets, a market exists to eliminate liquidity problems. Potential bunching problems from applying progressive tax rates to the deferred gain on all included assets could be addressed through averaging conventions. Moreover, unless immediate recognition is elected, the tax due for illiquid nonmarketable property would be deferred until that property is sold. There would be no forced sale of those assets.

2. Lack of basis records.

Whether one finds credible the claim that basis records somehow disappear at death (recall that they were necessary to establish gain or loss on a lifetime sale), the proposed system would put taxpayers on notice prospectively of the need for basis records, and it would provide relief for problematic asset areas — namely, personal residences and noncollectible tangible personal property. Also, a lookback rule would be provided to determine unknown basis. Under that rule, the basis of an asset could be determined by discounting its value at the relevant tax date back to its acquisition date.

3. Interaction with the estate tax.

The income tax payable on recognized gain would be deductible from the decedent’s gross estate for estate tax purposes. Thus, the total income and estate tax burden on property subject to the realization regime would be the same whether that property is sold before death or deemed to have been sold just before death.

4. Inadequate transition relief.

Transition has proven to be a particularly difficult issue. There is no agreement on whether the regime should apply to all realizations occurring after the effective date, only appreciation occurring after the effective date, or assets acquired after the effective date. As detailed later, each option has a different effect. Ideally, transition should not create winners and losers. But sometimes the search for equitable transition rules produces perceived inadequate relief or political impracticality and, as a consequence, a desirable law change is not enacted. In my view, we should not let the perfect be the enemy of the good. The most important objective is the enactment of the general rule; transition relief is, by definition, temporary.

D. Proposal

The challenge is to design a system that is comprehensive in coverage, administrable, perceived as fair, and not subject to abuse. The task involves not only identifying the assets that would be subject to the regime but also the exclusions and exemptions deemed necessary to promote administrability and perceptions of fairness. One must resolve the fundamental question of what constitutes the taxable unit to which the new regime would apply. The need for and design of antiabuse protections must be examined. Further, provisions to assist in valuation problems and the determination of unknown basis must be devised. And finally, one must select an effective date provision that will be viewed as acceptable by the affected community of taxpayers and advisers. The proposal reflects my judgment on the appropriate balance of conceptual purity, administrability, and political reality.

E. General Rule

Subject to specified exclusions and exemptions, all deathtime and lifetime transfers would be income tax realization events. All non-spousal deathtime transfers of marketable assets20 would be recognition events. All non-spousal lifetime transfers would be recognition events. Recognition could be elected by the taxpayer-transferor for any realized gain on an asset-by-asset basis. The tax on realized but unrecognized gain, together with a deferral charge to equate the total payment to what would have been due upon realization, would be payable when the subject property is later disposed of in a recognition event.

This proposal directly addresses the complete exemption of the bulk of the wealth accumulation of the superrich from income tax. The details are discussed in Section II.

II. Technical Explanation

A. Property Held at Death

1. Death is an income tax realization event.

The objective of the proposed regime is to capture all of a decedent’s accrued gains and losses at death and to ensure that the resulting total tax liability will be collected in present-value terms.

Subject to the effective date rule and the exemptions and exclusions noted later, death would be treated as an income tax realization event for all property owned by a decedent, as well as any property subject to a general power of appointment in the hands of the decedent.21 The fair market value of that property, which would become the tax basis in the hands of any recipient, would be determined at death (or at the estate tax alternate valuation date, if applicable), with the same choice being required for both estate tax and income tax purposes. Property to which special estate tax valuation rules would apply, without regard to whether the decedent is subject to the estate tax, would be valued in accordance with those rules.22

2. Recognized gain is limited to marketable assets.

The extent to which realized gain is to be recognized is a critical issue on which proposals have varied over the years. The 1969 Treasury proposal, Treasury’s 1977 “Blueprints for Basic Tax Reform” proposal, and the Obama administration proposals would have treated death as a recognition event for all assets, other than those subject to exemptions or exclusions. Each proposal provided some form of liquidity relief.23

In my judgment, the prior proposals failed to recognize the practical and — arguably more important — political issues surrounding the problem of liquidity. It is one thing to fund the tax payment through the sale of marketable property. By definition, that property is liquid. It is quite another to require a current tax payment associated with nonmarketable assets, such as a farm or small business, especially because a forced sale in those circumstances may materially affect the amount that could be realized at that time. While it is conceivable to borrow against the value of the asset to pay the tax, that is not an ideal solution. Nor is a deferred payment schedule, which is economically equivalent to an installment loan from the government and would still require raising cash to pay each installment. To alleviate this problem, I propose that the immediate recognition of gain be limited to marketable assets other than those transferred to a surviving spouse (and for which recognition is not elected) or subject to one of the exclusions described later.

Net recognized gain, determined after accounting for any net realized loss from nonmarketable property, would be taxed separately from all other income realized during the decedent’s final tax year, at the rate then applicable to capital gain. The gain would be spread over five years, so the effect of progressive rates would be moderated.24 A net recognized loss would first be used, on a pro rata basis, to offset net realized gain from nonmarketable property held at death. Any remaining loss may be used to offset net capital gain on the decedent’s final income tax return, and to the extent necessary, may be carried back to the three preceding tax years. Losses in excess of net capital gain for the preceding three years may be used to offset ordinary income commencing in the year immediately preceding death.

3. Treatment of nonmarketable property.

Nonmarketable property for which recognition has not been elected, other than property subject to the exclusions described later, would be valued at death in accordance with the valuation principles stated earlier. Net realized gain in excess of any applicable net recognized loss would be calculated. Net recognized loss would be allocated pro rata with realized gain. The tax payment for each nonmarketable asset would be deferred until the property was sold or transferred inter vivos in a recognition transaction that resulted in the transferee no longer being treated as the owner of the property for income tax purposes, at which time the gain would be taxed at the rate in effect on the date of death of the decedent, together with an interest charge that would equate the present value of the deferred tax with the tax that would have been paid at death.25 Realized losses in excess of realized gains on nonmarketable property (net realized losses) could be used to offset net recognized gain, with any excess treated in the same manner as net losses on marketable property.

The transferor of any nonmarketable asset may elect recognition treatment on an asset-by-asset basis.

The transferee of a nonmarketable asset would be liable for the deferred tax upon a taxable disposition (as well as any tax due on the disposition itself). For multiple nonrecognition transfers — such as if A dies and transfers the property to B, and then transferee B dies, and the property passes to C — a separate realization event would occur. The realized gain at B’s death would be equal to the difference between the value of the property at B’s death and the basis of the property (determined by reference to the value of the property when B received it). If C disposes of the property in a tax transaction, C would be responsible for the payment of deferred taxes on both the transfer from A to B and the transfer from B to C, as well as any tax due on the recognition event. Gain on all realization events would be reported on a separate information return that would also report the basis of the transferred property in the hands of the transferee. The tax attributable to the gain reported on that return, together with the appropriate deferral charge, would be due at the same time as the tax attributable to the gain realized when the property is disposed of in a taxable transaction.

The deferred tax liability would be secured by a lien on the subject property.

4. Examples.

Example 3: M dies owning marketable security A with a value of $100 and a basis of $10, marketable security B with a value of $200 and a basis of $100, and marketable security C with a value of $100 and a basis of $150. M recognizes gain of $140 ($90 + $100 - $50).

Example 4: N dies owning marketable security A with a value of $100 and a basis of $10, and nonmarketable asset B with a value of $100 and a basis of $50. No election is made to treat the $50 gain on asset B as a recognition event. N will recognize gain of $90. The realized gain of $50 attributable to asset B will be taxed according to the deferred gain rules upon its disposition in a recognition event by the transferee.

Example 5: O dies owning marketable security A with a value of $100 and a basis of $200. A has no unrealized gain. O recognizes a loss of $100, which may be used on O’s final income tax return.

Example 6: P dies owning marketable security A with a value of $100 and a basis of $300. P owns nonmarketable asset B with a value of $200 and a basis of $100. P also owns nonmarketable asset C with a value of $400 and a basis of $100. The net loss of $200 attributable to marketable security A is allocated $50 to asset B and $150 to asset C. Thus, the net realized gain for asset B is $50, and for asset C it is $150.

Example 7: Q dies owning marketable security A with a value of $200 and a basis of $50. Q owns a nonmarketable asset with a value of $50 and a basis of $100. The realized loss of $50 offsets the recognized gain of $150, leaving a net recognized gain of $100.

Example 8: R dies owning no marketable assets and nonmarketable asset A with a value of $100 and a basis of $200. The realized loss of $100 may be used on R’s final income tax return.

5. Marital transfers.

Marital transfers present special issues. If the objective of measuring all gains and losses at the decedent’s death is implemented, the traditional treatment of deathtime spousal transfers must be revisited.

Prior proposals have treated the marital unit as a single taxpayer for realization purposes. Thus, transfers to a surviving spouse would not be treated as realization events, and the decedent’s basis in the transferred property would carry over to the recipient spouse. However, the decision to treat the marital unit as a single taxpayer is a policy choice that introduces several issues that must be addressed.

First, a marital exemption gives an executor the incentive to transfer low-basis property to a surviving spouse and transfer high-basis property to other beneficiaries. This incentive can be mitigated by creating an average basis mechanism that allocates total basis of the decedent’s assets in accordance with their relative FMVs. However, an average basis mechanism could be difficult to administer because the reallocated basis would depend on both the basis and the value of every other asset, each of which might be uncertain. Second, the assets used to fund a marital bequest would have to be identified before the income tax liability of the decedent was calculated. Thus, an executor would be required to claim tentative exemptions on the decedent’s income tax return that reflected the property to be transferred after death. Amended returns would be required if the distribution differed from that originally claimed on the return.

An alternative, which would be consistent with the objective of determining all gain and loss at the decedent’s death, would recognize the special status of the marital unit but remove the gaming opportunities and administrative difficulties presented by treating the marital unit as a single taxpayer. Unless recognition was elected, all transfers to a surviving spouse in any form that would qualify for the estate tax marital deduction would be treated as realization-but-nonrecognition events in accordance with the rules governing outright transfers of nonmarketable property. The effect would be to determine the tax due for all property held by the decedent but defer the tax on appreciated property transferred to a surviving spouse until the earlier of the sale of the property, the inter vivos transfer of the property (whether marketable or not), the death of the surviving spouse in the case of marketable property, or the taxable disposition of nonmarketable property in the hands of a testamentary distributee of the surviving spouse. Immediate recognition could be elected for loss property.

6. Charitable transfers.

Charitable transfers of fee interests in property would not be treated as realization events. The conceptually correct result is to treat the transfer as an income tax realization event and allow a charitable contribution deduction in the amount of the value of the transferred property. Under current law, a charitable transfer is not a realization event. Political pragmatism would likely prevail, and attempting to overturn a long-standing rule seems a bridge too far at the moment.

There are alternative possibilities involving a split system in which gain would be realized for either lifetime or deathtime transfers but not both. Non-realization would, of course, provide an incentive to make transfers during the period to which that regime applied. There is no principled basis on which to make this determination, but it could provide a rationale for imposing the correct answer in at least some circumstances.

Transfers of partial interests in property to a charity would be excluded to the extent the value of the interest would be deductible for federal income tax purposes. In effect, therefore, remainder interests would generally be in the form of charitable remainder annuity trusts or unitrusts, or pooled income funds, and charitable income interests would be in the form of guaranteed annuity interests or unitrusts. This rule would avoid the difficult valuation issues presented by unusual types of charitable bequests.

When a qualified charitable interest either precedes or succeeds the interest of a noncharitable beneficiary, the amount realized that is attributable to each interest would be determined in accordance with the present value of the respective interests. The decedent’s basis would be allocated in the same proportion. Gain or loss would be realized to the extent of the difference between the portion of the amount realized allocated to each noncharitable interest and the basis allocated to that interest. Gain or loss would be recognized to the extent the transferred interest consists of marketable property. Special rules would have to be developed to determine when gain attributable to the transfer of nonmarketable assets would be recognized.

7. Miscellaneous items.

a. Life insurance.

Proceeds of life insurance on the death of the decedent would not be subject to realization.26

b. Income in respect of a decedent.

The regulations state:

In general, “income in respect of a decedent” refers to those amounts to which a decedent was entitled as gross income but which were not properly includible in computing his taxable income for the taxable year ending with the date of his death . . . under the method of accounting employed by the decedent.27

Examples of income in respect of a decedent (IRD) include compensation for services, accrued dividends and interest, amounts in qualified retirement plans, and amounts due on installment sales. Under current law, those items do not receive a basis step-up under section 1014.28 Rather, the item is taxable to the estate or beneficiary at the same time and in the same manner that it would have been taxed had the decedent lived.29

The issue is whether to continue the current-law treatment of IRD, tax all IRD at death, or adopt a hybrid approach. The 1969 Treasury proposal would have repealed section 691 and required recognition of IRD at death.30 The rationale for the repeal was that the IRD rules “were designed to avoid bunching of ordinary income in the decedent’s final return.” The 1969 proposal contained special rules to alleviate perceived bunching problems, so Treasury concluded that section 691 was no longer necessary.

Under my proposal, items of IRD would be valued at death, and the realized gain would be recognized as the deferred payments are received. Thus, inclusion under the proposal would not create a liquidity issue. The rule would require valuation of the deferred amount. In this respect, however, it would be no different from any other nonmarketable asset subject to this regime. Moreover, taxation of IRD items in this manner would be consistent with the underlying policy goal of determining the total tax liability of a decedent at death. IRD items are property of the decedent, and if liquidity issues are adequately addressed, there is no reason for the items to receive special treatment.

8. Exclusions.

a. Basic exclusion.

Any proposed regime must take into account the practical aspects of its implementation and should not impose undue burdens on decedents’ estates. On the other hand, the regime must be comprehensive enough to achieve its objective of subjecting unrealized appreciation to tax.

There are several ways to limit the applicability of the regime. One is to provide a minimum basis amount.31 A second is to provide an exclusion for a defined amount of gain.32 A third is to exempt all estates below a specified level.

A minimum basis proposal would affect only taxpayers whose aggregate asset basis is below the minimum. It is not immediately apparent why this limited form of relief would be appropriate. In any event, it would require a convention to allocate basis among assets.

A gain exclusion would provide relief to all taxpayers with aggregate gain. It too would require a convention to allocate the exclusion among assets. The simple solution would be to allocate the exclusion pro rata among the gain assets. Under either rule, the minimum basis adjustment or gain exclusion should be applied first to capital gain property and then to property whose disposition would result in ordinary income.33

The exemption of estates below a specified level is superficially attractive because it would eliminate the need to compute gain or loss for estates below the threshold, and it would avoid allocation problems. It would have disparate impacts depending on the amount of unrealized appreciation in each exempt estate. But more importantly, the cliff effect of cutting off the application of the rule at a specified dollar amount would simply be unfair.

An exclusion is the preferred result.

b. Nonbusiness tangible personal property.

All prior proposals have recognized that attempting to tax nonbusiness tangible personal property creates difficult issues. There are valuation and basis identification problems, and as a practical matter, most of that property (other than collectibles34) does not appreciate in value. The question is how to fashion an appropriately targeted exclusion.

There are at least three issues. The first is whether the exclusion should be limited by a dollar amount. The second is whether, if so limited, the exclusion should apply on an asset-by-asset or aggregate basis. The third is how to allocate the exclusion among affected assets.

I believe that the complexity and administrative burden connected with any dollar-limited or dollar-allocated exclusion cannot be justified. Thus, I would endorse Treasury’s 2016 budget proposal to exempt all nonbusiness tangible personal property other than collectibles as defined in section 408(m)(2).

c. Principal residence.

Consistent with the treatment accorded a lifetime sale, a $250,000 exclusion would apply to gain realized on a principal residence that qualifies under section 121. The exclusion would be portable for transfers to a surviving spouse.

9. Deductibility of tax.

Any currently payable tax would be deductible from the decedent’s gross estate. Deferred tax presents a different issue. The deferred tax is not a liability of the decedent; it is payable by the owner of the property at the time it is disposed of in a taxable transaction. However, the actual value of the property that passes to a recipient is net of the deferred tax liability. Thus, the deferred tax should reduce the value of the asset in the decedent’s estate.

10. Payment of tax.

The tax on recognized gain, calculated as described earlier, would be payable by the estate of the decedent. Deferred tax would be payable by the owner of the property at the time of recognition.

11. Periodic imposition of tax.

The question arises whether it would be necessary to require periodic payment of the deferred tax attributable to nonmarketable property. If the deferral charge on the deferred payment is accurate, this would be an unnecessary complication.

12. Foreign tax considerations.

The extent to which the deathtime tax could be offset as a result of deathtime recognition under the taxing system of another country must be further explored.

B. Lifetime Transfers

1. Generally, income tax recognition events.

The guiding principle of this proposal is that, to the extent practicable and consistent with other applicable income tax rules, all transfers of property should be accounted for in the tax return of the transferor at the time of the transfer. Consequently, apart from interspousal transfers, charitable transfers, and transfers of nonbusiness tangible personal property (other than collectibles), all gratuitous lifetime transfers would be recognition events. Recognition (rather than realization) is appropriate because the lifetime transfer is a voluntary event, and liquidity issues can be anticipated and resolved. An election to treat interspousal transfers as recognition events is provided in the proposal. As explained later, noncharitable transfers to trusts would be recognition events when and to the extent the transferor is no longer treated as the owner of the property for income tax purposes. Special rules would be provided for the creation of joint tenancies.

2. Outright transfers.

Except as provided below, an outright transfer of property other than to a spouse or charity would be a recognition event. Gain would be recognized to the extent of the difference between the FMV of the property and its basis. Losses would not be recognized for transfers of nonmarketable property or property transferred to related parties within the meaning of section 267. The recipients of such property would receive a carryover basis.

3. Spousal transfers.

Consistent with the general principle of the proposal and the suggested deathtime transfer rule, outright transfers to a spouse (as well as transfers that qualify for the gift tax marital deduction) would be treated as realization events. Recognition would occur upon the disposition of the property by the transferee spouse, with the tax determined in accordance with the rules set forth earlier.35 However, immediate recognition of gain property could be elected.

4. Charitable transfers.

Lifetime charitable transfers would be treated in the same manner as deathtime transfers. Thus, outright transfers would not be subject to the regime, and split interest or transfers in trust would be treated as described earlier.36

5. Transfers in trust.

Transfers in trust in which the transferor retains no powers or interests would be recognition events.37 Transfers in trust in which the transferor retains a power or interest pose a special problem of determining the taxable event. This is because under the transfer tax, a transfer of property may be subject to tax twice, as illustrated later, whereas under the income tax, there is only a single event that shifts the tax burden from the transferor. There are two options to resolve the problem. The first is to treat a transfer as a recognition event when and to the extent it is subject to transfer taxation. The second is to treat the transfer as a taxable event when and to the extent the transferor is no longer treated as the owner of the property for income tax purposes. The difference in treatment may be illustrated by the following simplified example:

Example 9: Assume that A transfers marketable property with a value of $100 and a basis of $10 in trust, with income payable to A for life and the remainder payable to unrelated-party B. The actuarial value of A’s retained interest is $80. Under current law, A has made a completed gift to B of $20.38 However, A is taxed on the income from the trust until A’s retained interest expires.39 If transfer tax principles govern the determination of the recognition event, A has made a taxable transfer of $20, and the gain recognized is $18 at the time of the transfer. However, the property is also included in A’s estate40 and consequently would be subject to further income taxation at that time. On the other hand, if income tax principles are used to determine the taxable event, recognition will occur only upon the expiration of A’s interest.

The proposal adopts the rule that realization generally occurs at the time and to the extent the transferor is not treated as the owner of the trust corpus for income tax purposes, with recognition dependent on the character of the trust assets. Thus, in some cases, the transfer itself or a subsequent lifetime transfer would constitute a realization event. However, when a transferor has until death been treated as the owner of all or a portion of a trust for income tax purposes, the death of the transferor would be a realization event, with recognition depending on the composition of the taxable interest. Realization could therefore occur, even if the trust corpus is not included in the transferor’s gross estate.

6. Joint tenancies.

In general, the creation of a joint tenancy or similar interest would be treated as a recognition event without regard to whether or to what extent a gift tax might be imposed.

Example 10: Assume stock is purchased by an individual in his name for $10,000. When the stock is worth $20,000, he creates a joint tenancy with another person. If the applicable state law is such that the new joint tenant can sever the property and thereby take stock worth $10,000, the first individual has made a transfer of $10,000 and will recognize gain of $5,000. The new joint tenant has a basis of $10,000, and the transferor has a basis of $5,000 in his remaining interest. If, as a result of applicable state law, the respective interests are valued actuarially, the actuarial values are used to determine the amount of recognition by the transferor and the basis of the transferee.

A transfer to a spouse that qualifies for the gift tax marital deduction would be a realization event in accordance with the spousal transfer rule described earlier.

If the creator of the joint tenancy subsequently transfers his interest to another person by gift, gain would be recognized to the extent of the difference between his basis and the FMV of his interest in the property at the time of the transfer. The recipient of the transferred interest would have a basis equal to the FMV of the interest at the time of transfer. If the joint tenant donee transfers her interest by gift, gain would be recognized in an amount equal to the difference between her basis and the FMV of her interest at the time of the transfer.

7. Exclusions.

a. Nonbusiness tangible personal property.

Transfers of nonbusiness tangible personal property other than collectibles would be excluded.

b. Principal residence.

As with deathtime transfers, a $250,000 exclusion would apply to gain realized on a principal residence that qualifies under section 121. The exclusion would be portable for a transfer to a spouse.

c. No present interest exclusion.

The gift tax present interest exclusion would not apply.

C. Effective Date/Transition

The proposed regime would be very different from current law. Perhaps the most important practical decisions in its implementation would be the choice of an effective date, and if necessary, the provision of transition relief. Those decisions will require a careful balancing of theoretical purity with a calculated appraisal of political reality.

1. Scope.

There are three basic choices regarding the scope of the deathtime regime. First, the regime could apply to all realization events occurring after a specified date. Thus, all pre-effective-date appreciation would be captured.41 Second, the regime could apply only to appreciation occurring after a specified date.42 Finally, it could apply only to assets acquired after a specified date.

Application of the regime to all realization events occurring after the effective date would treat all taxpayers similarly regardless of when the appreciated asset was acquired. As the JCT has noted, “One could argue that the absence of a transition rule raises a question of fairness for taxpayers who have made decisions based on present law to retain appreciated assets in anticipation of death.”43 It is, of course, arguable whether a taxpayer has a protected right to the tax treatment in effect when he acquires an asset. Moreover, such an approach would put pressure on basis identification and on the need to provide rules to establish basis when it is unknown. More practically, apart from the fiscal 2016 and 2017 Obama Treasury proposals, there has been general agreement that at the least, only appreciation occurring after the effective date should be subject to the new regime.

The 1969 Treasury proposal, Treasury’s 1977 “Blueprints for Basic Tax Reform,” and the 1976 carryover basis regime all provided that only appreciation occurring after the effective date would be subject to the regime. In practice, that solution foundered on the alleged difficulty of establishing a new basis for nonmarketable property on the effective date. One solution, of course, is to get an appraisal — a potentially costly exercise that would invite controversy with the IRS. Another is to devise a “discount back” approach in which the date of death value would be discounted back to the effective date at a prescribed rate of interest.44

The third alternative is to apply the new regime only to assets acquired after the effective date of the statute. For policy purists, this is an unacceptable solution precisely because it would codify the questionable expectancy of those who bought assets before the change in the law and would be uneven in its effect over the short term. That is a powerful policy argument. However, there are at least two counterarguments.

First, apart from issues specifying how the regime would apply for post-effective-date transfers to, and acquisitions by, pre-effective-date entities, the rules are clear: Basis must be kept for all post-effective-date acquisitions. Second — and most important in my mind — is the likelihood that this is the proposal that is most likely to survive the political process.45 As noted earlier, to me, the goal is the codification of the right rule. Transition relief is, by definition, temporal. The resulting revenue loss is the price to be paid to “purchase” the correct result.

2. Lifetime transfers.

All lifetime transfers other than to charity would be realization events. This rule would apply only to transfers of assets acquired after the effective date. Thus, current law would apply to the transfer of any pre-effective-date assets. However, pre-effective-date assets acquired after the effective date by lifetime transfer would be treated as post-effective-date acquisitions by the transferee. The basis of those assets would be the carryover basis of the transferor, as under current law. The gain realized upon the disposition of those assets by the donee would be equal to the difference between the FMV of the asset and its basis in the hands of the donor. The only difference from current law is that a subsequent lifetime transfer or retention at death would become a realization event to the donee.

3. Effective date.

To reduce planning opportunities, the new regime should apply to decedents dying and transfers made after the date on which the implementing legislation is introduced in Congress.

4. Antiabuse rules.

If the new regime applies only to property acquired after the effective date, a mechanism would have to be devised to capture the appreciation attributable to post-effective-date transfers of property transferred to pre-effective-date assets (such as partnerships or corporations) in nonrecognition events. The simplest rule would be to have a pro rata portion of the owner’s interest subject to the realization regime.

Example 11: Assume that A is the sole owner of X Corp., which has a value of $1,000. A makes a post-effective-date transfer of $100 to the corporation. One-eleventh of the value of the X stock would be subject to the new regime.

More significant would be the incentive for existing entities to make post-effective-date acquisitions that would otherwise be made by their owners. Although this would probably not be a major problem for publicly traded entities, it would be for closely held entities.

Example 12: Assume that A, the owner of X, a pre-effective-date closely held entity, would like to acquire Z Corp. after the effective date. If acquired outright by A, Z would be subject to the new regime. If acquired by X, absent a look-through rule, Z would not be subject to the new regime.

This is a potentially significant problem that does not have an easily administrable solution. Conceptually, post-effective-date acquisitions that occur outside the normal course of the entity’s business, narrowly defined, should be treated as post-effective-date acquisitions. However, defining the transactions that would be subject to this rule and devising a mechanism to ensure adequate tracing would be a daunting task.

5. Transition relief.

Application of the new regime only to assets acquired after the effective date would alleviate the need for general transition relief. Notwithstanding the fact that taxpayers would be on notice that they would be required to maintain basis records for post-effective-date transfers, there could still be circumstances in which basis records are unavailable or difficult to determine. In those cases, the “discount back” approach, in which the value at the date of the realization event is discounted back to the appropriate acquisition date at a prescribed rate of interest, could be used to determine basis.

III. Conclusion

The time to repeal section 1014 is long overdue. Replacement with the suggested regime would enhance horizontal equity, economic efficiency, and tax revenue. Most importantly, at a time when income inequality is growing to what many consider unacceptable levels, the proposal would directly address what is perhaps the most glaring loophole in the income tax: the complete exemption of the bulk of the superrich’s wealth accumulation from income tax. Hopefully, this article provides a roadmap to help accomplish those important objectives.


1 There is no official document that states this proposal. However, it has been widely reported in the media and has not been denied.

2 See, e.g., Harry L. Gutman, “Reforming Federal Wealth Transfer Taxes After ERTA,” 69 Va. L. Rev. 1183 (1983); Jerome Kurtz and Stanley S. Surrey, “Reform of Death and Gift Taxes: The 1969 Treasury Proposals, the Criticisms and a Rebuttal,” 70 Colum. L. Rev. 1365, 1381 (1970); and Michael J. Graetz, “Taxation of Unrealized Gains at Death — An Evaluation of the Current Proposals,” 58 Va. L. Rev. 830 (1973).

3 This article is dedicated to the memory of Stanley Surrey, who provided me with impeccable guidance and advice and for whom redressing the tax-free step-up problem was a major unfulfilled goal. It is also dedicated to Donald C. Lubick, my mentor at Treasury and my senior partner in the trenches of the carryover basis battles from 1977 to 1980. And I must acknowledge my friend Michael Graetz, who has encouraged me over the years to mount Rocinante and tilt at this particular windmill.

4 For previous discussions, see Treasury, “Tax Reform Studies and Proposals,” 91st Cong., 1st Sess., pt. 3, 331-351 (1969) (1969 Treasury proposals); Lawrence A. Zelenak, “Taxing Gains at Death,” 46 Vand. L. Rev. 361 (1993); and Graetz, supra note 2. For a more recent discussion, see American College of Trust and Estate Counsel, “Report by the Tax Policy Study Committee on Proposals to Tax the Deemed Realization of Gain on Gratuitous Transfers of Appreciated Property” (Oct. 2019).

5 JCT, “Estimates of Federal Tax Expenditures for Fiscal Years 2020-2024,” JCX-23-20, at 2 (Nov. 5, 2020); JCT, “Estimates of Federal Tax Expenditures for Fiscal Years 2020-2024,” JCX-23-20, at 2 (Nov. 5, 2020) (“Tax expenditures are defined under the Congressional Budget and Impoundment Control Act of 1974 . . . as ‘revenue losses attributable to provisions of the Federal tax laws which allow a special exclusion, exemption, or deduction from gross income.”).

6 Urban-Brookings Tax Policy Center, “How Many People Pay the Estate Tax?” in Tax Policy Center Briefing Book (2020).

7 See Gutman, supra note 2, at 1187-1197.

8 Assuming a 30 percent estate tax, A’s heirs receive $2.1 million, and B’s heirs receive $1.68 million.

9 And that isn’t all. If A borrowed money using the stock as collateral, A had the use of the appreciation during her life and the debt can be repaid from the tax-free proceeds of the sale of the stock after A’s death. There is nothing fair about that, either.

10 JCX-23-20, supra note 5, at 29.

11 See Hearings on Revenue Revision of 1942 Before the House Committee on Ways and Means, 77th Cong., 2d Sess., vol. 1 at 89 (1942) (statement of Randolph Paul).

12 See Hearings on President’s 1963 Tax Message Before the House Committee on Ways and Means, 88th Cong., 2d Sess., at 20, 49, and 122 (1963).

13 See 1969 Treasury proposals, supra note 4.

14 Richard B. Covey, “Possible Changes in the Basis Rule for Property Transferred by Gift or at Death,” 50 Taxes 831 (1972).

15 Carryover basis was enacted as section 2005 of TRA 1976. It was retroactively repealed by section 401 of the Crude Oil Windfall Profit Tax Act of 1980.

16 Treasury, “Blueprints for Basic Tax Reform,” at 204 (Jan. 17, 1977).

18 Treasury, “General Explanations of the Administration’s Fiscal Year 2016 Revenue Proposals,” at 156 (Feb. 2015) (2016 Treasury proposals); and Treasury, “General Explanations of the Administration’s Fiscal Year 2017 Revenue Proposals,” at 155 (Feb. 2016) (2017 Treasury proposals).

19 See Howard Hoffman, “The Role of the Bar in the Tax Legislative Process,” 37 Tax L. Rev. 413 (1982).

20 Marketable securities are defined in section 731(c). Actively traded personal property, as defined in reg. section 1.1092(d)-2, would also be subject to recognition.

21 This realization rule would not depend on estate tax inclusion rules.

22 See section 2032A.

23 For example, the 2017 Treasury proposals would have allowed the exclusion for capital gain on some small business stock (section 303), a postponement of the tax due until sale for some small family-owned-and-operated businesses, and a deferred payment plan for nonliquid appreciated assets. 2017 Treasury proposals, supra note 18, at 156.

24 Gain that would be taxed as ordinary income, such as original issue discount items, inventory, or depreciation recapture, would also have an averaging convention.

25 The design of the deferral tax is an important component of the proposal. The present value of the tax liability is known; it is the future value that must be calculated. The variables are (1) the period between the realization and recognition events and (2) the interest rate to apply. The latter will, of course, vary, perhaps considerably. The IRS could publish tables to use to approximate the actual interest rates over the appropriate period. (See the later discussion in the text for the treatment of successive nonrecognition transfers.) As an alternative, on an elective basis, interest on the deferred tax at then-current rates could be payable annually.

26 Although not part of this proposal, the internal interest buildup on permanent life insurance should be taxed currently to the policyholder.

27 Reg. section 1.691(a)-1(b).

28 Section 1014(c).

29 Reg. section 1.691(a)-3.

30 1969 Treasury proposals, supra note 4, at 338.

31 Id. at 342.

32 See, e.g., 2017 Treasury proposals, supra note 18, at 156 (proposing an indexed $100,000 portable exclusion per taxpayer).

33 E.g., inventory or depreciation recapture.

34 See section 408(m)(2).

35 An inter vivos spousal transfer does not raise the same administrative issues as a deathtime transfer. See supra Section II.A.5. Consequently, the decision could be made to exempt those transfers from the general rule and account for the property only upon disposition by the transferee spouse.

36 See supra Section II.A.6.

37 The proposal’s application to lifetime transfers in trust, as well as to deathtime transfers of property included in the decedent’s estate, invites a reexamination of the differing rules regarding the income and transfer tax treatment of transfers to trusts, with the goal of establishing identical rules for the taxable event for both. This is an issue that has been examined in the past. See, e.g., American Law Institute Federal Income Tax Project, “Subchapter J — Income Taxation of Estates, Trusts and Beneficiaries and Income in Respect of a Decedent,” tentative draft No. 12 (Mar. 28, 1984); and Treasury, “Tax Reform for Fairness, Simplicity and Economic Growth,” vol. 2, at 378-380 (Nov. 1984).

38 The transfer tax treatment of this disposition has several widely recognized mismeasurement possibilities, e.g., the transferor’s life expectancy, the interest rate used to determine the value of the retained interest, and the fact that the transferred corpus may be invested in a manner that differs from the interest rate assumed in the initial valuation. Adopting the income tax rule for a completed gift would eliminate these mismeasurement issues. Note that if B is a member of the transferor’s family within the meaning of section 2704(c), the entire transfer would be subject to gift tax. See section 2702.

39 Section 677.

40 Section 2036(a)(1).

41 2016 Treasury proposals, supra note 18, at 157; and 2017 Treasury proposals, supra note 18, at 156.

42 1969 Treasury proposals, supra note 4, at 335; and Treasury, “Blueprints,” supra note 16, at 204.

43 JCT, “Description of Certain Revenue Provisions Contained in the President’s Fiscal Year 2016 Budget Proposal,” JCS-2-15, at 191 (Sept. 2015).

44 This is the opposite of the mechanism to determine the tax due on a deferred sale of nonmarketable property. However, it introduces a new element for consideration: the rate of appreciation in the subject property. The simplest solution is to ignore fluctuations in the rate of appreciation and simply provide a discount rate by table. There are, as expected, more complicated ways to address the problem, each of which would have to be reversed to calculate the acquisition date value (basis). Section 1291 provides an existing model for deferred passive foreign investment company earnings. Others have been suggested by, e.g., Harry Grubert and Rosanne Altshuler, “Shifting the Burden of Taxation From the Corporate to the Personal Level and Getting the Corporate Tax Rate Down to 15 Percent,” 69 Nat’l Tax J. 643, 658-666 (2016); and Alan J. Auerbach, “Retrospective Capital Gains Taxation,” 81 Am. Econ. Rev. 167 (Mar. 1991). See also Senate Finance Committee Democrats, “Treat Wealth Like Wages,” Appendix I: Lookback Taxation of Income From Nontradable Property, at 23-24 (2019) (proposal by Finance Committee ranking member Ron Wyden, D-Ore.).

45 It is my understanding that had Treasury agreed to apply carryover basis only to assets acquired after the effective date, the provision would not have been repealed. And today, 40 years after its repeal, it would be virtually universally applicable.


Copy RID