Congress ought to adopt the Biden administration’s proposal to tax unrealized gains over $1 million per person at death—full stop. But with the Biden administration’s sensible proposal now facing resistance from congressional Democrats, eyes are turning toward carryover basis as a compromise. One political advantage of carryover basis is that it’s very hard to brand carryover basis as a “death tax” (though well-funded opponents will no doubt try). In a carryover basis regime, death is a non-event for income tax purposes. You can say it’s a “death tax” just as you can say that an apple is a banana, but the statement would have little relationship to reality.
If Democrats decide to go down the carryover basis path, though, it matters which carryover basis regime they would be adopting. Not all carryover basis regimes are created equal. All conceivable carryover basis regimes would be better than the status quo, but some would be considerably better than others.
“True” vs. Modified Carryover Basis
Congress first adopted carryover basis for gifts in 1921. Under the original carryover basis regime, the basis of property acquired by gift was the donor’s basis at the time of the gift—end of story. One might call this “true carryover basis.”
In 1958, Congress modified the original carryover basis regime to allow a step-up for gift tax paid with respect to gifted property. For property gifted on or after September 2, 1958, the donee’s basis was the donor’s basis at the time of the gift plus the amount of gift tax paid with respect to the gift (though never more than fair market value at the time of the gift). The Tax Reform Act of 1976 further modified the modified carryover basis regime so that the step-up for gift tax paid would apply only to the amount of gift tax attributable to unrealized gain.
The Tax Reform Act of 1976 also replaced stepped-up basis at death with a prospective carryover basis rule. Under the prospective carryover basis rule, an heir’s basis was stepped up for any appreciation that occurred before December 31, 1976, and then stepped up again for the portion of estate taxes attributable to post-1976 appreciation. In 1980, Congress retroactively repealed this part of the 1976 Act—though it left in place the 1976 Act’s regime for gifts.
Reading the tea leaves, the version of carryover basis at death that’s likeliest to make it through budget reconciliation is an estate tax equivalent of the current regime for gifts: modified carryover basis with a step-up for estate tax paid with respect to unrealized gain. But true carryover basis would raise meaningfully more revenue and would be significantly easier to administer.
To see how true carryover basis and modified carryover basis differ, consider the following example:
D holds MegaCorp stock at death. D’s basis in the stock at the time of death is $50 million. The fair market value of D’s stock at the time of death is $100 million. The estate tax rate is 40 percent. The capital gains tax rate is 30 percent. D has exhausted her unified credit through lifetime gifts.
True Carryover Basis. The basis of MegaCorp stock in the hands of D’s heirs is the same as D’s basis: $50 million. Therefore, D’s heirs will owe $15 million in income tax with respect to pre-death appreciation when they sell the inherited stock (30 percent capital gains rate x $50 million gain).
Modified Carryover Basis (i.e., Step-Up for Estate Tax Paid with Respect to Unrealized Gain). The basis of MegaCorp stock in the hands of D’s heirs is $70 million: D’s $50 million basis plus the $20 million of estate tax attributable to unrealized gain. D’s heirs will owe $9 million in income tax with respect to pre-death appreciation when they sell the inherited stock (30 percent capital gains rate x $30 million gain).
Under the 1958 version of modified carryover basis, the basis of MegaCorp stock in the hands of D’s heirs would be stepped up to $90 million, and D’s heirs would owe $3 million in income tax with respect to pre-death appreciation when they sell the inherited stock.
A few observations:
— There’s a lot of money at stake in the choice among carryover basis regimes. Robert Avery, Daniel Grodzicki, and Kevin Moore estimate that unrealized gains constitute 29 percent to 45 percent of the gross estate for households subject to the estate tax. Separately, the Tax Policy Center estimates that when the temporary 2018-2025 increase in the estate tax exemption expires, annual estate tax liabilities will be roughly $30 billion per year. So relative to true carryover basis, and ignoring behavioral responses for now, a modified carryover basis regime like the current law for gifts would result in roughly $9 billion to $14 billion per year in additional upward basis adjustments (more as time goes on).
— Under true carryover, individuals subject to the estate tax would potentially have an incentive to realize gains before death. In the running example, D would pay $15 million in income tax if she sold her MegaCorp stock right before death, and the estate tax on D’s estate would be 40 percent x ($100 million - $15 million) = $34 million. Thus, total income-plus-estate taxes would be $49 million. By contrast, if D held her MegaCorp stock until death, the estate tax would be $40 million, and an additional $15 million in income tax would be due upon the ultimate sale of MegaCorp stock. Thus, total income-plus-estate taxes on pre-death appreciation would be $55 million (with a portion of that deferrable).
Which would D prefer: $49 million in tax before/at death versus $40 million in tax at death plus $15 million in tax for her heirs sometime in the future? The answer depends upon interest rates and how long D expects her heirs to defer. If D realizes gains shortly before death in order to reduce estate tax, then true carryover basis would produce essentially the same result as deemed realization at death. (Under the Biden administration proposal for deemed realization at death, the income tax due on deemed realizations would be subtracted from the decedent’s gross estate.)
— Administratively, modified carryover basis would be complicated. Under a modified carryover basis regime, the basis of property acquired from a decedent wouldn’t be knowable until the decedent’s federal and state estate tax liabilities are resolved. Federal estate tax returns are due nine months after the date of death, with a six-month extension. So, for example, if D dies in mid-November 2022, the extended due date for D’s estate tax return would fall in mid-February 2024. If D’s estate or D’s heirs sold the MegaCorp stock before the end of December 2022, the capital gain would be reported on a tax year 2022 return with an extended due date of October 15, 2023. So the capital gain might have to be reported before basis is knowable.
Post-audit adjustments complicate matters further. The audit rate for large estates is high: 22 percent of estate tax returns reporting a gross estate of $10 million or more were audited in fiscal year 2019. The vast majority of those audits (82 percent) ended in a recommended change. Some of these cases take years (potentially even more than a decade) to reach a resolution. So under a modified carryover basis regime, basis may remain in flux for a while.
The administrative challenges of modified carryover basis aren’t insuperable. After all, we seem to have made this work—more or less—for inter vivos gifts. But differences between the gift tax and estate tax make modified carryover basis harder for the latter. In particular, gift tax uncertainty matters only for the basis of property gifted in the same calendar year. By contrast, estate tax uncertainty potentially affects the basis of all property in the decedent’s gross estate. A valuation dispute with respect to one item can affect the basis of all others.
— The rationale for modified carryover basis isn’t particularly compelling. The strongest argument for adjusting basis to reflect gift tax paid with respect to unrealized gain is that it eliminates the incentive under true carryover basis to sell property immediately before death. But why is that such a bad thing? If deemed realization at death is the normative ideal, and Democratic leaders are resorting to carryover basis only because deemed realization at death is politically unattainable, then shouldn’t Democrats try for the carryover basis regime that comes closest to deemed realization at death?
— How about true carryover basis with optional deemed realization? Still, maybe it’s sad to think about anyone spending her last moments of life calling her broker to initiate a sale. (Indeed, that is a sad thought!) One way to address this would be by allowing the executor of the decedent’s estate to opt into deemed realization at death (with the income tax due on the deemed realization then subtracted from the decedent’s taxable estate). Not all executors would make the election, but the existence of the option would undermine any argument that true carryover creates a horizontal inequity between individuals who realize gains right before death and individuals whose estates realize gains right afterwards. Everyone would be able to opt into treatment at least as favorable as if they had realized gains right before death.
— The choice between true carryover basis (with or without optional deemed realization) and modified carryover basis is probably too in-the-weeds to matter politically. It’s certainly hard to imagine many flesh-and-blood midterm voters caring whether the Democrats’ carryover basis regime does or doesn’t allow a step-up for estate tax attributable to unrealized gains. Going with true carryover (or true carryover with optional deemed realization) rather than modified carryover would allow Democrats to raise extra revenue—quite progressively—and reduce administrative burdens without a clear political cost.
— For political reasons, it may be necessary to preserve tax-free stepped-up basis for the first $1 million per person of unrealized gains. The executor of the decedent’s estate would then be responsible for allocating the $1 million allowance across assets. This, too, will raise administrative and compliance costs, but it’s otherwise hard to see a political path forward for carryover basis.
— Finally, carryover basis at death—whether true or modified—is a heck of a lot better than tax-free stepped-up basis at death. Well-advised high-net-worth individuals try to die without leaving large taxable estates, but they still benefit from tax-free stepped-up basis. One way is by borrowing against appreciated assets to fund lifetime consumption (“buy-borrow-die,” to borrow Ed McCaffery’s now-famous phrase). At death, the estate can sell the appreciated assets free of income tax and use the proceeds to repay the loan. Another way is through deathbed swaps (described in more detail in this recent white paper by Bob Lord and me on estate and gift tax loophole closers). In the deathbed swap strategy, a high-net-worth individual who is the deemed owner of an irrevocable grantor trust borrows as much as she can shortly before her death and then swaps the borrowed cash into the trust in exchange for appreciated assets of equivalent value. The swap takes unrealized gains out of the trust and leaves the trust with cash. Upon the individual’s death, her estate will sell the appreciated assets—again free of income tax—to repay the loan.
In both the buy-borrow-die and deathbed swap strategies, the estate sells the appreciated assets shortly after the individual’s death. Therefore, the difference between deemed realization and carryover basis isn’t large. (If carryover basis were the law, some estates might defer sale for longer, but deferral would be profitable under those circumstances only if assets grow at a faster rate than the interest rate on the loan.)
In short: Carryover basis could achieve some of the same objectives as deemed realization at death (albeit with less of the revenue coming within the 10-year budget window). Some carryover basis regimes would come closer to achieving the objectives of deemed realization than others. True carryover basis (potentially with a deemed realization option) would be quite a bit better, from a revenue-raising and progressivity perspective, than modified carryover basis with an estate tax step-up.
*Thanks to Dan Shaviro for exchanges that have informed my thinking on this subject. And if you’re interested in receiving these posts as e-mails, subscribe for free at the top of the page.