Examining the history of the back-and-forth between taxpayers and the Internal Revenue Service over income shifting helps us gain a clearer idea of what isn’t possible and what’s required to be able to shift income successfully today. Indeed, some techniques that seemingly were eliminated might not be dead yet, as we’ll see. I won’t be discussing any international tax or consolidated return issues, other than to acknowledge that there are income-shifting benefits in many of these arrangements.
What’s Income Shifting?
Income shifting is the reallocation of taxable income from a taxpayer who’s in a high income tax bracket to a taxpayer who’s in a lower income tax bracket. Typically, the shift of income is from parent to child but can be from grandparent to grandchild—or even to an unrelated partner or friend. The history of the modern federal income tax, originally enacted by the Tariff Act of 1913, and taxpayers’ responses to changes in tax law, is a fascinating one, especially with respect to efforts by taxpayers who shift income to others.
The old Shakespearean adage “what’s past is prologue” remains true today, especially in taxation and estate planning. Oliver Wendell Holmes Jr., while associate justice of the U.S. Supreme Court, once wrote “[a] page of history is worth a volume of logic.”1
Taxpayers have employed various means and schemes in income shifting—some successful and some not. I’ll review those methods using a historical perspective and conclude with an outline of viable types of income-shifting opportunities.
When it Started
One of the most common battlegrounds in taxation for over a century has pitted the IRS against taxpayers who attempt to divert taxable income to individuals who are in lower income tax brackets. While it no doubt began in 1913 with the enactment of the federal income tax, income shifting started in earnest in 1917, when the tax rates exploded to pay for World War I. (See “Original 1913 Income Tax Brackets,” p. 24 and “1917 Rate Explosion,” p. 24.)
As “Original 1913 Income Tax Brackets” illustrates, the 1913 tax rates topped out at 7%. However, once the Allies lured the United States into World War I, the tax rates grew exponentially, such that the former top rate of 7% was replaced by a top rate of 67%.2
In its nascency, the income tax was the only federal tax. The federal estate tax wasn’t enacted until 1916. At the outset, there also was no federal gift tax, which wasn’t enacted until 1924, but it was repealed in 1926, before being permanently added in 1932. There also was no concept of married filing jointly, which wouldn’t come on the scene until 1948; every taxpayer was considered separately.
Assignment-of-Income Doctrine
The first income-shifting skirmish occurred in attempts by working spouses to divert some of their earned income to their non-working spouses, which almost always reduced the couple’s overall income tax bill because the non-working spouse got to ride up the income tax brackets, too. This skirmish resulted in the Supreme Court’s creation of the assignment-of-income doctrine in Lucas v. Earl3 back in 1930.
In 1901 (well before the income tax ratification in 1913), a California lawyer and his wife entered into a contract whereby half of his earnings were to be “treated and considered . . . to be . . . owned by us [Earl and his wife] as joint tenants . . . with rights of survivorship.” In 1920 and 1921, Mr. Earl reported half of his earnings on his individual income tax return, and Mrs. Earl reported the other half of the income on her individual income tax return.
The predecessor to the IRS audited the couple’s returns and reallocated all of the income to Mr. Earl’s return, pushing him way up the graduated income tax brackets. The Board of Tax Appeals (BTA) sided with the government, but the U.S. Court of Appeals for the Ninth Circuit reversed.4
The Supreme Court reversed the Ninth Circuit and reinstated the BTA decision. In so doing, as organ for the Court, Justice Holmes laid out the assignment-of-income doctrine, with the now famous sentence that’s been a burr in the saddle for many a tax student ever since:
The fruits cannot be attributed to a different tree from that on which they grew.5
The assignment-of-income doctrine, which taxes the individual who performed the work with the income from his earnings, closed one loophole. Plenty of income-shifting loopholes remained, however, and taxpayers—not surprisingly—exploited them.
Community Property Income Shifting
In 1930, the same year that it decided Lucas v. Earl, the Supreme Court also decided Poe v. Seaborn,6 which held that income shifting was possible between spouses who were subject to a community property regime under applicable state law, because a community spouse has an undivided one-half interest in all community income of the working spouse under the laws of almost all community property jurisdictions. This created a firestorm in the states that didn’t have community property laws.
Joint Filing by Spouses
In the beginning, all taxpayers had to file separately, so the brackets applied separately to every taxpayer. In 1948, over two vetoes by President Harry Truman (more related to the tax cuts in the bill than this issue), spouses began to be able to file as married filing jointly, thereby eliminating at least that disparity between individuals who live in community property states and those who don’t. Thereafter, and up to the present, spouses may choose to combine their income and loss on one joint income tax return, which can give some significant income tax relief.
Income Shifting Passive Income
Because it became practically impossible to shift earned income due to the assignment-of-income doctrine except between spouses in community property states, the focus turned to unearned passive income.
In the beginning, this shifting of income could be done by creating a wholly revocable trust. However, Congress shut that down in 1924 with the enactment of the first grantor trust rules.7 This enactment turned the focus to short-term irrevocable trusts in which the grantor retained significant powers over the trust as well as a reversionary interest to get the property back at the end of the trust term. With no gift tax, people also used to swap income-producing property back and forth as needed for income reduction/reallocation purposes.
Congress Fights Back
Congress began to enact measures intended to curtail income shifting. The federal estate tax was adopted in 1916.8 In 1924, the first grantor trust rules came on the scene,9 which were intended to counteract income shifting through trusts. However, the original grantor trust rules only reached principal that the grantor could take back during the taxable year or when income was payable to the grantor, so they were easy to avoid.
The original gift tax was enacted in 1924,10 but it was repealed in 192611 before being reenacted for good in 1932.12 The federal gift tax exacted a toll for transferring property back and forth, thereby effectively eliminating this option back then. However, query whether in light of the current high applicable exclusion amount,13 gifting income-producing property back and forth has become viable again in some situations, at least until 2025, when the applicable exclusion amount is scheduled to revert to $5 million, indexed from 2012.
The purpose of the gift tax was to backstop the federal estate tax as well as the income tax by impeding the free flow of assets between related parties via inter vivos transfers. With the passage of the gift tax, the loophole of freely shifting income-producing property back and forth among family members was effectively closed, or at the very least, there was now the potential for a meaningful toll on such a transfer.
Short-Term Irrevocable Trusts
Given the ineffectiveness of the original grantor trust rules, the IRS fought short-term irrevocable trusts with little success in the courts. However, by 1940, the Supreme Court, which historically had been very resistant to the transfer taxes, had shifted. Appointments by President Franklin Delano Roosevelt transformed the Court into a much more activist, pro-tax court, setting the stage for Helvering v. Clifford.14
Helvering v. Clifford
The IRS fought taxpayers on a case-by-case basis on short-term irrevocable trusts. However, things came to a head in 1934, when a taxpayer set up an irrevocable trust that had a 5-year term over which the taxpayer was trustee with some marketable securities for the benefit of the taxpayer’s wife and children. At the end of the 5-year term, the trust corpus reverted to the taxpayer. The taxpayer retained significant power over the trust, including the power to vote securities in the trust as well as having complete discretion as trustee over whether the trust made any income distributions to the wife.
But, he didn’t retain any power that would run afoul of the two then-existing income tax grantor trust statutory rules. The taxpayer paid a gift tax on the 1934 transfer, and the trustee actually paid the trust income in 1934 to the wife, who declared the income and paid tax on it, albeit at a much lower tax rate because her income was far less than her husband’s.
In 1940, the Supreme Court, in a clear example of that court acting as “super legislature,” rendered its decision in Clifford.15 The majority found that, despite the express language of the grantor trust rules that were on the books, the taxpayer retained sufficient powers over the trust to treat him as the owner of the trust for income tax purposes. Justice John Roberts dissented, simply noting: “The decision of the court disregards the fundamental principle that legislation is not the function of the judiciary, but of Congress.” In my opinion, Justice Roberts was correct.
Unfortunately, the Supreme Court’s amorphous “facts-and-circumstances” test clogged the courts with more cases on this issue and really decided nothing but Clifford’s case. Query: Could a court use the extension of the substance-over-form principles enunciated in Clifford to turn back the grantor trust revolution, that is, ignoring a purported intentionally defective grantor trust?
Clifford Regulations
In 1946, in light of the unsatisfactory work by the Court in Clifford, and in an attempt to issue some meaningful guidance, the IRS promulgated the bright-line test by regulation (the so-called “Clifford regulations”), which provided that if the grantor’s retained reversionary right lasted for more than 10 years, the trust wouldn’t be a grantor trust. In the recodification of the Internal Revenue Code in 1954, Congress adopted the longer than 10-year bright line rule in IRC Section 673, which was the law until the Tax Reform Act of 1986.
Five Percent Test
In 1986, Congress amended Section 673 to provide for the current in excess of 5% test, which is the law to this day and had the effect of causing reversionary interests to have to be held so long that the grantor would risk inclusion of the trust in the gross estate under IRC Section 2037.
It’s noteworthy that retention of a reversionary interest in excess of 5% causes inclusion, but that asset will get a new fair market value at the taxpayer’s death.16 Therefore, there may be situations in which retention of a reversionary interest will allow for some basis planning.
Also in 1986, Congress amended Section 672 to provide that a grantor would be deemed to have the rights of his spouse,17 which eliminated the spousal reversionary Clifford trust that had been used.
Graduated Tax Brackets in Trusts
In the beginning, trusts and estates used the same graduated income tax tables as individuals. This permitted trusts and estates to take advantage of the graduated tax rates, which taxed undistributed income at usually low rates. In 1986, Congress ended this strategy by dramatically compressing the tax rates on undistributed income of an estate or trust, and in 2019, a trust or estate hits the top marginal income tax bracket of 37% (not counting the 3.8% Medicare surtax on net investment income) at $12,750, compared to $612,350 for married couples filing jointly in 2019.
For many years, taxpayers would create a large number of fairly small trusts for the same people that generated little income, which allowed each trust to ride up the brackets.
In 1984, Congress ended this game by adopting IRC Section 643(f), which consolidates trusts for income tax purposes when the same or substantially the same grantor forms identical trusts for the same or substantially the same beneficiaries as if they were one trust for income tax purposes unless the grantor can prove that tax avoidance wasn’t the driver for the creation of multiple trusts, that is, proving a negative, which is very difficult. Fast-forward to the Tax Cuts and Jobs Act in 2017 and IRC Section 199A. Some people act as if Section 643(f) is new law, but it isn’t. It just hasn’t had much utility since the Tax Reform Act of 1986. Look for litigation in this area.
Sale to Related Parties at a Loss
This was a common technique that permitted a seller to use a loss to offset other income even though the property usually stayed in the family, which generated lots of deductible losses after the beginning of the Great Depression and caused a lot of stock churning within families that never gave up ownership of the property “sold.” In 1934, Congress wised up to this scheme and adopted the precursor to IRC Section 267, which disallows losses on sales between related parties.18
Family Partnerships
This remains one of the most common income-shifting techniques. The IRS fought attempts to interpose partnerships in an attempt to shift earned income and was fairly successful, but the cases continued, particularly as to unearned, passive income. In the 1940s, the Tax Court began resolving the cases through a test of the bona fides of the entity,19 namely, whether the entity had any income-producing capital and whether each partner had contributed capital or meaningful services to the entity or whether the entity earned all of the income from the capital or efforts of just one partner.
The Tax Court’s bona fides test didn’t put an end to the endless stream of family partnership cases, so the Supreme Court had to get involved again, in Comm’r v. Culbertson20 in 1949, to address, in its words, “the family partnership problem,” which, of course, isn’t its role! Unfortunately, in Culbertson, the Supreme Court employed another facts-and-circumstances test that wasn’t particularly helpful.
In 1951, after having been pressured by angry constituents, Congress finally got involved, and it adopted the precursor to IRC Section 704(e), which respects partnerships if each partner has a capital interest (even if that interest is acquired by gift or bequest) in which capital is a material income-producing factor. This essentially solved the problem clogging the courts with these family partnership cases.
Below Market Loan
Despite the broad gift language of IRC Section 2511, taxpayers routinely made loans that bore little or no interest to related parties without gift or income tax consequences. The case law on this issue was split among the circuits, leaving the matter to the Supreme Court for ultimate resolution.
In Dickman v. Comm’r,21 which was decided in 1984, the Supreme Court held that an interest-free loan had gift tax consequences in an amount that was equal to the foregone interest that wasn’t charged. That very same year, Congress reacted to the Supreme Court’s invitation to legislate in the area and adopted the current IRC Sections 1274 and 7872, which give a safe harbor for loans that bear at least a minimum applicable federal rate of interest, reclassifying payments when there isn’t sufficient interest as part income to the recipient of the payments.
Unearned Income
Taxpayers routinely diverted unearned, passive income to related persons who were in lower income tax brackets. In 1986, Congress slammed the door on unearned income shifting to young children by adopting the so-called “kiddie tax” rules, which tax a child’s unearned income (children under 19 or students who are under 24) above a low threshold amount at the top earning parent’s top marginal income tax bracket.
The kiddie tax rules don’t apply to earned income of a minor or to income shifting of unearned, passive income to individuals who are outside of the kiddie tax age parameters, so this technique remains viable today.
Use of Anti-Taxpayer Provisions
The grantor trust rules require and mandate that the grantor pay income tax on trust income over which the grantor has any of the powers or rights spelled out in Subpart E of Part I of Subchapter J of the IRC of 1986. In Revenue Ruling 85-13, despite opposition in one court decision that found to the contrary,22 the IRS held that transactions between a grantor and a wholly grantor trust were properly ignored for federal income tax purposes.
The grantor trust provisions permitted grantors to pay the income tax on income going to others without gift tax consequences because the grantors were legally obligated to pay that income tax. This wasn’t a gift; it was required by the grantor trust rules that have been on the books in some form since 1924.
Now, because Rev. Rul. 85-13 stated that transactions between grantor trusts and grantors also were ignored, this permitted installment sales, which have always been an effective estate-freezing technique, to intentionally defective grantor trusts that weren’t includible in the grantor’s estate for federal estate tax purposes, creating a perfect storm of estate planning. In light of taxpayers’ creative turning of the grantor trust rules against the IRS, query whether the grantor trust rules remain desirable.23
Income Shifting Investments
This technique, which is still viable today, permits grantors to control the timing and amount of income that a trust or its beneficiaries pay. A trustee could invest in growth stocks that pay little or no dividends and harvest capital gains whenever it’s necessary to create some cash for distribution purposes at advantaged capital gains tax rates. Alternatively, a trustee could invest in tax-advantaged investments such as tax-free municipal bonds. Additionally, the trustee could balance gains and losses.
Income Shifting Personal Efforts
This technique is still viable today and permits income shifting down to the recipient while giving the payor of the wages a deduction for the amount paid. The amount paid must be reasonable for the time and energy spent, while any excess over this amount would be a taxable gift, and that excess wouldn’t be deductible by the payor. The payor must also respect child labor laws. The kiddie tax rules don’t apply to earned income.
Gifts of Opportunity
The IRS has never seriously attempted to treat transfers of opportunities between related parties as gifts, even though some academicians have called for that treatment,24 so this technique remains viable today. Until challenged, clients who have significant business contacts often come across great opportunities that they can divert to loved ones, who take the opportunity and run with it. Because the client never owns the property, there’s no transfer and, thus, no need to do any estate planning with the capital created by the opportunity coming into fruition.
Guarantees of Loans
Another potentially lucrative endeavor that remains available today is for a client to lend creditworthiness to a loved one by guaranteeing loans. If the loan is successful and the project makes money, the client owns nothing, and, hence, there’s nothing to estate plan around. If the loan goes south and the client is called on to make the loan good, this shouldn’t be a taxable gift because of the legal obligation to repay. This underscores the risks of loan guarantees: Sometimes the guarantor is called on to satisfy the loan, which essentially constitutes estate planning in reverse.
The gift tax issue is whether some consideration should be paid for use of the guarantee, and clients rarely charge for their guarantees to related parties. The problem here is quantifying the proper amount of such consideration in a particular situation, which is probably why the IRS has left it alone, but the Dickman principle still applies.25
U.S. Savings Bonds
The purchase of a U.S. savings bond for the benefit of another is still viable today and involves a situation in which the purchaser gives up the right to earn the interest himself, which is a form of opportunity cost. Properly timed as to its maturity, the purchase of a U.S. savings bond can avoid the kiddie tax by making sure that the bond doesn’t mature until after the recipient attains the age of 18, unless the recipient is a student, in which case it’s attaining the age of 24.
Gifts
Perhaps the simplest and most viable technique for income shifting that remains today involves making gifts of income-producing property. Whether the gift is of an amount that’s under the gift tax applicable exclusion amount ($15,000 in 2019) or the donor’s lifetime applicable exclusion amount ($11.4 million in 2019), such gifts can be made without having to come out of pocket with federal gift tax while making a substantial gift. Only one state (Connecticut) has a gift tax, so there usually isn’t any gift tax at all in these transactions. The applicable exclusion amount is so large that people of even modest wealth can swap income-producing property back and forth to switch receipt of income or loss.
Income-Shifting Requirements
The first requirement is unearned, passive income, because active, earned income won’t be able to be shared due to the assignment-of-income doctrine except for reasonable wages, which shouldn’t be overlooked. The recipient’s interest must be real, which often requires either effort by the recipient (for wages) or a gift of some sort to first be made. If the recipient of the income is supposed to have contributed capital or services to the project, that interest or those services must be real.
The parties generally must employ either a partnership/entity or a trust unless wages are paid for services actually performed. Most trusts used to shift income are grantor trusts for income tax purposes, which taxes to the grantor the income that’s payable to someone else or are beneficiary defective irrevocable trusts. Finally, the best recipients of shifted income are non-students who’ve attained the age of 18.
Endnotes
1. New York Trust Co. v. Eisner, 256 U.S. 345, 349 (1921).
2. In 1919, there were 49 different income tax brackets. Lucas v. Earl, 281 U.S. 111 (1930).
3. Lucas v. Earl, 281 U.S. 111 (1930).
4. Earl v. Commissioner, 30 F.2d 898 (9th Cir. 1929).
5. Lucas, supra note 3 at p. 115.
6. Poe v. Seaborn, 282 U.S. 101 (1930).
7. Revenue Act of 1924, Sections 219(g) and (h).
8. Revenue Act of 1916, Sections 200-212.
9. Ibid.
10. Revenue Act of 1924, Sections 319-324.
11. Revenue Act of 1926, Section 324.
12. Revenue Act of 1932, Sections 501-532.
13. In 2019, the applicable exclusion amount is $11.4 million.
14. Helvering v. Clifford, 281 U.S. 111 (1930).
15. Justice William O. Douglas’ decision also was criticized by a leading tax scholar, Dean Erwin Griswold, who wrote: “From an Olympian point of view, the result may well have seemed desirable. But there was little, if anything, in the statute to support it. Nor was it the culmination of a series of decisions, slowly etching out a new ground in the law. Even today, it seems to have been a rather strong case of judicial law-making.” Forward to Bernard Wolfman, Jonathan L.F. Silver and Marjorie A. Silver, Dissent Without Opinion: The Behavior Of Justice William O. Douglas In Federal Tax Cases, at pp. ix, xi (1975).
16. Internal Revenue Code Section 1014(a)(1).
17. IRC Section 672(e).
18. Revenue Act of 1934, Section 24(a)(6).
19. See, e.g., Comm’r v. Tower; 327 U.S. 280 (1946) and Lustig v Comm’r, 327 U.S. 293 (1946).
20. Comm’r v. Culbertson, 337 U.S. 733 (1949).
21. Dickman v. Comm’r, 465 U.S. 330 (1984).
22. Rothstein v. United States, 735 F.2d 704 (2d Cir. 1984).
23. See, e.g., Mark L. Ascher, “The Grantor Trust Rules Should be Repealed,” 96 Iowa Law Review 885 (2011).
24. Paul L. Caron, “Taxing Opportunity,” 14 Va. Tax Rev. 347 (1994) and Randall J. Gingiss, “The Gift of Opportunity,” 41 DePaul L. Rev. 395 (1992).
25. The closest that the Internal Revenue Service ever came was in Private Letter Ruling 9113009 (Dec. 21, 1990), which involved the gift tax consequences of a parent guaranteeing a loan for his children, in which the IRS ruled that the guarantee involved a transfer of valuable rights for purposes of IRC Section 2511 under a Dickman rationale. However, after the IRS was roundly criticized, it withdrew this aspect of the PLR. See PLR 9409018 (Dec. 1, 1993).