Lottery winnings as capital gains.
Pity J. Michael Maginnis. In 1991, he had the misfortune to win $9 million in the lottery. (1) Five years later, he sold his remaining winnings--fifteen annual payments of $450,000 each--to Woodbridge Financial Corporation for a $3.95 million lump sum. He reported this payment on his tax return as ordinary income, but he changed his mind several years later and sought a refund of some $305,000, claiming that the lottery payment was a capital gain. (2) Strangely, the IRS agreed and refunded his money. Then the IRS had its own change of heart--again several years later--and, in 2001, sued Maginnis, claiming that the refund was erroneous. An Oregon district court agreed with the Service, (3) the Ninth Circuit affirmed, (4) and poor Maginnis had to return his refund.
There is little debate that this is the right result: Maginnis's attempt to convert gambling income into capital gain was a fairly transparent ploy. (5) Nonetheless, Judge Fisher's opinion for the Ninth Circuit, which sets out a two-factor test for whether a gain is ordinary income under the "substitute for ordinary income" doctrine, is problematic. This Comment argues that an alternative approach that analyzes the transaction by which Maginnis received his lottery right may better explain and confine the use of the notoriously murky "substitute for ordinary income" doctrine.
Part I of the Comment discusses the "substitute for ordinary income" doctrine. Part II describes Maginnis's two-pronged test for applying the doctrine and points out the economic and doctrinal difficulties with that test. Part III proposes an alternate analysis that better achieves the policies of the "substitute for ordinary income" doctrine.
The Ninth Circuit sided with the IRS on the basis of the "substitute for ordinary income" doctrine, which holds that "'lump sum consideration [that] seems essentially a substitute for what would otherwise be received at a future time as ordinary income' may not be taxed as a capital gain." (6) A classic example is that of an employee who sells his rights to collect future wages: He will receive ordinary income, not capital gain, because the payment is a mere substitute for his right to receive ordinary income. (7)
This doctrine is usually traced to two leading cases: Hort v. Commissioner, which held that a payment to cancel a lease was ordinary income, (8) and Commissioner v. P.G. Lake, Inc., which held that the assignment of a right to receive (some of) the proceeds of future sales of oil also created ordinary income. (9) In both cases, the taxpayer attempted to secure capital gains treatment by selling future rights to receive ordinary income. If this were allowed, virtually no one would have to pay tax on ordinary income; any such income could be packaged, assigned, and transformed into capital gain. (10) The "substitute for ordinary income" doctrine sprung up to prevent this abuse.
The problem with the doctrine is that every capital asset is a substitute for ordinary income; read literally, the doctrine would completely swallow the concept of capital gains. A commercial building is worth only as much as the present value of its future leases--but those lease payments are ordinary income, while the building is a capital asset. The Fifth Circuit long ago noted this absurdity, explaining that "[t]he only commercial value of any property is the present worth of future earnings or usefulness," and quoting Lord Coke as asking, "[W]hat is land but the profits thereof?." (11)
The doctrine thus has little explanatory power. Instead, it lends itself to ad hoc decision making: "[C]ourts must locate the boundary case by case, a process that can yield few generalizations because there are so many relevant but imponderable criteria." (12) An overbroad "substitute for ordinary income" doctrine, besides being analytically unsatisfactory, would create the potential for the abuse of treating capital losses as ordinary. (13) The difficulty, then, is finding a way to appropriately cabin the doctrine to prevent abuses without allowing it to consume the entire notion of capital assets.
To limit the doctrine and avoid these difficulties, the Ninth Circuit in Maginnis identified two factors that characterize an asset as capital: first, that the taxpayer made "any underlying investment of capital" in exchange for the asset and, second, that the sale "reflect[ed] an accretion in value over cost to any underlying asset." (14) While the court qualified these factors, noting that "we do not hold that they will be dispositive in all cases," it nonetheless found them "crucial" to its decision. (15) At first glance, this test might seem to bring some clarity to the extremely murky theory of substitutes for ordinary income. But on closer examination, each prong of the Ninth Circuit's test proves to be untenable.
The first prong of the Maginnis test requires that the taxpayer make an "underlying investment in exchange for a right to future payments." (16) This, presumably, aims to distinguish capital assets, like stock, from noncapital assets, like assignments of future wages. But such a test is both over- and underinclusive.
First, this prong of the test does not explain the result in many standard cases. The taxpayer in P.G. Lake does appear to have invested in the underlying asset (a working interest in oil and gas leases), but the court nonetheless decided the case under the "substitute for ordinary income" doctrine, (17) Even clearer is the hornbook example of common stock: If a shareholder buys stock and then sells the right to future dividends, that sale (prior to dividend tax reform) is ordinary income despite the underlying investment. (18) Thus, the test does not account for the doctrine.
Second, this prong would treat as ordinary several types of assets that are clearly capital. If the test requires the taxpayer who sells the asset to have made the underlying investment of capital, then any taxpayer who inherits stock will be liable for ordinary income, rather than capital gain, if she sells it. The same would apply to a taxpayer whose parents gave her a plot of land. These are obviously wrong results; a capital asset does not become noncapital simply because it is received without consideration. (19)
Perhaps, though, instead of meaning that the taxpayer must have invested in the underlying asset, Judge Fisher meant only that someone must have invested in it. Once money is invested, the asset becomes capital and may then be sold, donated, or devised without a change in character. This seems much more sensible. Such a test accurately distinguishes a share of stock (which, once someone has bought it, can be bequeathed and remain capital) from a worker's right to receive future wages (which, because she has invested only labor and not capital, is not a capital asset).
But this reading of the first prong still doesn't work. Besides being unsupported by case law or statute, the revised first prong can't explain Maginnis itself. In this case, someone did invest capital in the underlying asset: the State of Oregon. (20) The asset was transferred to Maginnis in a gambling transaction, but it was not created out of thin air (or by labor). Thus, if Maginnis really means only that some underlying investment of capital is necessary to produce a capital asset, then it doesn't cover its own facts.
The second prong of the Maginnis test, asking whether the sale "reflect[s] an accretion in value over cost to any underlying asset," (21) is similarly unsatisfactory. It attempts to determine the character of a gain from its amount. An asset with zero basis will create more gain, when sold, than the same asset with a significant basis. This distinction should not, however, affect the character of the gain. The facts of Maginnis itself illustrate the problems that come from inferring the character of gain from the amount of basis. The Ninth Circuit noted that "the amount a purchaser such as Woodbridge might pay for the right might be subject to some uncertainty," (22) but it did not fully appreciate the source of this uncertainty. The uncertainty involves fluctuations that are exactly equivalent to those in the value of, say, a municipal bond. (23) If Maginnis acquired such a bond with negligible basis, then his gain upon selling it would be large--but it would all be capital gain. The difference between sale price and basis cannot be used to distinguish capital from ordinary gain.
The Maginnis test does not satisfactorily distinguish which substitutes for ordinary income are actually taxed as ordinary income and which are taxed as capital gains. But the result is right; what is needed is a way to achieve this outcome without unnecessarily expanding the reach of the "substitute for ordinary income" doctrine. One solution might be to focus not on the abstract question of whether a lottery right is a capital asset but on the transaction at issue: the receipt, as a gambling winning, of the lottery right.
In a sense, Maginnis received $9 million (24) in ordinary income as soon as he won the lottery: He immediately gained the right to receive $9 million, taxable at ordinary rates as gambling winnings. Now, in theory, the government could have taxed the entire amount as soon as Maginnis won the lottery. The lottery right is a right to future payments, but so are stocks and bonds, and if Maginnis won shares of stock in a lottery, he would be liable for tax (at ordinary rates) on their full value immediately. (25)
This is the best explanation for why the lottery right was not a capital asset when Maginnis sold it: not that the future income was ordinary, or that he had not made any investment, but that he only avoided tax on receipt via a kind of administrative grace. If he had been given an Oregon state bond rather than a lottery right, he might in fact have been held liable for taxes on the full amount as soon as he won (and any gain on a future sale of the bond would be capital). But he avoided immediate tax on his gambling winnings--and, therefore, when he eventually sold off those rights, it was sensible to hold him responsible for the taxation that he had avoided.
So why do we not tax the lottery right (as a gambling winning) as soon as Maginnis wins it? The answer may lie in an analogy to the realization requirement. It is a basic principle of tax law that economic income should not be taxed until it is "realized"--a policy, traceable at least to Eisner v. Macomber, (26) based on principles of liquidity and valuation. Appreciation in an asset's value is not typically realized (and thus taxed) until the asset is sold, because taxing appreciation as soon as it occurs would create difficulties of annual valuation and might require holders to sell some illiquid assets to pay tax on the appreciation. (27)
But similar policy concerns can arise in taxing rights to future payment upon receipt. Valuing a stock or bond on receipt is easy because it has a market price, but valuing a fixed-income item like a lottery right or a lease may involve difficulties in choosing a discount rate. Similarly, stocks and bonds are often reasonably liquid, and a portion of one's holdings can be sold to pay tax on them, while leases and lottery rights tend to exist in more illiquid markets and are more likely to be indivisible. (28)
This analysis might be viewed as a (sufficient, but not necessary) test for applying the "substitute for ordinary income" doctrine. Assets that are fully realized upon receipt--stocks, bonds, land--are capital assets, while assets that, for policy reasons of valuation and liquidity, are not realized immediately--leases, lottery rights--are substitutes for ordinary income in the hands of those who receive them in ordinary-income transactions (e.g., as gambling winnings).
A better view, however, might be that there is simply no such doctrine--instead, the term "substitute for ordinary income" is applied to a number of unrelated transactions that can be better explained in other ways. Maginnis himself proposed that the doctrine should be confined to two clear cases of abuse: "carve-outs" of part of a taxpayer's interest in property and rights to future income from personal services. (29) The first of these, carve-outs, formed the basis for Hort and Lake; conceptually, one can say that a carve-out of a capital asset is not a capital asset without using the phrase "substitute for ordinary income." (30) The second, rights to future income from personal services, seems arbitrary: Why should personal-services income be its own special category?
Rather than classify personal services as a case of a substitute for ordinary income, it makes more sense simply to note that the right to payment for future personal services, like the lottery right, is not fully taxable on receipt. (31) Thus, whenever it does become fully taxable, it is taxed as it would have been when received: as personal-service income, as gambling winnings. In a sense, the theory here is of an "open transaction": (32) Because the initial receipt of the right was not "closed," i.e., realized and taxed, when that right accrued, the transaction closed when the right was sold, and the entire amount was taxable based on the character of the original transaction (personal-service income, gambling winnings).
This approach has several benefits. It preserves parity between a winner who sells his lottery right and one who chooses to receive a lump sum payment from the state. (33) It clarifies that a future holder of the right, such as Woodbridge, may treat it as a capital asset. (34) And it is easily generalized beyond lottery rights, to personal-service income and even leases. (35) Most important, this approach provides a boundary--or, at least, a guidepost to finding the boundary--to the "substitute for ordinary income" doctrine. While defining that doctrine in its entirety is beyond the scope of this Comment, confining it ought to be an important goal. This is partly because of its economic indeterminacy: As long as the doctrine can be invoked for nearly any asset, it will promote uncertainty, abuse, and transactional complexity. But it is also because of the doctrine's uncertain foundation in the Code: The text of [section] 1221, defining "capital asset," does not provide any support for the doctrine, which is a pure judicial creation. The Supreme Court's decision in Arkansas Best appears to stand for the proposition that judicial creation of exceptions beyond those in [section] 1221 is disfavored. (36)
The Ninth Circuit's approach in Maginnis seems to set out a general test for applying the "substitute for ordinary income" doctrine, encouraging judicial expansion of that vague and economically indeterminate concept. This Comment advocates an approach that narrows the "substitute for ordinary income" doctrine considerably. It avoids reliance on conclusory statements that an asset is a substitute for ordinary income, and suggests reading that doctrine as merely a label for a collection of a few specific cases--principally, carve-outs and assets not realized immediately--that have historically presented an opportunity to abuse the capital-gains rate differential. This approach focuses not on a vague characterization of assets, but on a careful look at the tax and economic realities of the transactions involved. It may not be feasible to abandon the "substitute for ordinary income" doctrine wholesale, but we can take a step in the right direction by focusing on real issues in specific cases rather than on general statements of a doctrine with no basis in statutory text or economic reality.
(1.) The facts in this paragraph are drawn from United States v. Maginnis, 356 F.3d 1179, 1180-81 (9th Cir. 2004).
(2.) Capital gains are taxed at a significantly lower rate than ordinary income. Compare I.R.C. [section] 1(h)(1)(C) (2000) (capital gains), with id. [section] 1(a)-(d) (ordinary income).
(3.) United States v. Maginnis, No. CIV. 01-368-KI, 2002 WL 1482390 (D. Or. May 28, 2002).
(4.) Maginnis, 356 F.3d at 1179.
(5.) Several other recent cases have so held. See Clopton v. Comm'r, 87 T.C.M. (CCH) 1217 (2004); Simpson v. Comm'r, 85 T.C.M. (CCH) 1421 (2003); Johns v. Comm'r, 85 T.C.M. (CCH) 1318 (2003); Boehme v. Comm'r, 85 T.C.M. (CCH) 1039 (2003); Davis v. Comm'r, 119 T.C. 1 (2002).
(6.) Maginnis, 356 F.3d at 1182 (quoting Comm'r v. P.G. Lake, Inc., 356 U.S. 260, 265 (1958)) (alteration in original).
(7.) See 2 BORIS I. BITTKER & LAWRENCE LOKKEN, FEDERAL TAXATION OF INCOME, ESTATES AND GIFTS [paragraph] 47.1 (3d ed. 2000).
(8.) 313 U.S. 28 (1941).
(9.) 356 U.S. 260 (1958).
(10.) See MARVIN A. CHIRELSTEIN, FEDERAL INCOME TAXATION [paragraph] 17.03, at 368 (9th ed. 2002).
(11.) United States v. Dresser Indus., 324 F.2d 56, 59 (5th Cir. 1963) (emphasis and internal quotation marks omitted). The Ninth Circuit in Maginnis was well aware of the problem, explaining that
[m]any assets, including common stock, are typically valued on the basis of the present value of their future income stream, so an approach that took the substitute for ordinary income doctrine too far, and defined the term capital asset too narrowly, would hold that no sale of an asset that produces revenue.., could be taxed as a capital gain.
Maginnis, 356 F.3d at 1182.
(12.) Maginnis, 356 F.3d at 1182 (internal quotation marks omitted). Bittker and Lokken identify six types of transactions; of these, the most important are temporal divisions ("horizontal carve-outs"), which generally produce ordinary income unless the owner of a temporal division sells her entire estate. 2 BITTKER & LOKKEN, supra note 7, [paragraph] 47.9.5.
(13.) Cf. Ark. Best Corp. v. Comm'r, 485 U.S. 212 (1988) (limiting exceptions to the capital-asset definition and disallowing ordinary loss on sale of bank stock).
(14.) Maginnis, 356 F.3d at 1183.
(17.) Comm'r v. P.G. Lake, Inc., 356 U.S. 260, 261-62, 265 (1958).
(18.) This was clearly the result prior to dividend tax reform. See, e.g., CHIRELSTEIN, supra note 10, [paragraph] 17.03, at 367. Now that dividends are taxed like capital gains, see I.R.C. [section] 1(h)(11) (West 2004), it is less clear how courts would approach the dividend carve-out, though one assumes it would be taxed at dividend rates.
(19.) This is subject to a substantial caveat, which is that the characterization of certain special kinds of assets does depend on whether their basis is traceable to their creators. Thus, copyrights, letters, and similar property are capital in the hands of buyers but ordinary in the hands of their creators or those who receive them as gilts. I.R.C. [section] 1221(a)(3) (2000). See generally Jeffrey C. McCarthy, Federal Income Taxation of Fine Art, 2 CARDOZO ARTS & ENT. L.J. 1 (1983).
(20.) A state that owes a lottery prize either sets aside money or buys an annuity to fund it. See, e.g., Boehme v. Comm'r, 85 T.C.M. (CCH) 1039 (2003) (describing Colorado's annuity).
(21.) Maginnis, 356 F.3d at 1183.
(22.) Id. at 1184.
(23.) Maginnis had a right to a fixed series of payments from the State of Oregon--just like a state bondholder. As interest rates go up, the value of the right declines, and vice versa. If investors lose confidence in the state's ability to pay its debts, the right loses value.
(24.) In this Part, I ignore the issue of discounting to present value. Maginnis earl be thought of as winning either $9 million over twenty years or its present value now; I use "$9 million" as a shorthand for "the discounted present value of $9 million of payments over twenty years."
(25.) Cf. Treas. Reg. [section] 1.61-2(d) (as amended in 2003) (taxing compensation in the form of property at its fair market value).
(26.) 252 U.S. 189 (1920).
(27.) See CHIRELSTE1N, supra note 10, [paragraph] 5.02, at 75; MICHAEL J. GRAETZ & DEBORAH H. SCHENK, FEDERAL INCOME TAXATION: PRINCIPLES AND POLICIES 144 (rev. 4th ed. 2002).
(28.) These differences are not inevitable and should not be overstated. Leases may be assigned and bundled by financiers, and many stocks are illiquid and difficult to value. But as broad classes, it does seem likely that some rights to future payments are easy to think of as presently valuable instruments, while others are better conceived of as mere streams of future payments.
(29.) See Maginnis, 356 F.3d at 1185; Reply Brief for Appellant, Maginnis (No. 02-35664), reprinted in Lottery Winners Argue Right to Payment Is Capital Asset, TAX NOTES TODAY, Aug. 28, 2003, 2003 TNT 167-18 (LEXIS).
(30.) See CHIRELSTEIN, supra note 10, [paragraph] 17.03, at 369-70 ("The 'substitute' language [in Hort], in the view of most commentators, was merely a shorthand way of asserting that carved-out interests do not qualify as capital assets and do not absorb any portion of the taxpayer's property basis."). Further discussion of the carve-out branch of the "substitute" doctrine is beyond the scope of this Comment.
(31.) That is, if I enter into a five-year, $100,000-per-year employment contract, I am not taxed on the entire $500,000 at the moment of the contract. Even deferred compensation for already performed services is generally taxed when received, not when accrued. See Rev. Rul. 60-31, 1960-1 C.B. 174; see also CHIRELSTEIN, supra note 10, [paragraph] 11.01, at 270-73.
(32.) For the more common use of this term in tax law, see Arthur Fleischer, Jr. & William L. Cary, The Taxation of Convertible Bonds and Stock, 74 HARV. L. REV. 473, 478 (1961).
(33.) This was a goal of the Maginnis court. See Maginnis, 356 F.3d at 1184.
(34.) What makes it an ordinary asset in Maginnis's hands is how it was received, as untaxed gambling winnings; if it is purchased on the open market, then it becomes a capital asset.
This approach could also theoretically allow someone in Maginnis's position to capture the "gain over cost" that the court dismissed in the second prong of its opinion: If, when Maginnis won the lottery, his right was worth $3 million, but, before he sold it, interest rates declined and the right's value increased to $3.5 million, then he should have had $3 million in gambling income and $0.5 million in capital gain. This parallels what would have happened if Maginnis had won an Oregon state bond in the lottery. This is a satisfying result as a matter of tax theory, but I suspect that a court would not actually endorse this approach. The Maginnis court itself was unimpressed by the possibility of change in value to the lottery right. See supra text accompanying note 22. And, in Hort v. Commissioner, 313 U.S. 28 (1941), the Supreme Court refused to make a similar distinction where a lease, ordinary when signed, was later purchased (technically, cancelled by a lump sum payment) after having increased in value due to changes in economic circumstances.
(35.) When a landlord enters into a lease, the full value of the future payments is not taxable immediately; if he sells or cancels the lease, however, he owes taxes at ordinary rates on the income that he then receives. Leases can be considered as a carve-out, as they were in Hort. See supra text accompanying note 30. There are good reasons for treating leases under the carve-out prong of the "substitute" doctrine, rather than under the open-transaction prong, but the open-transaction approach is, I think, also illuminating.
(36.) See Ark. Best Corp. v. Comm'r, 485 U.S. 212, 218 (1988) ("The body of [section] 1221 establishes a general definition of the term 'capital asset,' and the phrase 'does not include' takes out of that broad definition only the classes of property that are specifically mentioned."); see also Patrick E. Hobbs, The Scope of the Inventory Exclusion Under L.R.C. [section] 1221(1), 26 LOY. LA. L. REV. 289, 317 (1993); Jay A. Soled, The Sale of Donors' Eggs: A Case Study of Why Congress Must Modify the Capital Asset Definition, 32 U.C. DAVIS L. REV. 919, 940 (1999). Maginnis himself relied on Arkansas Best to argue that the "substitute for ordinary income" doctrine should apply only in two types of eases. See supra note 29 and accompanying text.
|Printer friendly Cite/link Email Feedback|
|Author:||Levine, Matthew S.|
|Publication:||Yale Law Journal|
|Article Type:||Case Note|
|Date:||Oct 1, 2004|
|Previous Article:||Emergency power and the Militia Acts.|
|Next Article:||The federalist dimension of regulatory takings jurisprudence.|