As Lee Anne Fennell writes in Slices and Lumps: Division and Aggregation in Law and Life, “law must determine the proper unit of analysis—how widely or narrowly to set its viewfinder—in assessing whether a given line has been crossed or a given standard has been satisfied . . . . Second, law must decide how sharp or graduated its responses should be—whether to use a cliff-like or a sharply notched structure . . . or a more modulated set of legal responses.” Law engages aggregation and division in at least one additional, closely related way: law must sometimes decide the proper unit of analysis not just in deciding whether the law has been violated, but also to decide what body of law applies. Law must identify the right party to regulate—must determine what, or who, constitutes the subject of the law—and this, too, can be a question of aggregation or division. As an investigation of US tax law demonstrates, both taxpayers and the government are well aware of when separating an aggregation, or, alternatively, forcing an aggregation of individuals, can provide an advantage. Tax law shows the call and response of taxpayers’ and the government’s finding aggregation where there is apparent division and dividing what seems to be aggregated.
This response focuses on two examples of slicing and lumping in the subject of taxation in US federal tax law: first, choice of entity, and, second, regulation of transactions that involve related parties or parties who are for some other reason motivated by something other than market interests.
I. Choice of Entity
Perhaps the most obvious example of aggregation of the subject of tax law is choice of entity. Choice of entity implicates the aggregation question in at least two ways. First, an entity involves aggregating individuals, and the law then treats the entity as a single entity—a true aggregated whole—or, sometimes, notwithstanding the aggregation of individuals into the entity, the law treats the individuals separately. The law, that is, is imposed either on a lump (of individuals) or a slice (of an entity).
Second, the ways in which individuals are permitted to aggregate into entities is itself a question of aggregation. When the law provides a set of ways to aggregate for tax purposes, ranging from highly standardized to highly tailorable by the parties involved, the law essentially offers a set of goods. The continued importance of both standardized and tailorable entities provides an inroad to expanding Fennell’s discussion of standardization and customization.
Corporations taxed under Subchapter S of the Internal Revenue Code (“S corporations”) do not usually pay federal income tax, but they are still the subject of regulation by the tax law. S corporations are the Model T of entities for tax purposes: they were relatively early to the scene, they were widely adopted as the entity of choice for many privately held companies, and they are quite standardized. Bundled with flow-through taxation comes limits on the number and type of shareholders and, most interestingly for these purposes, the requirement that there be only one class of stock. While there can be variations in voting rights, each share of an S corporation must bring with it the same rights to distributions and liquidations proceeds. Either one takes the “package” of the S corporation or one does not.
In the late 1990s and early 2000s, once the Internal Revenue Service clarified that it would treat limited liability companies as partnerships, people predicted the imminent demise of the S corporation. The LLC also provides one level of taxation and limited liability, but the partnership tax rules, in contrast to the S corporation rules, allow for a huge amount of flexibility when it comes to structuring interests. Different rights to distributions and liquidation proceeds are only the beginning. Partnerships, and entities taxed as partnerships, are allowed a tremendous amount of customization.
But the S corporation did not disappear. To the contrary—it thrived. There are still more S corporations than there are entities taxed as partnerships. For the 2018 tax year, for example, there were 5.13 million S corporation returns filed, and 4.24 million partnership tax returns. This is in part an historical artifact, but it is not just a matter of history: the number of entities taxed as partnerships and the number of S corporations are growing at roughly the same rate: from 2017 to 2018, the number of S corporation tax returns grew by 5.9%, and the number of partnership tax returns grew by 4.8%.
One reason for the persistence of the S corporation is that partnerships provide a good deal more flexibility and customization, but at the cost of a good deal more complexity for the taxpayer (and for the IRS). It is apparently not enough that taxpayers could choose a simple approach even within the context of a partnership—there’s no requirement in a partnership that different interests have different rights to distributions. Rather, at least some taxpayers prefer opting into a regime in which they are required to follow rules that have the effect of simplifying their choices.
That said, many taxpayers do choose for their entities to be taxed as partnerships. Because of their flexibility, partnerships have presented an enticing arena in which taxpayers can practice creative tax planning. The question of whether partnerships are best considered an aggregation of individuals (a collection of slices) or a single entity (a lump) has been much discussed, and raising the topic remains a fun and easy way to provoke any partnership tax lawyer or professor. The current state of partnership tax law represents, among other things, an ongoing back-and-forth in which taxpayers figure out ways to take advantage of entity treatment, and Congress or the IRS responds by forcing taxpayers to be treated as individuals for some particular purpose, and vice versa. Taxpayers then create a new money-saving approach, and the conversation continues.
For example, Section 751(a) (which I will not describe here out of respect for the non-tax reader) represented a compromise between “full fragmentation,” which “reflects an aggregate view and is intended to reduce opportunities for using partnership to convert ordinary income into capital gain,” and “an entity view,” under which “sale of a partnership interest would give rise to capital gain or loss.” Section 751 exists in the form it does because in 1954, when Section 751 was originally enacted, Congress rejected “full fragmentation” as “too drastic for the evil sought to be remedied,” which relates to the sale of partnership. This was a back-and-forth between a certain conception of an entity, subsequent taxpayer abuse, Congressional response, and so on.
II. A Pair, or Two Ones?
Various rules throughout the Internal Revenue Code either aggregate two individuals that are otherwise separate for tax purposes, or separate two parties that are otherwise aggregated. This response considers four examples: the related-party loss disallowance rule; the gift dual basis rule; and the related-party installment sale rule, all of which treat two taxpayers as a single taxpayer; and the taxpayer-advantageous principal residence gain exclusion, which in a way allows a taxpayer to retain her single status even after she is married, thus treating a single taxpayer (a married couple filing jointly) as containing two separate individuals.
A. Disallowing Losses Between Related Parties
Gain or loss is realized (though it may or may not be recognized) when property is sold or exchanged. No loss is allowed, however, when property is transferred between related parties. Related parties include members of families, certain entities and their owners, among others. The idea here is that if the parties are close enough, then under Section 267, a transfer does not necessarily permanently disallow the loss. Rather, when the property is subsequently transferred, the transferor need recognize a gain only to the extent the gain exceeds the loss previously unrecognized.
For example, imagine that Alan uses property that he purchased for $120 (his basis) and that is now worth $100. If Alan sells the property to an unrelated party, he will have loss equal to the excess of the basis over the fair market value, that is, $120 – $100 = $20 of loss. If he sells the property to his mother, Bertha, however, he will not be permitted to recognize the loss, because his mother is a related party for purposes of Section 267. If Alan is an adult who lives separately from his mother, Alan’s taxes and his mother’s taxes are separate. The tax law does not aggregate them. But for the purpose of recognizing loss, they are treated as a single person. Just as Alan could not recognize loss if he sold the property to himself, so he cannot recognize loss if he sells the property to Bertha.
The law aggregates Alan and Bertha in another way as well: when Bertha later disposes of the property, she may be able to get the benefit of the loss that was disallowed to Alan. She must recognize gain “only to the extent that it exceeds so much of [the loss disallowed Alan] as is properly allocable to the property.” If she had bought the property for $100, and she later sold it for $130, she would have $30 of gain. But because she has to recognize that gain only to the extent it exceeds the loss that was disallowed to Alan, she must recognize only $10 of gain. (Her gain is $130 – $100 = $30, further reduced by Alan’s $20 loss, $30 – $20 = $10 gain recognized.) Alan did not get the benefit of the $20 loss, but Bertha did. Alan was, in other words, permitted to transfer the loss to Bertha (though only if the loss reduces gain—Alan’s loss won’t be permitted to increase a later loss by Bertha).
This loss disallowance rule was enacted in response to what Congress perceived to be “tax dodgers who transferred securities or other property from one member of a family to another in order to deduct a capital loss against ordinary income.” In response to a rule disallowing losses between family members—that is, in response to a rule that treated two separate taxpayers as one if they were family members—taxpayers simply found other related parties with whom they could enter into transactions but who were not (yet) covered by the loss disallowance rule. Congress expanded the definition of related parties, and the conversation continued.
B. Gifts and the Dual Basis Rule
A similar rule applies to the transfer of losses through gift. In general, a gift is not a realization event for the donor and is not income to the recipient. The recipient generally takes the gift with a transferred basis: the property has the same basis in the recipient’s hands as it did in the donor’s. However, if the property has a greater basis than fair market value at the time of the gift, the recipient’s basis for determining loss is the fair market value at the time of the gift.
For example, if Alan gives Bertha property worth $100 with a basis of $80 to Alan as a gift, Bertha’s basis in the property is $80. If she later sells it for $130, she will have $50 of gain: the $20 of gain that the property had in Alan’s hands, plus an additional $30 of gain that accrued while she held it. If Alan gives Bertha property worth $100 with a basis of $120 to Alan as a gift, though, the dual basis rule comes into play. Her basis depends on the price for which she sells it.
If Bertha later sells the property for $130, she will have $10 of gain. She can use the $120 transferred basis from Alan, because whether she uses the $100 fair market value basis or the $120 transferred basis, she will have gain. Bertha’s use of the transferred basis effectively permits Alan’s loss to offset Bertha’s gain.
If she sells the property for less than the property’s fair market value at transfer, say for $90, she will have $10 of loss. Whether she uses the $100 fair market value basis or the $120 transferred basis, she will have loss, so she is required to use the $100 basis. And, finally, if she sells the property for $105, she will have neither gain nor loss. If she used a fair market value basis she would have, but that is the loss basis. If she used the transferred basis, she would have loss, but that is the gain basis. (The regulations also confirm this result.) This is the same result as in the related party loss disallowance rules of Section 267: Alan’s loss can reduce her gain but cannot increase her loss. Thus, although implemented in a different way than the loss disallowance rule, the dual basis rule for gifts may also permit the loss to be recognized, to some extent, once the property moves outside of the donor-recipient relationship.
The dual basis rule is not a related party rule, as it applies to any gifts, even those between unrelated parties. But the concerns that motivate the dual basis rule are the same as those that motivate the related parties loss disallowance rule: concerns about an alignment of interest, and a sense in which the transfer is not truly an alienation of the property because the parties are in some way not truly separate. A transfer is a gift if it is made out of “detached and disinterested generosity,” “out of affection, respect, admiration, charity or like impulse.” Furthermore, the giver cannot receive consideration in exchange for the gift, the gift must occur outside the usual market setting, andthe parties’ interests cannot be at odds. Similarly, various subcommittee reports and the Senate Report regarding the dual basis rule describe the problem as involving “transfer [of loss property] by gift to a relative or close friend with a large income” notwithstanding that whether the dual basis rule applies does not depend on the relationship between the donor and the recipient.
C. Installment Sales Between Related Parties
The installment sale rules similarly lump together two or more parties that the tax code generally treats as separate. The installment sale rules generally allowed a seller to delay recognition of gain until he receives principal payments on a note. A seller who, for example, has $20 of gain on sale, and receives $10 cash and a $90 note can delay recognition of 90% of that gain until she receives payment on the note.
On the other side, though, the buyer gets immediate basis credit for the note. If the buyer buys the property for $10 cash and a $90 note, the buyer has a basis of $100. The buyer can now either take cost recovery deductions on that $100 basis or can immediately sell the property for no gain.
For example, imagine that Alan owns land with a basis of $50, and a buyer, Xavier, wants to buy the land for $200 cash, creating $150 of gain for Alan. Under the general installment sale rule, if Alan sells the land to his mother, Bertha, for a $200 installment note on which Bertha is personally liable, and Bertha in turn sells the land to Xavier, she will have no gain, because she has a basis of $200 in the property. Alan will have no gain until the note is paid off.
The statute does not permit such a transaction. Instead, because Alan and Bertha are, as son and mother, related parties (as defined in Section 267, the loss disallowance section), they are lumped together and treated as one taxpayer for purposes of the installment sale rules. If Bertha sells the property within two years of buying it from Alan, Alan will be treated as receiving a payment on the note equal to the payment that Bertha receives when she sells the property to Xavier. Thus, when Bertha sells the property to Xavier for $200, Alan will be treated as receiving a $200 payment on the note and will have to recognize the full $150 of gain. (If and when Bertha later pays off the note, Alan will not have to recognize additional gain.)
D. Principal Residence Gain Exclusion
The principal residence gain exclusion takes the opposite approach: rather than forcing two separate parties to be treated as a single aggregated taxpayer, the principal residence gain exclusion permits a “lump” of taxpayers, a married couple filing jointly, to retain certain characteristics from their pre-marital, single days.
Section 121 permits a single taxpayer to exclude $250,000 of gain from the sale of her principal residence, and permits a married couple filing jointly to exclude $500,000 of such gain. (The doubling, or failure to double, various limitations for taxpayers who are married filing jointly deserves its own discussion.) For example, imagine that Adam, a single taxpayer, has a basis of $600,000 in his house and sells it for $1,400,000. He realizes gain of $800,000 ($1,400,000 – $600,000) but can exclude $250,000 of that gain, so he must recognize only $550,000 (that is, $800,000 – $250,000) of gain from the sale. If Adam is not single, but instead is married to Brad and they file jointly, and they sell the house, they will be required to include only $300,000 of gain ($800,000 – $500,000).
Section 121 can be used only once every two years. If Adam sold his house in Year 1, bought a new house immediately, and then sold that new house in Year 2, he would not be permitted to exclude any of the gain from the sale of the new house under Section 121.
Now imagine the situation in which Adam sells his house in Year 1 and uses his $250,000 exclusion under Section 121. He then marries Brad. In Year 2, Adam and Brad sell the house in which they are living. Is any exclusion available to them? Adam has used the exclusion within the past two years, but Brad has not. One possible answer is that Adam’s use of the exclusion counts to make the exclusion unavailable to the couple under the two-year rule. Another possible answer is that Adam and Brad as a couple have not used the exclusion in the past two years, and so the $500,000 exclusion is available. But the IRS takes a third approach, one explicitly endorsed by the legislative history: Adam’s $250,000 exclusion has been used up, but Brad’s is still available, and so the couple may exclude $250,000 of their gain. Although they are filing jointly, Adam and Brad are not entirely merged for tax purposes; each keeps his own separate ledger of whether he individually has used the $250,000 exclusion within the past two years.
This response only begins to touch on the many ways that the concepts of aggregation and division can be used to understand various areas of tax law. One might also consider annual accounting, which chooses a year as the relevant period to consider for taxation, and when the tax code deviates from annual accounting (as when, for example, one entity is acquired by another); the realization requirement, which taxes all increases or decreases in value at a single time, instead of as they accrue; income-based phaseouts and thresholds, which can result in discontinuities in tax rates; taxation of financial instruments, including the tax treatment of stripped bonds; and more. Understanding tax law as a collection of rules that slice and lump may open the door to many new insights.
Sarah B. Lawsky is the Benjamin Mazur Summer Research Professor of Law at Northwestern Pritzker School of Law. The author wishes to thank Joshua Blank, Neil Buchanan, Erin Delaney, Michelle Falkoff, David Weisbach, and participants in the Slices and Lumps symposium for helpful discussions and comments.