ArticlesVol. 2, No. 2

Making Partnerships Work for Mom and Pop and Everyone Else

Emily Cauble

Associate Professor, DePaul University College of Law
Citation:2 Colum. J. Tax. L. 247

Abstract

Entities treated as partnerships for tax purposes cover the spectrum from small Mom and Pop operations to mammoth enterprises owned by sophisticated partners. Designing tax law to govern this diverse array of entities is a challenging exercise since rules that are well suited to provide simplicity for entities at the small, unsophisticated end of the continuum may be poorly designed for preventing manipulation by entities at the large, sophisticated end of the continuum. However, a careful examination of how tax rules can best promote simplicity reveals opportunities for reform that would make the law more appropriate for entities all along the continuum.

  • INTRODUCTION
  • I. EXISTING LAW UNDER § 704(c)
    • A. Traditional Method
    • B. Traditional Method with Curative Allocations
    • C. Remedial Method
    • D. Summary of § 704(c) Methods
    • E. Anti-Abuse Rule
    • F. Other Constraints
    • G. Reverse § 704(c) Allocations
    • H. Special Rules for Securities Partnerships 1. Facts of Example 3 2. Results Under an Asset-by-Asset Approach 3. Results Under an Aggregate Approach
  • II. HISTORY OF § 704(c) AND CONCERNS ABOUT UNSOPHISTICATED PARTNERS
    • A. Original Version of § 704(c)
    • B. Justifications Offered for the Original Version of § 704(c)
    • C. Evolution of § 704(c)
  • III. EXISTING LAW UNDER § 754
    • A. Section 754 Elections and Built-In Gain Assets
    • B. Section 754 Elections and Built-In Loss Assets
    • C. Other Constraints
  • IV. HISTORY OF § 754 AND CONCERNS ABOUT UNSOPHISTICATED PARTNERS
  • V. HOW REFORMS WOULD MAKE LAW LESS SUSCEPTIBLE TO MANIPULATION BY SOPHISTICATED PARTNERSHIPS
    • A. Section 704(c)
    • B. Section 754
  • VI. THE TRUTH ABOUT COMPLEXITY
    • A. Ease With Which Otherwise Complex Rules Are Avoided as a Mitigating Factor
    • B. Length and Technical Nature of Rules as Contributing Factors
    • C. Conformity to Expectations as a Mitigating Factor
    • D. Uncertainty as a Contributing Factor
    • E. Factual Information Needed as a Contributing Factor
  • VII. REFORMS WOULD SIMPLIFY LAW IN SOME RESPECTS AND, IN OTHER RESPECTS, MAKE LAW NO MORE COMPLEX

    • A. Requiring Partnerships to Use the Most Accurate Method Would Reduce, or at Least Not Increase, Ex Ante Complexity.
      • 1. Current law does not reduce Ex Ante Complexity by allowing taxpayers to easily avoid an otherwise complex rule.
      • 2. The proposed reforms would reduce Ex Ante Complexity by making applicable law shorter.
      • 3. The proposed reforms would not make the applicable rules meaningfully more technical for purposes of Ex Ante Complexity.
      • 4. The proposed reforms would lead to results that better conform to expectations.
      • 5. The proposed reforms would reduce uncertainty.
      • 6. The proposed reforms would not impose any incremental information gathering requirements.
    • B. Requiring partnerships to use the most accurate method would reduce, or at least not increase, Ex Post Complexity.
      • 1. Current law does not allow taxpayers to easily avoid an otherwise complex rule.
        • a. Section 704(c)
        • b. Section 754
      • 2. The proposed reforms would make the law shorter.
      • 3. The proposed reforms would not make the applicable rules meaningfully more technical.
      • 4. The proposed reforms would reduce uncertainty.
      • 5. The proposed reforms would not impose any incremental information gathering requirements.
    • C. Making the elections universally available ultimately bred more complexity.
  • VIII. CONCLUSION

 

 

 

Introduction

In the context of partnership tax, one frequently repeated storyline involves sophisticated parties exploiting rules designed to accommodate the needs of Mom and Pop partnerships. For example, in the 1990s, a subsidiary of General Electric Capital Corporation (“GECC”) undertook a transaction to shift the tax consequences of $310 million of taxable income from itself to parties not subject to tax on the income, reducing total tax paid by $62 million. The transaction did not involve shifting economic income from GECC’s subsidiary to the parties not subject to tax and involved minimal economic risk. The parties’ tax advisors crafted the transaction in order to take advantage of a tax provision that was designed to serve the needs of unsophisticated partners, as discussed in detail later in this paper.[1] Often transactions like this provoke responses from Congress or the Treasury designed to prevent taxpayers from undertaking similar transactions in the future. The responses of lawmakers invariably further complicate partnership tax law.

In 1954, when Congress enacted Subchapter K (the part of the Internal Revenue Code (“Code”) that governs the taxation of partners and partnerships),[2] Congress was aware of the fact that many small, unsophisticated entities were treated as partnerships for tax purposes.[3] Even today, many partnerships are small entities.[4] Thus, lawmakers designing rules to govern partnerships have justifiably expressed concern for the ability of small, unsophisticated partnerships to apply complex rules. However, as transactions like the one described in the preceding paragraph demonstrate, undue focus on the needs of unsophisticated partnerships leads to rules that are ripe for manipulation by sophisticated taxpayers. Moreover, it is not that sophisticated taxpayers use partnerships for only tax-motivated transactions. Sophisticated taxpayers also use partnerships to carry out transactions motivated by genuine economic goals and large partnerships earn significant income.[5] Use of large partnerships has grown, in part, because many hedge funds, real-estate funds, and private-equity funds are organized as partnerships for tax purposes. In addition, in recent years, sponsors that manage such funds, including Blackstone, Fortress, and KKR, publicly offered interests in entities treated as partnerships for tax purposes.[6] The entities are entitled to receive a portion of the economic return that the sponsors receive from the funds that they manage.[7]

As a result of the varied use of partnerships, lawmakers are left with the unenviable task of crafting law that can simultaneously address the needs of unsophisticated partnerships, evolve to combat abuses attempted by sophisticated partnerships, and provide flexibility demanded by sophisticated partnerships operating legitimate businesses.[8] In response to the predicaments that lawmakers face when attempting to accommodate the needs of diverse partnerships, other scholars have suggested bifurcating the law applicable to partnerships. Under such proposals, a simplified version of current law would be available only to uncomplicated partnerships, and a more complex version of current law (or, under some variations of the proposal, the rules that currently apply to corporations) would be imposed on all other partnerships.[9]

In this paper, I propose more modest reform within the confines of one body of law applicable to all partnerships.[10] I argue that useful improvement can be achieved within one set of rules because there are revisions that can simplify the law (making it more suitable for unsophisticated partnerships) and, at the same time, make the law more accurate and, as a consequence, less susceptible to manipulation by sophisticated partnerships.

It may seem surprising that there are opportunities to simultaneously make the law more suitable for unsophisticated partnerships and make the law less prone to exploitation by sophisticated partnerships. Yet, such opportunities exist. They exist because certain provisions that are intended to make the law simpler for unsophisticated partnerships, in fact, make the law more complicated, as an examination of theories related to legal and tax complexity will demonstrate.

In the rest of the paper, I will elaborate upon the points described above, using two specific tax provisions as examples.[11] Part I describes the first specific provision of the Internal Revenue Code, § 704(c), a provision that governs tax allocations made with respect to contributed property. Part II discusses the history of § 704(c) and how concern about complexity and unsophisticated partners affected the evolution of § 704(c). Part III describes the second specific provision, § 754, a provision that addresses certain adjustments made to a partnership’s basis in its assets. Part IV discusses the history of § 754 and how concern about imposing burdensome requirements on unsophisticated partners influenced the development of § 754. Part V discusses how reforms would make partnership tax law less susceptible to manipulation by sophisticated partnerships. Part VI examines considerations that should inform an understanding of what constitutes complexity. Part VII applies the considerations discussed in Part VI in order to demonstrate that the proposed reforms would simplify the law in some respects (and, in other respects, make it no more complex), resulting in law that is more suitable for unsophisticated partnerships. Part VIII concludes the paper.

I. Existing Law Under § 704(c)

Section 704(c) is a provision that governs how partners share tax items that a partnership recognizes with respect to certain assets. Section 704(c) applies to two different types of property. First, § 704(c) applies to property contributed to a partnership at a time when the value of the property is greater than (or less than) the contributing partner’s tax basis in the property. The contributing partner generally does not recognize the existing built-in gain (or existing built-in loss) at the time of the contribution. When the partnership later sells the property (or recognizes other tax items, such as depreciation, with respect to the property), § 704(c) requires that the partners share tax gain or loss (or other tax items) in a way that takes into account the built-in gain (or built-in loss) that existed when the property was contributed to the partnership. Second, if a new partner joins a partnership at a time when the partnership’s existing assets contain built-in gains or built-in losses and the partnership revalues its assets to reflect the fact that the new partner buys into the partnership based on the current value of the partnership’s assets, § 704(c)[12] requires that the partners share future tax gain or loss (or other tax items) recognized by the partnership in a way that takes into account the built-in gain (or built-in loss) that existed in the partnership’s assets at the time the new partner joined the partnership. This Part will describe the mechanics of § 704(c) in detail.

When a partner contributes property to a partnership, the partner and the partnership generally do not recognize any gain or loss for tax purposes as a result of the contribution.[13] In order to demonstrate, assume the following facts:

Example 1. An individual, A, owns a piece of land that A acquired some time ago for $5,000. The value of the land has increased over time so that the land is currently worth $15,000. A and B, another individual, form an entity that is treated as a partnership for tax purposes. A contributes the land and B contributes $15,000 cash to this newly formed AB partnership, each in exchange for a 50% interest in the AB partnership.

Under the facts of Example 1, A will not be required to recognize (and potentially pay tax on) the $10,000 built-in gain that exists in the land at the time of the contribution.[14] However, in order to ensure that the $10,000 built-in gain is preserved to potentially be recognized at a future point in time, the AB partnership will obtain a tax basis in the land equal to $5,000 (A’s basis in the land),[15] and A will obtain a tax basis in A’s interest in the partnership equal to $5,000 (A’s basis in the land).[16] As a result, if A sold his or her interest in the partnership for $15,000 after the contribution, A would recognize $10,000 of tax gain, and if the AB partnership sold the land for $15,000 after the contribution, the AB partnership would recognize $10,000 of tax gain. Moreover, a partnership does not itself pay tax at an entity level on income recognized by the partnership but, rather, allocates items of taxable income, gain, loss and deduction among its partners so that its partners will take such items into account for purposes of computing their taxable income.[17] Consequently, this $10,000 of tax gain recognized by the partnership would be allocated to the partners.[18]

Regarding how the $10,000 tax gain would be allocated, § 704(c)(1)(A) provides:

[I]ncome, gain, loss and deduction with respect to property contributed to the partnership by a partner shall be shared among the partners so as to take account of the variation between the basis of the property to the partnership and its fair market value at the time of the contribution.[19]

In other words, under the facts of Example 1, if the AB partnership sells the land for $15,000 so that the AB partnership recognizes $10,000 of tax gain, the $10,000 tax gain must be allocated between A and B in a manner that takes into account the difference between the basis of the land ($5,000) and the fair market value of the land ($15,000) at the time A contributed the land to the partnership.[20] Section 704(c)(1)(A) leaves to the Treasury Regulations the task of specifying how allocations should take into account built-in gain or built-in loss that exists in property at the time at which it is contributed to a partnership.[21]

The Treasury Regulations under § 704(c) provide that allocations must be made using a “reasonable method” that is consistent with the purpose of § 704(c). This purpose, according to the Treasury Regulations, is to “prevent the shifting of tax consequences among partners with respect to precontribution gain or loss.”[22] The Treasury Regulations describe three methods that are “generally reasonable.”[23] These three methods are the “traditional method,” the “traditional method with curative allocations,” and the “remedial method.”[24] The Treasury Regulations do not require that partnerships use any particular method and, in fact, the Treasury Regulations provide that a partnership may use different methods with respect to different items of contributed property as long as the “overall method or combination of methods are reasonable based on the facts and circumstances and consistent with the purpose of section 704(c).”[25] Finally, the Treasury Regulations contain an anti-abuse rule, discussed in more detail below, that places some constraints on the flexibility afforded by the Regulations.[26]

In order to illustrate the operation of the three methods, we return to the facts of Example 1 set forth above. If the partnership sells the land for at least $15,000, each of the three methods under § 704(c) would lead to the same result. Assume, for example, that the AB partnership sells the land for $20,000 one year after the partnership was formed, and two years after selling the land, the AB partnership liquidates and distributes the cash that it holds ($35,000) equally ($17,500 each) to A and B. When the partnership sells the land for $20,000, the partnership recognizes tax gain of $15,000 (the excess of the selling price over the partnership’s $5,000 basis in the land). Regardless of which of the three § 704(c) methods is used, the AB partnership will allocate this $15,000 tax gain as follows: $12,500 to A and $2,500 to B. The $12,500 of tax gain allocated to A represents the $10,000 of gain that accrued while the land was held by A prior to contributing the land to the partnership plus A’s 50% share of the $5,000 gain that accrued after the land was contributed to the partnership. The $2,500 of tax gain allocated to B represents B’s 50% share of the $5,000 gain that accrued after the land was contributed to the partnership. Finally, each partner’s basis in his or her interest in the partnership is increased by the amount of tax gain allocated to that partner upon sale of the land,[27] resulting in a $17,500 basis for each partner. Consequently, A and B do not recognize gain or loss as a result of receiving a $17,500 cash distribution on liquidation since the amount of the cash distribution received by each partner equals each partner’s basis in his or her interest in the partnership.[28]

As described above, the results under § 704(c) are not affected by the method elected by a partnership if the contributed land appreciated in value prior to the contribution and the partnership sells the land for an amount at least equal to the land’s value at the time of contribution. However, if the land appreciated in value prior to contribution and the partnership sells the land for an amount less than the land’s value at the time of the contribution, the tax results will depend on which method the partnership elects to use for making § 704(c) allocations with respect to the land, and use of the remedial method most reliably ensures that tax consequences will not be shifted from the contributing partner to other partners. Moreover, while partners do not expect land to decrease in value after it is contributed to a partnership, the value of contributed land often declines – a fact of which many real estate partnerships were reminded in the recent economic downturn. Furthermore, in the case of depreciable property, the method elected by the partnership can affect the allocation of depreciation deductions even if the value of the property does not decline, and in this area as well, the remedial method most consistently guarantees that tax consequences will not be shifted from the contributing partner to a non-contributing partner.[29]

In order to demonstrate the impact of the particular § 704(c) method elected in a case in which the value of appreciated land declines after contribution to a partnership, assume the following facts:

Example 1A. An individual, A, owns a piece of land that A acquired some time ago for $5,000. The value of the land has increased over time so that the land is currently worth $15,000. A and B, another individual, form an entity that is treated as a partnership for tax purposes. A contributes the land and B contributes $15,000 cash to this newly formed AB partnership, each in exchange for a 50% interest in the AB partnership. As a result of the contribution, A does not recognize any tax gain.[30] Following the contribution, the partnership’s basis in the land is $5,000,[31] A’s basis in his or her interest in the partnership is $5,000,[32] and B’s basis in his or her interest in the partnership is $15,000.[33] One year after the partnership was formed, the AB partnership sells the land for $10,000. Two years after selling the land, the AB partnership distributes the cash that it holds ($25,000) equally to A and B ($12,500 each).

The results under each of the § 704(c) methods are described below.

A. Traditional Method

Under the facts of Example 1A, because the partnership’s basis in the land is $5,000, the partnership recognizes $5,000 of tax gain upon sale of the land for $10,000. If the partnership uses the traditional method for allocating items under § 704(c) with respect to the land, $5,000 of tax gain will be allocated to A and no tax gain or loss will be allocated to B.

In effect, use of the traditional method, under the facts of Example 1A, results in a shift from A to B of $2,500 of the $10,000 tax gain attributable to the increase in value of the land that occurred while it was owned by A. Prior to A’s contribution of the land, the land increased in value by $10,000 (to $15,000), and A benefited economically from that increase in value in its entirety because A was able to exchange the land for a 50% interest in a partnership that held assets worth $30,000. Subsequent to A’s contribution of the land to the AB partnership, the value of the land declined by $5,000. Because A and B share the economic benefits and burdens of the AB partnership equally, this decline in value will be shared equally by A and B ($2,500 each), so that, for example, when the partnership distributes the cash that it holds ($25,000) in liquidation, each of A and B receive $12,500 ($2,500 less than the value that each contributed to the partnership). Consequently, if A and B were each allocated taxable gain and loss in an amount that matched economic gain and loss, A would be allocated $10,000 of tax gain from sale of the land and $2,500 of tax loss from sale of the land (or $7,500 of tax gain from sale of the land on net), and B would be allocated $2,500 of tax loss from sale of the land. However, because the only tax item recognized by the partnership as a result of sale of the land is $5,000 of tax gain, $5,000 of tax gain is allocated to A to get as close as possible to the result described above, while using only tax items actually recognized by the partnership from sale of the land.[34] Compared to what should have been allocated to the partners based on their economic gain and loss, A is allocated $2,500 less tax gain than what A should have been allocated, and B is allocated $2,500 less tax loss than what B should have been allocated. In effect, $2,500 of tax gain has been inappropriately shifted from A to B.[35]

Moreover, while this shift may be temporary, it can, nevertheless, significantly affect the tax consequences experienced by A and B. After the allocation of the $5,000 tax gain from sale of the land to A, A’s basis in his or her interest in the partnership will increase to $10,000,[36] and B’s basis in his or her interest in the partnership will remain $15,000. If the partnership distributes $12,500 cash to each of A and B in liquidation of the partnership, A will recognize $2,500 of tax gain (the excess of $12,500 cash over A’s $10,000 basis in his or her interest in the partnership), and B will recognize $2,500 of tax loss (the excess of B’s $15,000 basis in his or her interest in the partnership over $12,500).[37] The $2,500 tax loss recognized by B corresponds to the $2,500 tax loss from sale of the land that should have been but was not allocated to B at the time of the sale of the land, as a result of shifting $2,500 of tax gain from A to B. Likewise, the $2,500 tax gain recognized by A corresponds to the $2,500 tax gain from sale of the land that should have been allocated to A but was instead shifted from A to B. However, recognition of a $2,500 tax loss (or gain) on liquidation by B (or A) does not fully compensate for the earlier shift in tax gain from A to B. For one thing, tax gain and loss from sale of the land could be of a different character, with different resulting tax consequences, than tax gain or loss recognized on liquidation. For another, the tax loss (or gain) on liquidation may be recognized years later if substantial time elapses between sale of the land and liquidation of the partnership.[38]

B. Traditional Method with Curative Allocations

If the only tax item ever recognized by the partnership is the gain from sale of land contributed by A (as is the case under the facts of Example 1A), then the traditional method with curative allocations would lead to the same results as the traditional method. However, if the facts are complicated slightly, the two methods lead to different results. In order to demonstrate, assume the following set of facts:

Example 1B. An individual, A, owns a piece of land (“Land #1”) that A acquired some time ago for $5,000. The value of the land has increased over time so that the land is currently worth $15,000. A and B, another individual, form an entity that is treated as a partnership for tax purposes. A contributes the land and B contributes $15,000 cash to this newly formed AB partnership, each in exchange for a 50% interest in the AB partnership. As a result of the contribution, A does not recognize any tax gain.[39] Following the contribution, the partnership’s basis in Land #1 is $5,000,[40] A’s basis in his or her interest in the partnership is $5,000,[41] and B’s basis in his or her interest in the partnership is $15,000.[42] The partnership uses the $15,000 cash contributed by B to acquire a second parcel of land (“Land #2”) for $15,000. Consequently, the partnership’s basis in Land #2 is $15,000. One year after the partnership was formed, the AB partnership sells Land #1 for $10,000. Subsequently (and in the same tax year as the sale of Land #1 or in a subsequent tax year), the partnership sells Land #2 for $20,000. Two years after selling Land #2, the AB partnership distributes the cash that it holds ($30,000) equally to A and B ($15,000 each).

Under the facts of Example 1B, if the traditional method is used, the $5,000 tax gain that results from sale of Land #1 is allocated entirely to A, and the tax consequences of $2,500 of the pre-contribution increase in value of Land #1 are effectively shifted from A to B, as described above in connection with Example 1A. Upon sale of Land #2, the partnership recognizes $5,000 of tax gain and $5,000 of economic gain. Under the traditional method, because A and B share equally the $5,000 economic gain from Land #2,[43] they will also share equally the $5,000 tax gain from Land #2. After the allocation of $5,000 of tax gain from sale of Land #1 to A, $2,500 of tax gain from sale of Land #2 to A and $2,500 of tax gain from sale of Land #2 to B, A’s basis in his or her interest in the partnership will increase to $12,500, and B’s basis in his or her interest in the partnership will increase to $17,500.[44] When the partnership distributes $30,000 of cash equally to A and B on liquidation ($15,000 each), A will recognize $2,500 of tax gain and B will recognize $2,500 of tax loss.[45] As discussed above, while this $2,500 amount corresponds to the amount of tax gain attributable to pre-contribution appreciation in the value of Land #1 that was shifted from A to B, recognition of tax gain and loss on liquidation does not fully compensate for the earlier shift because the liquidation may occur in a later year than the year in which the land was sold and the tax gain and loss recognized on liquidation may be of a different character (leading to different tax consequences) than tax gain and loss from sale of the land.

If the traditional method with curative allocations is applied to the facts of Example 1B, the $5,000 of tax gain recognized on sale of Land #1 will be allocated in its entirety to A, like it was under the traditional method. However, unlike the traditional method, under the traditional method with curative allocations, the $5,000 of tax gain recognized upon the sale of Land #2 will be allocated in its entirety to A (rather than being allocated $2,500 to each of A and B to match how economic gain from the sale of Land #2 is shared).[46] In effect, $2,500 of tax gain from Land #2 is shifted from B to A on sale of Land #2 to counteract the fact that $2,500 of tax gain attributable to pre-contribution appreciation in the value of Land #1 was shifted from A to B on sale of Land #1. After the allocation of $5,000 of tax gain from sale of Land #1 to A and $5,000 of tax gain from sale of Land #2 to A, A’s basis in his or her interest in the partnership will increase to $15,000,[47] and B’s basis in his or her interest in the partnership will remain $15,000. When the partnership distributes $30,000 of cash equally to A and B on liquidation ($15,000 each), A and B will not recognize tax gain or loss as a result of the liquidation.[48]

In summary, rather than relying on tax gain and loss recognized on liquidation of the partnership to offset an earlier shift among the partners of tax gain or loss resulting from the sale of contributed property (like the traditional method), the traditional method with curative allocations utilizes tax gain and loss recognized from sale of other assets during the life of the partnership to adjust for a shift among the partners of tax gain or loss arising from the sale of contributed property. Consequently, the traditional method with curative allocations potentially mitigates a shift in tax consequences attributable to pre-contribution appreciation or depreciation in asset value more accurately than the traditional method for two reasons. First, the traditional method with curative allocations can potentially offset a shift of tax consequences attributable to pre-contribution change in asset value earlier in time.[49] In Example 1B, this is true since the sale of Land #2 occurs earlier than the liquidation of the partnership. Second, whereas tax gain or loss recognized on liquidation may be of a different character (with different resulting tax consequences) than the tax gain or loss attributable to pre-contribution change in asset value, the Treasury Regulations require that tax gain or loss that is allocated by the partnership under the traditional method with curative allocations “must be expected to have substantially the same effect on each partner’s tax liability” as the tax gain or loss attributable to pre-contribution change in asset value.[50] In other words, the allocations described above are only allowed if tax gain from sale of Land #2 is of the same character as tax gain from sale of Land #1.

However, the ability of the traditional method with curative allocations to more effectively mitigate a shift of tax consequences is limited because the partnership may not, in fact, recognize actual tax items that can be used to offset the shift. Thus, as mentioned in connection with Example 1A above, if the partnership’s only asset was Land #1 and the partnership never recognized a tax item other than the $5,000 tax gain recognized from sale of Land #1, the traditional method with curative allocations would lead to the same results as the traditional method.

C. Remedial Method

Unlike the traditional method, the remedial method does not rely on tax gain or loss recognized on liquidation of the partnership to correct for an earlier shift in tax consequences attributable to a pre-contribution change in asset value, and, unlike the traditional method with curative allocations, the remedial method does not rely on the allocation of actual tax items recognized by the partnership to potentially compensate for such a shift. Rather, under the remedial method, the partnership invents purely fictional tax items of precisely the right amount and the right character to ensure that any shift in tax consequences attributable to a pre-contribution change in asset value is completely offset at the exact time that it would otherwise occur.

In order to illustrate the operation of the remedial method, we return to the facts of Example 1A set forth above. In that example, when the partnership sells the land for $10,000, on net, A has realized $7,500 of economic gain with respect to the land. This $7,500 net economic gain can be separated into two components: (1) $10,000 of economic gain that accrued between the time A acquired the land for $5,000 and the time A exchanged the land for a 50% interest in a partnership that held assets worth $30,000 and (2) A’s 50% share of the $5,000 of economic loss that accrued between the time A contributed the land and the time the partnership sold the land. In total, B has realized $2,500 of economic loss with respect to the land, which represents B’s 50% share of the $5,000 economic loss that accrued after the land was contributed to the partnership. Therefore, if A and B were allocated tax gain and loss in connection with a sale of the land in an amount that precisely matched economic gain and loss realized by each partner, A would be allocated $7,500 of tax gain and B would be allocated $2,500 of tax loss.

However, under the traditional method or the traditional method with curative allocations, the partnership allocates $5,000 of tax gain to A since that is the only tax item actually recognized by the partnership. Thus, under these methods, A is allocated $2,500 less tax gain than what should be allocated to A, and B is allocated $2,500 less tax loss than what should be allocated to B because $2,500 of tax gain has effectively been shifted from A to B.

Under the remedial method, on the other hand, the partnership invents a fictional item of $2,500 of tax loss from sale of the land and an equal and offsetting fictional item of $2,500 of tax gain from sale of the land. In addition to allocating $5,000 of actual tax gain to A, the partnership allocates the $2,500 of fictional tax gain to A, so that A recognizes, in total, $7,500 of tax gain, an amount that precisely matches economic gain realized by A. Likewise, the partnership allocates the fictional item of $2,500 of tax loss to B, which precisely matches economic loss realized by B. As a result, the partnership shifts no tax gain from A to B. Furthermore, because the partnership has allocated tax items to A and B in an amount that matches economic gain and loss realized and the partnership has not shifted any tax gain from one partner to the other, A and B will not recognize any further tax gain or loss when they each receive $12,500 of cash on liquidation. In particular, after the partnership allocates: (i) $7,500 of tax gain from sale of the land to A and (ii) $2,500 of tax loss from sale of the land to B, A’s basis in his or her interest in the partnership will increase to $12,500,[51] and B’s basis in his or her interest in the partnership will decrease to $12,500.[52] Therefore, when the partnership distributes $25,000 of cash equally to A and B on liquidation ($12,500 each), A and B will not recognize gain or loss as a result of the liquidation.[53]

D. Summary of § 704(c) Methods

The results under the three methods, as shown by the examples above, are condensed in the following table.

Table 1: Summary of § 704(c) Methods

Table_1

In summary, the remedial method most reliably carries out the purpose of § 704(c), namely preventing the shifting of tax consequences among partners with respect to pre-contribution gain or loss.[54] In particular, unlike the other methods, the remedial method ensures that, regardless of the amount of actual tax items recognized by the partnership, no such shift will occur.

E. Anti-Abuse Rule

While the Treasury Regulations generally allow a taxpayer to select any reasonable method under § 704(c) with respect to each item of contributed property, they constrain flexibility to some extent by providing that:

An allocation method (or combination of methods) is not reasonable if the contribution of property . . . and the corresponding allocation of tax items with respect to the property are made with a view to shifting the tax consequences of built-in gain or loss among the partners in a manner that substantially reduces the present value of the partners’ aggregate tax liability.[55]

Based on the language of the regulation as well as the way § 704(c) operates, it would appear that a partnership will not run afoul of the anti-abuse rule solely because it selects the § 704(c) method that leads to the most favorable tax consequences. For one thing, if the contribution of a particular piece of property is taken as a given, the selection of one method versus another method would generally be motivated solely by the goal of generating the most favorable tax consequences since selecting one method instead of another method does not affect anything other than tax consequences. In other words, once a given piece of property has been contributed, use of the traditional method instead of the remedial method, for example, only affects how tax items are allocated with respect to the property and does not affect how the partners share in any economic benefits and burdens associated with the property. Therefore, if the contribution of property is taken as a given, the fact that the partnership opts for the traditional method because it might be expected to shift tax gain from a partner subject to a higher effective rate of tax to a partner subject to a lower effective rate of tax should not, in and of itself, run afoul of the anti-abuse rule. This is so because, in every case in which the selection of the traditional method is based on an analysis of the consequences that will follow from selecting that particular method, the selection must be based on the conclusion that the traditional method will lead to the most favorable tax consequences since the only results that vary across the different methods are tax results.   For another thing, the Treasury Regulations state, “An allocation method is not necessarily unreasonable merely because another allocation method would result in a higher aggregate tax liability.”[56]

Thus, instead of addressing the issue of which method is selected, it would seem that the anti-abuse rule envisions a situation in which shifting tax consequences to reduce aggregate tax liability was an important factor motivating the contributing partner to contribute the property to the partnership in the first place.[57] It is not clear how significant the goal of reducing tax liability must have been if it was one of multiple motives that influenced the partner’s decision to contribute property to the partnership.[58] It is also unclear what tax consequences should be used as a baseline for purposes of determining whether or not the contribution of the property and allocation method(s) used substantially reduce the present value of the partners’ aggregate tax liability.[59] One might think that the tax consequences that would result from using the remedial method should be used as a baseline since the remedial method is intended to ensure that tax consequences of built-in gain or built-in loss are not shifted among the partners. However, the preamble to the Treasury Regulations indicates otherwise, stating, “[T]he IRS and Treasury believe that it would be inappropriate to adopt the remedial allocation method as a baseline for measuring whether the partners’ aggregate tax liability has been reduced. Such a baseline would make the remedial allocation method preeminent, undercutting its elective nature.”[60] Moreover, the Treasury Regulations make explicit that, if a partnership is found to violate the anti-abuse rule, the Internal Revenue Service (“the Service”) may not cure the abuse by requiring that the partnership use the remedial method.[61]

F. Other Constraints

The Code contains other provisions designed to foreclose the ability of potential transactions to circumvent the purpose of § 704(c). For example, §§ 704(c)(1)(B) and 737 prevent partnerships from avoiding the consequences that follow from § 704(c) by distributing property to a different partner than the partner who contributed the property[62] or by liquidating a contributing partner’s interest prior to sale by the partnership of the contributed property.[63] Also, § 704(c)(1)(C) provides that a partnership cannot allocate any portion of a pre-contribution loss with respect to property contributed to a partnership to a partner other than the contributing partner.[64] Thus, § 704(c)(1)(C) limits the shifting of tax losses from the contributing partner to other partners even though § 704(c) is evidently less concerned about shifting tax gains from one partner to another partner despite the fact that shifting tax gains can also reduce tax revenue (as shown in Tables 7 and 8 below). This asymmetrical treatment of tax gains and losses is replicated in other areas of tax law.[65]

G. Reverse § 704(c) Allocations

If a new partner joins a partnership at a time when the partnership’s existing assets contain built-in gains or built-in losses and the partnership revalues its assets to reflect the fact that the new partner buys into the partnership based on the current value of the partnership’s assets, allocation methods similar to the allocation methods described above must be used to prevent shifting the tax consequences of built-in gains and built-in losses from the existing partners to the new partner.[66] In this context, the tax allocations are sometimes called “reverse § 704(c)” allocations.

To illustrate the operation of reverse § 704(c) allocations, assume the following set of facts:

Example 2. Two individuals, A and B, form an entity treated as a partnership for tax purposes. A and B each contribute $5,000 cash to the newly formed AB partnership. The AB partnership acquires a parcel of land for $10,000. The AB partnership’s basis in the land is $10,000. At a time when the value of the land is $30,000, another individual, C, contributes $30,000 cash to the AB partnership in order to acquire a 50% interest in the partnership. The partnership revalues the land to $30,000 to reflect the partners’ economic arrangement.

If the partnership subsequently sells the land for $30,000, the partnership will recognize a $20,000 tax gain. Regardless of which method the partnership uses for making reverse § 704(c) allocations with respect to the land, the partnership will allocate this tax gain to A and B (not C) in its entirety. Thus, the tax consequences of the existing built-in gain are not shifted to C. This result appropriately matches the partners’ economic arrangement because the economic benefit of the $20,000 increase in value of the land is captured by A and B and not C. In other words, if the partnership distributes the $60,000 cash it holds after sale of the land 50% to A and B and 50% to C, A and B collectively receive $20,000 more than they initially contributed to the partnership and C receives only a return of C’s initial contribution. Thus, it is appropriate that A and B, collectively, are allocated $20,000 of tax gain, while C is allocated no tax gain or loss.

H. Special Rules for Securities Partnerships

In general, § 704(c) allocations and reverse § 704(c) allocations must be determined on a property-by-property basis.[67] This requirement is relaxed in the case of reverse § 704(c) allocations made by a “securities partnership,” which is allowed to determine reverse § 704(c) allocations on an aggregate basis across all of its “qualified financial assets.”[68] A “securities partnership” is a partnership that shares all economic gain pro rata among partners based on capital invested (except for a disproportionate share that benefits a partner who provides investment management services) and is either (1) registered as a management company under the Investment Company Act of 1940 or (2) holds qualified financial assets that represent at least 90% of the value of its non-cash assets and revalues its assets at least annually.[69] A “qualified financial asset” is personal property (including stock) that is actively traded and, in some cases, certain other property.[70] A typical hedge fund, for example, would qualify as a “securities partnership.” The distinction between an asset-by-asset approach and an aggregate approach is demonstrated by the following example.

1. Facts of Example 3

The AB partnership qualifies as a securities partnership. The AB partnership owns two assets, both of which are qualified financial assets. Two individuals, A and B, formed the AB partnership by contributing $100 each to the partnership in exchange for a 50% interest. A third individual, C, subsequently acquires a 50% interest in the partnership (diluting A’s interest to 25% and B’s interest to 25%). The assets held by the partnership at the time C joins the partnership, along with the original cost of each asset, the value of each asset at the time C joins the partnership, the eventual sale price of each asset, and the timing of the subsequent sale of each asset are shown in the table below.

Table 2: Facts of Example 3

Table_2

2. Results Under an Asset-by-Asset Approach

Upon sale of Stock 1 for $200, the partnership recognizes $100 of taxable gain. Because $50 of built-in gain existed with respect to Stock 1 at the time C joined the partnership, the partnership must allocate $50 of the tax gain equally to A and B. The remaining $50 tax gain is attributable to gain that accrued after C joined the partnership and, therefore, is allocated 25% to A, 25% to B, and 50% to C, which corresponds to how they share in the $50 of gain economically.

Upon sale of Stock 2 for $110, the partnership recognizes $10 of taxable gain. Twenty five dollars of built-in gain existed with respect to Stock 2 at the time C joined the partnership and $15 of economic loss accrued with respect to Stock 2 after C joined the partnership. Tax allocations made upon sale of Stock 2 will depend on which method the partnership uses for making reverse § 704(c) allocations with respect to Stock 2. If the partnership uses the traditional method, for example, the partnership will allocate $10 of taxable gain equally to A and B and none to C. If the partnership uses the remedial method, though, the partnership will allocate a notional item of 50% times $15 (or $7.50) of tax loss to C (to correspond to the economic loss realized by C) and a notional item of $7.50 of tax gain plus $10 of actual tax gain (or $17.50 of tax gain) equally to A and B (which corresponds to the net economic gain realized by A and B).

3. Results Under an Aggregate Approach

Rather than tracking existing built-in gain asset-by-asset, the partnership simply takes into account the fact that an aggregate built-in gain existed when C joined the partnership. Furthermore, upon sale of a given asset, the partnership need not consider the amount of built-in gain that existed with respect to that particular asset when C joined the partnership. These facets of the aggregate approach can simplify the calculations involved, particularly if the partnership holds hundreds of qualified financial assets.

If an aggregate approach is used in the example above, the partnership will maintain a revaluation account for each partner reflecting the amount of tax gain that should be allocated to that partner. For each of A and B, at the end of year 1, the balance of the revaluation account will be 50% times $75 (their share of built-in gain that exists when C joins, calculated as $150 value Stock 1 when C joins minus $100 acquisition cost Stock 1 plus $125 value Stock 2 when C joins minus $100 acquisition cost Stock 2) plus 25% times $35 (their share of net gain that accrues after C joins, calculated as $200 eventual sale price Stock 1 minus $150 value Stock 1 when C joins plus $110 eventual sale price Stock 2 minus $125 value Stock 2 when C joins) or $46.25 in total. For C, at the end of year 1, the balance of the revaluation account will be 50% times $35 (C’s share of net gain that accrues after C joins) or $17.50. Upon sale of Stock 1 for $200, the partnership recognizes $100 of taxable gain. This $100 taxable gain will be allocated pro rata to A, B, and C based on the balance of each partner’s revaluation account or $42 to A, $42 to B, and $15.90 to C. Upon sale of Stock 2 for $110, the partnership recognizes $10 of tax gain. This $10 taxable gain will be allocated pro rata to A, B, and C based on the balance of each partner’s revaluation account or $4.25 to A, $4.25 to B, and $1.60 to C.

As a result, total allocations of tax gain to each partner from both assets are: (1) $46.25 for A, (2) $46.25 for B, and (3) $17.50 for C. For A and B this corresponds to their 50% share of the $75 of built-in gain that existed at the time C joined plus their 25% share of the additional $35 of net gain that accrued after C joined. For C, this corresponds to 50% of the $35 of net gain that accrued after C joined the partnership.

II. History of § 704(c) and Concerns about Unsophisticated Partners

Section 704(c) did not always exist in the form described above. In fact, when the part of the Code governing the taxation of partners and partnerships was adopted in 1954, applying the principles described above was entirely optional. This part will describe the original version of § 704(c), justifications offered for the original version of § 704(c), and the evolution of § 704(c) from its original form to its current form. As this discussion will show, concerns about imposing complex rules on unsophisticated partners played a key role in the development of § 704(c).

A. Original Version of § 704(c)

As originally enacted, § 704(c) provided that tax items recognized by a partnership with respect to property contributed to the partnership would be allocated “in the same manner as if such property had been purchased by the partnership.”[71] This general rule applied unless the partnership elected to make allocations in a manner that took into account the difference between the property’s fair market value and its basis at the time of the contribution.[72]

In order to illustrate the operation of the original rule, assume the following set of facts:

Example 4. A owns a piece of land that A acquired some time ago for $5,000. The land is currently worth $15,000. A contributes the land and B contributes $15,000 cash to the newly formed AB partnership, each in exchange for a 50% interest in the AB partnership. A does not recognize any gain as a result of the contribution. The AB partnership’s basis in the land is $5,000. A’s initial basis in his or her interest in the partnership is $5,000, and B’s initial basis in his or her interest in the partnership is $15,000. The partnership sells the land for $15,000. As a result, the partnership recognizes $10,000 of tax gain from sale of the land.

 Based on the original version of § 704(c), the partnership had two choices. Under the general rule, the partnership would allocate the $10,000 of tax gain as if the land had been purchased by the partnership. In other words, the partnership would allocate the tax gain equally ($5,000 each) to A and B since A and B each own a 50% interest in the partnership.[73] This would shift $5,000 of tax gain from A to B since, economically, A has realized a $10,000 economic gain (attributable to precontribution appreciation) and B has realized no economic gain. Following the allocation of tax gain of $5,000 to each partner, A’s basis in his or her interest in the partnership would be $10,000, and B’s basis in his or her interest in the partnership would be $20,000. If the partnership distributed $30,000 cash equally to A and B on liquidation ($15,000 each), A would recognize $5,000 of tax gain on liquidation and B would recognize $5,000 of tax loss on liquidation. But for character and timing differences (which could be significant), the tax gains and losses recognized on liquidation offset the earlier shift of tax gain from A to B.

Alternatively, if it chose to do so, the partnership could apply the principles of current § 704(c) by allocating tax gain in a manner that takes into account the difference between the fair market value of the land and the basis of the land at the time of contribution. In that case, all $10,000 of tax gain recognized on sale of the land would be allocated to A, so that no tax gain would be shifted from A to B.

B. Justifications Offered for the Original Version of § 704(c)

When discussing why it adopted the original version of § 704(c), the Senate Finance Committee stated that the general rule (under which partnerships did not take into account built-in gains and losses) was adopted because “of its extreme simplicity” compared to alternative methods.[74] The general rule was viewed as simple because tax items recognized with respect to contributed property could be allocated in the same manner as all other tax items.[75] The Committee’s concerns about complexity likely arose because of its observation that many partnerships were unsophisticated. As an indication of this underlying rationale, the Senate Finance Committee lamented that uncertainty in partnership tax was “particularly unfortunate” given the large number of partnerships and given that “the partnership form is much more commonly employed by small businesses and in farming operations than the corporate form.”[76]

Finally, the Committee seemed to assume that partners in a partnership would generally be subject to similar effective rates of tax so that shifting tax consequences from one partner to another partner should not raise serious concerns. Therefore, the Committee was not overly worried about whether or not partnerships employed a method that took into account built-in gains and losses, and, consequently, provided partnerships with the flexibility to decide whether to use such a method. Thus, the Committee stated, “partners should be free to choose this more complicated rule for dividing basis of property [in other words, the method that takes into account built-in gains and built-in losses] if they desire the more accurate tax results it brings.”[77] However, use of the more accurate method was elective because, in the Committee’s view, whether a partnership used the more complicated and accurate method was “not a matter involving revenue considerations to the Government.”[78]

C. Evolution of § 704(c)

Over time it became clear that shifting tax consequences from one partner to another was, in fact, a “matter involving revenue considerations to the Government.”[79] When partners have significantly different tax profiles, shifting tax consequences among partners can reduce total tax revenue collected, at least if the time value of money is taken into account and, in some cases, even if it is not.

In order to illustrate, we return to the facts of Example 4 given above and add some additional facts about the partners. In particular, assume that B is subject to a 0% effective rate of tax on tax gain from sale of the land and on tax gain and loss from liquidation of B’s interest in the partnership. Perhaps B is a tax-exempt entity and does not take into account tax gain or loss from sale of the land or liquidation of B’s interest in the partnership for purposes of computing any tax liability B might owe. Alternatively, perhaps B has significant tax losses from other sources that would offset tax gain from sale of the land and make tax loss from liquidation of B’s interest in the partnership not particularly valuable. Assume A, on the other hand, would be subject to a 35% effective tax rate on gain from sale of the land but a 15% effective tax rate on gain from liquidation of A’s interest in the partnership. Under this set of facts, the tax revenue collected using the general rule (under which tax gain from sale of the land is allocated equally to each partner) is shown in the following table.

Table 3: General Rule (Tax Gain From Sale of Land Allocated Equally to A and B)

Table_3By contrast, the tax revenue collected using the rule that applies current § 704(c) principles (under which tax gain from sale of the land is allocated entirely to A) is shown in the following table.

Table 4: Built-In Gain Taken Into Account (Tax Gain From Sale of Land Allocated Entirely to A)

Table 4

Thus, under the facts assumed above, the method used would affect timing of tax revenue collection as well as the total dollar amount of tax revenue collected. The facts of the example could be altered so that, for example, A is subject to a 15% effective tax rate on gain from sale of the land and a 15% effective tax rate on gain from liquidation of A’s interest in the partnership, while B is subject to a 0% effective tax rate on gain or loss from sale of the land or liquidation of B’s interest in the partnership. Under this set of facts, the tax revenue collected using the general rule (under which tax gain from sale of the land is allocated equally to each partner) is shown in the following table.

Table 5: General Rule (Tax Gain From Sale of Land Allocated Equally to A and B)

Table_5

By contrast, the tax revenue collected using the rule that applies current § 704(c) principles is shown in the following table.

Table 6: Built-In Gain Taken Into Account (Tax Gain From Sale of Land Allocated Entirely to A)

Table_6Thus, under this second set of facts, the same amount of total tax revenue is collected over the life of the partnership ($1500 in total) under either method. However, if the partnership is required to apply current § 704(c) principles, more tax revenue is collected in an earlier year. Assuming a 10% discount rate (compounded annually) for purposes of illustration, $750 of tax revenue collected in year 5 plus $750 of tax revenue collected in year 1 is the equivalent of $1262 of tax revenue collected in year 1. Therefore, if the partnership is required to apply current § 704(c) principles, the value of additional tax revenue collected in year 1 is $1500 minus $1262 or $238.

Because a shift in tax consequences among partners can affect the amount of tax revenue collected and the timing of tax revenue collection, in 1984, § 704(c) was amended to largely resemble its current form. Thus, allocating tax items so as to take into account built-in gain or built-in loss that existed when property was contributed to a partnership was no longer optional. However, the Treasury Regulations that implement the current rules allow partnerships to select among various methods for making § 704(c) allocations subject to an anti-abuse rule, as described above. Moreover, as described above, despite the fact that § 704(c) now requires a partnership to take into account built-in gain or built-in loss when making allocations with respect to contributed property, unless the partnership uses the remedial method, it is still possible for tax consequences attributable to pre-contribution changes in the value of assets to be shifted from the contributing partner to the other partners.

The Treasury did not require partnerships to use the remedial method (or something similar), in part because the Treasury apparently believed it lacked authority to do so.[80] However, even if the Treasury was correct in this belief, mandatory use of the remedial method could have been required by Congress. Aside from the possible limitations on the authority of the Treasury, resistance to mandatory use of the remedial method was driven by the same type of concerns that drove Congress to adopt the original version of § 704(c). Namely, concerns about avoiding complexity impeded the adoption of a mandatory remedial method.

One method under consideration leading up to the promulgation of final regulations under § 704(c) was called the “deferred sales method.” The “deferred sales method” was similar to the remedial method and, like the remedial method, it would have ensured that tax consequences attributable to pre-contribution changes in asset value were not shifted away from the contributing partner. In its March 1979 report, the American Law Institute (“A.L.I.”) did not recommend adopting the deferred sales method as a mandatory method under § 704(c) because of concerns about the method’s complexity, even though the A.L.I. acknowledged that many partnerships in 1979 were sufficiently sophisticated to handle the complexities of the deferred sales approach.[81] The A.L.I. was concerned about two aspects of complexity. First, the A.L.I. was concerned that applying the method would require partnerships to value property at the time it was contributed to a partnership which could be difficult for properties that lacked readily ascertainable values.[82] Moreover, the A.L.I. indicated that a method like the traditional method could avoid some of the difficulties of valuing a property when it was contributed to a partnership.[83] Second, the A.L.I. feared that applying the deferred-sales method would be computationally complex.[84]

Because the § 704(c) regulations do not require use of the remedial method, tax consequences can still be shifted in a manner that affects timing (and, in some cases, the total dollar amount) of tax revenue collected. To illustrate, assume the following facts:

Example 5. A owns a piece of land that A acquired some time ago for $5,000. The land is currently worth $15,000. A contributes the land and B contributes $15,000 cash to the newly formed AB partnership, each in exchange for a 50% interest in the AB partnership. The partnership sells the land for $10,000 in year 1. The partnership recognizes $5,000 of tax gain from sale of the land. In year 5, the partnership distributes $25,000 cash equally ($12,500 each) to A and B in liquidation. B is subject to a 0% effective rate of tax on tax gain from sale of the land and on tax gain and loss from liquidation of B’s interest in the partnership. A, on the other hand, is subject to a 35% rate of tax on tax gain from sale of the land but a 15% rate of tax on tax gain from liquidation of A’s interest in the partnership.

Based on the facts of Example 5, the tables below show the tax revenue collected under the traditional method and under the remedial method. As the tables below illustrate, selection of the traditional method would result in lower total tax revenue and a delay in collection of tax revenue compared to use of the remedial method.[85]

Table 7: Traditional Method

Table 7

Table 8: Remedial Method

Table 8

III. Existing Law Under § 754

Section 754 is a provision that allows a partnership to make an election that will generally dictate whether or not the partnership will adjust its basis in its assets following a transfer of an interest in the partnership or following certain distributions made by the partnership. This Part discusses the mechanics of § 754 in detail.

When one partner sells his or her interest in a partnership to another person, unless a § 754 election is in effect, generally the sale will have no impact on the partnership’s basis in its assets.[86] As a result, if a § 754 election is not in effect, gain attributable to an increase in the value of a partnership’s assets potentially will be recognized twice for tax purposes – first on sale of an interest in the partnership and second on sale by the partnership of its assets. By making a § 754 election, the partnership effectively eliminates the second, duplicative tax gain.[87] Special rules apply in the case of partnerships that hold assets that have declined in value.[88] Finally, once a partnership files a § 754 election, the election generally will apply to all transfers that occur in the year with respect to which the election was filed or any subsequent year.[89]

A. Section 754 Elections and Built-In Gain Assets

The mechanics of a § 754 election in the context of a partnership that holds assets that have appreciated in value can be more fully demonstrated by an example. For purposes of the example, assume the following facts:

Example 6. Two individuals, A and B, each contribute $100 in exchange for a 50% interest in a newly formed entity treated as a partnership for tax purposes. As a result, each partner will have a basis in his or her interest in the partnership equal to $100.[90] The partnership uses the cash contributed by the partners to acquire a parcel of land for $200, so that the partnership’s initial tax basis in the land is $200. Over time, the value of the land increases. Two years after the formation of the partnership, when the value of the land is $300 and the partnership holds no other assets and owes no liabilities, A sells his or her interest in the partnership to C for $150. As a result of the sale, A will recognize $50 of gain for tax purposes, which equals the excess of the amount received from C over A’s basis in his or her interest in the partnership.[91] C’s initial basis in his or her interest in the partnership will be $150.[92]

If the partnership does not have a § 754 election in effect, the transfer will have no impact on the partnership’s basis in the land, and, thus, under the facts of Example 6, the partnership’s basis in the land will remain $200.[93] Assume, one year after the sale of A’s interest in the partnership to C, the partnership sells the land for $300. Because the partnership’s basis in the land remained $200, the partnership recognizes $100 of tax gain on sale of the land which is allocated $50 to C and $50 to B. The $50 of tax gain that is allocated to C is effectively a duplication of the tax gain recognized by A on sale of his or her interest in the partnership to C because both items of $50 tax gain are attributable to A’s share of the increase in value of the land that occurred prior to sale of the partnership interest by A to C.

This duplication of tax gain may be temporary. In particular, as a result of the allocation of $50 of tax gain to each partner, each partner’s basis in his or her interest in the partnership will increase by $50 so that C’s basis in his or her interest in the partnership becomes $200 and B’s becomes $150.[94] As a result, if the partnership distributes $150 cash to B and C in liquidation, C will recognize a $50 tax loss at the time of the liquidation (the excess of C’s basis in his or her interest in the partnership over the amount of cash received) and B will recognize no tax gain or loss.[95] The tax loss recognized by C on liquidation is equal in amount to the earlier, duplicative tax gain recognized by C on sale of the land. However, the tax loss recognized by C may not fully offset the effects of the $50 of tax gain recognized by C on sale of the land if the liquidation of the partnership occurs in a later year than the year in which the land was sold. For one thing, if the $50 of tax loss does reduce C’s tax liability, it does so in a later year than the year in which C may have incurred tax liability as a result of the sale of the land. Therefore, taking into account the time value of money and the fact that applicable tax rates may have changed since the time at which the land is sold, the tax loss may not fully offset the consequences of the tax gain. For another thing, the tax loss is likely a capital loss[96] and, assuming C is an individual, may result in only a limited reduction of tax liability if C does not recognize capital gains in the year in which the partnership is liquidated or in a subsequent year. This is so because capital losses of non-corporate taxpayers can generally only be used against capital gains and up to $3,000 of ordinary income,[97] and excess capital losses of non-corporate taxpayers can generally only be carried forward to succeeding taxable years to be used in those years subject to the restrictions just described.[98]

By contrast, if the partnership has made a § 754 election, the transfer of the partnership interest from A to C will have an impact on the partnership’s basis in the land. In particular, under the facts of Example 6, the partnership will increase its basis in the land by $50, which is calculated based on the excess of C’s basis in his or her interest in the partnership (the $150 paid by C for the interest) over C’s share of the partnership’s basis in its assets ($100 or 50% of the $200 basis in the land).[99] However, this increase in basis will be taken into account solely for purposes of determining C’s tax consequences and will not affect tax gain or loss allocated to B.[100] Assume, one year after the sale of A’s interest in the partnership to C, the partnership sells the land for $300. Absent the $50 increase in basis of the land, the partnership’s basis in the land would have been $200, so that the partnership would have recognized $100 of tax gain. Fifty percent of this amount or $50 would be allocated to B, and, even taking into account the basis adjustment, $50 is in fact allocated to B because the basis adjustment affects C and not B. On the other hand, the $50 of tax gain that would have been allocated to C absent an upward basis adjustment of $50 is entirely eliminated by the $50 upward basis adjustment, so that no tax gain or loss is allocated to C. Thus, unlike what occurs in the absence of a § 754 election, the $50 of tax gain attributable to A’s share of the increase in the value of the land that occurred prior to A’s sale of his or her partnership interest to C (and that was recognized by A on sale of his or her interest in the partnership to C) is not recognized a second time by C when the land is sold by the partnership. After the partnership recognizes $50 of tax gain and allocates it to B, B’s basis in his or her interest in the partnership becomes $150 and C’s remains $150. Thus, if the partnership distributes $150 of cash to each partner on liquidation, neither partner recognizes gain or loss as a result of the liquidation.[101]

The tax consequences just described are summarized in the following table.[102]

Table 9: Section 754 Election and Built-In Gain Assets

Table 9

B. Section 754 Elections and Built-In Loss Assets

If a partnership’s assets have a “substantial built-in loss” immediately after the transfer of an interest in the partnership, then the results that would follow from making a § 754 election are mandatory regardless of whether the partnership has made such an election.[103] In other words, in such a case, the partnership is required to take steps to avoid the recognition of the same tax loss a second time. A “substantial built-in loss” exists if a partnership’s total basis in its assets, in aggregate, exceeds the total value of the partnership’s assets by more than $250,000.[104] Certain large partnerships, for which making basis adjustments would be particularly difficult, are allowed to carry out the purpose of mandatory downward basis adjustments through alternative methods.[105]

The following example demonstrates the results of a mandatory downward basis adjustment and the limits on when such an adjustment is required. For purposes of the example, assume the following facts:

Example 7. Two individuals, A and B, each contribute $500,000 in exchange for a 50% interest in a newly formed entity treated as a partnership for tax purposes. The partnership uses the cash contributed by the partners to acquire a parcel of land for $1,000,000. Over time the value of the land decreases. Two years after the formation of the partnership, when the value of the land is $700,000 and the partnership holds no other assets and owes no liabilities, A sells his or her interest in the partnership to C for $350,000. One year later, the partnership sells the land for $700,000. Two years after that, the partnership distributes $350,000 cash to each of B and C in liquidation.

Because the partnership’s basis in its assets exceeds the value of the partnership’s assets by $300,000 immediately after the transfer under the facts of Example 7, a substantial built-in loss exists, and the partnership must reduce the basis in the land by $150,000 for purposes of determining C’s tax consequences. The resulting tax consequences are summarized in the following table.

Table 10: Substantial Built-In Loss

Table 10

 By contrast, if the special rule regarding duplication of built-in losses did not apply and the partnership had not made a § 754 election under the facts of Example 7, the resulting tax consequences would be those shown in the following table.

Table 11: Tax Consequences Absent Substantial Built-In Loss Rules and Absent § 754 Election

Table 11

In this case, the $150,000 tax loss recognized by A on sale of his or her interest in the partnership to C is duplicated, at least temporarily, because, upon sale of the land, C recognizes tax loss attributable to the same decrease in value of the land that led to A’s tax loss. The tax consequences of this tax loss recognized by C may be offset when C recognizes $150,000 tax gain on liquidation of the partnership. However, the offset may not be perfect because the tax gain occurs in a later year than the year of the tax loss, and the tax gain could be of a different character, with different resulting tax consequences, than the tax loss.

Finally, while mandatory downward basis adjustments prevent duplication of losses in cases like the example just described (as shown in Table 10), duplication of losses can, nevertheless, occur in some circumstances. For example, in the case just presented, if the built-in loss that existed in the land at the time of the sale of the partnership interest from A to C had been $250,000 or less and a § 754 election was not in effect, then the built-in loss could be duplicated along the lines shown in Table 11. A basis adjustment also would not be required if, for example, the partnership held land with a built-in loss of $300,000 at the time of the transfer but also held another asset with a built-in gain of $50,000 at the time of the transfer since the existence of a substantial built-in loss is determined on an aggregate basis across all assets. Therefore, a partnership could duplicate a $300,000 loss if it also duplicated a $50,000 gain. In the case of a partnership owned by individuals subject to high marginal rates of tax on ordinary income, for example, this could be a desirable outcome if the built-in gain asset was a capital asset (so that the gain that is duplicated is a capital gain subject to preferential tax rates) and the built-in loss asset is inventory (so that the loss that is duplicated is an ordinary loss that can be used to offset ordinary income subject to higher tax rates).

C. Other Constraints

In addition to the rules that apply in cases in which a partnership’s assets have a “substantial built-in loss”, the anti-abuse regulations that apply for purposes of partnership tax law generally contain examples regarding potentially abusive transactions that involve the failure to make a § 754 election.[106] The examples in the Treasury Regulations suggest that, if a partnership is formed largely for the purpose of duplicating a tax loss that results from the failure to make a § 754 election, the transaction can be successfully challenged. However, the mere failure to make a § 754 election, in and of itself, will not be considered abusive simply because it results in the duplication of tax loss in a particular situation. As the Treasury Regulations state, “The electivity of section 754 is intended to provide administrative convenience for bona fide partnerships that are engaged in transactions for a substantial business purpose. . . . Congress clearly recognized that if the section 754 election were not made, basis distortions may result.”[107]

IV. History of § 754 and Concerns about Unsophisticated Partners

As originally enacted, the results that follow from a § 754 election were not mandatory even in cases involving potential duplication of substantial built-in losses. In 2004, Congress enacted the rules mandating downward basis adjustments in cases in which a partnership’s assets have a substantial built-in loss, in order to address concerns about tax-shelter transactions involving intentional duplication of losses.[108]

Concerns about computational complexity have influenced the hesitation to require basis adjustments in all cases.[109] For example, in a 1980 report recommending making basis adjustments mandatory in all cases, the A.L.I. acknowledged objections that could be made to such a change in the law. One objection was that “unsophisticated partnerships might have to make complex basis adjustments.”[110] The A.L.I. also expressed concern for partnerships at the other end of the spectrum, stating that large partnerships would face complexity if they had to make basis adjustments for “innumerable partners.”[111]

Similarly, in 2004, the House Committee Report described the limitation of mandatory basis adjustments to cases involving substantial built-in losses as a feature that would preserve “the simplification aspects of the current partnership rules for transactions involving smaller amounts.”[112] Interestingly, while the changes ultimately enacted principally resembled the House version of the bill, the Senate’s original version of the bill would have required basis adjustments in all cases. When discussing its proposed bill, the Senate Report stated:

The electivity of these adjustments has become anachronistic and should be eliminated, the Committee believes. Therefore, this provision makes these partnership basis adjustments mandatory, addressing both loss and gain situations. The bill provides that the partnership basis adjustments remain elective in the limited case of transfers of a partnership interest by reason of the death of a partner because that situation may involve unsophisticated taxpayers and constitutes only a narrow, limited set of transfers.[113]

V. How Reforms Would Make Law Less Susceptible to Manipulation by Sophisticated Partnerships

As discussed below, partnership tax law could be reformed in ways that would make the rules more suitable for sophisticated partnerships and unsophisticated partnerships at the same time. The way to achieve this goal is adoption of rules that would require more accurate income measurement (such as requiring all partnerships to use the remedial method under § 704(c) and making the adjustments that follow from a § 754 election mandatory in all cases). These reforms would make the law more suitable for sophisticated partnerships because more accurate income measurement makes it more difficult for taxpayers to manipulate tax consequences.[114] This Part elaborates on this conclusion as applied to §§ 704(c) and 754.

A. Section 704(c)

While the anti-abuse rule in the § 704(c) Regulations likely addresses the most flagrant cases of abuse,[115] it leaves to partnerships the freedom to select a particular § 704(c) method based purely on the goal of reducing tax liability. Selecting the method that minimizes tax liability may not constitute abuse, but it may, nevertheless, be undesirable from the standpoint of tax revenue collection and fairness. Unlike the anti-abuse rule, a mandatory remedial method would bring to an end the ability of partnerships to select a particular method in order to reduce tax liability.

In order to illustrate, we can assume the following facts:

Example 8. A owns a piece of land that A acquired some time ago for $5,000. The land is currently worth $15,000. A contributes the land and B contributes $15,000 cash to the newly formed AB partnership, each in exchange for a 50% interest in the AB partnership. B is subject to a 0% effective rate of tax on tax gain from sale of the land and on tax gain and loss from liquidation of B’s interest in the partnership. A, on the other hand, is subject to a 35% rate of tax on tax gain from sale of the land but a 15% rate of tax on tax gain from liquidation of A’s interest in the partnership.

Further, assume A and B formed the partnership and A contributed the land entirely for valid business reasons. Regardless of the resulting tax consequences, A and B would have decided to form the partnership and have A contribute the land to the partnership. Moreover, when the partnership was formed, the partners expected that the land would maintain its value or increase in value. Therefore, they expected that the partnership would sell the land for at least $15,000. If the land was sold for at least $15,000, the partnership would recognize at least $10,000 of tax gain on sale of the land, and, regardless of the method selected under § 704(c), the first $10,000 of tax gain would be allocated to A and any additional tax gain would be shared equally. Thus, if the land was sold for at least $15,000, tax gain would be allocated in the same manner as economic gain and no shifting of tax consequences would result. Therefore, the partners did not expect that the contribution of the land would result in a shift of tax gain from A (who is subject to a high effective rate of tax) to B (who is subject to a 0% rate of tax).[116] Consequently, since the land was contributed entirely for business reasons and since the partners did not expect the land to be sold at a price that would allow for a shift of tax consequences, the anti-abuse rule should not apply.

That said, assume A and B seek advice regarding which § 704(c) method should be used with respect to the land. Furthermore, assume that the land is in fact sold for less than $15,000, contrary to the partners’ expectations. Their advisors could recommend selecting the traditional method in order to shift tax gain from A (who is subject to a high rate of tax) to B (who is subject to a 0% rate of tax). For example, if the land is sold for $10,000, the results under the traditional method versus the remedial method are shown in Tables 7 and 8 above. Therefore, A and B select the traditional method. They do so not because it is a computationally easier method to apply (even though that appears to be the reason that partnerships were given the flexibility to opt for methods other than the remedial method) but because it leads to more favorable tax consequences.

In summary, while the anti-abuse rule may foreclose the most blatantly tax-motivated transactions, the Treasury Regulations nevertheless give taxpayers the freedom to select among various allowable methods under § 704(c). This election will generally be used by well-advised taxpayers to reduce potential tax liability.[117] Moreover, in some cases, sophisticated partnerships will likely seek tax advice when designing transactions and may still use the flexibility afforded by § 704(c) to carry out tax-motivated transactions but do so with enough business purpose window dressing to escape the anti-abuse rules. Therefore, offering flexibility is undesirable from the standpoint of tax revenue collection and fairness, and, as discussed below, not well suited to the needs of unsophisticated partnerships.

B. Section 754

Existing rules mandating downward basis adjustments in some cases foreclose the possibility that losses will be duplicated in situations in which a particularly large built-in loss exists, in aggregate. In addition, examples in the anti-abuse regulations may prevent the duplication of losses in cases not covered by the mandatory downward basis adjustment rules if the primary reason for forming a partnership is duplication of losses. Nevertheless, the failure to universally mandate basis adjustments may be undesirable from the standpoint of tax revenue collection and fairness. The current rules leave open the possibility that losses can still be duplicated in a way that reduces tax revenue if valid business reasons exist for forming the partnership in the first place (or if taxpayers are able to disguise their tax-avoidance purpose with an engineered business purpose). Moreover, well-advised taxpayers will tend to benefit from the elective nature of the rules more often than other taxpayers.[118] Therefore, particularly given the other concerns following from the elective nature of the rules described below, mandatory basis adjustments would be preferable to the existing system.

VI. The Truth about Complexity

As discussed above, adoption of mandatory rules that would require more accurate income measurement would make partnership tax law less susceptible to manipulation by sophisticated partnerships. Historically lawmakers have hesitated to enact these reforms on the grounds that they make the law too complex for unsophisticated partnerships.[119] Fortunately, this is not the case, and, in fact, the opposite is true as these reforms would simplify the law in some respects and, in other respects, make the law no more complex. The alternative conclusion (that the proposed reforms would make the law too complex) results from certain misconceptions about complexity. In the next Part, I will discuss those misconceptions, but, before doing so, in this Part, I will discuss how complexity should be understood, drawing, in part, from literature related to tax complexity and legal complexity generally.

Legal complexity can be defined to include any features of law that cause costs to be incurred because time, effort, and expertise are required to understand and apply law. As a threshold matter, it is useful to differentiate between two types of complexity, which I will label “Ex Ante Complexity” and “Ex Post Complexity.” Costs that follow from Ex Ante Complexity include: (1) the costs that result from steps taken by affected persons to ensure that the decisions they make concerning future activities are the same decisions that they would make if they fully understood the implications of relevant law; and (2) the costs that result because the decisions that affected persons make concerning future activities are not the same decisions that they would make if they fully understood the implications of relevant law. Others have noted that rules that are not well understood are unlikely to create the incentives that they are intended to produce.[120] The failure of misunderstood rules to create intended incentives is an example of the second type of cost just listed. Costs that follow from Ex Post Complexity include costs of assessing consequences of actions already undertaken or complying with any requirements to report actions already undertaken.

In the tax area, one key difference exists between Ex Ante Complexity and Ex Post Complexity. With respect to Ex Ante Complexity, it may not be clear to unsophisticated taxpayers how altering their behavior could affect their tax liability. Therefore, unsophisticated taxpayers may not be likely to seek expert advice prior to undertaking activities. With respect to Ex Post Complexity, it generally will be clear to taxpayers that it is necessary to file tax returns to report the results of activities already undertaken. Thus, unsophisticated taxpayers are likely to be aware of the need to seek expert advice when preparing tax returns.

Consequently, while there may be a material difference between unsophisticated partnerships and sophisticated partnerships with respect to Ex Ante Complexity (and therefore a reason to have particular concern for unsophisticated partnerships with respect to this type of complexity), there is less likely to be a material difference between unsophisticated partnerships (who seek expert advice) and sophisticated partnerships with respect to Ex Post Complexity. Therefore, there is no reason why Ex Post Complexity would burden unsophisticated partnerships more severely than sophisticated partnerships.[121] If anything, sophisticated partnerships may incur higher costs related to Ex Post Complexity because it may be more difficult to determine the tax consequences of complicated fact patterns typical of sophisticated partnerships.[122] This part elaborates on the foregoing discussion by describing factors that contribute to each type of complexity or mitigate each type of complexity.[123]

A. Ease With Which Otherwise Complex Rules Are Avoided as a Mitigating Factor

The ease with which otherwise complex rules can be avoided can mitigate both Ex Ante Complexity and Ex Post Complexity. In other words, a rule that raises thorny issues when it applies does not meaningfully contribute to complexity if simple steps can be taken to ensure that the rule does not apply.[124] This principle is relevant generally as well as in tax.

With regard to Ex Ante Complexity, if it were difficult to determine the effects of a rule on future activities but simple to adjust the future activities to ensure that a clearer rule applies, overall Ex Ante Complexity would be fairly low. In tax, for example, under the Treasury Regulations that govern how entities are classified for tax purposes (the “Check-the-Box Regulations”), certain entities are eligible to elect whether they will be treated as pass-through entities or corporations for U.S. tax purposes.[125] If an entity that is eligible to elect its classification does not file an election, it will be classified based on specified default rules.[126] Unless a contrary election is filed, a U.S. domestic entity eligible to make an election is classified as a pass-through entity (i.e., a partnership or a disregarded entity) by default.[127] For non-U.S. entities, the default rule is less straightforward. A non-U.S. entity that is eligible to elect its classification is classified as a pass-through entity by default if at least one of its owners does not have limited liability but is classified as a corporation by default if all of its owners have limited liability.[128] Determining with certainty whether or not applicable non-U.S. law provides owners of a non-U.S. entity with “limited liability,” as the term is defined in the U.S. Treasury Regulations, is not an easy proposition. At the very least, making such a determination would require consulting with and likely obtaining a costly opinion from non-U.S. legal counsel. However, it is unnecessary to incur costs to determine an entity’s default classification because taxpayers can, instead, take the easy step of filing an entity classification election specifying the desired classification.[129] Thus, even when the parties are fairly confident that their intended classification is consistent with the default classification based on their understanding of whether or not the entity offers limited liability, they often file an entity classification election on a protective basis.[130]

With regard to Ex Post Complexity, for example, if a given rule required onerous reporting requirements with respect to activities already undertaken but affected persons were allowed to take easy steps to avoid the reporting requirements even after the activities have been undertaken, overall Ex Post Complexity would be fairly low.

B. Length and Technical Nature of Rules as Contributing Factors

More technical rules increase Ex Ante Complexity and Ex Post Complexity, all else being equal. Likewise, the length of potentially applicable law contributes to both types of complexity, all else being equal. The length and technical nature of rules contribute to Ex Ante Complexity by making it more difficult to predict the effects of potential future actions. Similarly, these factors contribute to Ex Post Complexity by making it more difficult to assess the effects of past actions.

Other scholars have categorized rules as “technical” when understanding the rules requires special expertise.[131] I use the term in a similar way. Thus, rules tend to be more technical when they are described (in statutes, regulations, case law, or other applicable sources) using language that includes terms of art or words defined in particular, unique ways. Furthermore, the computational difficulty of rules contributes to their technical nature.

In tax, Ex Post Complexity related to computational difficulty can be significantly mitigated through use of computer programs.[132] However, use of computer programs does not necessarily reduce the likelihood that computational difficulty will result in certain costs associated with Ex Ante Complexity. In particular, computationally difficult rules could lead to an increased likelihood of incurring costs that result because the decisions that affected persons make regarding future actions are not the same decisions that they would make if they fully understood the implications of relevant law.[133] Nevertheless, as discussed below in Part VI.A, computational difficulty does not invariably increase the likelihood of incurring such costs.

C. Conformity to Expectations as a Mitigating Factor

Conformity to expectations can mitigate Ex Ante Complexity. In particular, even if a rule is long and technical, as long as the rule conforms to expectations, the rule does not significantly increase the likelihood of incurring costs that result because the decisions that affected persons make regarding future actions are not the same decisions that they would make if they fully understood the implications of the rule. In other words, if a rule conforms to expectations, affected persons should make the same decisions that they would make if they fully understood the implications of the rule, even if the length and technical nature of the rule prevent affected persons from obtaining actual knowledge of the rule’s specific terms.

In order to illustrate, we can make the assumption that a person who runs a business treated as a partnership for tax purposes has some cursory knowledge of partnership tax even though he or she lacks expertise.[134] Assume that his or her cursory knowledge encompasses awareness of the fact that partnerships are not taxed at an entity level but instead allocate tax items to their partners. However, assume that the person has no knowledge of the highly technical rules that specify how tax items are shared among partners.

Under the facts assumed, it seems likely that the person would expect that tax items of a partnership would be shared among partners in a way that is consistent with how the partners share in economic gains and losses of the partnership,[135] and the person should have a fairly solid understanding of how economic gains and losses are shared since that is a business matter. The person’s expectation about the sharing of tax items would work well as a rule of thumb since the consistency between tax and economics is the overall objective of the rules regarding allocation of tax items. Thus, the person has a workable understanding of the results of the tax rules.

Moreover, the person understands the results of the rules fairly well even though the Treasury Regulations regarding partnership tax allocations are long and technical.[136] Because of the length and technical nature of the Regulations, the precise parameters of the rules remain unknown to the business person who leaves the computational exercise of applying the rules (i.e. coping with Ex Post Complexity) to his or her tax accountants. Nevertheless, the person has an operational understanding of the rules since the result of the rules conforms to the person’s expectations.

As Professor Zelenak argues, computers cannot solve everything that is wrong with computationally difficult rules because, while computers handle the arithmetic associated with computational difficulty (and thus mitigate Ex Post Complexity), computers do not solve the potential problem of a lack of general understanding of technical rules (and thus have no effect on whether or not affected persons will take the rules into account when planning their activities).[137] As an example of a computationally difficult rule, Professor Zelenak discusses the phase-out that reduces otherwise allowable itemized deductions by the lesser of: (1) 3% of the excess of adjusted gross income over $100,000; or (2) 80% of otherwise allowable itemized deductions.[138]

I would note that the reason computers cannot address the resulting lack of understanding (which can distort decision-making) is that: (1) the rule is formulated in a way that is specific and arbitrary; and (2) the rule is technical. Because it is formulated in a way that is specific and arbitrary, the rule cannot be consistent with the expectations of an uninformed person. In other words, there is no reason why an uninformed person would have any intuitive estimate of the exact dollar amount at which a given phase-out sets in or the exact rate at which deductions are phased out because these parameters of the rule are inherently arbitrary. Therefore, for a person to be assured of making the decisions he or she would make if he or she understood the rule, he or she would need to have knowledge of the exact specifications of the rule. This knowledge is difficult to obtain since it requires wading through technical rules.

D. Uncertainty as a Contributing Factor

Uncertainty contributes to Ex Ante Complexity and Ex Post Complexity, all else being equal. Uncertainty arises when a person who has a clear and correct understanding of all available legal authority still cannot confidently predict how the rules apply to a given set of facts.

With respect to Ex Ante Complexity, a person affected by uncertain rules has three choices: (1) leave his or her intended future actions unchanged but strive to change applicable law in a way that reduces uncertainty, (2) leave his or her intended future actions and the law unchanged and accept a lack of understanding of the outcome of legal rules (and therefore accept the risk of incurring costs that follow from a failure to design future activities with the relevant legal rules in mind), or (3) leave applicable law unchanged but modify intended future actions so that the consequences of the actions are no longer uncertain. Uncertainty contributes to Ex Ante Complexity because, under any of the three options just described, a person governed by uncertain law incurs costs.

Under the first option, a person incurs costs in order to change applicable law. In tax, such costs could include the expenses associated with obtaining a private letter ruling from the Service. Under the second option, a person accepts the risk of incurring costs that follow from a failure to design future activities as if the person had the relevant legal rules in mind. Under the third option, a person changes his or her actions so that the legal result of his or her actions becomes certain. If the costs of changing a person’s actions are low, this strategy involves easily avoiding an otherwise complex rule. In tax, for example, as discussed above,[139] a person could modify his or her actions by taking the inexpensive step of filing a protective entity classification election when the results that would follow from a failure to do so are uncertain. However, this third strategy can be costly if the costs of modifying future activities are high. In tax, for example, a taxpayer might face uncertainty regarding whether a given financial interest would be treated as debt or equity for tax purposes. If the taxpayer wants more certainty regarding the tax outcome, he or she may have to modify the economic terms of the instrument, which could be costly if the original terms were optimal for non-tax reasons.

With respect to Ex Post Complexity, a person affected by uncertain rules has two choices:[140] (1) work to change applicable law in a way that reduces uncertainty or (2) accept a lack of understanding of the outcome of legal rules (and therefore accept the costs that follow from a failure to report activities properly or insure against such costs). Uncertainty contributes to Ex Post Complexity because, under either of these options, a person governed by uncertain law incurs costs.

Under the first option, a person incurs costs in order to change applicable law. Under the second option, a person either accepts costs that follow from a failure to report activities properly or incurs insurance costs. In tax, for example, a person could either accept the risk that his or her reported tax consequences will be successfully challenged (which could lead to penalties) or insure against that risk by, for example, obtaining an expensive legal opinion that could be one factor tending to provide some measure of protection against penalties in some circumstances.

E. Factual Information Needed as a Contributing Factor

A legal rule can contribute to Ex Ante Complexity or Ex Post Complexity if determining the proper application of the rule requires persons to gather information. This is only true, however, if the persons would not otherwise have the information readily available.

VII. Reforms would Simplify Law in Some Respects and, in Other Respects, Make Law No More Complex

Requiring all partnerships to use the remedial method under § 704(c) and making basis adjustments mandatory regardless of whether a partnership has made an election under § 754 would reduce or at least not increase Ex Ante Complexity and Ex Post Complexity.[141] Lawmakers have resisted these reforms on the grounds that the reforms would impose burdensome complexity on unsophisticated partnerships.[142] Thus, current law generally allows all partnerships to elect between using the most accurate method (in other words, the remedial method in the context of § 704(c) and basis adjustments in the context of § 754) and using a less accurate method.

Current law is flawed for three reasons. First, requiring partnerships to use the most accurate method would reduce, or at least not increase, Ex Ante Complexity. Second, requiring partnerships to use the most accurate method would reduce, or at least not increase, Ex Post Complexity. Third, making the elections universally available ultimately bred more complexity.

A. Requiring Partnerships to Use the Most Accurate Method Would Reduce, or at Least Not Increase, Ex Ante Complexity

Judged under each of the factors that affect Ex Ante Complexity described above in Part VI, mandatory use of the most accurate method would reduce, or at least not increase, Ex Ante Complexity. In terms of the first factor that affects Ex Ante Complexity, the current elective regime does not serve the purpose of allowing taxpayers to easily avoid an otherwise complex rule. Regarding length, the proposed reforms would make law shorter. With respect to the technical nature of rules, for unsophisticated taxpayers who do not consult tax advisors prior to engaging in transactions, the effect on Ex Ante Complexity is mitigated by the fact that the proposed reforms would lead to tax results more consistent with expectations. For unsophisticated taxpayers who do consult tax advisors prior to engaging in transactions or for sophisticated taxpayers, the reforms do not increase the technical nature of rules. The proposed reforms also reduce Ex Ante Complexity by reducing uncertainty. Finally, the proposed reforms would not impose any additional information gathering requirements and, thus, would not increase Ex Ante Complexity in that respect.

1. Current law does not reduce Ex Ante Complexity by allowing taxpayers to easily avoid an otherwise complex rule

Ex Ante Complexity plays a role at the time when taxpayers are deciding whether or not to engage in a given transaction. In the context of §§ 704(c) and 754, for example, Ex Ante Complexity plays a role at the time that taxpayers are deciding whether or not to contribute property to a partnership, at the time that taxpayers are deciding whether or not a partnership should sell property, and at the time that a person is deciding whether or not to acquire an interest in a partnership. Ex Ante Complexity is a measure of: (1) the costs that result from steps taken by taxpayers to ensure that the decisions they reach with respect to these transactions are the same decisions that they would reach if they fully understood the implications of relevant partnership tax law, and (2) the costs that result if the decisions that taxpayers reach with respect to these transactions are not the same decisions that they would reach if they fully understood the implications of relevant partnership tax law.

The proposed reforms to § 704(c) could increase Ex Ante Complexity if either of the two types of costs listed above were lower under the current regime in which taxpayers are allowed to elect to use the traditional method (thereby avoiding the remedial method) than under a regime in which taxpayers were required to use the remedial method. However, this is not the case. Likewise, it is not the case that either of the two types of costs listed above would be lower under the current regime in which taxpayers are allowed to elect to avoid basis adjustments than under a regime in which taxpayers were required to make basis adjustments.

Some taxpayers will obtain actual knowledge (either directly or through an advisor) of the consequences that will result from a prospective transaction under applicable partnership tax law. These taxpayers incur the first type of cost related to Ex Ante Complexity described above (in other words, costs that result from steps taken by taxpayers to ensure that the decisions they reach are the same decisions that they would reach if they fully understood the implications of relevant partnership tax law). These taxpayers are likely to be sophisticated taxpayers since they seek tax advice prior to engaging in a transaction and, therefore, they are not the type of taxpayers for whom lawmakers designed supposedly simple rules. In any event, the proposed reforms would, if anything, reduce the costs incurred by these taxpayers. In particular, if the remedial method under § 704(c) is required (or if partnerships are required to make basis adjustments), taxpayers or their advisors only need to predict the tax consequences of a prospective transaction that result from use of the remedial method (or that result if basis adjustments are made). If multiple methods are allowed, taxpayers or their advisors need to predict the tax consequences that would follow from a prospective transaction under each method (or with and without basis adjustments).[143] Consequently, the elective nature of current law does not allow these taxpayers to reduce costs by avoiding the determination of results under the remedial method (or by avoiding the determination of results that follow when basis adjustments are made).

Other taxpayers will make a decision regarding future transactions without obtaining actual knowledge (either directly or through an advisor) of the consequences that will result from the transactions under applicable partnership tax law. Taxpayers that make decisions in this way are likely to be unsophisticated taxpayers and, therefore, precisely the type of taxpayers for whom lawmakers designed supposedly simple rules. Such taxpayers avoid the first type of cost related to Ex Ante Complexity (namely, costs that result from steps taken by taxpayers to ensure that the decisions they reach are the same decisions that they would reach if they fully understood the implications of relevant partnership tax law). However, such taxpayers potentially incur the second type of cost related to Ex Ante Complexity (namely, costs that result if the decisions that taxpayers reach are not the same decisions that they would reach if they fully understood the implications of relevant partnership tax law). A taxpayer who makes a decision without actual knowledge of partnership tax law will make the decision based on what he or she expects partnership tax law would provide. Therefore, if partnership tax law provides for results that are more consistent with his or her expectations, he or she is less likely to incur the second type of cost related to Ex Ante Complexity. As described below in Part VII.A.4, consistency with expectations is more likely to occur under the proposed reforms.

2. The proposed reforms would reduce Ex Ante Complexity by making applicable law shorter.

Shortening applicable law reduces Ex Ante Complexity by making the tax consequences of future transactions easier to predict (at least for taxpayers who make such predictions based on actual knowledge of partnership tax law obtained directly or through an advisor). The proposed reforms would reduce the length of applicable law. In particular, requiring all partnerships to use the remedial method and requiring all partnerships to make basis adjustments in all cases would reduce the length of applicable statutes and regulations. This is the case since the reforms would obviate the need for provisions that: (1) describe other § 704(c) methods, (2) set forth the process for making a § 754 election, and (3) describe special rules that apply in the case of substantial built-in losses. Thus, the proposed reforms would reduce Ex Ante Complexity by making applicable law shorter.

3. The proposed reforms would not make the applicable rules meaningfully more technical for purposes of Ex Ante Complexity.

Regarding the technical nature of rules, the remedial method may be slightly more computationally difficult than the traditional method. Making basis adjustments is, admittedly, more computationally difficult than not making basis adjustments. Nevertheless, the proposed reforms would not increase Ex Ante Complexity.

As described above in Part VII.A.1, even under current law, a taxpayer who obtains actual knowledge of partnership tax law (either directly or through an advisor) in order to fully evaluate a prospective transaction must cope with the computational difficulty of the remedial method or the computational difficulty of making basis adjustments as part of the process of obtaining a complete picture of future tax consequences. In other words, such a taxpayer would evaluate the tax consequences of the transaction under each of the available methods under § 704(c), including the remedial method. Likewise, such a taxpayer would evaluate the tax consequences of a prospective transaction by assuming two alternative possibilities – one possibility that involves making basis adjustments and one possibility that involves not making basis adjustments.  Because such a taxpayer already evaluates the tax consequences that would follow from a transaction under the remedial method and under the assumption that basis adjustments are made, making the remedial method and basis adjustments mandatory would not exacerbate Ex Ante Complexity for such a taxpayer.

Likewise, the proposed reforms would not exacerbate Ex Ante Complexity for a taxpayer who makes decisions without actual knowledge of applicable law. This is so because the computational difficulty of applicable law has no effect on such a taxpayer’s evaluation of future transactions.

4. The proposed reforms would lead to results that better conform to expectations

Whether or not rules conform to the expectations of affected persons is most relevant with respect to taxpayers who make decisions regarding future transactions without acquiring actual knowledge of applicable law (directly or through an adviser). When taxpayers have no actual knowledge of applicable law, their decisions about future transactions will be based on their uninformed expectations about what applicable law provides. If these expectations are consistent with actual law, the decisions that the taxpayers make based on their expectations will be the same as decisions they would make if they had a complete and actual understanding of applicable law. Therefore, if applicable law conforms to their expectations, these taxpayers will not incur the second type of cost related to Ex Ante Complexity (namely, costs that result if the decisions that taxpayers reach are not the same decisions that they would reach if they fully understood the implications of relevant law). In addition, taxpayers who make decisions with respect to future transactions without acquiring actual knowledge of applicable law do not incur the first type of cost related to Ex Ante Complexity (namely, costs that result from steps taken by taxpayers to ensure that the decisions they reach are the same decisions that they would reach if they fully understood the implications of relevant law). Consequently, Ex Ante Complexity will be low for these taxpayers as long as applicable law conforms to their expectations.

As discussed above,[144] it seems fair to conclude that most people who own partnership interests, if they gave the matter any thought, would expect that partners in a partnership would share tax items in a way that coincides with how they share economic gains and losses. Compared to the traditional method, the remedial method leads to results more consistent with this expectation since the remedial method provides for sharing tax items in a way that better coincides with how partners share economic gain and loss, as demonstrated by the examples in Part I. Similarly, basis adjustments lead to results more consistent with this expectation since basis adjustments result in each economic gain or loss generating only one matching tax gain or loss, as demonstrated by the examples in Part III. Thus, the proposed reforms would lead to tax results that better conform to people’s expectations and consequently would reduce Ex Ante Complexity for taxpayers who make decisions with respect to future transactions without acquiring actual knowledge of applicable law. These taxpayers are more likely to be unsophisticated taxpayers and, therefore, precisely the type of taxpayers for whom lawmakers designed supposedly simple rules.

5. The proposed reforms would reduce uncertainty.

The uncertainty inherent in applicable law can increase the costs associated with predicting the tax consequences of potential future transactions. This effect will be most pronounced for taxpayers whose predictions are based on actual knowledge (obtained either directly or through an adviser) of applicable law. Such taxpayers will tend to be sophisticated and, consequently, not the type of taxpayers for whom lawmakers designed supposedly simple rules. In any event, the proposed reforms would decrease uncertainty, and therefore simplify the process of predicting the tax consequences of potential future transactions.

Regarding § 704(c), in some cases, ambiguities can arise under the traditional method in situations in which they do not arise under the remedial method. In other words, there are cases in which a complete and correct understanding of the Treasury Regulations does not lead to a clear tax result under the traditional method but does lead to a clear tax result under the remedial method.[145] Regarding both §§ 704(c) and 754, the proposed reforms would mitigate uncertainty by reducing the number of instances in which taxpayers might engage in transactions that raise the question of whether or not an inherently uncertain anti-abuse rule applies.[146]

6. The proposed reforms would not impose any incremental information gathering requirements.

The factual information needed to predict the tax consequences of a potential transaction under the remedial method (or if basis adjustments are made) is either the same factual information needed to predict tax consequences of the potential transaction under the traditional method (or without basis adjustments) or is information that the parties would have in any event. Consequently, the proposed reforms would not increase Ex Ante Complexity by imposing any additional information gathering requirements.

B. Requiring partnerships to use the most accurate method would reduce, or at least not increase, Ex Post Complexity.

Judged under each of the factors that affect Ex Post Complexity described above in Part VI,[147] mandatory use of the remedial method and mandatory basis adjustments would reduce, or at least not increase, Ex Post Complexity. In terms of the first factor that affects Ex Post Complexity, the current elective regime does not serve the purpose of allowing taxpayers to easily avoid an otherwise complex rule. Regarding length, the proposed reforms would make law shorter. With respect to the technical nature of rules, the effect on Ex Post Complexity of any increase in computational difficulty would be mitigated by technological advances. The proposed reforms also reduce Ex Post Complexity by reducing uncertainty. Finally, the proposed reforms would not impose any additional information gathering requirements and, thus, would not increase Ex Post Complexity in that respect.

1. Current law does not allow taxpayers to easily avoid an otherwise complex rule.

Current law appears to be designed in an attempt to reduce Ex Post Complexity by reducing the frequency with which computationally difficult rules apply. In particular, the goal of current law is apparently to allow taxpayers to avoid the computational difficulty associated with the remedial method in favor of the supposedly lesser computational difficulty associated with the traditional method and to allow taxpayers to avoid the computational difficulty associated with basis adjustments.[148] However, current law fails to meet this goal for two reasons. First, as a practical matter, the computational difficulty associated with the remedial method is no greater than the computational difficulty associated with the traditional method, and the computational difficulty associated with making basis adjustments is no greater than the computational difficulty of failing to do so for the reasons discussed in Part VII.B.3 below. Second, the elective nature of the rules does not result in taxpayers avoiding the computational difficulty that would be undertaken if taxpayers were required to use the remedial method or make basis adjustments.

This second point requires further elaboration. As discussed above,[149] a rule that might otherwise be complex is not, in fact, complex when people can take straightforward steps to avoid application of the rule. In theory, this proposition would suggest that, if the remedial method truly led to more onerous computations than the traditional method, then giving taxpayers the option to elect the traditional method would reduce Ex Post Complexity. In practice, however, providing for an elective regime exacerbates Ex Post Complexity because not only do taxpayers (or more accurately, their advisors) not avoid the work that would be required to compute results under the remedial method, but they also take on the work that would be required to compute results under the traditional method and the work of choosing between the two. In other words, if one method is required, partnerships (or their advisors) only need to apply that method. If multiple methods are allowed and if partnerships desire to make an informed choice about which method to use, partnerships (or their advisors) need to understand how all of the methods operate in order to select the method under which the results will be reported.[150] Furthermore, under an elective regime, the Service must devote resources necessary to cope with taxpayers using different methods rather than one method.

The conclusion that the elective regime does not produce simplification can be more fully illustrated in the context of § 704(c) and § 754.

a. Section 704(c)

It is possible to imagine a scenario under which the elective regime would produce simplification, if the computational difficulty associated with the traditional method truly was lower than the computational difficulty associated with the remedial method. In particular, one might imagine that a partnership that values simplification over the possibility of achieving an optimal tax outcome would attain simplicity by automatically opting for the traditional method, if it were associated with lower computational difficulty, rather than analyzing the results under all possible methods. However, as a practical matter, this scenario is highly unlikely to occur.

First, properly reporting tax consequences that follow under the traditional method (like many other aspects of partnership tax law) requires sufficient technical expertise such that it is unlikely that an unsophisticated partnership would apply the traditional method without accounting or legal assistance. Second, once an unsophisticated partnership seeks accounting or legal advice, it is likely that the accountant or lawyer will analyze the impact of all available methods. For one thing, the expert may be aware of the fact that reporting under the remedial method is no more difficult than doing so under the traditional method, so the remedial method should not be discarded out of hand. For another, the expert may be motivated by other factors such as a concern that a failure to provide advice about all realistic, available options could lead to the risk of a malpractice claim.

b. Section 754

As is the case regarding § 704(c), an unsophisticated partnership may rely on an expert to analyze the question of whether or not to make a § 754 election.[151] In order to mitigate the risk of a malpractice claim, for example, the expert likely assesses both options.[152]

2.The proposed reforms would make the law shorter.

Regarding length, as discussed above,[153] requiring all partnerships to use the remedial method and requiring all partnerships to make basis adjustments in all cases would reduce the length of applicable statutes and regulations. Thus, the proposed reforms would reduce Ex Post Complexity by shortening applicable law that must be analyzed by persons in order to properly report tax consequences.

3. The proposed reforms would not make the applicable rules meaningfully more technical.

Regarding the technical nature of rules, the remedial method may be slightly more computationally difficult than the traditional method in that it can involve more extensive bookkeeping. Making basis adjustments is, admittedly, more computationally difficult than not making basis adjustments. However, with respect to Ex Post Complexity, these effects are mitigated because any additional computations can be automated. Likely, the fact that any computations can now be automated through the use of computer programs is a fact that the Senate had in mind when it stated that the elective nature of basis adjustments is “anachronistic.”[154] Thus, the proposed reforms would not increase the technical nature of rules in a way that meaningfully increases Ex Post Complexity.

4. The proposed reforms would reduce uncertainty.

As described above,[155] mandatory use of the remedial method and mandatory basis adjustments would decrease uncertainty. With respect to the remedial method, this is true because there are cases in which ambiguities can arise under the traditional method when they would not arise under the remedial method.[156] In addition, both mandatory use of the remedial method and mandatory basis adjustments would mitigate uncertainty by reducing the number of instances in which taxpayers undertake transactions that raise the question of whether or not an inherently uncertain anti-abuse rule applies.[157]

As discussed above, reduced uncertainty decreases Ex Ante Complexity because it is easier to decide whether or not to engage in a transaction when it is easier to predict the tax consequences of that transaction. Likewise, reduced uncertainty decreases Ex Post Complexity by making it easier to evaluate how to report a transaction once it has occurred.

5. The proposed reforms would not impose any incremental information gathering requirements.

Regarding §§ 704(c) and 754, partnerships should already have all the information they would need to report tax results under the proposed reforms. Therefore, the proposed reforms would not affect this factor contributing to Ex Post Complexity.

In the context of § 704(c), one of the reasons the A.L.I. expressed concern about adopting a method similar to the remedial method was that it would require partnerships to value assets when they are contributed to a partnership. However, contrary to the view the A.L.I. seemed to espouse,[158] a partnership would need to value an asset at the time of contribution to the partnership even if the partnership was using the traditional method for making § 704(c) allocations.[159] In addition, a partnership needs to value an asset at the time it is contributed to a partnership in order to effectuate the economic deal among the partners. If A and B form a partnership, where A contributes land to the partnership, B contributes cash to the partnership, and the partners intend to share equally in the economic returns of the partnership, the partners will need to value the land in order to determine how much cash B should contribute.[160] Likewise, in the case of reverse § 704(c) allocations, a partnership needs to value its assets at the time new partners join the partnership in order to determine how much the new partners must contribute for a given interest in the partnership.

6. Making the elections universally available ultimately bred more complexity.

As discussed above, current law does not promote simplicity for unsophisticated partnerships relative to what would result if the law required all partnerships to use the remedial method and make basis adjustments. However, even if current law did promote simplicity for unsophisticated partnerships, only unsophisticated partnerships should be allowed to elect the less accurate (supposedly less complex) method. Instead of limiting use of the less accurate method to unsophisticated partnerships, the ability to elect the less accurate method is universally available. In the context of § 704(c), until 1984, universal availability meant allowing all partnerships to avoid applying current § 704(c) principles if they so chose. After 1984, it meant allowing all partnerships to use the traditional method under § 704(c) rather than the more accurate remedial method. In the context of § 754, until 2004, universal availability meant granting all partnerships the option of avoiding making basis adjustments. After 2004, it meant giving all partnerships the flexibility to avoid making basis adjustments except in cases where a substantial built-in loss exists.

The universal availability of the elections may reflect oversimplification. In particular, perhaps lawmakers took the view that it would be simpler to make the elections universally available than to attempt to distinguish between sophisticated and unsophisticated partnerships. Moreover, lawmakers may have felt justified in doing so since partnerships were historically viewed as entities used by unsophisticated taxpayers.[161]

However, as Professor David Weisbach has argued, lawmakers cannot design tax law based on the most common fact pattern.[162] Professor Weisbach explains why doing so would be problematic: “Uncommon transactions that are taxed inappropriately become common as taxpayers discover how to take advantage of them.”[163] In other words, designing tax law with the typical case in mind results in an unstable system since what was the atypical case will become more typical as taxpayers adjust their transactions to take the rules into account.

In the context of §§ 704(c) and 754, partnerships might have originally been used by unsophisticated taxpayers who would use the elections, if at all, for the intended purpose of simplification. However, partnership tax rules designed with unsophisticated taxpayers in mind encouraged sophisticated taxpayers to form partnerships so that they could use the elections for the unintended purpose of tax liability reduction. Over time, the unintended purpose became the more typical use of the rules. Lawmakers took action to limit the ability to use the rules for the unintended purpose (by adopting anti-abuse rules), and anti-abuse rules increased complexity by increasing uncertainty and the length of applicable law. The end result is a system that may be more complex than what would have existed if lawmakers had limited the availability of the election to unsophisticated partnerships. Furthermore, because the election does not further simplicity for unsophisticated partnerships, current law is more complex that what would have resulted if lawmakers had required use of the remedial method and required basis adjustments in all cases.

VIII. Conclusion

Rules intended to simplify partnership tax law in order to cater to Mom and Pop partnerships have sometimes had the opposite effect – complicating partnership tax law and inadvertently catering to sophisticated partnerships by increasing opportunities for tax planning and abuse. In place of current law, reforms suggested by this paper would simplify partnership tax law, as shown by an examination of how tax rules can best promote simplicity. In addition, the reforms would reduce the potential for tax revenue loss and unfairness inherent under current rules.

Acknowledgments: My thanks go to Amitai Aviram, Ronald Blasi, John Colombo, Dhammika Dharmapala, Christine Hurt, David Hyman, Douglas Kahn, Richard Kaplan, Amy McCormick, Larry Ribstein, Arden Rowell, Glen Staszewski, Suja Thomas, and Verity Winship. I would also like to thank the organizers of the Illinois Academic Fellowship Program for their support.


 

[1] TIFD III-E, Inc. v. United States, 459 F.3d 220 (2d Cir. 2006). The mechanics of the transaction, in the context of a simplified version of the facts of the actual case, are described below. See infra note 29. The Service challenged the transaction arguing, among other things, that the parties to whom the taxable income was allocated were not really partners since they bore very limited economic risk. Most recently, the District Court, on remand, held in favor of the taxpayer. TIFD III-E, Inc. v. United States, 660 F. Supp. 2d 367 (D. Conn. 2009).

[2] In this paper, I use the term “partnership” to refer to an entity treated as a partnership for tax purposes. Even if an entity is treated, for example, as a limited liability company for purposes of business organization law, it could be treated as a partnership for tax purposes provided that it has two or more owners, it has not elected to be treated as a corporation for tax purposes, and it is not required to be treated as a corporation under the publicly traded partnership rules or any other provision. See Treas. Reg. § 301.7701-2 (as amended in 2006).

[3] Thus, for example, the Senate Finance Committee Report accompanying the adoption of Subchapter K observed that uncertainty in partnership tax was “particularly unfortunate” given the large number of partnerships and given that “the partnership form is much more commonly employed by small businesses and in farming operations than the corporate form.” S. Rep. No. 83-1622, at 89 (1954). See also Jeffrey L. Kwall, Taxing Private Enterprise in the New Millennium, 51 Tax Law. 229, 235–36 (1998) (discussing how the use of partnerships has changed substantially since 1954 when “the prototypical partnership was seen as a small, simple enterprise”).

[4] For example, in 2003, 93% of entities treated as partnerships for tax purposes each individually earned less than $1,000,000 of business receipts. See Table 2.–Number of Businesses, Business Receipts, Net Income, Deficit, and Other Selected Items, by Form of Business, Industry, and Business Receipt Size, Tax Year 2003, I.R.S., Statistics of Income Division, http://www.irs.gov/pub/irs-soi/03ot2busbr.xls (last visited Apr. 22, 2011).

[5] For example, in 2003, only 5.5% of business receipts earned by all tax partnerships was earned by tax partnerships that each, individually, earned less than $1,000,000. Id.

[6] The entities avoid treatment as corporations for tax purposes under I.R.C. § 7704(a) (2010) (treating certain publicly traded partnerships as corporations for tax purposes) by complying with I.R.C. § 7704(c) (2010) (setting forth an exception for certain publicly traded partnerships if 90% or more of their gross income consists of specified types of qualifying income).

[7] For further discussion of the changing use of partnerships, see Larry E. Ribstein, The Rise of the Uncorporation (2010). See also Kwall, supra note 3.

[8] Partnerships are not unique in this regard. However, the challenges faced in the partnership tax context may be more pronounced because opportunities for abuse are plentiful under a set of tax rules that provide for pass-through taxation (unlike the tax rules that apply to C corporations) and allow for flexible economic arrangements (unlike the tax rules that apply to S corporations).

[9] See, e.g., George K. Yin, The Future Taxation of Private Business Firms, 4 Fla. Tax Rev. 141 (1999). See also Curtis J. Berger, W(h)ither Partnership Taxation?, 47 Tax L. Rev. 105 (1991) (proposing restructuring the taxation of entities so that all large businesses (determined based on gross revenues) would be taxed like corporations and all small businesses would be subject to a pass-through taxation system); Mark P. Gergen, The End of the Revolution in Partnership Tax?, 56 SMU L. Rev. 343, 345 (2003) (“[Professor Yin] proposes the creation of a simplified pass-through regime – call it K-lite – along the lines of current subchapter S….Whatever its precise contours, once K-lite is available, we may demand more from entities that opt for the freedom of K-heavy.”). Along similar lines, recent reports have indicated that the current administration may propose reforms under which certain businesses that earn more than $50 million in gross receipts would no longer be eligible for treatment as partnerships for tax purposes and, instead, would be treated as corporations for tax purposes.  See, e.g., Kim Dixon, U.S. Mulls Making More Firms Pay Corporate, http://www.msnbc.msn.com/id/42856826/ns/business-us_business/ (last visited May 9, 2011).

[10] This approach avoids practical difficulties inherent in the bifurcation approach, such as selecting a proxy to accurately distinguish between sophisticated and unsophisticated partnerships. In addition, this approach may be more feasible in that it would involve less drastic reform. The fact that this approach would be less disruptive is particularly appealing given the large number of entities treated as partnerships for tax purposes.

[11] Other examples could include current law regarding the treatment of the transferor of a partnership interest.

[12] Treas. Reg. § 1.704-3(a)(6)(i) (as amended in 2010). Allocations in this context are sometimes referred to as “reverse § 704(c)” allocations.

[13] I.R.C. § 721 (2010).

[14] I.R.C. § 721(a) (2010).

[15] I.R.C. § 723 (2010).

[16] I.R.C. § 722 (2010).

[17] I.R.C. §§ 701, 702 (2010).

[18] Likewise, if the value of the land had declined in value over time prior to the contribution so that, while A acquired the land for $25,000, the land was worth $15,000 at the time of the contribution, A would not be allowed to recognize the $10,000 built-in loss that existed in the land at the time of the contribution. I.R.C. § 721 (2010). In order to preserve the built-in loss, the AB partnership would obtain a tax basis in the land equal to $25,000 (A’s basis in the land), and A would obtain a tax basis in A’s interest in the partnership equal to $25,000 (A’s basis in the land). I.R.C. §§ 722, 723 (2010). As a result, if A sold his or her interest in the partnership for $15,000 after the contribution, A would recognize $10,000 of tax loss, and if the AB partnership sold the land for $15,000 after the contribution, the AB partnership would recognize $10,000 of tax loss to be allocated to A. I.R.C. § 704(c)(1)(C) (2010).

[19]I.R.C. § 704(c)(1)(A) (2010).

[20] If, instead of land, a partner contributes depreciable property to a partnership, § 704(c) also governs the allocation of tax depreciation deductions with respect to the property with the same goal of preventing a shift of tax consequences among partners with respect to pre-contribution gain or loss. See infra note 29.

[21] I.R.C. § 704(c)(1)(A) (2010).

[22] Treas. Reg. § 1.704-3(a)(1) (2010).

[23] Id.

[24] Id.

[25] Treas. Reg. § 1.704-3(a)(2) (2010).

[26] Treas. Reg. § 1.704-3(a)(10) (2010).

[27] I.R.C. § 705(a)(1)(A) (2010).

[28] I.R.C. § 731(a) (2010).

[29] This can be illustrated with an example that is a simplified version of the facts of Castle Harbour. TIFD III-E, Inc. v. U.S. (Castle Harbour), 459 F.3d 220 (2d Cir. 2006). Assume T and TE form an entity that is treated as a partnership for tax purposes. T is subject to U.S. tax on income allocated to T from the partnership, but TE is not subject to U.S. tax on income allocated to TE from the partnership. T contributes airplanes to the partnership. The airplanes are depreciable. At the time of the contribution, the airplanes are worth $1,000, but the airplanes have a tax basis of $100. Furthermore, the airplanes have a remaining depreciation recovery period of one year. TE contributes $1 of cash to the partnership. The partnership leases the airplanes to a third party and earns $1,000 of rental income for one year. The partnership sells the airplanes for $0 and liquidates in year 10. Assume the partnership agrees that, for purposes of determining the partners’ book capital accounts that will measure the amount each partner is entitled to receive on liquidation of the partnership, all book income and loss will be allocated 99% to TE and 1% to T. In year 1, the book items that are allocated 99% to TE and 1% to T consist of: (1) book depreciation of $1,000 (measured as: $1,000 beginning book value of airplanes x ($100 tax depreciation/$100 beginning tax basis of airplanes) per Treas. Reg. § 1.704-1(b)(2)(iv)(g)(3)) and (2) $1,000 of rent received. Thus, on net $0 of book gain is allocated 99% ($0) to TE and 1% ($0) to T. Consequently, capital account balances of T and TE remain $1,000 for T (T’s initial capital account balance since T contributed property worth $1,000) and $1 for TE (TE’s initial capital account balance since TE contributed $1 of cash) at the end of year 1. As a result, even though 99% of book items are allocated to TE, when the partnership distributes the $1,001 of cash that it holds on liquidation, TE receives $1 and T receives $1,000. Regarding the allocation of tax items in year 1, if the partnership uses the traditional method for making § 704(c) allocations with respect to the airplanes (as the taxpayer did in Castle Harbour), the results will be as follows. First, because $1,000 of book income attributable to rent received by the partnership is allocated 99% to TE and 1% to T, $1,000 of taxable income recognized by the partnership as a result of rent received by the partnership will be allocated 99% ($990) to TE and 1% ($10) to T. Second, because $1,000 of book depreciation is allocated $990 to TE and $10 to T, tax depreciation from the airplanes would be allocated in the same manner if it was available. However, the only tax depreciation available is $100, which is allocated in its entirety to TE to get as close as possible to matching the allocation of book depreciation. Thus, in total, taxable income is allocated $890 to TE who is not subject to tax and $10 to T. If the partnership instead uses the remedial method for making § 704(c) allocations with respect to the airplanes, for one thing, book depreciation of the airplanes would be spread over a longer period of time per Treas. Reg. § 1.704-3(d)(2). Also, the partnership would invent notional tax items of depreciation to match book depreciation allocated to TE and equal, offsetting notional tax items of operating income from the airplanes to allocate to T. For the sake of simplicity, if we focus on the second modification made by the remedial method, then total tax items allocated in year 1 would be: $0 to TE (which consists of $990 of rental income, $100 of actual tax depreciation, and $890 of notional tax depreciation) and $900 to T (which consists of $10 of actual rental income and $890 of notional operating income from the airplanes). Consequently, T’s taxable income effectively equals $1000 of rental income minus $100 of remaining tax depreciation on the airplanes, and no taxable income is shifted from T (who is subject to tax) to TE (who is not subject to tax). If both modifications made by the remedial method are applied, then even under the remedial method, some taxable income could be shifted temporarily from T to TE, but the shift would be less drastic than what occurs under the traditional method.

[30] I.R.C. § 721(a) (2010).

[31] I.R.C. § 723 (2010).

[32] I.R.C. § 722 (2010).

[33] Id.

[34] Limiting the partnership to using tax items actually recognized from sale of the land is called the “ceiling rule.” Abiding by the “ceiling rule” is the distinguishing feature of the traditional method.

[35] This shift occurs because the $5,000 tax gain recognized by the partnership is effectively the net result of the $10,000 of gain that accrued prior to contribution of the land and the $5,000 of loss that accrued after contribution of the land. Netting the two figures results in $2,500 of the tax gain attributable to the pre-contribution increase in value of the land that economically benefited A offsetting the portion of the tax loss attributable to the post-contribution decline in value of the land that economically burdened B.

[36] I.R.C. § 705(a)(1)(A) (2010).

[37] I.R.C. § 731(a) (2010).

[38] For a more complete demonstration of the potential character and timing differences, see Tables 7 and 8, supra Part II.C.

[39] I.R.C. § 721(a) (2010).

[40] I.R.C. § 723 (2010).

[41] I.R.C. § 722 (2010).

[42] Id.

[43] This is because the amount each partner will receive on liquidation of the partnership is increased by $2,500 as a result of the fact that Land #2 is sold for a $5,000 economic gain.

[44] I.R.C. § 705(a)(1)(A) (2010).

[45] I.R.C. § 731(a) (2010).

[46] Allocating the tax gain from Land #2 differently than the economic gain (or, more precisely, “book gain,” which, in this case, is the same as economic gain) from Land #2 in order to offset the shift in tax gain from Land #1 is referred to as a “curative allocation.” See Treas. Reg. § 1.704-3(c)(1) (2010).

[47] I.R.C. § 705(a)(1)(A).

[48] I.R.C. § 731(a).

[49] Regarding the timing of making curative allocations, the Treasury Regulations provide that, “the period of time over which the curative allocations are made is a factor in determining whether the allocations are reasonable.” Treas. Reg. § 1.704-3(c)(3)(ii) (2010).

[50] Treas. Reg. § 1.704-3(c)(3)(iii) (2010).

[51] I.R.C. § 705(a)(1)(A).

[52] I.R.C. § 705(a)(2)(A) (2010).

[53] I.R.C. § 731(a).

[54] Treas. Reg. § 1.704-3(a)(1) (2010) (describing the purpose of § 704(c) (2010)).

[55] Treas. Reg. § 1.704-3(a)(10) (2010). Other constraints on the flexibility afforded by § 704(c) are discussed below. See infra notes 62-64 and accompanying text. Furthermore, the remedial method has been used in abusive transactions (including transactions undertaken by Enron), prompting amendments to the Regulations under § 704(c). Those amendments, finalized in June 2010 provide: “even though a partnership’s allocation method may be described in the literal language of [the Regulations under § 704(c)], based on the particular facts and circumstances, the Commissioner can recast the contribution as appropriate to avoid tax results inconsistent with the intent of subchapter K. One factor that may be considered by the Commissioner is the use of the remedial allocation method by related partners in which allocations of remedial items of income, gain, loss or deduction are made to one partner and the allocations of offsetting remedial items are made to a related partner.” T.D. 9485, 2010-26 I.R.B. 771. It should be noted that the abuse in these transactions did not involve shifting tax consequences in a way that was inconsistent with economic consequences, since the remedial method ensures consistency between the two. Rather, because the partners were related parties, the partners were willing to shift economic consequences in order to shift tax consequences and obtain a favorable tax result. Unrelated parties would not ordinarily be willing to engage in such a transaction. For discussion of the Enron transaction, see the discussion of “Project Condor” in Staff of Joint Comm. on Tax’n, 107th Cong., Report of the Investigation of Enron Corporation and Related Entities Regarding Federal Tax and Compensation Issues, and Policy Recommendations (Joint Comm. Print 2003).

[56] Treas. Reg. § 1.704-3(a)(1).

[57] This is also suggested by the language of the anti-abuse rule itself, since the Treasury Regulation states that it applies when the contribution of the property “and” the allocation of tax items are made with a view to shifting tax consequences. Treas. Reg. § 1.704-3(a)(10). In addition, the Treasury Regulations provide examples in which the use of a particular method is unreasonable because the contribution of the property in the first place is made with a view to shifting tax consequences. See Treas. Reg. § 1.704-3(b)(2), Example 2 (2010); Treas. Reg. § 1.704-3(c)(4), Example 3 (2010).

[58] For further discussion, see Joel Scharfstein, An Analysis of the Section 704(c) Regulations, 48 Tax Law. 71, 88 (1994) (“Although the regulations do not indicate what ‘with a view’ means, it must mean something less than ‘the principal purpose.’ Is a ‘principal purpose’ or a ‘significant purpose’ required, or will the proscribed view be considered to exist whenever there are tax advantages (or the possibility of tax advantages), and the partners knew about them?”).

[59] For further discussion, see Id. at 87; Laura Cunningham, Use and Abuse of Section 704(c), 3 Fla. Tax Rev. 93, 116–17 (1996).

[60] T.D. 8585, 1995-1 C.B. 120.

[61] Treas. Reg. § 1.704-3(d)(5)(ii) (2010).

[62] To illustrate the operation of § 704(c)(1)(B), assume, for example, A owns a piece of land that A acquired some time ago for $5,000. The land is currently worth $15,000. A contributes the land, B contributes $15,000 cash, and C contributes $15,000 cash to the newly formed ABC partnership, each in exchange for a one-third interest in the ABC partnership. A does not recognize any gain as a result of the contribution. I.R.C. § 721(a) (2010). The ABC partnership’s basis in the land is $5,000. I.R.C. § 723 (2010). A’s initial basis in his or her interest in the partnership is $5,000, B’s initial basis in his or her interest in the partnership is $15,000, and C’s initial basis in his or her interest in the partnership is $15,000. I.R.C. § 722 (2010). If the partnership sold the land for $15,000, the partnership would recognize $10,000 of tax gain which would be allocated, in its entirety, to A regardless of which method the partnership uses for making § 704(c) allocations. Instead of selling the land, assume the partnership distributes the land to B in liquidation of B’s interest in the partnership, and, in a subsequent year, distributes $15,000 cash to each of A and B in liquidation of the partnership. Absent a special rule (and assuming § 751(b) does not apply and assuming that the land is not inventory), no gain or loss would be recognized by the ABC partnership, A, B, or C as a result of the distribution. I.R.C. §§ 731(a), 731(b) (2010). B’s basis in the land would equal $15,000 (B’s basis in his or her interest in the partnership prior to liquidation). I.R.C. § 732(b) (2010). A recognizes $10,000 of gain upon receipt of $15,000 cash on liquidation of the partnership. Therefore, A eventually recognizes an amount of gain equal to the built-in gain that existed when the property was contributed to the partnership. However, the gain is not recognized until liquidation. If the land is distributed to B within 7 years of A’s contribution of the land to the partnership, however, § 704(c)(1)(B) requires that A recognize $10,000 of gain from a deemed sale of the land at the time at which the land is distributed to B.

[63] To illustrate the operation of § 737, assume, for example, A owns a piece of land (Land #1) that A acquired some time ago for $5,000. The land is currently worth $15,000. A contributes Land #1, B contributes $15,000 cash, and C contributes $15,000 cash to the newly formed ABC partnership, each in exchange for a one-third interest in the ABC partnership. A does not recognize any gain as a result of the contribution. I.R.C. § 721(a). The ABC partnership’s basis in Land #1 is $5,000. I.R.C. § 723. A’s initial basis in his or her interest in the partnership is $5,000, B’s initial basis in his or her interest in the partnership is $15,000, and C’s initial basis in his or her interest in the partnership is $15,000. I.R.C. § 722. The ABC partnership acquires an additional parcel of land (“Land #2”) for $15,000. If the partnership sold Land #1 (the land contributed by A) for $15,000, the partnership would recognize $10,000 of tax gain which would be allocated, in its entirety, to A regardless of which method the partnership uses for making § 704(c) allocations. Assume, rather than selling Land #1, at a time when the value of each parcel of land is still $15,000, the partnership distributes Land #2 to A in liquidation of A’s interest in the partnership. Absent a special rule (and assuming § 751(b) does not apply), no gain or loss would be recognized by the ABC partnership, A, B, or C as a result of the distribution. I.R.C. §§ 731(a), 731(b). A’s basis in Land #2 would equal $5,000 (A’s basis in his or her interest in the partnership prior to liquidation). I.R.C. § 732(b). Therefore, upon a sale of Land #2 for $15,000, A would recognize an amount of gain equal to the built-in gain that existed when Land #1 was contributed by A to the partnership. I.R.C. § 737 alters the results just described. If the distribution to A occurs within 7 years of A’s contribution of Land #1 to the partnership, § 737 requires that A recognize $10,000 of gain from a deemed sale of Land #1 at the time at which the distribution is made to A, and, as a result of the application of § 737, A would obtain a $15,000 basis Land #2.

[64] It is unclear how this restriction should be reconciled with the statement in the Treasury Regulations that the Service will not require a partnership to use the remedial method. See Treas. Reg. § 1.704-3(d)(5)(ii). It appears that, in the case of property contributed with a built-in loss, § 704(c)(1)(C) leads to the same result as the remedial method for all partners other than the partner who contributed the property by providing that, with respect to such partners, the property will be treated as if its tax basis equals its fair market value at the time of the contribution. See I.R.C. § 704(c)(1)(C)(ii) (2010). However, it may be the case that, with respect to the contributing partner, the results are not required to match the results that would follow from the remedial method. For example, assume two individuals, A and B, form a partnership. A contributes land with a basis of $200 and a fair market value of $150 to the partnership in exchange for a 50% interest in the partnership. B contributes $150 cash to the partnership in exchange for a 50% interest in the partnership. The partnership subsequently sells the land for $200. As a result of § 704(c)(1)(C), the partnership will be treated as if its tax basis in the land is $150 for purposes of determining tax gain or loss allocated to B. Consequently, B will be allocated $25 of tax gain as a result of the sale. The same result would follow absent § 704(c)(1)(C) if the partnership used the remedial method. In particular, while the partnership would not recognize any actual tax gain or loss (since its basis in the land is $200), the partnership would invent a notional item of $25 of tax gain to allocate to B to match the $25 of economic gain realized by B. Section 704(c)(1)(C) has no effect on the basis of the land for purposes of determining tax gain or loss allocated to A, the non-contributing partner. Therefore, with respect to A, it seems that, if the partnership uses the traditional method, A would be allocated no tax gain or loss (since the partnership recognizes no tax gain or loss). However, if the partnership uses the remedial method, it seems A would be allocated $25 of tax loss (which consists of $25 tax loss invented to offset the $25 tax gain that is invented to allocate to B).

[65] For example, in the case of inter vivos gifts, as a result of the basis rules in § 1015(a), the tax consequences of existing built-in gain can be shifted from the donor to the donee but the tax consequences of existing built-in loss cannot be shifted (for purposes of recognizing a loss). Along similar lines, when a shareholder contributes property to a corporation in a transaction that does not lead to recognition of gain or loss as a result of § 351, if the property has a built-in gain, the built-in gain will be preserved at two levels (at the shareholder level, by giving the shareholder a basis in stock received that is less than the fair market value of the stock and, at the corporate level, by giving the corporation a basis in the property contributed that is less than the fair market value of the property). I.R.C. §§ 358(a)(1), 362(a) (2010). However, if property contributed by a shareholder has, in aggregate, a built-in loss, the built-in loss will be preserved either at the shareholder level (by giving the shareholder a basis in stock received that is greater than the fair market value of the stock) or at the corporate level (by giving the corporation a basis in the property contributed that is greater than the fair market value of the property) but not at both levels. I.R.C. §§ 358(a)(1), 362(e)(2) (2010).

[66] Treas. Reg. § 1.704-3(a)(6)(i) (2010).

[67] Treas. Reg. § 1.704-3(a)(2) (2010).

[68] Treas. Reg. § 1.704-3(e)(3) (2010).

[69] Treas. Reg. § 1.704-3(e)(3)(iii) (2010).

[70] Treas. Reg. § 1.704-3(e)(3)(ii)(A), (B) (2010).

[71] I.R.C. § 704(c)(1) (1954).

[72] I.R.C. § 704(c)(2) (1954).

[73] If the partnership, in fact, planned to use this method, B would likely be willing to contribute less than $15,000 if B were taxable on gain from sale of the land. For simplicity, the example above assumes that B is not taxable on gain from sale of the land.

[74] S. Rep. No. 83-1622, at 90 (1954).

[75] Id. at 90 (discussing the fact that the sharing of tax items with respect to contributed property is identical to the sharing of tax items with respect to non-contributed property under the general rule).

[76] Id. at 89. For current statistics regarding size of partnerships, see supra notes 4, 5.

[77] Id. at 93.

[78] Id.

[79] For further discussion of the evolution of § 704(c), see, e.g., Gergen, supra note 9, at 348.

[80] For further discussion, see Cunningham, supra note 59, at 116 – 17.

[81] See American Law Inst., Federal Income Tax Project, Subchapter K, Proposals for Changes in the Rules for Taxation of Partners 153-54 (Tent. Draft No. 3, 1979) (“The concern with complexity [discussed in 1954] . . . has continuing relevance. Nevertheless, it is recognized that many of the partnerships that use Subchapter K today are more sophisticated than the partnerships which existed in 1954 and would therefore be better able to handle the complexities of a deferred sale approach.”). Id. at 156 (“[T]he valuation difficulties inherent in a deferred sale approach . . . together with the complexity which such an approach was viewed as adding to Subchapter K, result in the conclusion that the Project would not recommend such a rule.”).

[82] Id. at 146 (“The most important difficulty with the deferred sale approach is that it requires valuation of all property contributed to a partnership.”). For discussion of why this concern should not preclude adoption of the remedial method (or something similar such as the deferred sales method), see infra notes 158 to 160 and accompanying text.

[83] Id. at 155 (“[T]he ceiling limitation [effectively the traditional method] is to some extent inconsistent with the more correct theoretical result under the deferred sale approach [effectively the remedial method]. . . . Despite this illogical element, the ceiling limitation has one important advantage. It restricts any allocation to readily ascertainable factors – the property’s basis and its sales price. . . . For this reason, a credited value or deferred sale approach with a ceiling limitation [effectively the traditional method] may avoid some of the valuation difficulties that would result under a pure credited value approach [effectively the remedial method].”). For discussion of why this is incorrect, see infra note 159 and accompanying text.

[84] Id. at 151 (“A deferred sale or credited value approach will probably be perceived as more complex than present law by most taxpayers since it will require more computations.”). For further discussion of the impediments to adopting the deferred sales approach, see Andrea Monroe, Saving Subchapter K: Substance, Shattered Ceilings, and the Problem of Contributed Property, 74 Brook. L. Rev. 1381 passim (2009). For an alternative explanation of why taxpayers are allowed to elect among different methods under § 704(c), see Heather M. Field, Choosing Tax: Explicit Elections as an Element of Design in the Federal Income Tax System, 47 Harv. J. on Legis. 21, 60-61 (2010) (arguing that § 704(c) represents an area in which lawmakers allowed for an election because all of the methods are reasonable approaches for implementing Congress’s policy goal but also stating, “Nevertheless, Congress could have, and possibly should have, simply selected a single method for handling precontribution gains and losses.” Id. at 61.).

[85] Under a different set of facts, the remedial method could result in less tax revenue than the traditional method.  This would be the case if, for example, the facts were the same except that A was subject to a lower rate of tax than B.  However, in such a case, the parties would likely opt for the remedial method even under current law, so mandating use of the remedial method should not result in tax revenue loss.  Moreover, because the remedial method results in tax consequences that more accurately reflect economic consequences, the remedial method still results in a more accurate measure of tax liability even in cases in which it results in a lower amount of tax liability.

[86] I.R.C. § 743(a) (2010).

[87] I.R.C. § 743(b) (2010).

[88] See infra notes 103-105 and accompanying text.

[89] I.R.C. § 754 (2010).

[90] I.R.C. § 722 (2010).

[91] This gain will be capital gain assuming the partnership does not hold the land as inventory and holds no unrealized receivables. I.R.C. §§ 741, 751(a) (2010).

[92] I.R.C. § 742 (2010).

[93] I.R.C. § 743(a) (2010).

[94] I.R.C. § 705(a) (2010).

[95] I.R.C. § 731(a) (2010).

[96] Id. (flush language) (stating that the tax loss recognized on liquidation will be treated as tax loss recognized by C from a sale of the partnership interest); I.R.C. § 741 (2010) (stating that, except as provided in § 751, tax loss recognized from the sale of a partnership interest will be treated as capital loss).

[97] I.R.C. § 1211 (2010).

[98] I.R.C. § 1212 (2010).

[99] I.R.C. § 743(b) (2010).

[100]Id.

[101] I.R.C. § 731 (2010).

[102] A similar issue of potential duplication of built-in gain arises when property is distributed by a partnership to a partner. As is true in the case of the transfer of interests in a partnership, the built-in gain will or will not be duplicated depending on whether the partnership has a § 754 election in effect. For example, assume three individuals, A, B, and C each contribute $100 in exchange for a one-third interest in a newly formed entity treated as a partnership for tax purposes. As a result, each partner will have a basis in his or her interest in the partnership equal to $100. I.R.C. § 722 (2010). Assume the partnership uses $100 of the amount contributed by the partners to acquire a parcel of land for $100, so that the partnership’s basis in the land is $100. The partnership holds the remaining $200 contributed by the partners in cash. At a time when the value of the land is $400, the partnership distributes $200 cash to A in liquidation of A’s interest in the partnership. A recognizes $100 of gain as a result of the distribution (the excess of the amount of cash received over A’s basis in his or her interest in the partnership). I.R.C. § 731(a) (2010). This $100 tax gain is attributable to A’s one-third share of the $300 of economic gain that has accrued with respect to the land. In other words, because the land is worth $300 more than its initial acquisition cost, the amount distributed to A in liquidation of A’s interest must exceed the amount A contributed by one-third of $300 or $100, and because A receives $100 more in cash than the amount A contributed, A recognizes $100 of tax gain as a result of the distribution. If no § 754 election is in effect, the distribution has no impact on the partnership’s basis in the land. I.R.C. § 734(a) (2010). As a result, if the partnership sells the land for $400 in a subsequent year, the partnership will recognize $300 of tax gain, allocated $150 to each of B and C, and $50 of the amount allocated to each partner represents a duplication of the tax gain recognized by A when A received the cash distribution. If a § 754 election is in effect, then the partnership will increase its basis in the land by $100 to $200. I.R.C. § 734(b) (2010). As a result, if the partnership sells the land for $400 in a subsequent year, the partnership will recognize $200 of tax gain, allocated $100 to each of B and C, and none of the tax gain recognized by A is duplicated.

[103] I.R.C. § 743(b) (2010).

[104] I.R.C. § 743(d)(1) (2010). A similar rule applies in the case of downward basis adjustments in connection with the distribution of property. I.R.C. §§ 734(b), (d).

[105] Making basis adjustments becomes computationally more difficult if a partnership holds multiple assets and if interests in the partnership change hands frequently. This is the case since a separate basis adjustment has to be taken into account for each asset with respect to each transfer. In recognition of this fact, § 743(e) of the Code provides that the mandatory downward basis adjustments that would otherwise apply in the case of a “substantial built-in loss” do not apply to an “electing investment partnership,” provided that the partnership establishes that losses are only allocated to the transferee partner to the extent they exceed the loss recognized by the transferor. Several open questions remain regarding which entities can qualify as electing investment partnerships and regarding how partnerships can comply with the requirement of establishing that losses are only allocated to the transferee partner to the extent they exceed the loss recognized by the transferor. For further discussion, see Kristen J. Hangen, Economic Crisis Renews Interest in Electing Investment Partnerships, 126 Tax Notes 945 (2010). Regarding the reason for adopting a special rule for electing investment partnerships, the House Committee Report stated, “The Committee was made aware that certain types of investment partnerships would incur administrative difficulties in making partnership-level basis adjustments.” H.R. Rep. No. 108-548, pt. 1, at 283 (2004).

[106] See Treas. Reg. § 1.701-2(d), Example 8 (2010); c.f. Treas. Reg. § 1.701-2(d), Example 9 (2010).

[107] Treas. Reg. § 1.701-2(d), Example 9 (2010).

[108] See, e.g., H.R. Rep. No. 108-548, pt. 1, at 283 (2004) (“The Committee believes that the partnership rules currently allow for the inappropriate transfer of losses among partners. This has allowed partnerships to be created and used to aid tax-shelter transactions.”); S Rep. No. 108-192, at 151 (2003) (“The Committee believes that the present-law electivity of partnership basis adjustments upon transfers and distributions leads to anomalous tax results, causes inaccurate income measurement, and gives rise to opportunities for tax sheltering. In particular, the failure to make partnership basis adjustments permits partners to duplicate losses and to transfer losses among partners, creating an inappropriate incentive to use partnerships as tax shelter vehicles.”).

[109] For an alternative explanation for the § 754 election, see Field, supra note 84, at 35-36 (arguing that the purpose of allowing taxpayers to make the § 754 election is to reconcile the differences between tax consequences that result from a sale of an interest in a partnership and tax consequences that result from an economically similar sale by a partnership of an interest in its assets). Yet, as Professor Field observes, this purpose could be served equally well if the results following from a § 754 election were mandatory. Id. at 42. Professor Field also notes that mandatory adjustments may be undesirable because of complexity. Id. at 43.

[110] American Law Institute, Federal Income Tax Project 81-82 (Tentative Draft No. 4, 1980).

[111] Id.

[112] H.R. Rep. No. 108-548, pt. 1, at 283 (2004).

[113] S. Rep. No. 108-192, at 189-90 (2004).

[114] This is not intended as a broad claim that more accurate income measurement is always desirable or that taxable income should always match economic income as closely as possible. The claim is limited to the context of §§ 704(c) and 754 where more accurate income measurement reduces potential for manipulation by sophisticated taxpayers, which is desirable from the standpoint of tax revenue collection and fairness. For discussion of the impact of tax elections on fairness and tax revenue collection generally, see infra note 117.

[115] For a discussion of the potential impact of the anti-abuse rule on a fairly blatant case of abuse, see Karen C. Burke, Castle Harbour: Economic Substance and the Overall-Tax-Effect Test, 107 Tax Notes 1163 (2005). For discussion of a simplified version of the facts of the Castle Harbour case, see supra note 29.

[116] It is likely that taxpayers would rarely set up a transaction to shift tax consequences when the shift in tax consequences would only occur if an asset is sold for a sales price different than its current value. This may explain why the examples of abuse given in the Treasury Regulations are all examples where the expected shift in tax consequences results from the allocation of depreciation. For discussion of these examples, see supra note 57. See also Cunningham, supra note 59 at 117–18 (arguing that the examples in the Treasury Regulations suggest that the defining feature of cases involving abuse is designing a transaction to take advantage of significant differences between the remaining economic life of an asset and the remaining cost recovery period).

[117] Any time the tax law allows taxpayers to explicitly make an election that affects tax consequences, one concern is that the election will be used in a manner that only reduces tax revenue collected, particularly by well-advised, sophisticated taxpayers. See, e.g., Field, supra note 84, at 26, 31 (“The availability of tax planning opportunities is criticized as complex, costly, wasteful, revenue reducing, and inequitable, and these critiques may resonate particularly strongly in the context of explicit elections. . . . [A] well-advised rational taxpayer will almost always exercise the election in a way that minimizes its tax liability, at the expense of the fisc. . . . Additionally, an election, while technically available to all eligible taxpayers, may be functionally available only to the wealthiest, most sophisticated group of taxpayers, who can best navigate the complexity of the election process.”); George K. Yin, The Taxation of Private Business Enterprises: Some Policy Questions Stimulated by the “Check-the-Box” Regulations, 51 SMU L. Rev. 125, 130 (1997) (“If the taxpayer is well-advised, the election, which has ramifications for tax purposes only, will always be to the detriment of the fisc.”).

[118] This concern arises in connection with tax elections generally. See supra note 117 and accompanying text. In addition, this concern was noted by the A.L.I. When discussing the § 754 election, the A.L.I. stated, “The present elective system is often a trap for the unwary.” American Law Institute, Federal Income Tax Project 80 (Tentative Draft No. 4, 1980).

[119] See supra Parts II, IV.

[120] In other words, if people do not understand a rule, the rule is not likely to affect their behavior in the intended manner. See, e.g., Lawrence Zelenak, Complex Tax Legislation in the TurboTax Era, 1 Colum. J. Tax L. 91, 103 (2010) (making the argument in the tax context); Susan Dynarski & Judith Scott-Clayton, The Cost of Complexity in Federal Student Aid: Lessons from Optimal Tax Theory and Behavioral Economics 2, (Harv. Kennedy Sch. Fac. Research Working Paper Series, Paper No. RWP06-013, 2006) (making the argument in the context of rules governing federal student aid); Austan Goolsbee, The TurboTax Revolution: Can Technology Solve Tax Complexity?, in The Crisis In Tax Administration 124, 138 (Henry J. Aaron & Joel Slemrod eds., 2004) (making the argument in the tax context). In the tax area, scholars have argued that this effect may, at times, be a positive consequence of opaque rules. Sometimes tax rules are intended to influence behavior, in which case it is crucial for the rules to be well understood. Sometimes tax rules are intended to generate revenue with as little distortion of behavior as possible, in which case opaque rules might reduce distortions. See, e.g., Brian Galle, Hidden Taxes, 87 Wash. U. L. Rev. 59 passim (2009); Goolsbee, supra at 138; Deborah Schenk, Exploiting the Salience Bias in Designing Taxes 22-23 (N.Y.U. Law & Econ., Working Paper No. 233, 2010).

[121] Ex Post Complexity might represent a special concern for a particular type of partnership, namely a small partnership, since, given the relatively fixed nature of costs related to Ex Post Complexity, such costs could impose a larger relative burden on small partnerships. However, small partnerships are not necessarily unsophisticated partnerships.

[122] In particular, while many people have remarked on the admitted complexity of partnership tax, some of the most technically difficult rules can be avoided by simple partnerships. See, e.g., Philip F. Postlewaite, I Come to Bury Subchapter K, Not to Praise It, 54 Tax Law. 451, 474 (2001) (stating that an A.L.I. report regarding taxation of private business that criticizes the complexity of partnership tax law “ignores the simplicity of Subchapter K in simple business arrangements” and citing to the fact that very simple partnerships can avoid analysis of technical rules specifying when allocations have “economic effect” if they comply with the “economic effect equivalence” test). Sections 704(c) and 754 are additional examples of partnership tax rules that tend to become more computationally complex when applied to complicated partnerships. Section 704(c) (or, more precisely, reverse § 704(c)) is most difficult to apply to a partnership, such as a hedge fund, that holds a large number of assets and that admits different partners over time who buy into the partnership based on the value of its assets at the time they join. Relief for this complexity is granted by allowing securities partnerships to apply reverse § 704(c) on an aggregate basis. See supra notes 67 – 70 and accompanying text. However, as a demonstration of their ability to cope with complexity, many securities partnerships use an asset-by-asset method even though they would be allowed to use the theoretically simpler aggregate method. See Scharfstein, supra note 58, at 94 (“[A] substantial number of securities partnerships now use asset-by-asset tracking. This is made feasible by the availability of computer programs designed for this purpose and the declining cost of computer usage. In particular, asset-by-asset tracking is rapidly becoming the norm among major securities partnerships with high minimum investments (e.g., $1 million or more).”). Likewise, basis adjustments are the most difficult to apply to a partnership that holds a large number of assets and in which interests are transferred frequently. The A.L.I. and Congress have recognized this fact. See supra notes 105 and 111 and accompanying text.

[123] A number of factors contribute to complexity or mitigate complexity. Factors of particular importance for the analysis in Part VII are discussed below. It is possible to label factors that influence complexity differently or categorize concepts relevant to complexity under the heading of different factors. See, e.g., Peter H. Schuck, Legal Complexity: Some Causes, Consequences, and Cures, 42 Duke L. J. 1 passim (1992). The factors listed in this paper are used as a helpful framework to demonstrate each of the ways in which the proposed reforms might affect complexity. Using a different framework should not alter the ultimate conclusion that the proposed reforms would promote simplicity, in some respects, and at least not further complexity, in other respects. For example, Professor Schuck lists four factors that contribute to legal complexity: (1) density (which exists when understanding the rules that apply in a given context requires the expertise of a wide variety of legal specialties), (2) technicality (which exists when understanding and applying rules requires special sophistication and expertise), (3) differentiation (which exists when applicable rules are provided by different sources that have different bases for legitimacy – such as statutes and common law), and (4) indeterminacy or uncertainty. Id. at 3-4. If the proposed reforms were analyzed under this framework, they would have no meaningful impact on factors (1) through (3), and they would reduce uncertainty.

[124] See, e.g., Richard A. Epstein, Simple Rules for a Complex World 25-27 (1995). The example Professor Epstein provides is the rule against perpetuities. Thoughtfully examining whether the rule would apply to any given fact pattern can be time consuming, and a person taking such an approach risks reaching the incorrect result at the end of his or her analysis. However, a standard savings clause can be used as an expedient to ensure that the rule will not apply even if it would have applied absent inclusion of the savings clause. Id. As Professor Epstein observes, the same principles apply any time parties can easily contract around a default rule. Id. In particular, if contracting around a rule is easy, the rule is not complex as a practical matter even if it would take significant time and effort to understand the rule.

[125] Treas. Reg. § 301.7701-3 (2010).

[126] Id.

[127] Id.

[128] Id.

[129] This is not to say that the entity classification rules allow taxpayers to avoid the work of understanding the pass-through tax rules (or the corporate tax rules) because taxpayers can elect to treat an entity as a corporation (or as a pass-through entity). In order to evaluate which classification is desirable, taxpayers need to understand the tax consequences of each classification. However, once a taxpayer decides on the desired classification, the taxpayer can make a protective election in order to avoid the work of determining whether the entity would be treated as a corporation or a pass-through entity in the absence of making an election.

[130] A similar example involves filing a protective election to apply the qualified electing fund (“QEF”) rules to a non-U.S. corporation if it turns out to be classified as a passive foreign investment company (“PFIC”). Such elections are frequently filed when it is difficult to determine whether or not the corporation could become a PFIC in the future. See, e.g., William M. Funk, On and Over the Horizon: Emerging Issues in U.S. Taxation of Investments, 10 Hous. Bus. & Tax L. J. 1, 22 (2010).

[131] Schuck, supra note 123, at 3.

[132] See, e.g., Zelenak, supra note 120.

[133] Id. at 103 (arguing that computationally difficult rules turn the tax system into a black box, which is problematic for a number of reasons including the fact that it interferes with the “ability of taxpayers to engage in well-informed basic tax planning, and to respond appropriately to the many incentives Congress has embedded in the tax laws.”).

[134] This assumption is likely consistent with reality in many cases.

[135] Such an expectation would be consistent with a general understanding that, subject to numerous exceptions, taxable income is intended to measure economic well being.

[136] The applicable Regulations under § 704(b) are approximately 70 pages long, and they use specialized terms of art like “capital account” and “substantiality.”

[137] Zelenak, supra note 120.

[138] Id. at 106 (citing § 68).

[139] See supra notes 125-130 and accompanying text.

[140] Since the relevant actions have already been undertaken, modifying future activities is no longer an option.

[141] Laura E. Cunningham & Noel B. Cunningham, Simplifying Subchapter K: the Deferred Sale Method, 51 SMU L. Rev. 1 (1997) (arguing that the deferred sale method, similar to the remedial method, would be preferable to current law and could significantly simplify current law). See also Monroe, supra note 84 (making a similar argument). For additional criticism of the current approach under § 754, see Karen C. Burke, Repairing Inside Basis Adjustments, 58 Tax Law. 639 (2004). This discussion of complexity is not intended to suggest that tax simplification is a goal that should be pursued without regard to other policy considerations, and, often tax simplification cannot be achieved without undermining other policy goals like fairness and prevention of tax abuse. However, it is also not the case that tax simplification is always at odds with these other goals. In the context of §§ 704(c) and 754, for example, the proposed reforms could simultaneously promote simplicity (or at least not make the law more complex) and further the goals of fairness and prevention of tax abuse. For discussion of the impact of tax elections on these other goals, see supra note 117 and for discussion of the impact of the proposed reforms on these other goals, see supra Part V.

[142] See supra Parts II and IV.

[143] This concern is true of many explicit elections that affect tax consequences. See, e.g., Field, supra note 84, at 27–30 (describing how elections create complexity because taxpayers must analyze the benefits and burdens of various alternatives); Yin, supra note 117, at 130 (“[E]lections are inherently costly and complex for the taxpayer. The taxpayer must incur the transaction cost of evaluating all tax consequences of available options before making an informed choice.”); Edward Yorio, The Revocability of Federal Tax Elections, 44 Fordham L. Rev. 463, 463-65 (1975) (describing how elections generally create complexity). See also Lawrence Lokken, As the World of Partnership Taxation Turns, 56 SMU L. Rev. 365, 371 (arguing that requiring partnerships to use the remedial method would “simplify the law by eliminating both any need for taxpayers to evaluate competing options and any need for any anti-abuse rule to disallow overly aggressive uses of the alternative rules.”). See also supra note 129.

[144] See supra notes 134-135 and accompanying text.

[145] For example, assume two individuals, A and B, form a new entity treated as a partnership for tax purposes. A contributes land with a tax basis of $750 and a fair market value of $1000, and B contributes $1000 cash. In a subsequent year, when the value of the land has increased to $1200, C contributes $2200 cash for a 50% interest in the partnership (diluting A’s interest to 25% and B’s interest to 25%). In a later year, the land is sold for $1100. If the partnership uses the remedial method for making § 704(c) allocations and reverse § 704(c) allocations with respect to the land, then the amount of tax gain and loss that must be allocated to each partner is clear. Under the remedial method, each partner is allocated tax gain and loss in an amount that matches net economic gain and loss. Net economic gain realized by A is $325 which is made up of (i) $250 gain realized prior to contributing the land to the partnership (ii) 50% of the $200 gain that accrued subsequent to contributing the land to the partnership but prior to C joining the partnership and (iii) 25% of the $100 loss that accrued after C joined the partnership. Net economic gain realized by B is $75 which is made up of (i) 50% of the $200 gain that accrued after the partnership was formed but prior to C joining the partnership and (ii) 25% of the $100 loss that accrued after C joined the partnership. Net economic loss realized by C is $50 which consists of 50% of the $100 loss that accrued after C joined the partnership. Consequently, tax items allocated to each partner under the remedial method are: $325 tax gain to A, $75 tax gain to B, and $50 tax loss to C. The only item actually recognized by the partnership is $350 of tax gain from sale of the land, but the partnership invents a notional item of $50 of tax loss and an offsetting notional item of $50 of tax gain, so that the partnership has, in total, the proper amounts of tax items to allocate ($400 of tax gain and $50 of tax loss). The results if the partnership uses the traditional method for making § 704(c) and reverse § 704(c) allocations with respect to the land are not entirely clear. The only item that the partnership can allocate is $350 of tax gain. None of the gain should be allocated to C since C realized a net economic loss. However, it is not clear whether the tax gain should be allocated: (i) $75 to B (to fully match B’s net economic gain) and $275 to A ($50 less than net economic gain realized), (ii) $25 to B ($50 less than net economic gain realized) and $325 to A (to fully match A’s net economic gain) or (iii) in some combination that falls in between options (i) and (ii).

[146] Anti-abuse rules are inherently uncertain because they are, by their nature, open-ended. Also, there are a number of unanswered questions regarding the anti-abuse rule in the § 704(c) Regulations. See supra notes 58-61 and accompanying text.

[147] Conformity to expectations is not discussed in connection with Ex Post Complexity because Ex Post Complexity involves evaluating and reporting tax consequences of transactions that have already occurred. Accurately reporting tax consequences requires actual knowledge of applicable tax law. Therefore, expectations of uninformed taxpayers are not particularly relevant, except to the extent that applicable tax law is easier to interpret when it conforms to expectations.

[148] See supra Parts II, IV.

[149] See supra notes 124-130 and accompanying text.

[150] This concern is true of many explicit elections that affect tax consequences. See supra note 143.

[151] The question of whether to make a § 754 election to some extent involves Ex Ante Complexity. Once the election is made, it will generally apply to all transfers of partnership interests that occur during the taxable year with respect to which the election was filed and any future taxable years. I.R.C. § 754 (2010). Therefore, making the election affects tax consequences of transactions that will happen in future years. However, despite the fact that the decision to make the election is, to some extent, forward-looking, even unsophisticated taxpayers may benefit from expert advice in connection with making the decision. This is the case because, at the time that the expert is consulted to evaluate how to report the tax consequences of the sale of a partnership interest, the expert may alert the unsophisticated partnership to the need to make a decision about whether or not to make a § 754 election.

[152] Moreover, if the expert does take a short cut in the analysis, it is likely that the expert would opt for automatically making the § 754 election (even though that is more computationally difficult). Under current law, if the election is made then: (i) tax gains will not be duplicated and (ii) tax losses will not be duplicated. See supra Part III. If the election is not made then (i) tax gains will be duplicated and (ii) tax losses will not be duplicated (unless they are not captured by the substantial built-in loss rules). See supra Part III. Therefore, if a shortcut is used in evaluating whether or not to make the election, the result will likely be making the election since (i) it avoids duplication of gains (which is better than what happens if the election is not made) and (ii) losses might not be duplicated in any event.

[153] See supra Part VII.A.2.

[154] See supra note 113 and accompanying text.

[155] See supra Part VII.A.5.

[156] See supra note 145 and accompanying text.

[157] See supra note 146 and accompanying text.

[158] See supra note 83 and accompanying text.

[159] The A.L.I. indicated that, under a method like the traditional method, the partnership only needs to know the sales price and the basis of an asset to properly allocate tax items. Id. This is simply untrue. To illustrate, assume A owns a piece of land that A acquired some time ago for $5,000. A contributes the land and B contributes cash to the newly formed AB partnership, each in exchange for a 50% interest in the AB partnership. Assume the partnership sells the land for $10,000 in year 1. The partnership would recognize $5,000 of tax gain from sale of the land. Assume the partnership uses the traditional method under § 704(c) for purposes of making tax allocations with respect to the land. If the land was worth $7,000 when A contributed it to the partnership, then the $5,000 tax gain should be allocated $1,500 to B (B’s 50% share of the $3,000 increase in value of the land that occurred after it was contributed to the partnership) and $3,500 to A (A’s 50% share of the $3,000 increase in value of the land that occurred after it was contributed to the partnership plus the $2,000 increase in value that occurred before it was contributed to the partnership). If the land was worth $12,000 when A contributed it to the partnership, then the proper tax allocations are quite different. In particular, all $5,000 of the tax gain should be allocated to A. In summary, simply knowing the sales price and the basis of the land is not enough to determine the proper tax allocations under the traditional method because the proper tax allocations also depend on the value of the property at the time it was contributed to the partnership.

[160] Professor Monroe observes that, following the promulgation of the substantial economic effect regulations (which occurred after the A.L.I. issued its report), valuing property at the time of contribution was necessary in order for a partnership to maintain capital accounts in accordance with the substantial economic effect regulations. Monroe, supra note 84, at 1433–34. Yet, even prior to the issuance of these regulations, partnerships would need to value contributed property in order to properly carry out their economic deal. The A.L.I. report posited a hypothetical in which multiple partners each contribute property to a partnership that is not easy to value but the partners agree to treat the value of each parcel of property as equal. See American Law Institute Federal Income Tax Project, Subchapter K Proposals for Changes in the Rules for Taxation of Partners pt. F(D) (Tent. Draft No. 3, 1979). However, it is not clear how the partners would agree that the value of the properties were equal without having some reliable estimate of the value of each piece of property.

[161] See supra note 3 and accompanying text.

[162] David A. Weisbach, Formalism in the Tax Law, 66 U. Chi. L. Rev. 860 (1999).

[163] Id .at 869.