by Edward A. Zelinsky

Edward A. Zelinsky is professor of law at the Benjamin N. Cardozo School of Law of Yeshiva University.

In this article, Professor Zelinsky criticizes the Ninth Circuit's most recent decision in the Albertson's litigation. That decision reversed the appeals court's initial judgment and denied the taxpayer a 1983 deduction for interest accrued, but not yet paid, under nonqualified deferred compensation plans for Albertson's executives.

In reaching this pro-government result, Professor Zelinsky contends, the Ninth Circuit improperly overrode highly specific statutory terminology that grants Albertson's the 1983 interest deduction it claimed. While the appeals panel purported to implement legislative purpose, the court, Professor Zelinsky maintains, was merely substituting its own policy preferences for those embodied in the code.

If exacting code terminology is to be overridden, Professor Zelinsky concludes, that task belongs to Congress, not to the courts.

Professor Zelinsky acknowledges the help of colleagues who reviewed earlier drafts of this article: Professors Laura Cunningham, Noel Cunningham, James B. Lewis, Paul Shupack, Stewart Sterk, and Lawrence Zelenak. In addition, he thanks Stephen C. Wallach of the Cardozo Law School Class of 1995 for research assistance and Jacoba Zelinsky of the Johns Hopkins University Class of 1998 for several (but not all) of her editorial suggestions.

Professor Zelinsky has no connection -- personal, professional, financial, or otherwise -- with Albertson's, Inc. or with any other taxpayer affected by the outcome of the Albertson's litigation.

* * * * *

Table of Contents

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . .1691

Background. . . . . . . . . . . . . . . . . . . . . . . . . . . .1692

Analyzing Albertson's II. . . . . . . . . . . . . . . . . . . . .1693

Albertson's Paths Not Taken. . . . . . . . . . . . . . . . . . . 1697

Who Should Override Statutory Terminology?. . . . . . . . . . . .1699

Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . .1701

Introduction

In a much-anticipated decision, the Ninth Circuit Court of Appeals has reversed its earlier ruling in Albertson's Inc. v. Commissioner ¹ (Albertson's I) and has held that Albertson's, an accrual-basis taxpayer, may not deduct currently accrued but unpaid interest it owed for 1983 under nonqualified deferred compensation agreements with its executives. While Albertson's II ² is important as both a compensation and an accounting decision, it is most significant for the broader issue of statutory interpretation it raises: In ruling for the government, the Albertson's II court overrode the explicit terminology of the Internal Revenue Code in the name of perceived, but unexpressed, legislative purpose. While the Ninth Circuit openly acknowledged its reservations about this approach, in the end, it suppressed both its misgivings and the language of the code.

Albertson's I was much too casual about the highly specific language of the code; Albertson's II is even more nonchalant about the exacting terminology of the statute, terminology that, the appeals court acknowledged, for the year at issue literally grants Albertson's the interest deduction it sought.

The Ninth Circuit apparently felt compelled to avoid the harmful consequences predicted by the government and many commentators ³ on the case: if Albertson's I and its pro-taxpayer ruling were not reversed, the court was told, the federal fisc would hemorrhage and the code's elaborate scheme for qualified plans would be disrupted. In fact, the appeals court could have enforced the relevant language of the statute without opening a fiscal Pandora's box, as the exacting code terminology granting Albertson's its 1983 deduction for accrued but unpaid interest was changed by Congress in 1984 ⁴ to deny that deduction for future years.

The fundamental issue raised by Albertson's II is thus the propriety of the judiciary ignoring exacting terminology in the tax statute under the banner of unarticulated legislative purpose. In Albertson's II, that purpose, as understood by the court, was so general and imprecise that it ultimately was nothing more than the appeals court's own preference from among competing tax policy criteria. That preference is, as a matter of policy, perfectly plausible and, indeed, is the preference that Congress embraced prospectively when it amended the relevant statute in 1984. However, the court's policy predilections are not those embodied in the code for the year at issue in Albertson's II.

Hence, in the final analysis, the appeals panel merely substituted its judgment for Congress's, labelled that substitution the implementation of legislative purpose, and thereby overrode quite precise language that awarded the taxpayer the deduction it sought.

If highly specific terminology in the tax code is to be overturned, that task belongs to Congress, not to the courts.

With commendable candor, the Albertson's II panel openly expressed its qualms about the nonliteral course it felt impelled to pursue. However, rather than revealing its qualms, the court should have acted on them.

Background ⁵

Albertson's, Inc. is an accrual-basis taxpayer that entered into nonqualified deferred compensation agreements with its executives. Those agreements provided for the accrual of interest annually during the executives' working careers and the ultimate payment at retirement of these interest amounts along with the underlying deferred compensation on which the interest is earned.

In 1983, Albertson's deducted the accrued but unpaid interest that arose in that year. For 1983, section 404 provided as to nonqualified deferred compensation that "such...compensation shall not be deductible under section 162 (relating to trade or business expenses)" but shall instead be deductible on a delayed basis when subsequently "includible in the gross income of employees" receiving it. ⁶ In 1984, Congress amended these provisions prospectively to state that nonqualified deferred compensation "shall not be deductible under this chapter" but shall instead be deductible later when "includible in the gross income of employees." ⁷

Reading these provisions in a straightforward fashion, an understandable and workable rule emerges: for 1983 and earlier years, employers maintaining deferred compensation programs cannot deduct currently amounts otherwise covered by section 162, i.e., the underlying compensation itself. There is, however, for these years no barrier to current deduction of amounts described in other code sections, e.g., interest governed by section 163. On the other hand, going forward from 1984, all income tax deductions arising from nonqualified deferred compensation plans are subject to the rule of delayed deductibility, deductible only upon receipt by (and taxability to) the employee. ⁸

Hence, as a statutory matter, Albertson's should have won the battle but lost the war, deducting currently its accrued 1983 interest but, by virtue of the 1984 statutory change, delaying any further interest deductions until employees actually receive those amounts on retirement.

Had Albertson's I applied the 1983 ⁹ version of the statute in this fashion, it would have been a decision of historical significance only: Albertson's would have had its 1983 interest deduction but the Treasury would have had the newer, pro-fisc version of the statute for all subsequent years. However, the Albertson's I panel was not content with this modest resolution of the matter but quite unnecessarily indicated that the 1984 variant of the statute, the version in the code today, also permits current deductions for accrued interest stemming from nonqualified deferred compensation arrangements. It was this dictum that transformed Albertson's I from historical curiosity to live controversy.

The 1984 version of section 404, the appeals court stated, subjects only nonqualified deferred "compensation" to the rule of delayed deductibility. That provision makes no reference to "interest" as such and, hence, does not apply to it.

That reading, however, nullifies to an important degree Congress's alteration of the statute. It would have been better for Congress to have amended section 404 more precisely; on the other hand, the import of the 1984 changes is clear, expanding the rule of delayed deductibility from those amounts otherwise governed by section 162 to all deductions "under this chapter," i.e., under the entire income tax statute including the interest provisions of the statute. ¹⁰

Because of its unnecessary comments about current law, Albertson's I created the proverbial firestorm: the IRS, asserting a potential loss to the Treasury of $7 billion, requested a rehearing; numerous commentators lambasted the Ninth Circuit's decision /11; legislation was introduced to overturn Albertson's I. ¹² In the face of this legal and political maelstrom, the appeals court agreed to reopen the case ¹³ and in Albertson's II reversed its earlier ruling, holding that Albertson's could not deduct in 1983 accrued but unpaid interest stemming from its nonqualified deferred compensation arrangements.

Albertson's II articulates three basic themes: that the literal terminology of the statute grants Albertson's its 1983 interest deduction, that the legislative purpose animating section 404 nevertheless compels denial of that deduction, and that the appeals court was not sure it was doing the right thing.

Early in its opinion, the Ninth Circuit panel candidly noted that it was "not entirely certain that the result we reached [in Albertson's I] is wrong." Underlying this forthright admission was the court's acknowledgement that the "plain language" of section 404 supported Albertson's' position: for 1983, section 404 does not delay interest deductions allowable under section 163 but only the deductions under section 162 for the basic deferred compensation itself. And the Albertson's II panel explicitly noted that the controlling version of section 404 was the more specific 1983 version. ¹⁴

Nevertheless, the panel held, Albertson's must lose because section 404 embodies a "matching principle" that employer deductions in the context of nonqualified deferred compensation be delayed until subsequent taxability of deferred amounts to the employee. To buttress its conclusion, the court relied heavily on the distinction between qualified and nonqualified deferred compensation and on scholarly commentary describing and endorsing the matching principle. Ultimately, the court concluded that permitting Albertson's its 1983 interest deduction is an unreasonable result that justifies disregard of the literal terms of the code.

Analyzing Albertson's II

As a practical matter, the precise holding of Albertson's II is of modest import: for the Treasury and most taxpayers, the 1983 version of section 404 is now of limited significance. Of greater practical importance is the Ninth Circuit's withdrawal of its improper construction of revised section 404. While that portion of Albertson's I was dictum, the Treasury was right to believe that it would lead employers to claim immediate deductions for accrued interest when the statute delays such deductions.

However, the fundamental issue raised by Albertson's II is the propriety of the court's methodology, i.e., ignoring exacting statutory terminology directly on point to implement a legislative purpose perceived (but not explicitly expressed) in the statute. Indeed, it is hard to envision a purer example of that methodology than Albertson's II. Much attention has been paid in recent years to the role in statutory interpretation of legislative history materials; ¹⁵ many tax commentators have similarly focused upon the use of Treasury regulations in the construction of the code; ¹⁶ yet other legal scholars have probed the limits of language and judicial responses to statutory ambiguity. ¹⁷ None of these issues complicates Albertson's II, a decision that, to implement perceived but unarticulated legislative purpose, overrides statutory terminology the court itself treats as unambiguous without benefit of legislative history or regulatory authority.

It is also hard to envision a better example of the pitfalls of that approach, i.e., that the legislative purpose discerned by the court is so general and imprecise that, in the final analysis, that alleged legislative purpose is nothing more than the court's own policy preferences.

Consider initially the distinction between qualified and nonqualified deferred compensation plans. The former, heavily regulated, generate immediate deductions for employer contributions; the latter, essentially unregulated, yield deductions on a delayed basis, when the employee receives his deferred compensation. ¹⁸ This rule of delayed deductibility is frequently denoted a matching principle in that, temporally, the employer's deductions are matched to the later period in which the employee recognizes offsetting income.

Under strong pressure from the government and some commentators, the Albertson's II panel concluded that an immediate deduction for accrued but unpaid interest under a nonqualified plan violates this matching principle and undercuts the distinction between qualified and nonqualified deferral: if employers can obtain current deductions through nonqualified plans, they will eschew qualified plans with their onerous rules designed to extend coverage to rank-and-file personnel and will instead utilize nonqualified deferred compensation arrangements to obtain current deductions.

These generalized observations, however, do not resolve the statutory issue in Albertson's, i.e., the scope of the matching principle for 1983. The 1983 version of section 404, with its explicit reference to section 162, specifies a limited matching rule under which employers must delay only deductions for the underlying deferred compensation itself; the 1984 (and current) version of section 404, with its more expansive reference to all deductions under the income tax, embodies a broader matching rule under which all deductions (including interest deductions) are postponed.

When the Albertson's II panel said that a current deduction for Albertson's' 1983 interest violates the matching rule, it was simply imposing its own bent for the expanded matching rule adopted by Congress in 1984 for nonqualified deferred compensation over the more circumscribed rule found in the earlier version of section 404. In effect, the Albertson's II panel treated the inconvenient and highly precise language of the 1983 statute as a scrivener's error that had gone unnoticed for decades and thereby imposed for 1983 its own predilection for a broader matching rule when the statute implements such a rule starting in 1984.

In effect, the Albertson's II panel treated the inconvenient and highly precise language of the 1983 statute as a scrivener's error that had gone unnoticed for decades.

To buttress its invocation of the matching rule, the Albertson's II court cited several commentators on the income tax. However, a careful review of these commentators indicates that they generate less support for the Ninth Circuit's decision than the appeals court recognized.

For example, the Albertson's II panel cited the current version of the Bittker/Lokken tax treatise for the proposition that nonqualified deferred compensation is taxed under the matching principle, which delays the employer's deduction until the compensation is subsequently taxed to the employee. ¹⁹ However, the treatise's discussion of section 404 does not address the particular issue in Albertson's, i.e., whether the 1983 version of the principle applies to accrued interest or just to the underlying deferred compensation. ²⁰

Likewise, the Albertson's II panel cited Professor Halperin's influential article on time-value-of-money issues in which he identifies section 404 as exemplifying the matching rule. ²¹ As a normative matter, Professor Halperin's analysis does indicate that employers paying nonqualified deferred compensation should not deduct interest until it is actually paid to the employee. However, as a matter of positive law, the Halperin article does not address the statutory reach of the matching rule as it actually appears in the code, i.e., whether for 1983 accrued but unpaid interest is subject statutorily to the rule of delayed deductibility in the context of nonqualified deferred compensation. ²²

Similarly, the other commentaries cited in Albertson's II provide no specific guidance as a matter of positive law on the breadth of the matching principle established by section 404 before its revision in 1984. For example, the Ninth Circuit quoted David S. Davenport's 1992 article for the proposition that "(c)urrent law" as embodied in section 404 establishes a matching rule. However, the court failed to recognize that this article's reference to "(c)urrent law" alludes to the revised, 1984 version of section 404, not the earlier version controlling in the Albertson's litigation. ²³

In short, the code in some places uses a matching principle and thereby delays a deduction until there is corresponding inclusion in the income of another taxpayer; many distinguished commentators believe this principle to be an important one which should be used more extensively in the tax law. However, these generalities do not resolve the particular issue in Albertson's, the scope of the matching principle in 1983 for nonqualified deferred compensation. On the other hand, the statute itself specifically limits delayed deductibility to those amounts otherwise deductible under section 162 and thus imposes no matching requirement upon the deductibility of interest governed by section 163.

Any doubt that generalized notions of a matching principle do not answer the specific issue in Albertson's should be dispelled by the Tax Court's recent decision under section 267, Tate & Lyle, Inc. v. Commissioner. ²⁴ Section 267 delays certain deductions for accrual-basis taxpayers when the payee is a related person using the cash method of accounting. While section 267 is quite detailed in many respects, in the context of foreign payees that provision broadly commands the Treasury to develop regulations implementing the "matching principle." In Tate & Lyle, the judges of the Tax Court could not agree among themselves what the matching principle is. The narrow majority held that the matching principle does not preclude the immediate deductibility of interest accrued by a U.S. subsidiary if the U.K. parent receiving such interest excludes it from income under the U.S.-U.K. income tax treaty. The Tax Court minority, on the other hand, concluded that the Treasury regulation delaying interest deductions in these circumstances properly applies the matching principle.

In Tate & Lyle, the Tax Court itself had to demarcate the scope of the matching principle since the statutory directive is to implement that principle. On the other hand, section 404 for 1983 does not impel such an open-ended inquiry but, rather, defines precisely the matching rule for deferred compensation as applying only to those amounts otherwise immediately deductible under section 162. The problems of the Tate & Lyle court belie the notion that the matching principle is a sufficiently precise concept to justify the decision of the Albertson's II court to ignore specific and controlling statutory language.

The Albertson's II panel finally declared that a current interest deduction for Albertson's is such an unreasonable result that the panel must disregard the literal terms of the statute to prevent that outcome. It is an unexceptional idea that courts should ignore statutory terminology when it leads to absurd results in particular cases. The relevant issue is whether a 1983 interest deduction for Albertson's is so unreasonable as to trigger this standard.

If time-value-of-money considerations are to be exalted above all other concerns, a 1983 interest deduction for Albertson's is indeed illogical. And there is no denying that a powerful current in contemporary tax scholarship and tax policy analysis extols time- value-of-money considerations to the exclusion of all competing tax policy criteria. On the other hand, both the code itself and more traditional modes of analysis suggest that a 1983 interest deduction for Albertson's reasonably compromises time-value-of-money concerns with at least two other competing tax policy criteria.

First, a 1983 interest deduction for Albertson's respects the taxpayer's normal method of accounting. The code indicates that a taxpayer like Albertson's, which consistently uses accrual accounting for its financial records, should presumptively use accrual accounting for tax purposes. ²⁵ In recent years, there has been much controversy defining the circumstances in which that presumption is overcome. ²⁶ Neverthless, there is no doubt that statutorily such a presumption exists.

The old version of section 404 can reasonably be viewed as compromising this presumption with the matching principle: To implement the policies of the matching principle, the underlying nonqualified deferred compensation is deductible under that provision only when included in the employee's income; to respect the taxpayer's method of accounting, accrued but unpaid interest is deductible currently under accrual principles. Hence, a 1983 interest deduction for Albertson's is not an unreasonable outcome but arguably reflects competing tax policy considerations and Congress's accommodation of them.

Moreover, an immediate deduction for Albertson's' 1983 accrued interest helps to establish vertical equity between an employer who pays his employee current salary and a competing employer who pays to his employee the same current salary plus deferred compensation. The latter employer has a lower net worth and thus less capacity to pay since, in addition to current salary, he has an accrued liability for future compensation. A tax system concerned about the proper tax treatment of these two employers vis-a-vis each other would grant an immediate deduction to the employer incurring that extra liability for deferred compensation.

Again, old section 404 can be interpreted as a compromise, balancing the matching principle with concern about vertical equity, honoring the former by delaying a deduction for the basic deferred compensation, implementing the latter by granting a deduction for accrued but unpaid interest. ²⁷ Like all accommodations of this sort, such a compromise is imperfect. The point, however, is not that old section 404 is ideal, but that it can be understood as a reasonable balance of conflicting considerations and hence not as easily dismissed as the Ninth Circuit concluded.

Professor Halperin, in a comment on the Albertson's litigation, rejects this line of thought as "unrealistic." ²⁸ If, he suggests, a company institutes deferred compensation for its employees, its competitor will match that move, either by creating its own deferred compensation program or by increasing current salary to offset its lack of deferred remuneration. Thus, there can be no vertical equity concerns between these two competitors as both will pay the same remuneration to their respective employees. Hence, there is no justification for old section 404 as partially responsive to vertical equity considerations. ²⁹

I agree that market forces would tend to create a single level of compensation for comparable employees working for similar companies. Employees would migrate from the employer offering current salary only to the employer also paying deferred compensation until a single standard of remuneration were established for both businesses. Nevertheless, these market adjustments would in practice be neither perfect nor instantaneous.

Some employees, particularly high-level executives of the sort who enter into deferred compensation arrangements, often have firm- specific skills ³⁰ and entitlements that impede their market mobility and thus preclude them from migrating to new employers as part of the process of market equilibration. For example, the president of Albertson's' chief competitor cannot simply move to the presidency of Albertson's in response to Albertson's' deferred compensation program. He might forfeit some entitlements (e.g., stock options, severance payments) if he leaves his current employer; much of his human capital may consist of information specific to his current employer and thus of no value to Albertson's; if that information is of potential benefit to Albertson's, the executive may be constrained legally from going to Albertson's by contractual, statutory, or common law restraints on the use of confidential and proprietary information or by noncompetition covenants into which the executive entered with his current firm.

Moreover, even when the labor market for executives tends to a single level of compensation, the adjustment is not automatic but occurs through informal channels and employer-employee negotiations spreading over more than a single year.

Hence, those who drafted and enacted old section 404 may reasonably have been concerned about vertical equity considerations during the period of market equilibration among those firms that offer their executives deferred compensation and those that do not and may have crafted old section 404 as a compromise between such considerations and the matching principle. From this perspective, an immediate deduction for accrued interest, while the deduction for the underlying deferred compensation is delayed, represents a plausible accommodation of competing concerns, an accommodation that the Ninth Circuit incorrectly dismissed as unreasonable.

Of course, we do not actually know why the drafters of old section 404 wrote the statute as they did. Perhaps they were thinking in the terms I have outlined; perhaps they had other considerations in mind; perhaps their reference to section 162 was an oversight, the significance of which they did not fully comprehend. We will never know.

It is, however, not necessary that we know how old section 404's specific reference to section 162 came about: for a statutory result to be dismissed as unreasonable, all of the possible justifications for that result must be found wanting; that is not the case for old section 404 as applied to the 1983 interest deduction claimed by Albertson's.

In short, Albertson's II does not explain satisfactorily why the literal and highly specific terms of old section 404 should be disregarded. The Albertson's II panel was commendably candid about its reservations. Ultimately, however, the court should have given greater weight to those reservations. The notion of a matching principle is too generalized and open-ended to justify repudiation of exacting and controlling statutory terminology; the result obtained by applying old section 404 as written -- immediate deductibility for accrued interest, delayed deductibility for the underlying deferred compensation -- is a plausible compromise between the considerations animating a matching rule and competing tax policy criteria. The Ninth Circuit was wrong to dismiss that compromise as unreasonable. In the final analysis, the legislative purpose perceived by the appeals court is nothing more than the court's own preference among competing policy considerations and the court's construction of old section 404 is simply the substitution of the court's policy preferences for those embodied in the statute in highly specific terms.

Albertson's Paths Not Taken

Could the Ninth Circuit have denied Albertson's a 1983 interest deduction in a different manner and thereby avoided the need to override what the court itself characterized as explicit statutory terminology? Several of the colleagues who reviewed an earlier draft of this article pointed to such an alternative: Time-value-of-money issues were not well-understood in 1942 when old section 404 was added to the code, this argument begins. Consequently, the drafters of that section would have used the term "compensation" to include both the initial salary deferred and what today is clearly understood as interest accruing on that initial amount. Hence, it is not significant that old section 404 refers only to section 162 and ignores section 163 since those who drafted old section 404, less adept than we are today at identifying interest, thought that the additional deferred compensation that is today characterized as interest governed by section 163 was part and parcel of the amounts governed by section 162. Thus, in subjecting to old section 404's matching rule accrued interest under the Albertson's nonqualified plan, the Ninth Circuit was, under this theory, not overriding an explicit statutory command to the contrary but instead was implementing the statute as its drafters intended.

This is a more plausible approach to the result reached by the Albertson's II panel than the court's own reasoning. However, for two reasons I am on the merits ultimately unpersuaded by this alternative argument and, in any event, I don't believe this alternative justification absolves the Ninth Circuit from its troubling methodology.

First, the argument's basic premise is unconvincing, i.e., that the nature of interest was so primitively comprehended in 1942 that the drafters of old section 404 could not have discerned interest as a distinct element of nonqualified deferred compensation plans like the Albertson's arrangement.

It is true that contemporary tax commentators and policymakers are far more concerned about time-value-of-money issues than were their predecessors. It is also true that many aspects of those issues are much better understood today than they were in 1942 and that implicit forms of compensating for the use of money, not recognized as interest in the past, are today understood as such.

It would, however, be wrong to overstate the ignorance of the tax community in 1942 and to assume that it could not recognize interest in the context of deferred compensation. The interest deduction had been a feature of the code since its inception and, hence, the dichotomy between trade and business expenses, on the one hand, and interest, on the other, was a well-established distinction when old section 404 was drafted. ³¹ The annual interest credited to participants in nonqualified deferred compensation arrangements like the Albertson's plan is typically labelled as interest and, when not explicitly denominated as interest, is clearly understandable as such -- as opposed to the more esoteric forms of interest that the courts and policymakers have in the past had greater trouble recognizing.

In short, those who would assume that the drafters of old section 404 could not comprehend the interest involved in deferred compensation arrangements need to prove their case.

Second, old section 404 was molded by two sets of drafters, the 1942 set and the 1984 set. ³² Even if those who originally authored section 404 in 1942 failed to distinguish between underlying deferred compensation and the interest earned on such amounts, their 1984 counterparts can hardly be charged with such ignorance of time-value- of-money concerns: The Tax Reform Act of 1984 is generally understood as the critical legislation that, for better or worse, elevated such concerns to a central place in the tax statute. ³³

The 1984 drafters clearly understood the difference between section 162 and section 163. Nevertheless, they chose to amend section 404 prospectively, leaving intact for taxpayers' open years old section 404 and its exacting reference to section 162. While we might charge the 1942 drafters of the code with not comprehending the import of such a reference, we cannot so charge their successors four decades later who were changing that reference for the future and who chose to leave intact for open years the specific reference to section 162.

Finally, even if this alternative argument is persuasive, it does not vindicate the Ninth Circuit's methodology. The fundamental problem with Albertson's II is not its holding but its rationale. That other arguments might justify the outcome in Albertson's II does not excuse the reasoning by which the Ninth Circuit denied Albertson's its 1983 interest deduction, i.e., ignoring what the court acknowledged to be controlling terminology in the name of unexpressed legislative purpose.

Conversations with yet other colleagues, as well as the commentary of Jasper L. Cummings Jr., ³⁴ raise the possibility that, even under new section 404, accrued interest is currently deductible under nonqualified deferred compensation plans and not subject to a matching rule that delays deductibility until its receipt by the employee. ³⁵ That argument starts with the 1984 legislative history, cited by the Albertson's II panel, which describes new section 404, not as delaying interest deductions, but as postponing depreciation deductions. Moreover, section 404 (both old and new) refers to "compensation," not to "compensation and inter- est." Hence, the argument runs, the broader matching rule of new section 404 extends only to underlying compensation properly capitalized and depreciated, not to interest which was and still is unmentioned in section 404.

I confess that, as I have rethought the Albertson's controversy since my initial statement on the case, ³⁶ I have at times persuaded myself along these lines, i.e., that, as a result of the failure of new section 404 to refer explicitly to interest, even under that provision the matching rule of delayed deductibility applies only to the underlying deferred compensation and not to interest subsequently accrued on those initially deferred amounts. However, in the end, I return to the conclusion that current law, as embodied in new section 404, subjects interest, as well as the underlying deferred compensation, to the matching rule.

The controlling feature of old section 404 is the specificity of its language, in particular its explicit incorporation of section 162; section 162 is not section 163. On the other hand, the amendment of section 404, which deleted the reference to section 162, markedly decreases the specificity of section 404 and thus properly permits other considerations in the construction of the statute. In such a context, it is sensible to search, inter alia, for broader congressional purposes since such a search seeks to discern the meaning of potentially equivocal language, not to override highly specific terminology. And, as noted earlier, the Tax Reform Act of 1984 was a clarion call to infuse time-value-of-money considerations into the tax law. ³⁷

Hence, I conclude that the better interpretation of new section 404 is that the term "compensation" inartfully includes interest and subjects such interest to the rule of delayed deductibility. I admit, however, that the question appears closer to me today than it did a year ago.

Who Should Override Statutory Terminology?

The discussion so far suggests why Albertson's II is of concern to those interested in compensation and accounting, i.e., timing issues. However, of greater significance are the implications of Albertson's II for the continuing debate about methods of statutory interpretation in general and approaches to the code in particular. For purposes of this debate, Albertson's II is a remarkably clean decision, uncomplicated by such factors as legislative history materials or regulatory authority. Nor is Albertson's a case where statutory ambiguity justifies the search for underlying legislative purpose. For the Ninth Circuit, the Albertson's saga ultimately reduced to the proposition that perceived, if unexpressed, legislative purpose should override explicit statutory terminology.

I suggest that Albertson's II reveals the soft underbelly of that proposition and indicates why overriding specific terminology in the tax statute is a task for Congress, not the judiciary. For four reasons, courts should be reluctant to override exacting language in the code in the name of unarticulated legislative purpose:

(a) The code is rarely a single purpose document. T he principle that a court can implement legislative purpose nonliterally assumes that the statute in question has a single purpose or has multiple purposes that point in the same direction. The classic case for overriding explicit statutory terminology in the name of underlying legislative purpose is when a simple statute embodies a single, uncomplicated policy, e.g., an ordinance that forbids spilling blood in the streets does not apply to a doctor performing an emergency procedure since the legislature's sole and evident objective was to stop criminal violence, not medical care. ³⁸

The code, in contrast, is a melange of compromises, some grossly political, some reflecting the accommodation of competing policy criteria. Rarely does a particular tax provision implement a single statutory mission and thus seldom can a court convincingly claim that a nonliteral approach to the code implements legislative purpose since there are many such purposes in the tax statute and often they conflict with one another.

It is, for example, reasonable to characterize the matching rule as a purpose of old section 404(a)(5); however, it is unconvincing to label matching as the section's only purpose: respect for the taxpayer's method of accounting ³⁹ and the search for vertical equity among differently situated taxpayers ⁴⁰ are well-recognized missions of the tax law and quite plausibly animated the compromise embodied in the 1983 version of section 404(a)(5).

The accommodation of conflicting policy objectives is a time- honored function of the legislature. It is thus troubling when a court (like the Albertson's II panel) identifies one among many possible statutory objectives and declares that to be the legislature's only purpose.

The Albertson's II court may have obliquely acknowledged the force of this concern. The appeals court did not say that no offsetting legislative purposes counterbalance the matching rule; the court carefully stated that Albertson's had not advanced any such countervailing purposes, thereby suggesting that there might be some but that the court had not been informed of them. ⁴¹

This reinforces the image of the code as a complex bundle of conflicting principles and counsels against courts concluding that they have discerned a single legislative purpose in the tax law that they can implement through a nonliteral application of the code when more likely there are conflicting considerations, even if the litigants have not managed to raise them.

(b) The code is correctable legislatively. It is a little too facile to say that the legislature can correct an improper judicial construction of a statute. The side that wins in court has an advantage in any subsequent legislative struggle since that side has the easier task of defending the status quo and prevails merely by blocking revision of the statute. Hence, which side wins in court may affect the results subsequently emanating from the legislature.

Nevertheless, statutes are correctable legislatively and some more so than others. The code is highly correctable legislatively. Indeed, in Albertson's, the Treasury's problem with section 404 had already been remedied prospectively before the litigation began.

While there are some statutes Congress revisits infrequently, near-continuous revision of the code has, for better or worse, become a durable feature of the federal legislative process. Moreover, because the tax law impacts on so many and such varied interest groups, tax policymakers (in contrast to legislators, administrators, and professional staff with more limited jurisdictions) are relatively less capturable by special interests and thus have greater ability to make independent and professional judgments to reform the tax statute. /42

Hence, if the balance of conflicting policies embodied in the current terminology of the code is not the balance desired by the current Congress, it is reasonable to rely on legislative revision to put the statute right.

(c) Nonliteral means pro-government. As Professor Zelenak has noted, nonliteral interpretations of the code tend to be pro- government interpretations of the code. ⁴³ A taxpayer who ignores statutory language in the name of an unexpressed legislative purpose is likely to be slammed with a wide range of penalties designed to protect the integrity of the code; ⁴⁴ when the government disregards the terminology of the tax statute, it is likely to be rewarded with a victory like Albertson's II.

At a minimum, the courts should be honest about this rather than invoking bromides about the disregard of statutory text to avoid unreasonable results, bromides that imply a nonexistent evenhandedness between taxpayer and fisc.

(d) Specificity counts. Ultimately, Albertson's is about statutory specificity, i.e., old section 404's quite precise reference to section 162.

Some tax statutes, by their very nature, require courts to explore and determine legislative purpose: When section 267(a)(3) directs that payments to foreign payees be subject to "the matching principle," the courts are thereby impelled to discern and implement the legislative objective embodied in this open-ended statutory language; when section 162(a)(1) restricts deductibility to "reasonable" compensation, the statute compels the courts to determine Congress's purpose in utilizing such flexible terminology. Sometimes, such terminology reflects a congressional failure to agree and a consequent delegation to the courts and administrative agencies of the controversy Congress itself could not resolve; at other times, elastic language reflects a considered determination that judges and administrators can better provide the details of the law than can the legislature.

On the other hand, much of the code is highly specific or, to employ Professor Levinson's useful phrase, "hard." ⁴⁵ It is difficult to imagine language less elastic than old section 404's reference to section 162. Such language has its own purpose: to convey meaning with as little room for reader discretion as possible.

It is not easy, and perhaps not possible, to reduce to a formula the circumstances in which courts should disregard statutory terminology to implement unarticulated legislative intent. However, it seems clear that, when statutory language is as specific as old section 404, a very high threshhold should be met before such language is overridden judicially and that the Albertson's II court did not meet that threshold. /46

Conclusion

It is not simply hard cases that make for bad law. The Albertson's court always had an easy path available to it: to honor the highly specific terms of old section 404 (thus granting Albertson's its 1983 deduction for accrued but unpaid interest) while leaving intact revised section 404 (thus signalling to Albertson's and other accrual method payors of nonqualified deferred compensation the future unavailability of such a deduction). Instead, in Albertson's II the Ninth Circuit was persuaded to deny the 1983 interest deduction as a matter of policy and to characterize that policy-based denial as an implementation of underlying, but unexpressed, legislative purpose.

The Albertson's II court is to be commended for the candor with which it revealed its reservations about the nonliteral course it ultimately took. However, instead of disclosing its reservations, the appeals court would have done better to have acted on them.

FOOTNOTES:

¹ 12 F.3d 1529 (1993), West's has withdrawn Albertson's I from the bound volume of the Federal Reporter.

² 42 F. 3d 537 (1994), Albertson's has asked the Ninth Circuit to rehear the case en banc. See "Thrasher: Welfare Plan Guidance Likely Soon," H&D, Jan. 13, 1995, p. 862.

³ See, e.g., Lee A. Sheppard, "The Ninth Circuit Creates a New Interest Deduction," Tax Notes, Jan. 24, 1994, p. 405; Daniel Halperin, "Ninth Circuit's Decision in Albertson's Is Outrageous," Tax Notes, Feb. 21, 1994, p. 1083. While much of the commentary on the Albertson's litigation has been pro-government in outlook, there have been those supporting the taxpayer's claim for a 1983 interest deduction. See, e.g., Jasper L. Cummings Jr., "Statutory Interpretation and Albertson's," Tax Notes, Jan. 23, 1995, p. 559.

⁴ On the actual timing of Congress's 1984 changes, see note 32, infra.

⁵ In addition to the deferred compensation issue, Albertson's I addressed Albertson's' eligibility for WIN and investment tax credits. The Ninth Circuit's resolution of these questions has not engendered any controversy. See Albertson's I, supra note 1.

⁶ Section 404(a)(5) of the code as in effect in 1983 (emphasis added)

⁷ Section 404(a)(5) (emphasis added).

⁸ In the case of nonqualified deferred compensation arrangements like those at issue in the Albertson's litigation, the code provides for taxability to the employee when such compensation is vested in the employee. See section 83. It is, however, the prevailing practice to vest such compensation in the employee when he actually receives it. Hence, in practice, deferred compensation is taxed to employees upon receipt.

⁹ Section 404 has been amended numerous times in ways not relevant to the Albertson's litigation and contains many provisions not pertinent to that litigation. By "1983" or "old" section 404, I mean section 404 as in effect for 1983 as it pertains to the deductibility of nonqualified deferrred compensation; by "new" or "1984" section 404, I mean the provisions of section 404 germane to nonqualified deferred compensation as such provisions were amended in 1984 and remain in effect today.

¹⁰ The other noteworthy effect of the 1984 changes is to subject to section 404 nonqualified deferred compensation outlays which would otherwise be depreciable under sections 167, 168 and 197.

¹¹ See, e.g., Sheppard, supra note 3; Halperin, supra note 3.

¹² See Barbara Kirchheimer, "Pryor Introduces Bill To Reverse Decision in Albertson's," Tax Notes, Mar. 7, 1994, p. 1309. S. 1877, 103rd Cong., 2d Session (1994).

¹³ 12 F. 3d 1539 (1994).

¹⁴ See footnote 5 of Albertson's II.

¹⁵ See Michael Livingston, "Congress, the Courts and the Code: Legislative History and the Interpretation of Tax Statutes," 69 Tex. L. Rev. 819 (1991).

¹⁶ See, e.g., Toni Robinson, "Retroactivity: The Case for Better Regulation of Federal Tax Regulators," 48 Ohio St. L.J. 773 (1987).

¹⁷ Much interesting legal scholarship has addressed the relationships among texts, authors, and readers and suggests that many statutes do not yield a single meaning. This issue does not arise as to Albertson's II since the appeals court acknowledged that the 1983 version of section 404 literally grants Albertson's a current interest deduction and thus there is no disagreement between the court and its critics about the meaning of the statutory text. For a sample of the burgeoning legal literature on texts and indeterminacy, see Levinson, infra note 45.

¹⁸ While section 404 has, over the years, become quite detailed, its basic timing rules have remained intact: employer contributions to qualified plans are, within certain limits, deductible "when paid" by the employer; employer contributions to nonqualified deferred compensation arrangements are deductible when taxable to the employee, which typically occurs later in the employee's career. Compare section 404(a)(1)(A) with section 404(a)(5).

The code's current deduction for employer contributions to qualified plans is conventionally labelled a tax expenditure. The Albertson's II opinion demonstrates the potency of that label, despite the questionable premises on which it is based. See Edward A. Zelinsky, "Tax Policy v. Revenue Policy: Qualified Plans, Tax Expenditures and the Flat, Plan Level Tax," 13 Va. Tax Rev. 591 (1994).

¹⁹ See Boris I. Bittker and Lawrence Lokken, 2 Federal Taxation of Income, Estates and Gifts (2d. ed., 1990) Para. 60.1.

²⁰ Id. at para. 60.2.2.

²¹ Daniel I. Halperin, "Interest in Disguise: Taxing the 'Time Value of Money,'" 95 Yale L.J. 506 (1986) at 520.

²² Indeed, Professor Halperin's discussion of the matching principle is more complex and subtle than the appeals court recognizes. As Professor Halperin demonstrates, under certain assumptions, a matching rule is not necessary to achieve a proper outcome from a time-value-of-money perspective and, under other conditions, a matching rule is not sufficient by itself to achieve such a result.

²³ David S. Davenport, "Education and Human Capital: Pursuing an Ideal Income Tax and a Sensible Tax Policy," 42 Case W. Res. L. Rev. 793 (1992) at 865.

The Ninth Circuit also cited Jasper L. Cummings' article on basis and the corporate income tax, which article characterizes section 404 as implementing the matching principle. However, the Cummings article neither addresses the Albertson's issue -- the deductibility of 1983 accrued interest -- nor does it express strong support for the matching principle. Jasper L. Cummings Jr., "The Silent Policies of Conservation and Cloning of Tax Basis and Their Corporate Applications," 48 Tax L. Rev. 113 (1992) at 162 ("a general matching theory...is more of an observation than a reasoned justification.")

Similarly, the Albertson's II opinion cites Professor Cunningham's article on the proper tax treatment of future costs. However, that article also does not discuss the particular statutory terminology of either the old or the revised section 404. Noel B. Cunningham, "A Theoretical Analysis of the Tax Treatment of Future Costs," 40 Tax L. Rev. 577 (1985) at 610.

Professor Gergen's reference to section 404 specifically discusses the 1984 alteration of the statute and the impact of that alteration on deferred compensation otherwise deductible pursuant to the code's depreciation provisions. While the appeals court cited the relevant statement in Professor Gergen's article, it failed to consider the implications of that statement for the Albertson's litigation: if section 404 had to be amended to cover items otherwise deductible under sections 167 and 168, that section also required amendment to cover items otherwise deductible under section 163. See Mark P. Gergen, "Reforming Subchapter K: Compensating Service Partners," 48 Tax L. Rev. 69 (1992).

One authority invoked by the Ninth Circuit indirectly alludes to the Albertson's controversy. While the text of the Merten's income tax treatise does not discuss the case or the precise wording of section 404, the treatise refers to LTR 8619006, which appears to be the private letter ruling in which the IRS reversed its earlier approval of Albertson's, 1983 deduction of its accrued but unpaid interest. See Merten's Law of Federal Income Taxation, Code Commentary (1990) at section 404(a):5 (footnote 2).

That ruling underscores the extent to which the IRS's position in the Albertson's litigation, ultimately accepted by the appeals court, disregards the controlling statutory terminology. In LTR 8619006, the IRS, in a passage labelled "applicable law," correctly cited 1983 section 404 and its quite specific reference to section 162. However, in its later discussion of the case, entitled "rationale," the IRS ignored the statutory reference to section 162 altogether: had the IRS grappled with the implications of this reference, it could not have ruled as easily as it did.

²⁴ 103 T.C. No. 37.

²⁵ Section 446(a) provides: "Taxable income shall be computed under the method of accounting on the basis of which the taxpayer regularly computes his income in keeping his books."

²⁶ See, e.g., Thor Power Tool Co. v. Commissioner, 439 U.S. 522 (1979).

²⁷ A simple example illustrates this point. Consider two accrual-basis taxpayers, A and B, which (except for nonqualified deferred compensation) both have income of $100 for a particular year. Assume further that A does not provide its executives nonqualified deferred compensation but that B does and that, for the year in question, B credits $10 to its executives' deferred compensation accounts on the first day of such year. Suppose also that, under B's nonqualified plan for its executives, such $10 earns interest of 8 percent at year's end.

If the tax system is concerned about vertical equity between A and B, the system must give B some kind of deduction both for the additional $10 it owes its executives as deferred compensation and the $0.80 it owes its executives as interest. On the other hand, the matching principle, at least as understood by some, indicates that B should not deduct the $10.80 until that amount is included in the executives' respective incomes.

Old section 404 is plausibly understood as compromising these two alternatives, delaying the $10 deduction to implement the matching principle but permitting the deduction of the $.80 interest as partial recognition that B has greater expenses for the year than does A.

²⁸ Daniel Halperin, "Halperin Reiterates His Complaints About Albertson's," Tax Notes, Apr. 25, 1994, p. 483 (letter to editor). Jasper L. Cummings Jr. characterizes Professor Halperin's criticism as "unconvincing." See Cummings, supra note 3, at 575 (footnote 141).

²⁹ If Professor Halperin is correct, my example is wrong. See note 27, supra. His comments suggest that competitive forces will either force A to institute a deferred compensation program like B's or will impel B to abolish its program to produce a competitive profit for its owners. In the former case, A will duplicate B's $10.80 expense; hence, there is no vertical equity issue since the incomes of A and B are identical. In the latter case, B will not undertake the $10.80 expense; again, there will be no vertical equity problem to remedy since A and B will have identical incomes.

³⁰ The distinction between employees' general training and skills in contrast to their firm-specific training and skills is a concept central to contemporary labor economics. See, e.g., Lloyd G. Reynolds, Stanley H. Masters, Colletta H. Moser, Economics of Labor (1987) at 103.

³¹ See Bittker and Lokken, supra note 19, at para. 31.1.1 (footnote 1) ("The interest deduction has been a fixture of the modern federal income tax since its inception.").

³² Technically, the drafters of new section 404 did their work in 1986 with Congress making their efforts retroactive to 1984. See Zelinsky, infra note 36 at 233-234. For ease of exposition, I describe these events as having occurred in 1984 since Congress retrospectively placed them there.

³³ See Cunningham, supra note 23, at 577 (characterizing the Tax Reform Act of 1984 as a congressional reaction to time-value-of- money problems "with a vengeance.").

³⁴ See Cummings, supra note 3.

³⁵ The Albertson's II panel correctly recognized that for 1983, the year at issue in the case, the controlling version of the statute is old section 404. See footnote 5 of Albertson's II. Were the Ninth Circuit now to hear the case en banc, new section 404 would not properly be considered by the court since that section does not apply to 1983, the year at issue in Albertson's, but applies to later years only.

³⁶ Edward A. Zelinsky, "The Ninth Circuit's Albertson's Decision: Right for 1983, Wrong for Today," Tax Notes, Apr. 11, 1994, p. 231.

³⁷ See note 33 supra, and accompanying text.

³⁸ This oft-cited example comes from the U.S. Supreme Court's seminal decision in Church of the Holy Trinity v. United States, 143 U.S. 457 (1892) at 461.

³⁹ See notes 25 and 26 supra, and accompanying text.

⁴⁰ See notes 27 through 30, inclusive, supra, and accompanying text.

⁴¹ In fact, the Albertson's brief on rehearing alerted the court to several articles which suggest possible policy justifications for old section 404. See Albertson's, Inc., Appellant/Cross-Appellee, v. Commissioner of Internal Revenue, Appellee/Cross-Appellant (dated: May 25, 1994), 94 TNT 115-25 , at footnote 3.

⁴² I develop this theme in greater detail in Edward A. Zelinsky, "James Madison and Public Choice at Gucci Gulch: A Procedural Defense of Tax Expenditures and Tax Institutions," 102 Yale L. J. 1165 (1993).

⁴³ Lawrence Zelenak, "Thinking About Nonliteral Interpretations of the Internal Revenue Code," 64 N. Car. L. Rev. 623 (1986) at 666.

⁴⁴ For example, the taxpayer is subject to a 20-percent penalty on any underpayment of tax attributable to the "intentional disregard" of "the provisions of" the code. Section 6662(a) and 6662(c).

⁴⁵ Sanford Levinson, "Law as Literature," 60 Texas L. R. 373 (1982) at 375-376.

⁴⁶ The question of statutory specificity, and the constraints such specificity imposes on courts, arises in the context of other highly detailed statutes. See, e.g., Knesz v. The Central Jersey Bank and Trust Company, 97 N.J. 1, 477 A. 2d 806 (1984) (Specific provision of the Uniform Commercial Code enforced literally despite countervailing policy considerations).

END OF FOOTNOTES

ALBERTSON'S: WHY COURTS SHOULDN'T OVERRIDE CLEAR STATUTORY LANGUAGE.