Menu
Tax Notes logo

Sense and Sensibility and Creditability: Redefining an Income Tax ‘in the U.S. Sense’

Posted on Oct. 11, 2021
Robert E. Culbertson
Robert E. Culbertson

Robert E. Culbertson is a senior counsel in the Washington office of Covington & Burling LLP. He thanks his Covington colleagues Allison Baker and Michael Caballero for their thoughtful comments and other assistance. He also thanks Mary Bennett, Bob Stack, and other members of the Washington International Tax Study Group for their useful input.

In this report, Culbertson argues that the foreign tax credit regulations proposed in November 2020 are inconsistent with the legislative purposes reflected in section 901 and with the statute’s long-standing interpretation.

The views expressed herein and any errors are the author’s own.

Copyright 2021 Robert E. Culbertson.
All rights reserved.

Table of Contents
  1. I. Introduction
    1. A. History Isn’t Bunk When a Code Section Remains Amazingly Unchanged for 100 Years
    2. B. Scope and Implications of the Principal Changes Proposed by the November 2020 NPRM
  2. II. Congressional Intent Reflected in the 100-Year History of Section 901
    1. A. Section 901: Eternally Youthful and Unchanged
    2. B. Normative Tax Policy Principles Embodied in the Initial Enactment and Amendment of the FTC in the Revenue Acts of 1918 and 1921
      1. 1. Congress sought horizontal equity between taxpayers with and without foreign income, enhanced international competitiveness, and the prevention of double nontaxation.
      2. 2. To protect U.S. taxation of U.S.-source income, Congress repealed a jurisdictional test and instead placed full reliance on an FTC limitation.
    3. C. Decades of Legislative and Regulatory Actions Have Reinforced the Original Tax Policy Decisions Reflected in Section 901
      1. 1. Many dogs, no barking; not even a growl.
      2. 2. Enactment of section 903 in 1942.
      3. 3. Enactment of subpart F in 1962.
      4. 4. Development of detailed regulations under section 901, 1979-1983.
      5. 5. Enactment of the TCJA in 2017.
  3. III. Many Changes Proposed by the November 2020 NPRM Are Inconsistent With the History and Purposes of Section 901
    1. A. The Jurisdictional Nexus Requirement
      1. 1. The proposed requirement.
      2. 2. Apparent purpose and likely double taxation impact of the rule.
      3. 3. Interaction of the jurisdictional nexus requirement with tax treaty creditability rules.
    2. B. Principal Changes to the Net Gain Requirement
      1. 1. Deletions of predominant character rules affect all three tests under the net gain rule.
      2. 2. Proposed changes to the realization requirement.
      3. 3. Proposed changes to the gross receipts requirement.
        1. a. Summary of proposed changes.
        2. b. Confusion regarding cost-based measures of receipts.
        3. c. Revival of rejected regulatory proposal.
        4. d. Concluding comments on proposed changes to the gross receipts requirement.
      4. 4. Proposed changes to the net income/cost recovery requirement.
        1. a. Required recovery of significant costs and expenses.
        2. b. Restriction (or repeal?) of the alternative allowance rule.
        3. c. Concluding comments on proposed changes to the net income requirement.
    3. C. Proposed Changes to the Section 903 Rules on the Creditability of Taxes Imposed in Lieu of Income Taxes
      1. 1. Overview of the changes.
      2. 2. Changes to reg. section 1.901-2 would have broad knock-on effects on the creditability of in-lieu taxes under section 903.
      3. 3. Rewriting of the substitution requirement.
        1. a. Overall complication of regulations implementing the eight-letter term ‘in lieu of.’
          1. i. Underlying income tax.
          2. ii. Non-duplication.
          3. iii. Close connection.
          4. iv. Jurisdictional nexus.
        2. b. The non-duplication test seeks to reverse Metropolitan Life.
        3. c. The close connection rule is unsupported by the language of section 903 and again inconsistent with Metropolitan Life.
    4. D. Concluding Observations on the Surpassing Strangeness of Treating the IRC as the Platonic Ideal of an Income Tax
  4. IV. Conclusions
    1. A. The November 2020 NPRM Should Be Withdrawn
    2. B. Future Changes to Section 901 or Its Implementing Regulations
  5. V. Appendix

“The subject of double taxation is rather prosaic. It is not dramatic or thrilling, and I shall not try to endow it with a fictitious interest.”

Thomas S. Adams1

I. Introduction

A. History Isn’t Bunk When a Code Section Remains Amazingly Unchanged for 100 Years

History is more or less bunk. It’s tradition. We don’t want tradition. We want to live in the present, and the only history that is worth a tinker’s dam is the history we make today.”

Henry Ford2

Compared with the other startling events of November 2020, the publication of proposed foreign tax credit regulations may seem rather prosaic.3 But from the standpoint of its direct revenue impact and its relevance for ongoing international negotiations, the notice of proposed rulemaking (NPRM) published on November 12, 2020, may well contain the most important international tax regulations in many years.4 The NPRM proposes an ambitious package of broad changes, apparently seeking to gain leverage for the United States in multilateral negotiations over the taxation of cross-border e-commerce. It also revisits many existing regulatory rules that had remained untouched for several decades.

Section 901 generally provides a credit against U.S. tax for foreign “income, war profits, and excess profits taxes.” The statutory text defining a creditable foreign tax was not one of the many international tax rules modified by the Tax Cuts and Jobs Act in 2017.5 Nor was it changed during the major rewrite of the international tax rules in the Tax Reform Act of 1986. In fact, the relevant text has remained remarkably unchanged for a century. The foreign tax credit was first enacted in the Revenue Act of 1918 and promptly modified in the Revenue Act of 1921. But since 1921, through the enactment of multiple revenue acts and not one, not two, but three IRCs, Congress has seen fit to add two words and delete one in the definition of a creditable tax, steadily allowing a credit for “any income, war profits, and excess-profits taxes paid or accrued during the same taxable year to any foreign country or to any possession of the United States.”6

Thus, by proposing substantial modifications to the regulations implementing a long-standing and unrevised statute, Treasury and the IRS have brought renewed focus to its original enactment and its 100-year history. Public comments on the proposed regulations have recognized the relevance of the legislative history of the FTC, demonstrating impressive erudition in explaining that history and emphasizing a congressional focus on the role of the FTC in protecting the international competitiveness of American business.7 And indeed, in a long, studious, and carefully argued preamble, Treasury and the IRS themselves use appeals to history to justify many of the proposed changes, referring repeatedly to long-standing “international norms” and to previously proposed (albeit rejected) regulatory innovations.8

The preamble does not, however, take an equally close look at the congressional purpose in enacting the FTC and in repeatedly amending its ancillary rules without ever changing the statutory definition of a creditable tax. This report aspires to fill that gap by reviewing, in Section II below, the relevant history and its implications for the proposed regulatory changes. Although 100 years of history sounds like it could be tedious to recount, in this case the task is simplified by the fact that much of this history features the famous dog that didn’t bark. That is, in the absence of relevant congressional activity (apart from repeated reenactments without meaningful change), the history of the 1918-1921 period is the legislative history that remains primarily relevant today, with little in the way of intervening activity to distract from that focus.

This is, of course, very different from the normal experience of tax practitioners — and reg drafters. The code is filled with rules that Congress has tinkered with, tweaked, or radically rewritten periodically, and indeed sometimes seemingly annually. Thus, tax work most often focuses on the meaning of recently enacted legislative changes. As a result, the fact that Congress has seen no need to substantively amend the relevant language of section 901 since 1921 leaves us in the unusual territory of having to consider congressional intent that was formed not in recent years or even recent decades, but a century ago, at the dawn of the modern income tax and a time when some modern interpretive tools (such as the blue book) had not yet been invented.9 And while the absence in that earlier period of the legislative logorrhea that usually accompanies a modern tax act may make it somewhat harder to ascertain the policies and purposes that Congress sought to achieve, it does not make those policies and purposes any less relevant to the implementation of the statute. It certainly doesn’t justify an analysis limited to a simplistic repetition of the avoidance-of-double-taxation mantra without considering what Congress actually meant when it referred to double taxation and why Congress thought it was important to prevent the double taxation of international income in the first place.

I do not argue that mere antiquity limits the government’s power to reinterpret a statute under its continuing rulemaking authority, particularly if a change in circumstances requires a response. But I suggest that in exercising that power, the government remains obligated to implement the legislative purposes reflected in the language of the statute, regardless of how long ago it was enacted. And if changed circumstances require fundamental changes to the statute (a proposition not yet demonstrated here, I submit), those changes must, of course, be enacted legislatively, not by regulatory fiat.

Chock-full of insights gleaned from our historical excursion into the origins of the FTC, we will review in Section III below the changes proposed by the November 2020 NPRM. Based on that analysis, I respectfully argue that the proposed regulations are inconsistent with the legislative purposes reflected in section 901 and with its long-standing interpretation. Section 901 has long been read to provide a credit for a foreign tax that is an income tax “in the U.S. sense,” and a tax may plainly be an income tax in the U.S. sense without being the U.S. income tax. Yet the practical impact of the November 2020 NPRM would be to deny income tax treatment to many foreign taxes merely because they are insufficiently similar to the detailed provisions of the IRC, presumably as of the moment when the foreign tax’s creditability is being determined. Nothing in the language or history of section 901 justifies this narrowing of the scope of creditable foreign taxes to those that emulate the particularities of the code as it stands at any given moment. As we will see below, this approach is problematic both in principle and practically.

To energize the reader for the hard work to come, I will round out this introduction with a brief overview of the surprising scope and alarming implications of the regulatory changes proposed by the November 2020 NPRM.

Figure 1. Creditability of Foreign Income Taxes Under Reg. Section 1.901-2: So Many Ways to Fail

B. Scope and Implications of the Principal Changes Proposed by the November 2020 NPRM

The November 2020 NPRM would most significantly add a new requirement to the definition of a creditable foreign tax.10 To be creditable, a foreign tax would have to satisfy one of three new tests of “jurisdictional nexus.” The proposed regulations would also modify many of the rules and standards set forth in the three existing requirements that define a net income tax: the realization requirement, the gross receipts requirement, and the newly renamed cost recovery requirement (nèe the net income requirement). Moreover, the regulations would substantially modify the definition of a tax imposed in lieu of an income tax under section 903.

Statutorily, section 901 provides a credit for a foreign “income tax.” As noted, that term has long been held to mean an income tax under U.S. standards, or “in the U.S. sense.”11 The proposed regulations appear to extrapolate from that gloss a new approach that depends not on whether a foreign tax is an income tax in the U.S. sense but rather on whether the foreign tax substantially conforms to the structure and detailed operation of the IRC as it may be amended from time to time. In fact, because the proposed regulations require foreign law to conform to the code (rather than to a more general concept of income taxation), a foreign tax imposed on economic income measured more accurately than under the code may not be creditable. For example, if the code provides for accelerated depreciation of an asset that foreign law depreciates economically, the foreign tax could be non-creditable because it departs from the computation of the income tax under the IRC. That approach is hard to square with the plain meaning of the term “income tax” or with much of the history of what it means to be an income tax in the U.S. sense. Instead, the proposed rules would demand greater conformity between U.S. and foreign law than has historically been required for a foreign tax to qualify as an income tax.

Figure 1 provides an overview of the many ways in which a foreign tax could flunk the revised creditability standards of the proposed regulations under section 901.12

Under the proposed regulations, the unamended statutory term “income tax” would be redefined to encompass only foreign taxes that substantially conform to the provisions of the IRC in terms of realization, recognition, and deductibility rules, and that also conform to the U.S. view of cross-border taxing jurisdiction (including the U.S. approach to sourcing rules). Overall, these changes would likely have at least three dramatic impacts on U.S. taxpayers.

First, as a substantive matter, any foreign tax that deviated from code rules could readily become non-creditable. The degree of permissible variance is unclear and seems to vary among the four principal tests of creditability — jurisdictional nexus, realization, gross receipts, and cost recovery — with particularly close conformity required for cost recovery. But the effect of these changes, and apparently their intent, will be to deny the creditability of many foreign taxes that from a plain-English standpoint would readily be viewed as income taxes and that have routinely been treated as such for purposes of section 901 by taxpayers, the IRS, and (when the IRS was reluctant) courts, including the Supreme Court.

Second, as a compliance matter, the proposed rules would impose substantial new administrative burdens because any significant changes in either U.S. or foreign taxing rules would require a reassessment of whether the two sets of rules conform to the extent now required (whatever that may be). Thus, regular reassessments of the creditability of all foreign taxes would become a feature of FTC compliance by taxpayers and of FTC audits by the IRS, with some non-creditable foreign taxes potentially becoming creditable and other creditable foreign taxes potentially becoming non-creditable, depending on the magnitude of the changes to both the U.S. and the foreign rules. Certainly, substantial changes in a foreign tax could always have required a reevaluation of its creditability, but the new level of conformity that would be required under the November 2020 NPRM would necessitate those reevaluations far more regularly, including after virtually all meaningful changes to any provision of the IRC.

Third, as a controversy matter, the proposed regulations would combine vague standards requiring some ineffable level of conformity between U.S. and foreign rules, with a deliberate effort to eliminate any aspect of the rules that permitted a practical or flexible inquiry into whether a foreign tax was in fact an income tax. Instead, the new rules would require formal comparisons couched in absolute terms, with no room to consider the actual impact of the rules as applied (touted by the preamble as avoiding a toilsome inquiry into mere “empirical” information). This combination of vague standards with formalistic tests will almost certainly produce FTC controversies at a pace to gladden the hearts of tax litigators everywhere.

In sum, the approach set forth in the proposed rules is difficult to square with relevant statutory language, would depart in fundamental ways from the tax policy views implemented by Congress throughout the 100-year history of the credit, would be the source of profound administrative and compliance problems, would likely result in substantial litigation, and in the final analysis, would almost certainly increase the incidence of double taxation, possibly exponentially. For all those reasons, and as discussed in more detail in Section III below, I respectfully suggest that the November 2020 NPRM be withdrawn. Renewed regulatory efforts should be guided by the policy imperatives that flow both from Congress’s long-ago enactment of the statute and from its 100-year history of restraint in leaving the provision substantively unamended.

II. Congressional Intent Reflected in the 100-Year History of Section 901

A. Section 901: Eternally Youthful and Unchanged

The starting point of our historical excursion is the recognition that the provisions of section 901 relevant to the definition of a creditable foreign tax have not been meaningfully altered since 1921. This antiquity of the underlying statute presents an unusual scenario for regulatory drafters, who as noted ordinarily are working to implement recently enacted statutes with extensive regulatory history, as well as copious public commentary and debate in the tax press, at tax conferences, etc. This situation is quite different, with much of the relevant history and legislative purposes having been formed a solid 100 years ago. As a result, the congressional purposes embedded in the statutory language that was drafted so long ago may be less readily apparent than in the case of more recent legislation. Further, stating that the statute was intended to prevent double taxation, while true, should be the beginning of the required analysis, not the end. Congress’s understanding of what double taxation meant and why it mattered should be just as relevant to the interpretation of the statute as the bare fact that it mattered, particularly when the relevant question is whether a foreign levy results in double taxation of the type that Congress sought to relieve.

Accordingly, I discuss below the implications of the initial enactment and early history of the FTC under section 901, as well as the importance of the intervening decades of change that have swirled all around the credit while the statute identifying what taxes are creditable remained a constant.13 This history provides still-relevant guidance on the interpretation and implementation of the FTC, and as shown below, the November 2020 NPRM is difficult to square with that guidance.

Before that interesting historical discussion, however, I still owe the reader some proof of my predicate that the statute has remained substantively unchanged for many decades. Fortunately the historical progression of the relevant language can be quickly recounted.14 As initially enacted in 1918, a credit against otherwise-applicable U.S. tax liability was provided to U.S. corporations for:

the amount of any income, war-profits and excess-profits taxes paid during the taxable year to any foreign country, upon income derived from sources therein, or to any possession of the United States.15

In 1921 this language was modified to refer to:

the amount of any income, war-profits, and excess-profits taxes paid during the same taxable year to any foreign country, or to any possession of the United States.16

Finally, fast-forwarding to 2021, we find that in the intervening century, Congress has made no substantive changes to the relevant language, still providing a credit for:

the amount of any income, war profits, and excess profits taxes paid or accrued during the taxable year to any foreign country or to any possession of the United States.17

A redline between the 1921 and 2021 versions of the language highlights how limited the changes have been, as noted above, adding two words and deleting one while consistently granting a credit for “the amount of any income, war profits, and excess profits taxes paid or accrued during the same taxable year to any foreign country or to any possession of the United States.”18 The three word changes highlighted above do not substantively change the definition of a creditable foreign tax. The addition of “accrued” merely addressed the accounting method for claiming the tax, and the word “same” appears to have been deleted as surplusage, or in connection with the enactment of FTC carryovers.

Given that the statute the government must implement today is a statute that Congress first enacted in 1918 and modified in 1921 and not thereafter, what Congress sought to achieve when it wrote the legislative language remains relevant to the interpretation of that language today. In the next section, I thus address the legislative purposes reflected in the initial enactment and early amendment of section 901, exploring why Congress thought it was important to prevent the double taxation of international income in the first place, and thereby identify the policy imperatives that continue to flow from those original legislative purposes.

B. Normative Tax Policy Principles Embodied in the Initial Enactment and Amendment of the FTC in the Revenue Acts of 1918 and 1921

1. Congress sought horizontal equity between taxpayers with and without foreign income, enhanced international competitiveness, and the prevention of double nontaxation.

Under the modern income tax first enacted before World War I, U.S. taxpayers were subject to tax on their worldwide income and were permitted to deduct foreign taxes paid, like any other cost of doing business.19 Given that state taxes paid have always been treated as deductible, not creditable, this treatment was not surprising, even though it meant that international income would likely be subject to double taxation. After all, double taxation by federal and state governments has long been viewed as an acceptable outcome, and the early drafters of the federal income tax, to the extent they considered the question, may have concluded that overlapping but distinct taxation by two sovereigns was simply how the fiscal cookie crumbled.

This is not to say that double taxation went unnoticed. As far back as the 13th and 14th centuries, problems of double taxation were addressed by fiscal authorities and commentators.20 While that early history is (regrettably) beyond the scope of this report, it is worth stating that the prevention of double taxation does not appear to have constituted a core principle of jurisprudential thinking, common law commentary, constitutional drafting, or any other wellspring of authority, until the development of cross-border economic activity in the 19th and 20th centuries brought increasing prominence to the problem. Although various approaches to the problem were considered starting in the mid-19th century,21 a full-fledged FTC does not appear to have been adopted anywhere until the U.S. FTC sprang fully armed from the brow of Congress in 1918.

What happened between 1913 and 1918 that inspired Congress to adopt this apparently unprecedented fiscal invention? The short answer would be World War I and Dr. Thomas S. Adams. World War I played a fundamental part because funding their war efforts required the warring countries to impose much higher rates of tax than had ordinarily been imposed during peacetime. But Adams may be the critical factor because, as has been recounted in detail elsewhere, he provided the intellectual impetus for the design, proposal, and enactment of the FTC in the Revenue Act of 1918, as well as its refinement into substantially its present form in 1921.22

The longer answer requires explaining why Congress, with the help of Treasury (through Adams), decided that taxation of the same income by two countries was something that needed to be prevented and that the right way to do so was to grant an FTC. The greatly increased rates of income tax certainly made this a more acute problem for taxpayers. After all, if the same $100 of income is subject to taxation by two sovereigns and each tax is imposed at a 10 percent rate, the resulting tax burden of no more than $20 would generally be viewed as tolerable (and somewhat more so if one of the sovereigns allowed a deduction for the other’s tax, reducing the total tax burden to $19). But if the two sovereigns’ tax rates approach or exceed 50 percent, the combined level of taxation quickly becomes ruinously intolerable.23 And that was precisely what was happening by 1918.

Thus, it is readily understandable that Treasury and Congress would decide that something must be done to prevent international double taxation at levels more problematic than had traditionally been tolerated. What is less obvious is why that something was the FTC. After all, the problem of double taxation arises because two sovereigns have competing claims to tax the income, with both claims generally being recognized as legitimate: A taxpayer’s country of residence may seek to tax on the basis of residence, while the country in which the income arises may seek to tax on the basis of source.

Most countries seek to impose residence-based taxation on the taxpayers residing within their borders, while also imposing source-based taxation on income arising within their borders. If a taxpayer resides in Country A but derives income from sources in Country B, both sovereigns may legitimately seek to tax the income based on those separate jurisdictional grounds. Thus, it was not a self-evident truth that double taxation is most appropriately resolved by causing the residence country (where the taxpayer resides) to cede taxing jurisdiction to the source country (where the income arises). In principle, it could be an equally elegant solution to have the source country cede primary taxing jurisdiction to the residence country. And indeed, in 1918 and the years immediately thereafter, many thought that would be the superior solution and advocated for that alternative approach through discussions at the International Chamber of Commerce and the League of Nations.24 But that was not the approach the United States adopted coming out of the blocks, and it was not the approach that ultimately prevailed in the ensuing international dialogue. Why then was the approach that was enacted here, and ultimately reflected in the League of Nations model tax treaty, one in which the residence country cedes primary taxing jurisdiction to the source country?

Again, the short answer would be largely because of the persuasive influence of Adams, who strenuously opposed the alternative approach of giving primary taxing jurisdiction to the residence country. But what did Adams argue, and why was he persuasive to Congress and ultimately on the international stage? And how has the understanding of this approach evolved over the past century, as the United States has repeatedly revisited the detailed operation of its FTC rules without ever altering the fundamental dynamics of its system (or the language of section 901)?

To begin with, although the FTC was first adopted in the context of a much larger war-funding bill, both proponents and opponents of the approach recognized the importance of the decision. That is, it was recognized at the time — and in fact the point seems reasonably self-evident — that by granting an FTC to its residents, the United States was ceding primary taxing jurisdiction to the country of source.25 But the apparently decisive consideration was the recognition that countries of source simply will not, as a practical matter, give up their jurisdiction to tax nonresidents on income arising within their borders. All that income is an attractive target for taxation, given both administrative convenience (the tax can more readily be collected in the country where the business activity earning the income is located) and the fact that the nonresident taxpayer is not a voter in the country and not otherwise likely to be as influential as residents. Certainly, the United States has only rarely limited its own claims to tax the U.S. income of franchiseless foreigners. Thus, Adams argued persuasively that countries of source cannot be counted on to give up the right to tax foreigners26 and that preventing double taxation must therefore become the primary responsibility of the country of residence.27

Thus far we have learned the apparent reason why, if double taxation is to be avoided, the United States decided in 1918 that the country of residence had to provide the mechanism to do so. And as for the mechanism that was chosen to provide that relief, the reason for preferring a credit over an exemption seems reasonably straightforward: Providing a credit enables the residence country to collect a residual tax whenever foreign tax is paid at a rate less than the residence country’s tax, whereas an exemption system permanently surrenders any residence country taxing rights over foreign-source income, regardless of the level of foreign tax imposed. Thus, adoption of the credit was recognized by Adams as necessary to prevent what today would be called double nontaxation. In particular, one of his four salient conclusions about double taxation “reached on the basis of wide observation and experience” was that:

the jurisdiction of domicile should usually grant an exemption only through the tax credit, by which the taxpayer is exempted at domicile only when he has proved payment of the tax in some other jurisdiction. The modern habit of living or incorporating in one jurisdiction and holding property or doing business in other jurisdictions had led to much unjust double taxation, but it has also led to a large volume of tax evasion, and the state which with a fine regard for the rights of the taxpayer takes pains to relieve double taxation, may fairly take measures to ensure that the person or property pays at least one tax.28

But we have not yet reached the more fundamental question why Congress decided that it was important to prevent double taxation in the first place. That is, what led Congress to conclude that high levels of taxation on cross-border income (even after giving a deduction for foreign taxes) were a problem that had to be fixed? Although the legislative history is limited, it supports the view, as emphasized by some of the comments on the proposed regulations, that Congress was concerned about the competitive impact that those levels of taxation would have on U.S.-based international businesses. For example, the House Ways and Means Committee report explained that under prior law, a U.S. taxpayer could “only deduct income, war, or excess profits taxes paid to a foreign country,” and stated that the imposition of high foreign tax rates in addition to U.S. taxes “places a very severe burden” on U.S. taxpayers.29 Perhaps more tellingly, during the floor debate on the bill, the chair of the Ways and Means Committee, Claude Kitchin, described the provision of an FTC as “not only a just provision, but a very wise one.”30 He went on to explain its wisdom in terms of its impact on the country’s international commerce:

It is wise from the standpoint of the commerce of the United States, of the expansion of business of the United States. There are thousands of citizens of the United States now going to South America, and they have been going for years, and we have thousands of citizens in Canada. We would discourage men from going out after commerce and business in different countries . . . if we maintained this double taxation.31

Kitchin’s explanation of the reasons for the enactment of the FTC appear in the context of a two-day debate in which he was on his feet for over nine hours, presenting detailed explanations of the provisions of the bill and of the reasons for the committee’s adoption of its provisions, and responding in detail to the questions and comments of other members.32 Thus, while the legislative history is not extensive, it does articulate a concern with the impact of unrelieved double taxation on international business activity.

In addition to reflecting a concern with the competitive impact of unrelieved double taxation on international business activity, there is also evidence suggesting that broader economic goals may have increased the priority of encouraging U.S. capital investment in businesses overseas. In particular, those investments were needed for the postwar reconstruction of European economies and to support European debtor nations’ ability to repay war-related debts to the United States.33

However, while these economic considerations were certainly an important part of the mix, there is also reason to believe that the enactment of the credit was even more fundamentally grounded in notions of horizontal equity, or the fair distribution of tax burdens among similarly situated taxpayers. As just noted, Kitchin first described the FTC as “a just provision” before going on to explain why it was also wise. Further, as the principal architect of the proposal to adopt an FTC, Adams seems to have been motivated in large measure by the normative view that double taxation would violate horizontal equity by leaving taxpayers with radically different tax burdens depending on the source of their income:

If each state utilized its full powers, multiple taxation would be rampant among those who derived any income from sources without the jurisdiction in which they were domiciled, whereas those whose income was derived within the jurisdiction in which they lived would be subject to but one tax.

That is, a taxpayer earning income only in the United States would pay tax at the U.S. rate, while another taxpayer earning the same amount of income in a foreign country would, after paying and deducting foreign taxes, face a higher net tax burden than a stay-at-home compatriot (who might be a business competitor). And ultimately this normative consideration leads to a view that double taxation is not just bad tax policy, but an injustice:

Equity in taxation is not always clear and plain. I see my own equities through a telescope, but the other fellow’s equities through a microscope. This is true of me and you as well as the legislator. But the worst forms of double taxation are clearly and plainly inequitable.34

Indeed, the Supreme Court had recognized the injustice of double taxation long before Adams came along:

Justice requires that the burdens of government shall, as far as is practicable, be laid equally on all, and if property is taxed once in one way, it would ordinarily be wrong to tax it again in another way when the burden of both taxes falls on the same person. Sometimes tax laws have that effect, but if they do it is because the legislature has unmistakably so enacted. All presumptions are against such an imposition.35

Further, Adams’s writings suggest that this normative aspect of the proposed FTC persuaded Congress to adopt it with an alacrity that surprised even him. In a paper presented 10 years later, Adams first explained why, “both theoretically and practically,” a taxpayer’s country of residence must “inevitably” bear the responsibility to prevent double taxation:

The explanation is simple. Every state insists upon taxing the non-resident alien who derives income from sources within that country, and rightly so, or at least inevitably so. . . . But [the average state] refuses to recognize when one of its own citizens or nationals gets income from a foreign source that he inevitably will be taxed abroad. As a necessary corollary of the principle of taxing at source or origin which it has adopted, the home state owes an exemption of some kind to its own citizen or resident who derives income from a foreign source or sources.36

He then explained the connection between this equitable principle and the enactment of the FTC in the 1918 act:

In the midst of the war, when the financial burden upon the United States was greater than it had ever been, I proposed to the Congress that we should recognize the equities which I have just noted, by including in the federal income tax the so-called credit for foreign taxes paid.37

But when Adams made that proposal to “recognize the equities,” he was not optimistic about its success:

I had no notion, ladies and gentlemen, when I proposed it, that it would ever receive serious consideration. I expected it to be turned down with the reply which I have received so often from legislative committees: “Oh, yes, Doctor, that is pretty good, but the finances won’t permit it.” But to my surprise, the credit for foreign taxes was accepted and approved, because it touched the equitable chord or sense, and because double taxation under the heavy war rates might not only cause injustice but the actual bankruptcy of the taxpayer.38

Thus, Adams attributed the enactment of the FTC to the fact that it resonated with Congress’s “equitable chord or sense.” And finally, he noted that while it may often be difficult to achieve equity in tax legislation, the horizontal inequity arising from double taxation is likely to inspire a legislative response:

Now, according to my experience, there is no means or instrument by which to get recognition of the equitable, disinterested, scientific point of view in taxation as potent as this argument based on the desirability of eliminating double taxation. There is something in the legislative mind which recognizes that if one taxpayer is being taxed twice while the majority of men similarly situated are being taxed only once, by the same tax, something wrong or inequitable is being done which, other things being equal, the legislator should correct if he can.39

Accordingly, while the relevant legislative history is limited, there is reason to believe that Congress sought to relieve double taxation because it accepted Treasury’s view (as presented by Adams) that double taxation would violate horizontal equity and work a substantive injustice. That original legislative purpose should continue to be taken into account in the implementation of the statute. A crabbed mechanical reading of section 901 would be at odds with a statutory purpose founded in normative considerations of equitable treatment and the prevention of injustice and with the recognition that countries of residence must bear the primary responsibility for preventing double taxation, given the inevitability of source-country taxation.40 Further, considering those normative purposes in the context of the other broad goals that motivated Congress, such as protecting the competitive position of U.S.-based international businesses, reinforces the conclusion that the FTC should be implemented consistently with its big-picture goals. Although narrow readings may be justifiable for narrowly drafted technical statutes, when the language of the statute and its history both demonstrate a broad purpose to prevent injustice, a narrow reading that promotes injustice should not be preferred.

2. To protect U.S. taxation of U.S.-source income, Congress repealed a jurisdictional test and instead placed full reliance on an FTC limitation.

In 1921, just two years after the surprisingly sudden enactment of the FTC, Congress returned to the subject to enact the only substantive modification of the scope of the creditability rule that it has ever seen fit to adopt.41 This modification, whose principal architect and advocate was again Adams on behalf of the Treasury Department, brought section 901 to substantially its current form.42 The substantive changes enacted in 1921 are notable, not only because they were the last changes to the statute but also because they strongly suggest that the jurisdictional nexus rule in the November 2020 NPRM may lack authority and is in all events deeply misguided. This follows from the simple fact that in 1921 Congress deleted the jurisdictional nexus rule that had appeared in the original FTC enacted in the 1918 act.

The FTC as first enacted had applied to taxes imposed by a foreign country “upon income derived from sources therein.”43 Under that formulation, a foreign tax was creditable only to the extent imposed on income arising in that country. Thus, the initial formulation of the U.S. FTC included the self-contained equivalent of a jurisdictional nexus rule — although obviously not referred to using that term, its effect was the same because it would have denied credits for extraterritorial foreign taxes.

It is therefore notable that when Congress undertook its first and last substantial modification of the newly enacted FTC rule just two years later, it deleted the reference to income derived from sources within the country imposing the tax.44 Given that Congress in 1921 deleted a statutory requirement that conditioned the creditability of a foreign tax on whether the foreign country was imposing that tax on income it had appropriate jurisdiction to tax, the scope of Treasury’s authority to reimpose a similar requirement by regulatory fiat seems doubtful at best.

Further, this conclusion is reinforced by the fact that Congress did not simply delete the jurisdictional nexus requirement, but rather replaced it with a source-based FTC limitation.45 That mechanism makes it clear that Congress did not delete the 1918 statute’s jurisdictional nexus rule because it was indifferent to whether a foreign country might overreach in its assertion of taxing jurisdiction. To the contrary, Congress recognized that such overreaching could operate to the detriment of the United States’ ability to collect its own residence-based tax on income arising in the United States, even under the jurisdictional nexus rule adopted in 1918. Therefore, the issue of the proper scope of foreign taxing jurisdiction had to be addressed by a distinct and more effective approach in the revised, 1921 version of the FTC. The FTC limitation enacted at that time thus took the place of the jurisdictional nexus rule and better ensured that the U.S. tax on U.S. income could not be reduced by excess credits generated by overbroad assertions of taxing jurisdiction by foreign governments.

By choosing an FTC limitation mechanism with an overall limitation, Congress in 1921 demonstrated no concern about whether a foreign government might choose to tax any income that the United States might view as outside the scope of that country’s taxing jurisdiction. Rather, the FTC limitation as adopted in 1921 readily permitted cross-crediting between high- and low-taxed foreign income, which meant that even taxes imposed by a foreign country on income that it arguably had no jurisdiction to tax could be credited for U.S. tax purposes — subject only to the taxpayer having sufficient foreign-source income and therefore sufficient FTC limitation to support a credit for that tax.

Indeed, Congress’s decision not to condition the FTC on a jurisdictional inquiry is confirmed by the fact that although the limitation is source-driven, the statute was not drafted to deny the creditability of a foreign tax imposed on income that the United States considers to be U.S.-source income. In that case, no FTC limitation would be available for the U.S. income taxed by the foreign country, but the foreign tax would still be recognized as a potentially creditable income tax. Thus, if the taxpayer had other foreign-source income in the relevant category taxed at rates lower than the U.S. rate (and thus giving rise to excess limitation), the foreign tax on the U.S.-source income could be fully credited, even though from a U.S. perspective, the foreign tax was imposed extraterritorially. This outcome may in part reflect early congressional recognition that sourcing rules for cross-border income are not revealed wisdom delivered on stone tablets and, in fact, vary substantially from country to country.46

Accordingly, it is notable that in 1921 Congress not only deleted the jurisdictional nexus rule it had adopted two years before, but it also replaced that rule with an FTC limitation that made clear that Congress sought to prevent FTCs from offsetting U.S. tax on U.S. income but not to police the limits of foreign countries’ taxing jurisdiction over other income (including income that was U.S.-sourced from a U.S. standpoint). Certainly nothing in that history suggests that a foreign tax is not an income tax merely because it was imposed on income that is viewed as extraterritorial to the jurisdiction imposing the tax. Thus, as long as a foreign tax has been imposed on income, the basis for the foreign country’s assertion of taxing jurisdiction over that income has never been relevant to the creditability analysis. This conclusion follows directly from the language of section 901’s reference to an “income” tax, as the jurisdictional scope of a tax is logically irrelevant to whether it is a tax imposed on income.

Further, the proper sources, bases, and scope of national taxing jurisdiction over international income were a matter of considerable uncertainty and active controversy at that time, much like today.47 Thus, if Congress had sought to hinge the creditability of a foreign tax on U.S. agreement with the scope of taxing jurisdiction asserted by a foreign government, it would presumably have said so in the form of a rule (like the 1918 rule) that focused on the country’s jurisdiction to tax. Congress chose instead to rely on the source-based limitation to protect its own tax on U.S.-source income. It may well have done so in recognition of the futility of trying to become the world’s arbiter of the scope of taxing jurisdiction asserted by scores of sovereign nations, each with plenary authority to determine what income it sought to tax. The far more practical approach that Congress adopted in 1921 was to cede primary taxing jurisdiction over all non-U.S.-source income to all non-U.S. sovereigns. Rather than try to condition the creditability of foreign taxes based on shifting concepts of jurisdictional nexus, Congress cut that knot by providing an FTC limitation driven only by the relative amounts of U.S.- and foreign-source income earned by the taxpayer.

The subsequent flip-flops in the design of the FTC limitation do not affect this analysis. Congress has periodically careened among overall, per-country, income-category, and other limitations on the FTC,48 but none of these changes have sought to revert to a jurisdictional analysis. Rather, each iteration of the limitation has been driven solely by the amount of income in each then-applicable limitation category, with the limitation determined by applying the U.S. tax rate to the amount of income found in that category. Thus, the post-1921 history of the FTC limitation provides no basis for questioning the fundamental conclusion that in 1921 Congress rejected an approach to the FTC that would depend on substantively evaluating the basis for a foreign country’s assertion of taxing jurisdiction over income and instead adopted a far simpler mathematical solution that accepts the amount of the foreign income tax as determined by the foreign sovereign but limits the credit to the amount of U.S. income tax on the relevant category of foreign income, as determined by the United States. By proposing to adopt the approach that Congress rejected in 1921, the November 2020 NPRM ignores dispositive history and raises the question whether Treasury has the authority to promulgate regulations adopting a limitation that not only does not appear in the statute but was, in fact, repealed by Congress in its only significant change to the definition of what constitutes a creditable tax in the 103-year history of the FTC.

C. Decades of Legislative and Regulatory Actions Have Reinforced the Original Tax Policy Decisions Reflected in Section 901

1. Many dogs, no barking; not even a growl.

The conclusion that the early history of the FTC and its limitation do not support the principal changes advanced in the November 2020 proposed regulations is reinforced by a century of congressional activity in relation to these rules. Congress has returned again and again (and again) to revise the FTC limitation — no less frequently than once a generation, and sometimes much more frequently than that, even revising the rules biennially in some periods. It is thus clear that Congress (frequently with extensive help from Treasury) has paid close attention to the proper scope and operation of the FTC and has repeatedly taken up the drafting pen to tweak, re-tweak, and re-re-tweak (etc.) the operation of those rules.49 Congress has thus had repeated opportunities to revisit section 901’s simple formulation providing credits for income taxes paid to foreign countries. Yet despite those many years of close focus and frequent amendment, Congress has never chosen to revisit the basic standard of creditability set forth in section 901, let alone reverse its 1921 decision by switching back to a rule concerning nexus. This strongly suggests that Congress has never questioned its 1921 decision to use the limitation as the means of protecting U.S. tax on U.S. income without seeking to evaluate the merits of any sovereign’s claim of jurisdiction to tax specific income.

While the absence of congressional activity may ordinarily be a thin reed on which to base a statutory interpretation, Congress leaving an enacted provision unchanged for 100 years should comfortably support an inference that Congress has not sought to change that rule — particularly when contrasted with frequent and extensive change in closely related provisions. Indeed, the fact that Congress hasn’t changed a statute that it hasn’t changed seems less an inference than a self-evident proposition approaching tautology. Thus, the absence of changes to the definition of a creditable tax in section 901, together with frequent changes in the FTC limitation, strongly supports the conclusion that despite its antiquity, the foreign tax creditability standard enacted by Congress in 1921 remains unchanged today.50

2. Enactment of section 903 in 1942.

In 1942 Congress enacted the only significant post-1921 change to the standards for creditability of foreign taxes, not quite 25 years after the initial adoption of the FTC. Notably, Congress added a new provision that, far from narrowing the scope of what constitutes an income tax, expanded the credit to permit some non-income taxes to be treated as creditable income taxes for purposes of section 901. As explained by the Senate Finance Committee, Congress acted to provide a credit for in-lieu taxes because the term “income tax” had been narrowly interpreted to encompass only “a concept of income tax rather closely related to our own”:

In the interpretation of the term “income tax,” the Commissioner, the Board [of Tax Appeals], and the courts have consistently adhered to a concept of income tax rather closely related to our own, and if such foreign tax was not imposed upon a basis corresponding approximately to net income it was not recognized as a basis for such credit. Thus if a foreign country in imposing income taxation authorized, for reasons growing out of the administrative difficulties of determining net income or taxable basis within that country, a United States domestic corporation doing business in such country to pay a tax in lieu of such income tax but measured, for example, by gross income, gross sales or a number of units produced within the country, such tax has not heretofore been recognized as a basis for a credit. Your committee has deemed it desirable to extend the scope of this section.51

In other words, when previously confronted with novel foreign taxes that were not clearly within the scope of a traditional income tax, Congress prioritized the policy of preventing double taxation rather than insisting on a strict technical reading of the term “income tax.” Congress thus chose to broaden the term to include foreign levies that are not, um, income taxes. In particular, it broadened the creditability of foreign taxes to include taxes calculated under computational methods (such as gross income or gross receipts taxes) that had not been found to meet the narrow U.S. definition of an income tax, if those taxes were imposed in lieu of a foreign country’s otherwise-applicable income tax.

The enactment of section 903 thus addressed a congressional concern that U.S. taxpayers were being subjected to both U.S. and foreign taxes on the same economic activity, but the existing FTC mechanism was ineffective to prevent double taxation because the foreign tax in question was not “rather closely related to our own” income tax.52 The fact that Congress acted to prevent double taxation, despite the revenue pressures of World War II, shows that it was more concerned about ensuring that U.S. taxpayers were protected against international double taxation than it was with the niceties of limiting FTCs to foreign taxes imposed upon net income, versus other forms of taxation commonly imposed by foreign governments that were equally burdensome to cross-border economic activity. Congress thus clearly viewed the imposition of those taxes on a basis other than net income as resulting in double taxation that the U.S. FTC should operate to prevent.

Accordingly, the enactment of section 903 reinforces the view that the FTC has consistently been intended by Congress to relieve double taxation of cross-border economic activity, based on some combination of normative and economic rationales, in a manner that supports a broad reading of the credit rules. Indeed, one view of this history is that Congress acted in 1942 to counter the overly narrow IRS interpretation of the term “income tax,” choosing to short-circuit disputes over the income tax status of particular foreign taxes by broadening the rules to focus not on the income tax status of a tax but rather on whether that tax is imposed in lieu of an income tax otherwise imposed by the foreign country. In combination with the separate levy rules that call for the creditability of foreign taxes to be analyzed separately when their taxable bases differ, this rule broadly ensured that if any country generally imposed an income tax on its residents but imposed some different tax on a U.S. taxpayer that did not otherwise qualify as an income tax, double taxation of the taxpayer would still be prevented by permitting that non-income tax to be treated as an income tax for section 901 purposes.

The legislative history sheds no light on the purpose of the requirement that the foreign tax be imposed in lieu of an otherwise generally applicable income tax. That is, the history does not address why the foreign country must generally impose an income tax, rather than permitting a credit if a country imposes a non-income tax on all taxpayers in lieu of imposing any form of income tax. One potential reason for the limitation is that the common factual pattern at that time (and today) was for many countries to impose gross-basis withholding taxes on nonresidents as a more readily collected alternative to the regular income taxes they impose on their residents. Given that the creditability of those taxes had proven controversial, Congress may have thought it sufficient to resolve the problem at hand without seeking to ensure the creditability of non-income taxes imposed in other circumstances. Another possibility is that Congress considered the requirement of an underlying income tax as a means of limiting the scope and amount of credits granted for non-income taxes, because the amount of a tax imposed in lieu of an income tax could generally be presumed to bear some relationship to the amount of the income tax.53

It could be argued that requiring a foreign country generally to impose an income tax is misguided because that requirement can readily lead to double taxation in the case of a jurisdiction that for whatever reason chooses to impose a differently computed tax (such as a low-rate gross income tax) on all taxpayers. Further, that tax would seem unlikely to be imposed at confiscatory rates if it applies to all taxpayers, so the absence of an underlying income tax imposed on resident taxpayers seems unlikely to result in alarmingly high rates of in-lieu taxation. But given that the requirement of a generally imposed income tax is embedded in the code itself, I leave the wisdom of that requirement for another day and focus here on the November 2020 NPRM’s proposed implementation of the requirement through the newly aggressive substitution requirement discussed in Section III.C below.

Given the self-evident liberalizing purpose of section 903, and the absence of any subsequent legislative changes to the rule (which has stood unchanged for its full 79-year history and thus comes close to matching the immutability of section 901 itself), the November 2020 NPRM’s changes to the implementing regulations discussed below seem out of step with the purpose of the statute. Two aspects of the proposed rules that are discussed in Section III.C below will make it difficult for many heretofore creditable in-lieu taxes to continue to qualify under section 903. In particular, the overall restriction of the definition of an income tax for which an in-lieu tax must substitute and the addition of a so-called non-duplication rule would severely restrict the creditability of many foreign taxes that, as a matter of plain English, are imposed in lieu of income taxes.

3. Enactment of subpart F in 1962.

Congress enacted subpart F in 1962 in response to proposals advanced by the Kennedy administration in 1961. While the administration’s proposals would have imposed current U.S. tax on the earnings of all U.S.-controlled foreign corporations, the enacted compromise reflected in subpart F imposed current U.S. tax only on specific categories of passive or otherwise movable income while generally leaving the “deferral privilege” in place for active foreign business income.54

By imposing immediate U.S. tax on several categories of previously deferrable income, subpart F in principle increased the importance of the FTC as the means to avoid double taxation of that income. While much of the affected income may have been subject to limited foreign tax, the design of subpart F included an indirect FTC based on section 902, ensuring protection against such double taxation. In the context of designing the credit for foreign income taxes imposed on CFCs, Congress during 1961 and 1962 once again had an opportunity to consider the scope of creditable foreign taxes, as part of a large and ambitious revenue bill making numerous substantial changes to the U.S. international tax regime. Further, the competitive impact of those changes was hotly debated during the legislative process and led to the adoption of the compromise approach reflected in subpart F rather than the more trenchant changes initially proposed by the Kennedy administration. Thus, Congress once again did nothing to alter existing law, just as it did nothing in that regard every single time it rewrote the FTC limitation rules. So in the congressional consideration of subpart F, we find yet another soundly sleeping hound, suggesting once again that Congress has never considered modifying the basic creditability standard that it adopted in the 1918-1921 period by referring simply to foreign income taxes.

4. Development of detailed regulations under section 901, 1979-1983.

The history of administrative guidance under section 901 likewise provides little support for the November 2020 NPRM’s radical proposed changes to the definition of a creditable income tax. For over half a century after the enactment of section 901, Treasury and the IRS saw no need to issue any regulations further elaborating on the statute’s straightforward reference to income taxes. The impetus for issuing those regulations arose only after the oil price shocks of the 1970s and concerns emerging at that time about the need for further guidance to distinguish between actual foreign income taxes and amounts paid to foreign sovereigns as disguised royalties for the right to extract minerals. Addressing the economic benefits obtained by those “dual capacity taxpayers” became the central focus of the regulatory guidance, which necessarily placed those issues into the broader context of guidance defining a foreign income tax.

The ensuing regulatory effort was highly controversial and concomitantly convoluted, as reflected in multiple reg packages starting with an initial NPRM in 1979 and continuing through reproposed regulations and temporary regulations issued in 1980 and re-reproposed regulations issued in 1983, and concluding with final regulations promulgated later in 1983.55 However, apart from the new regime applicable to dual capacity taxpayers, once the tumult and the shouting were finally over, the final regulations’ other guidance on the definition of a creditable foreign tax largely put into regulatory form concepts and principles that had been developed through judicial and administrative rulings over the previous 60 years. In particular, the final regulations reflected the courts’ long-standing substance-based approach to determining whether a foreign levy was an income tax in the U.S. sense, with multiple rules looking to the “predominant character” of a foreign tax and other rules building in additional flexibility.56

Given the scope of the regulatory effort previously undertaken, and the extent to which the resulting guidance reflected principles developed in prior decades, the 2020 NPRM departs from that guidance in some surprisingly radical ways. For example, as discussed in more detail below, the 2020 NPRM moves forcefully toward a formalistic analysis rather than a substantive one. Further, in some cases, proposals set forth in the November 2020 NPRM appear to be based on rules that were proposed during the prior regulatory effort and rejected at that time based on the comments received. As also discussed further below, it is unclear from the preamble to what extent the reasons for the prior rejection of these proposals were considered in renewing them 40 years later.57

5. Enactment of the TCJA in 2017.

The TCJA in 2017 represented the single most fundamental set of changes to the U.S. international tax system since its enactment early in the 20th century. Somehow adopting simultaneously both a purported participation exemption system for CFC earnings and rules imposing immediate reduced-rate U.S. taxation on most of those earnings, the TCJA ushered in an era of profound change and fundamental obscurity. Yet amid all that change and confusion, one provision remained notably untouched, yet again. You know by now that the island of calm amid the coprostorm of changes was the language defining creditable foreign taxes under section 901, which remained untouched while everything around it collapsed into incoherence with embarrassing acronyms.

Despite the scope of TCJA changes that gave Congress yet another opportunity to revisit the scope of the taxes that are treated as creditable under section 901, Congress once again showed no interest in revising section 901, just as it had failed to change those standards when it enacted subpart F in 1962 and when it made multitudinous amendments to the FTC limitation between 1932 and 2010. By imposing reduced-rate worldwide taxation on U.S. shareholders — taxing them under the GILTI regime on virtually all the income of a CFC, thereby ending the indefinite deferral of U.S. tax on active foreign earnings that had previously applied — while retaining full-rate current U.S. taxation of subpart F income, the TCJA structurally heightened the importance of the FTC. And Congress once again clearly focused on the operation of the FTC rules, not only repealing the indirect credit under section 902 (as no longer needed because CFC dividends are in principle exempt) but also enacting a new indirect credit rule under section 960(d) that “haircuts” the amount of foreign taxes deemed paid on GILTI inclusions. But once again, Congress saw no need to revise the definition of a creditable tax under section 901.

Nonetheless, despite the century of placid waters under section 901 that we have just been paddling through, the November 2020 NPRM proposes a much stormier course. We thus turn next to a detailed review of the proposed changes and will find that they are difficult to square with the legislative purposes reflected in the long history of section 901.

III. Many Changes Proposed by the November 2020 NPRM Are Inconsistent With the History and Purposes of Section 901

A. The Jurisdictional Nexus Requirement

1. The proposed requirement.

The November 2020 NPRM would modify the definition of a net income tax by adding to the long-standing net gain requirement a second definitional requirement.58 This test would provide that a foreign country’s tax on nonresidents must meet any one of the following three jurisdictional nexus requirements:

  • Activities nexus: Taxable income under foreign law must be limited to that attributable “under reasonable principles” to a nonresident’s activities in that country (including the nonresident’s “functions, assets, and risks” located in-country). The rule is intended to incorporate principles similar to those for determining effectively connected income under section 864(c); the location of customers, users, or other destination-based criterion must not be a “significant factor” in the operation of the tax. For example, a tax on electronic services provided by a nonresident company to users located in the taxing country, when the tax base is determined based on the percentage of the nonresident company’s users located in the taxing country, would lack jurisdictional nexus under this rule.

  • Source of income nexus: Sourcing rules must be “reasonably similar” to the sourcing rules that apply for U.S. federal income tax purposes, including that services must be sourced based on where they are performed, not based on the location of the service recipient.

  • Situs of property nexus: The foreign tax must be imposed only on dispositions of real property located in the foreign country or on movable property forming part of the business property of a taxable presence in the foreign country. This includes dispositions of interests in a company or entity holding that property.

In addition to the three jurisdictional nexus requirements, the country’s taxes on resident taxpayers (generally based on place of incorporation or management) must be imposed in conformity with arm’s-length transfer pricing principles. Thus, any allocation of income to or from a related entity under transfer pricing rules must be based on transfer pricing rules reflecting the arm’s-length standard. And again, the location of customers, users, or other similar destination-based criteria may not be a significant factor in the operation of the foreign tax. Foreign tax regimes that impose tax on worldwide income of residents (such as through a CFC regime) do not violate the nexus requirement. However, taxes paid by a locally regarded entity would not be creditable if the foreign country’s transfer pricing rules deviate from the traditional view of the arm’s-length principle, a concern that has been raised historically regarding some jurisdictions with significant U.S. investment.59

The proposed regulations would also add a jurisdictional nexus requirement to the definition of a creditable in-lieu tax under section 903.60

2. Apparent purpose and likely double taxation impact of the rule.

The preamble to the November 2020 NPRM discusses the jurisdictional nexus requirement at length, emphasizing the need to impose the nexus test as a response to digital services taxes and other “novel extraterritorial taxes” that are the subject of ongoing discussions at the OECD.61 Digital commerce has, of course, enabled some companies to derive substantial profits from international business activities without creating a taxable nexus in many countries under historical international norms.

Those norms, developed under the auspices of the League of Nations and the International Chamber of Commerce in the first half of the 20th century (with the extensive participation of our old friend Adams), provide a framework for dividing taxing jurisdiction between a source country and a residence country.62 Those norms generally cede primary taxing jurisdiction to a source country if the taxpayer maintains some form of physical presence in that country. That approach remained stable for many decades. Because most cross-border trade involved transfers of physical goods or services physically provided by human beings, jurisdictional determinations based on physical presence proved generally workable.63

However, the consensus around the norms has been challenged by the advent of significant forms of commerce with little or nothing in the way of physical attributes. Given the scope of that economic activity, and concerns about it falling through the mesh of a tax net that was not woven finely enough to catch many of the dollars flowing through modern cross-border trade, several countries have indeed turned to novel approaches for taxing market activity that does not give rise to any form of physical presence.

If many of the taxpayers swimming through the holes in the traditional nexus net are U.S.-based companies, the U.S. fisc could well be a net loser if (1) significant new taxing jurisdiction is asserted over nonphysical cross-border transactions; and (2) an FTC arises under section 901 for those new taxes. Thus, at one level the jurisdictional nexus proposal is presumably intended to defend the U.S. fisc by rejecting the creditability of those taxes, rationalized under a view that they violate traditional norms and understandings regarding source-versus-residence taxing jurisdiction and therefore are not income taxes.64

And at another level, by confronting U.S.-based companies with the specter of significant double taxation of cross-border income, the proposal may have been intended to put pressure on foreign governments, based either on lobbying activities by those companies or on foreign governmental concerns about the potential impact on their economies if U.S. companies quit their markets under the threat of unrelieved double taxation. Thus, the proposal may have made U.S.-based multinational corporations reluctant participants in a game of fiscal chicken. It raises three distinct concerns:

  • First, if U.S. companies actually suffer significant double taxation upon finalization of the proposed regulations, this may ultimately have adverse economic effects in the United States by impairing the competitive positions of those companies in global markets.

  • Second, this line of reasoning assumes, without analysis, that the traditional division between source and residence taxing jurisdictions is woven into the definition of an income tax within the meaning of section 901. As discussed above, the history of the provision suggests that it is not, so it is unclear whether the language of section 901 read in light of its history can support conditioning the creditability of a foreign tax on the foreign country adhering to a traditional view of the jurisdiction to tax cross-border income. It is far from clear how a country’s view of the scope of its taxing jurisdiction affects whether it is actually imposing an income tax, either as a matter of plain English or under any of the long-applicable criteria that have been used to judge whether a foreign tax is an income tax. The fact that a tax is imposed on income that we consider extraterritorial to the country imposing the tax seems logically unrelated to whether the tax is imposed on the basis of realization, gross receipts, etc.65

  • And third, the double taxation flowing from the jurisdictional nexus rule would by no means be limited to digital double taxation. As one example, the rule would also double tax any company that operates in a jurisdiction that does not apply traditional arm’s-length transfer pricing methods.66 Accordingly, while the jurisdictional nexus rule may have started out with a focus on the creditability of DSTs and other new market-based taxing regimes, it ended up casting a much broader net.

Under the jurisdictional nexus rule, then, it seems highly likely that many so-called digital taxes, along with other taxes computed in nontraditional ways, would become non-creditable. Further, while pre-2018 one might have thought that the indefinite deferral of U.S. taxation of CFC earnings could minimize the real incidence of double taxation flowing from the non-creditability of all those foreign taxes, the changes wrought by the TCJA mean that double taxation would happen immediately and endemically. In particular, the U.S. adoption of a quasi-worldwide tax system (albeit mysteriously coupled with a narrowly applicable participation exemption) would mean that the foreign profits of many U.S. companies would be subject to both foreign and U.S. taxes. The proposed regulations try to define that problem away by simply maintaining that the foreign taxes are not income taxes, and therefore no double income taxation has occurred.

But that semantic game is hardly satisfactory. First, most of those foreign digital taxes will in fact be imposed on a company’s income, so it is only the most arid of formalisms to claim that double taxation of income is not occurring. And second, such a result is antithetical to the policies that motivated Congress when it enacted a credit for foreign income taxes in 1918 and then for good measure came back and expanded that credit to non-income taxes in 1942. While Congress could rewrite the FTC today in response to digital taxes, the authority of Treasury and the IRS to reinterpret the statutory language is constrained both by the plain meaning of that language and by the legislative history and purpose reflected in its enactment as discussed above. And on those bases, it is far from clear that the double taxation that would arise under the jurisdictional nexus rule is consistent with either the language or purpose of section 901.

As a coda to the strangeness of the proposed regulations’ apparent equanimity about causing substantial cross-border double taxation, it is worth noting the preamble’s observation that if a digital tax deal is successfully negotiated at the OECD, the proposed rules would need to be revisited.67 But if the jurisdictional nexus rule is properly part of section 901’s definition of an income tax, it is difficult to see how logrolling at the OECD over source and residence taxing jurisdiction could alter that definition. Either jurisdictional nexus is part of the normative definition or it isn’t, and the preamble’s hint that it could be jettisoned if a deal is struck could be read as a tacit admission that it isn’t.

Further, even if the strategy underlying the November 2020 NPRM is successfully executed, it is not clear how the U.S. fisc would come out ahead. Although it might have temporarily forestalled the creditability of a handful of early DSTs that grabbed a modest amount of tax revenue, Treasury would have agreed to a standard for DSTs that the entire world could adopt, ultimately resulting in a much larger number of countries imposing OECD-blessed DSTs and those DSTs all being treated as creditable (assuming that Treasury would at that point drop its jurisdictional nexus innovation as hinted by the preamble).68

To wrap up our discussion of the proposed jurisdictional nexus requirement, I next address its interaction with tax treaty rules that confirm the creditability of treaty-covered foreign taxes.

3. Interaction of the jurisdictional nexus requirement with tax treaty creditability rules.

In its discussion of the jurisdictional nexus requirement, the preamble to the November 2020 NPRM states comfortingly that:

the proposed regulations, when finalized, would not affect the application of existing income tax treaties to which the United States is a party with respect to covered taxes (including any specifically identified taxes) that are creditable under the treaty.69

Treasury and the IRS have thus acknowledged that while a later-enacted statute can, under U.S. law, override a treaty, the same thing is not true of a later-promulgated regulation.70 Accordingly, a U.S. company that pays a treaty-covered tax will have nothing to worry about — at least not until the treaty is renegotiated and Treasury in its negotiating strategy (or the Senate in the process of advising and consenting on ratification) objects to continuing to allow credits for any taxes that are not otherwise creditable under then-applicable U.S. rules.

But even short of a treaty being renegotiated, there may be less than meets the eye to the treaty protection referenced by the preamble. To begin with, it is unclear whether any of the “extraterritorial” digital taxes targeted by the rule will be covered taxes under a treaty. Further, taxes that flunk the resident-taxpayer prong of the jurisdictional nexus rule (that is, taxes based on income allocation principles other than the arm’s-length standard) may also be unlikely to be covered taxes. Thus, treaty protection may well be unavailable for taxes whose creditability is denied under the jurisdictional nexus rule — although it may instead be available for more traditional foreign taxes that meet jurisdictional nexus but flunk the substantially tighter net gain requirement discussed in Section III.C below.

Second, given that treaty creditability rules were historically viewed as merely confirming the availability of FTCs otherwise provided by section 901, the treaties vary in the specificity of their references to covered foreign taxes, and there is little guidance addressing the application of these treaty rules. For example, if a treaty specifies that it covers a particular tax, but the treaty partner substantially modifies that tax, it might be questioned whether the treaty is sufficiently ambulatory to cover the revised tax. Given its proclivities, the IRS might well argue that the revised tax is no longer the same tax that the United States agreed was covered and that it is therefore no longer creditable — particularly if the foreign tax is changed in a manner that the proposed regulations may treat as problematic from a nexus standpoint.

Third, treaties generally address only the creditability of foreign taxes directly paid by a U.S. taxpayer and thus do not clearly address the creditability of taxes paid by foreign affiliates of U.S. taxpayers that might be creditable as indirect or deemed paid taxes under section 960. The preamble language seems to carefully preserve this distinction by referring only to covered taxes that are “creditable under the treaty.” Thus, it may be that foreign taxes paid by CFCs will not benefit from any treaty protection, even if the tax paid is otherwise a covered tax under the U.S. tax treaty with the CFC’s country of residence. Such taxes paid by CFCs constitute a substantial portion of the foreign taxes paid by U.S.-based multinationals.71

Finally, because the November 2020 NPRM would exponentially increase the level of uncertainty that taxpayers face regarding the creditability of any particular foreign tax, it would correspondingly increase the motivation for treaty partners to seek language ensuring the creditability of their taxes. That is, treaty partners may cease to be satisfied with a somewhat perfunctory confirmation of code creditability and seek to negotiate treaty provisions that provide more explicit confirmation of creditability. This dynamic would not necessarily be problematic per se, but it could slow the process of negotiating treaties, and given the need for Senate advice and consent on the ratification of any new tax treaties or protocols, it could add to the pressure on an already dysfunctional U.S. tax treaty ratification process.

I turn now from the bombshell jurisdictional nexus limit on creditability to the much more granular set of proposed changes to the rules of the net gain requirement. While these changes may be more like the proverbial death by a thousand cuts, they could prove just as fatal to the creditability of many foreign taxes.

B. Principal Changes to the Net Gain Requirement

Apart from adding the new jurisdictional nexus requirement, the November 2020 NPRM would substantially modify the current final regulations’ tests that define an income tax. Those regulations require that the predominant character of a foreign tax be that of income tax in the U.S. sense; they define the term “income tax” by reference to whether the foreign tax is imposed on “net gain”; and they test that net gain requirement by reference to whether the foreign tax has three features characteristic of an income tax: (1) imposing tax when income has been realized; (2) starting from actual gross receipts; and (3) allowing the deduction of relevant costs to reach net income. All those rules would operate very differently under the proposed regulations. Some of the principal changes are summarized below.

1. Deletions of predominant character rules affect all three tests under the net gain rule.

As noted, the current final regulations require that the predominant character of a foreign tax be that of an income tax in the U.S. sense, which, under the net gain requirement, is a tax that is “likely to reach net gain in the normal circumstances in which it applies.”72 The net gain requirement is in turn tested by reference to whether the foreign tax satisfies the realization, gross receipts, and net income requirements, “judged on the basis of its predominant character.”73 Those three detailed requirements in turn reiterate the predominant character standard. In fact, each rule does so multiple times.

The current regulations under reg. section 1.901-2 thus refer to the predominant character of a foreign tax a total of 13 times. In combination with the regulations’ other similarly softening language, such as the reference to the foreign tax being “likely to reach net gain in the normal circumstances in which it applies,”74 these references build in ample flexibility to adopt a reality-based view of the operation of the foreign tax and, importantly, one that decides creditability based on the foreign tax’s overall economic effect rather than on whether it punctiliously avoids committing any foot faults under several tightly drawn requirements. The current regulations’ focus on the predominant character of a tax, and on the normal circumstances in which it applies, reflects numerous judicial decisions requiring that the creditability of a foreign tax be determined based on its substantive similarity to an income tax in the U.S. sense.75

The November 2020 NPRM would delete all 13 references to the predominant character of a foreign tax, as well as the net gain rule’s reference to normal circumstances, and instead require a formalistic analysis based on the letter of foreign law as it applies in the abstract. Thus, under the revised tests, even if a taxpayer pays a foreign tax that is demonstrably imposed on net income, credits may be denied based on a formal difference between the code and the foreign law affecting the realization, gross receipts, or net income/cost recovery requirement.

2. Proposed changes to the realization requirement.

The November 2020 NPRM would substantially revise the drafting of the current final regulations’ realization requirement, but the substantive impact of the redrafting would be relatively limited compared with the scope of the changes to the gross receipts and net income/cost recovery rules. The realization requirement in the existing final regulations already provides as its primary test that the foreign tax is “imposed upon or subsequent to the occurrence of events (‘realization events’) that would result in the realization of income under the income tax provisions of the Internal Revenue Code,” plus enumerated “prerealization events” (such as mark-to-market rules).76

The principal changes that the proposed regulations would make to the realization requirement are thus limited to the deletion of the “predominant character” references throughout the rule. In this portion of the regulations, a total of five references to the predominant character of the foreign tax would be vaporized.

3. Proposed changes to the gross receipts requirement.

a. Summary of proposed changes.

The current final regulations generally provide that a foreign tax “satisfies the gross receipts requirement if, judged on the basis of its predominant character, it is imposed on the basis of gross receipts.”77 However, they also allow gross receipts to be estimated using an alternative method “that is likely to produce an amount that is not greater than fair market value.” The regulations further provide that a foreign tax may satisfy this requirement based on its predominant character, “even if it is also imposed on the basis of some amounts not” based on gross receipts. By contrast, the proposed regulations would permit the use of an alternative gross receipts test only in instances involving pre-realization timing differences or insignificant non-realization events. The proposed gross receipts test could thus deny the creditability of foreign taxes based on minor differences from the U.S. measure of gross receipts, even if the foreign country’s method reasonably measures income, even if it results in no meaningful deviation from the U.S. measure of gross receipts, and even if the foreign tax is demonstrably imposed on income.

b. Confusion regarding cost-based measures of receipts.

The proposed regulations flatly reject cost-based measures of gross receipts. The preamble states that “a foreign tax that requires gross receipts to be calculated by applying a markup to costs, fundamentally diverges from the measurement of realized gross receipts” under the code and is therefore non-creditable.78

Figure 2. Gross Receipts vs. Transfer Pricing

The regulations themselves provide an example illustrating this point, which is actually carried over from the existing final regulations, except with the conclusion reversed (despite the absence of any change in either the stated facts or the applicable statutory rule). The example says that “due to the difficulty of determining on a case-by-case basis the arm’s length gross receipts that headquarters companies would charge affiliates for such services,” the gross receipts of those companies are deemed to equal “110 percent of the business expenses incurred by the headquarters company.” The example then concludes that “because the cost-plus tax is based on costs and not on gross receipts,” the tax does not satisfy the gross receipts test.79

But the proposed regulations then sow some confusion about whether or how such measures differ from cost-based transfer pricing determinations of gross income, which are said to be acceptable. The preamble states that amounts allocated to a taxpayer “pursuant to transfer pricing rules that properly allocate income to a taxpayer on the basis of costs incurred by that entity, are treated as the taxpayer’s actual gross receipts.”80

Remarkably, then, under the November 2020 NPRM, a transfer pricing markup based on costs is treated as “actual gross receipts,” while a measurement of receipts as a markup based on costs “fundamentally diverges from the measurement of realized gross receipts” under U.S. law and thus fails the gross receipts test. The two positions cannot logically be reconciled. At best they can be read to suggest an incoherent distinction based on whether foreign law’s cost-plus measurement of income is phrased as a transfer pricing rule rather than as a measure of realized income. But there is no logical reason for basing the creditability of a foreign tax on such an empty formalistic distinction, as is readily shown by the example in Figure 2.

Thus, the proposed regulations’ self-contradictory references to cost-plus calculations of income would give rise to endless confusion and controversy about whether a particular foreign law calculation is a permissible transfer pricing cost-plus rule or an impermissible gross receipts cost-plus rule. Moreover, drawing such an incoherent and wholly form-based distinction, if that is really what Treasury and the IRS intend, would permit a country to convert what is otherwise a non-creditable tax to a creditable one by simply revising the form of the tax without altering the amount collected by even a single penny. This may not be surprising, given how the proposed regulations generally seek to apply a form-driven analysis, reversing the normal substance-based approach of U.S. tax law and thereby opening the door to manipulations of the formal aspects of foreign tax regimes with little substantive effect.81 But why that’s a good idea is far from obvious.

c. Revival of rejected regulatory proposal.

A notable aspect of the proposed changes to the gross receipts requirement is that they revive a rejected regulatory proposal. Temporary and proposed regulations issued in 1980,82 as well as reproposed regulations issued in 1983,83 would have denied creditability to a foreign tax that used an alternative method of computing gross receipts. Specifically, those rules would have permitted a credit for a tax imposed on estimated gross receipts only in the case of (1) transactions for which it was reasonable to believe that gross receipts may not otherwise be clearly reflected or (2) some pre-realization events. But after describing the 1983 proposed regulations, the preamble to the 1983 final regulations stated that “in response to comments made by the public, these restrictions have been deleted.”84 Although the preamble to the 2020 proposed regulations mentions some of this regulatory history, it does not discuss the government’s prior rejection of the rule that is being proposed, or whether the reasons for that rejection were considered when drafting the proposal.

It is also notable that during its brief tenure as a temporary regulation between 1980 and 1983, the approach that would be revived by the proposed regulations managed to generate a litigated case in which the IRS sought to use the rule to deny the creditability of a foreign tax. The agency’s effort was rejected by the Tax Court.85 The November 2020 preamble states that the current regulations’ substance-based rule causes complexity and controversy, yet the November 2020 NPRM would reinstate a rule that triggered one of the few litigated cases under the gross receipts rule. In fact, apart from that case, only one other foreign tax has resulted in litigation under the gross receipts test in the 40 years of its existence in its current form. That was the unusual U.K. windfall profits tax, which spawned two litigated cases.86 This does not sound like a level of controversy and confusion that justifies disregarding the substance of foreign taxes in the interest of administrative convenience.

d. Concluding comments on proposed changes to the gross receipts requirement.

The November 2020 NPRM would require that foreign law conform with the U.S. measure of gross receipts, even if a different method reasonably measures receipts and even if that method results in no meaningful deviation from the U.S. measure of gross receipts. Thus, consistent with the overall pattern of the proposed regulations, the proposed changes to the gross receipts test would deny the creditability of foreign taxes based on minor differences from the U.S. measure of gross receipts, even when those differences result in a tax demonstrably imposed on income.

By generally looking to “actual” gross receipts, the proposed rule fails to consider that actual gross receipts are not an absolute amount that can be scientifically measured but rather are the product of accounting methods and conventions that may reasonably vary from tax system to tax system. Indeed, the code itself has over time significantly altered the way in which gross receipts have been measured for U.S. tax purposes,87 and it has also used rules of administrative convenience to obviate the need to measure actual gross receipts.88 It would make little sense to bar the creditability of a foreign tax merely because the jurisdiction adopted a different rule of convenience or took a different path in the evolution of its approach to tax accounting methods. It seems particularly puzzling to treat a foreign tax as non-creditable based solely on the foreign tax system’s adoption of a different method, without regard to its reasonableness or the extent to which it approximates actual gross receipts.

Further, when combined with the abandonment of the “predominant character” tests, the proposed rules would deny credit for a foreign levy because of the mere possibility that the base of the foreign tax could depart from the base under U.S. law because of the foreign law’s method for measuring gross receipts, regardless of how unlikely that possibility might be, and indeed even when that outcome does not in fact occur. As noted above and discussed further below, that departure disregards decades of case law under section 901 emphasizing that the substance of a foreign tax determines its creditability, which, of course, is a subset of the broader substance-over-form principle that governs most of the U.S. federal income tax system — and for good reasons.89

4. Proposed changes to the net income/cost recovery requirement.

The November 2020 NPRM would essentially discard the existing net income requirement and replace it with an entirely rewritten and renamed cost recovery requirement. Running a redline between the two provisions generates a flood of red and blue ink.90 The changes to the rules would tighten, to the point of strangulation, the long-standing requirement that foreign law permit the recovery of significant costs and expenses. The proposed restriction of that rule would likely result in the denial of credits for many conventional corporate income taxes. Moreover, the proposed regulations create substantial confusion regarding the acceptability of allowances that are provided under foreign law as alternatives to the deduction of actual expenses.

a. Required recovery of significant costs and expenses.

A foreign tax meets the cost recovery requirement if the base of the tax is computed by reducing gross receipts by the “significant” costs and expenses attributable to those gross receipts. For this purpose, the significance of costs and expenses is defined rather inscrutably as follows:

Whether a cost or expense is significant for purposes of this paragraph (b)(4)(i) is determined based on whether, for all taxpayers in the aggregate to which the foreign tax applies, the item of cost or expense constitutes a significant portion of the taxpayers’ total costs and expenses.91

Several aspects of this definition are notable. First, it defines significant costs and expenses as costs and expenses that “are a significant portion of the taxpayers’ total costs and expenses.” In other words, significant costs and expenses are, um, significant costs and expenses.92 OK, I admit that it isn’t 100 percent fair to call this a tautology because at least the rule tells us to measure the significance of any given expense in relation to the total amount of expenses, rather than comparing an expense with, say, gross receipts. So at least we know the denominator of the relevant fraction — but our only clue to what size numerator will be deemed significant is that it must be “significant.”

Second, by defining significance by reference to whether an expense is significant “for all taxpayers in the aggregate to which the foreign tax applies,” the rule calls for a remarkable empirical-data exercise in which, for example, the prevalence of business meal expenses among all the taxpayers in a country must be surveyed and analyzed to determine whether those expenses are “a significant portion” of all taxpayers’ expenses. Given the preamble’s denigration of basing the creditability of a foreign tax on empirical data, this is hard to fathom.

Third, under this approach, the nondeductibility of an expense that is determined to be significant for all taxpayers in the aggregate will render the foreign tax non-creditable, regardless of whether the relevant taxpayer bears any such expense, so the tax will be non-creditable even if the relevant taxpayer is actually taxed on an amount of net income identical to the net income that it has under the IRC. Thus, for example, an interest deduction limitation may make a foreign income tax entirely non-creditable, even for taxpayers that borrow no money and pay no interest.

While these aspects of the rule seem remarkable, their importance may be largely mooted in practice by a per se rule that automatically treats six common expenses as significant, regardless of their actual significance to the relevant taxpayer or to all taxpayers in the aggregate. This per se list consists of costs and expenses related to:

  • capital expenditures;

  • interest;

  • rents;

  • royalties;

  • services; and

  • research and experimentation.93

Any foreign restriction on the deductibility of any one of these expenses would in principle render the relevant foreign tax non-creditable in its entirety. However, the potential scope of that denial is narrowed, though only somewhat, by a rule that preserves the creditability of a foreign tax, despite a disallowance of the deductibility of a significant expense (whether the expense is on the per se list or otherwise found to be significant), “if such disallowance is consistent with the types of disallowances required under the Internal Revenue Code.”94 To illustrate the scope of this code-consistency exception, the proposed regulation states the following:

For example, foreign tax law is considered to permit recovery of significant costs and expenses if such law disallows interest deductions equal to a certain percentage of adjusted taxable income similar to the limitation under section 163(j), disallows interest and royalty deductions in connection with hybrid transactions similar to those described in section 267A, or disallows certain expenses based on public policy considerations similar to those disallowances contained in section 162.95

The regulation’s acceptance of deduction limitations that are “consistent with the types of disallowance” required under the code is helpful, and it is likewise helpful that limitations “similar to” deduction limitations under sections 162, 163(j), and 267A may not be fatal. But a vast sea of uncertainty would engulf the question whether any given foreign deduction is sufficiently “consistent with” or “similar to” code rules to preserve the creditability of the foreign tax. Moreover, the regulation’s description of analogous code deduction limitations could be read to suggest a fairly narrow intended reading of the rule. For example, the regulation seems to limit the category of acceptable limitations on the deductibility of interest to those that are computed by disallowing “interest deductions equal to a certain percentage of adjusted taxable income.” By specifying one acceptable computational method, the regulation implies that no other methods are acceptable.

Thus, alternative computational approaches to thin capitalization concerns may well flunk this test and would result in the non-creditability of the foreign tax, regardless of how reasonable the foreign jurisdiction’s rule might be, regardless of whether it disallowed more or less interest expense than section 163(j) (for this taxpayer or in the aggregate), and regardless of whether the taxpayer even had any interest expense to deduct in the first place. The proposed regulations’ focus on matching current code rules is particularly perplexing in the context of the interest deduction limitation under section 163(j), given that varying approaches to thin capitalization have historically been deployed here, and yet others have been given serious legislative consideration. For example, during the consideration of the TCJA, both the House and the Senate passed versions of a proposed new section 163(n) that would have imposed an interest expense deduction limit based on the taxpayer’s worldwide debt-equity ratio.96

While no version of section 163(n) was included in the enacted legislation, the fact that interest deduction limitation rules based on worldwide capitalization ratios passed both houses of Congress suggests that section 163(j) might not be the ne plus ultra of such rules, against which all others must perforce be measured. If a foreign government were to enact a provision based on the House or Senate version of section 163(n), the proposed regulations would seem to require the anomalous but inescapable conclusion that the foreign tax is not an income tax because it fails to disallow “interest deductions equal to a certain percentage of adjusted taxable income” under a rule similar to section 163(j). Thus, it appears that under the proposed regulations, another country’s enactment of an interest deduction limitation based on any computational method other than a percentage of ATI could be fatal to the creditability of the country’s income tax.97

Given the wide variety of national solutions to common tax policy issues, including varying approaches to the appropriate design and operation of limitations on deductions for many expenses (whether on the per se list or otherwise), the cost recovery rule as drafted would likely require that many foreign income taxes be treated as non-creditable. This follows because a departure from “similarity” to U.S. deduction-denial rules for any one significant expense will result in the foreign tax being treated as non-creditable, full stop. Thus, unless a foreign country enacts the IRC of 1986 (and keeps up with its frequent amendment), there is likely little hope that every deduction-limiting rule in its national tax code will be similar to the design and operation of the code’s deduction-limiting rules, some of which are highly idiosyncratic, and all of which evolve over time.

For example, given that the regulations treat section 163(j) as an acceptable deduction limitation rule, it is notable that the provision itself was radically changed in 2017 under the TCJA and is scheduled to change again in 2023, in a way that more than doubles the estimated revenue effect of the TCJA’s changes to the provision. And while it is hardly a badge of shame that a revenue act may be concerned with — or even driven by — the legislation’s revenue effects, that reality means that not every change to the code necessarily reflects a normatively pure approach to tax policy. The 2023 change to section 163(j) that will shift the definition of net income from earnings before interest, taxes, depreciation, and amortization to earnings before interest and taxes, for example, appears to be more about revenue than policy. Thus, limiting creditable foreign taxes to those that closely conform to the frequently revenue-driven and non-normative provisions of the code rather inexplicably limits the definition of an income tax to the hot mess embodied in the code at any given moment, which is certainly unlikely to be imitated by many other sovereigns.98

Thus, while this may sound hyperbolic (just a little), if the November 2020 NPRM’s cost recovery rule is finalized in the form proposed, there is a real chance that it would result in the effective repeal of the FTC. Presumably, the drafters would maintain that this was not their intent. But if that is the case, they should clarify that by drafting a different rule in the final regulations.

b. Restriction (or repeal?) of the alternative allowance rule.

The final regulations under section 901 provide that “a foreign tax law that does not permit recovery of one or more significant costs or expenses, but that provides allowances that effectively compensate for nonrecovery of such significant costs or expenses, is considered to permit recovery of such costs or expenses.”99 This alternative allowance rule recognizes that tax systems frequently adopt rules of administrative convenience and that these rules of convenience do not necessarily depart from a reasonable measurement of net income.

Similar to the November 2020 proposed regulations’ confused treatment of cost-based determinations of gross income, the proposed rules in the cost recovery context set forth two irreconcilable statements regarding the treatment of alternative allowances. The proposed regulation first states that “a foreign tax satisfies the cost recovery requirement if the foreign tax law permits recovery of an amount that by its terms may be greater, but can never be less, than the actual amounts of such costs and expenses (for example, under a provision identical to percentage depletion allowed under section 613).”100 However, a little further along, the regulation states what appears to be the opposite rule:

A foreign tax law that does not permit recovery of one or more significant costs or expenses does not meet the cost recovery requirement, even if it provides alternative allowances that in practice equal or exceed the amount of nonrecovered costs or expenses.101

This confusion will need to be resolved, but it is notable that either formulation of the rule would result in the treatment of a foreign tax as non-creditable, even if it resulted in the imposition of an amount of foreign tax that was less than or equal to the amount that would be imposed based on full code-based deductions. The first formulation of the rule is nominally more liberal than the second in that it would preserve creditability if the foreign allowance would by its terms equal or exceed the actual deduction. But in the absence of that guarantee, the foreign tax would be just as non-creditable as under the second formulation, which would deny creditability even if the alternative allowance is reliably more generous than an actual deduction. Thus, even under the less restrictive version of the rule, a foreign country’s use of alternative allowances would foreclose creditability unless the foreign law provides “by its terms” that the alternative allowance will equal or exceed the actual cost. This rule seems unreasonably narrow in several respects:

  • The code itself sometimes uses alternative computations in place of strict adherence to actual expense numbers.102 But because the proposed test would demand that foreign deductions always be equal to or greater than the costs themselves, even a foreign system that provided rules identical to those U.S. rules would appear to fail the test.

  • A foreign provision that produces a lower amount of cost recovery in a single case for a single taxpayer (regardless of how small the difference) would technically fail the cost recovery requirement as drafted and thus presumably would deny creditability of the entire foreign levy for all taxpayers.

  • The proposed rule could thus deny a credit even if the allowance equals or exceeds the taxpayer’s actual deductions, and even though the general standard under the cost recovery requirement is only that it permit the recovery of significant costs and expenses.

  • Under the more liberal version of the rule described above, it would require foreign law “by its terms” to guarantee that the alternative allowance will equal or exceed actual deductions. Such a guarantee is likely nonexistent in the real world, given that the point of an alternative allowance is generally to avoid compliance burdens related to the determination of actual deductions. Thus, even that version of the rule is tantamount to a deletion of the current final regulations’ alternative allowance rule.

c. Concluding comments on proposed changes to the net income requirement.

In a proposed rule full of alarming innovations, the deductibility standard under the cost recovery rule would be the single most alarming change. This is because deduction limitations are a common feature of all national tax systems, and the potential is thus particularly high for mismatches between the idiosyncratic deduction limitations of the code and similar deduction limitations under foreign law. Based on the rigid standards set forth in the November 2020 NPRM, there is a significant likelihood that many corporate income taxes will become non-creditable because at least one of their deduction rules is insufficiently similar to analogous code deduction rules.

This approach is particularly puzzling given that expense deductions are hardly treated as sacrosanct by the code. In the U.S. tax system, expense deductions are viewed as a matter of legislative grace and are routinely granted or limited to implement a variety of goals that may be tax related (such as income measurement, administrative convenience, etc.) or serve other economic policies or political objectives (such as the promotion of homeownership). Further, the extent to which deductions are granted or denied may undergo years of technical refinement or political change. Consider, for example, the accretion of rules regarding the deductibility of home mortgage expenses or home office expenses.103 Thus, requiring that a foreign tax conform to the idiosyncrasies of the U.S. system by requiring all those systems to apply expense deduction rules similar to those that the code applies at any given moment would impose a dubiously narrow new reading of the statute’s unchanged reference to foreign income taxes.

Such a requirement of conformity would also increase compliance burdens under the FTC by requiring frequent (if not annual) reevaluations of the creditability of foreign taxes based on changes in either the U.S. or foreign deduction rules and potentially producing disruptive year-by-year changes in the creditability of foreign taxes. For example, a foreign tax that includes rules identical to current section 163(j) might well have failed the cost recovery requirement in the proposed regulations if those rules had been effective in 2017. But that tax then would have become suddenly creditable in 2018 once the TCJA changes to section 163(j) were enacted, creating the conformity between U.S. and foreign deductibility limitations that the proposed regulations would require.

The evolution of section 163(j) also illustrates the oddness of the proposed regulations’ myopic worldview in treating U.S. deductibility rules as reflecting the only acceptable view of a properly computed income tax. Notably, the TCJA’s changes to section 163(j) were adopted to bring our rules into better conformity with the interest limitations adopted in recent years by many of our trading partners.104 Given this tacit acknowledgement that the non-U.S. rules had become the international norm for those limitations, it seems anomalous to insist that only the United States knows how to properly design an income tax.

C. Proposed Changes to the Section 903 Rules on the Creditability of Taxes Imposed in Lieu of Income Taxes

1. Overview of the changes.

Under section 903, a foreign tax that is not an income tax may still be credited as an income tax under section 901 if it is imposed in lieu of a generally imposed income tax.105 The regulations under section 903 have long implemented the “in lieu” requirement under rules requiring that the foreign tax be imposed in substitution for the country’s generally imposed income tax, consistent with the plain meaning of the statutory language. The 2020 proposed regulations would tie the ability to meet this substitution requirement to a determination that the underlying foreign tax is a creditable income tax under the revised standards of reg. section 1.901-2.

Further, the proposed regulations would tighten the regulatory implementation of the statute’s requirement that the foreign tax be imposed in lieu of a generally applicable foreign tax by adding new tests to the regulations’ long-standing substitution requirement. In particular, the November 2020 NPRM would add “non-duplication” and “close connection” rules that appear to be designed to reverse the results of cases that the IRS lost decades ago. The result would be that some foreign taxes that are in fact imposed in lieu of income taxes would likely fail to qualify as such under the standards of the proposes rules, resulting in the non-creditability of many of those taxes and a further increase in the incidence of double taxation of cross-border income.

Figure 3 provides an overview of the many ways in which a foreign tax could flunk the revised creditability standards of the proposed regulations under section 903.

Figure 3. Creditability via Section 903 of Foreign Taxes in Lieu of Inocme Taxes: Even More Ways to Fail

2. Changes to reg. section 1.901-2 would have broad knock-on effects on the creditability of in-lieu taxes under section 903.

The fact that the proposed regulations would interpret the term “income tax” in section 903 to refer to a foreign tax that is determined to be an income tax for purposes of section 901 is hardly surprising; indeed, absent some evidence of a need to read the terms differently, it would be surprising if they weren’t read to mean the same thing. However, the cross-reference, while unsurprising, makes it clear that the dramatic narrowing of the creditability of foreign taxes that would arise under the proposed changes to reg. section 1.901-2 may also have dramatic effects on the creditability of in-lieu taxes through section 903, even before the tightening measures added to the section 903 regulations are considered. This follows from the fact that the jurisdictional nexus rule and the changes to the net gain requirements (particularly under the cost recovery rule) will mean that many foreign corporate income taxes will not qualify as creditable because few foreign corporate income taxes seem likely to match code deduction, gross receipts, and realization rules to the extent that the proposed regulations would require. As a result, if any of the affected foreign tax systems also impose an in-lieu tax in particular sectors, those in-lieu taxes will likewise be non-creditable, even if they otherwise meet the substitution requirement under the provisions of revised reg. section 1.903-1.

3. Rewriting of the substitution requirement.

a. Overall complication of regulations implementing the eight-letter term ‘in lieu of.’

Section 903 allows a non-income tax to be treated as creditable under section 902 if it is imposed “in lieu of” a generally imposed income tax. The dictionary (or even distant memories of high school French class) would quickly tell us that in lieu of means instead of or in place of.106 So section 903 grants a credit for any tax imposed instead of an income tax.

Admittedly, the existing regulations slightly complicate the analysis by turning the simple statutory test into a more elaborate “substitution” requirement. But the proposed regulations would vastly increase the rule’s complexity. As redrafted, the substitution requirement would impose four distinct tests.

i. Underlying income tax.

As noted above, the foreign country must have a generally imposed income tax as defined in reg. section 1.901-2(a)(3).

ii. Non-duplication.

No income tax may be imposed, with respect to any taxpayer, on any portion of the tax base of the in-lieu tax.

iii. Close connection.

The generally imposed income tax would apply “but for” the in-lieu tax. This requirement is satisfied if:

  • the income subject to the in-lieu tax is expressly excluded (for example, express exclusion of insurance income from the general income tax regime); or

  • the foreign country made a “cognizant and deliberate choice” to impose tax on the excluded category (based, for example, on the foreign legislative history). The rules rebuttably presume that a later-in-time tax does not qualify.

iv. Jurisdictional nexus.

If the generally imposed net income tax was applied to the excluded income, it would satisfy the jurisdictional nexus requirements of reg. section 1.901-2(c).

The preamble to the November 2020 NPRM states that all this complexity is needed to reflect the outcomes of decided cases,107 but the proposed regulations actually seek to reverse the outcome of the one case cited by the preamble.

b. The non-duplication test seeks to reverse Metropolitan Life.

The new non-duplication test would revive an IRS reading of section 903 that was long ago rejected by courts, and it would create substantial new administrative complexity. The preamble states that the proposed rules modify the regulatory guidance on the statute’s substitution requirement “by more specifically defining the circumstances in which a foreign tax is considered ‘in lieu of’ a generally-imposed income tax, consistent with the interpretation of the substitution requirement in prior judicial decisions,” and cites “for example” a 1967 case involving the Metropolitan Life Insurance Co.108

However, the proposed rules directly contradict the Metropolitan Life decision cited in the preamble, attempting to revive the IRS reading of the statute that the court rejected. In particular, the court rejected an IRS argument that had tried to deny the creditability of an in-lieu tax based on the fact that other taxpayers, not the taxpayer at issue, were subject to both the in-lieu tax and the foreign income tax.109 Yet the proposed regulations would insert precisely this rule by requiring that no generally imposed income tax be imposed in addition to the in-lieu tax, “on any persons with respect to any portion of the income” related to the base of the in-lieu tax, and concluding that an in-lieu tax therefore fails the substitution requirement “if a net income tax imposed by the same foreign country applies to the excluded income of any persons that are subject to the tested foreign tax, even if not all of the persons subject to the tested foreign tax are subject to the net income tax.”110 Under this rule, then, an in-lieu tax is non-creditable if “any person” subject to that tax is also subject to the general income tax, even if the relevant taxpayer is subject to only the in-lieu tax. Thus, contrary to the preamble’s claim, the proposed rule is not “consistent with the interpretation of the substitution requirement in prior judicial decisions,” because that argument was squarely rejected in Metropolitan Life.

Under the proposed non-duplication rule, any degree of overlap between an income tax and an in-lieu tax would disqualify the in-lieu tax. For example, assume Country X generally imposes an income tax that meets the creditability requirements of section 901, while imposing a gross income tax on some inbound inventory sales in lieu of the income tax. Further assume that Country X decides to apply both taxes to one industry, such as consumer electronics, in light of perceived profit margins on electronic devices. The gross income tax will fail the non-duplication requirement and thus not be creditable for any market participant, regardless of its product lines and thus regardless of whether it was actually subject to the income tax as well as the in-lieu tax.

Also, it is unclear as a compliance matter how taxpayers could ever prove the negative that the regulations would in principle require to be shown — that no taxpayer in the country is subject to both an in-lieu tax and an underlying income tax. Presumably, the testimony of local tax experts could help supply this empirical data, but it is disheartening that the creditability of a foreign in-lieu tax may ultimately depend on a battle of foreign tax experts,111 and particularly disheartening given that a non-duplication requirement has already been (repeatedly) held to be inconsistent with section 903.

c. The close connection rule is unsupported by the language of section 903 and again inconsistent with Metropolitan Life.

Similar issues are presented by the November 2020 NPRM’s new close connection rule, requiring that “but for the existence of the tested foreign tax, the generally-imposed net income tax would be imposed on the excluded income.” The proposed test is hard to square with either the language of the statute or precedent interpreting that language. The proposed rule’s requirement of proof that a foreign income tax “would be imposed” on the taxpayer absent the in-lieu tax goes beyond the language of the statute, which by its terms requires only that the non-income tax be imposed “in lieu” of the income tax, not that the income tax would otherwise apply to the taxpayer. That is, the statute requires only that the in-lieu tax apply instead of the income tax, not that it function to replace an otherwise-applicable income tax. The proposed regulations would require that a foreign government first design an income tax, impose that income tax on all taxpayers, and then exclude some taxpayers from that income tax while subjecting them to an in-lieu tax. The regulations would thus treat as non-creditable any in-lieu tax imposed by a foreign government that determines ab initio that its income tax will be inapplicable to specified categories of taxpayers and that it will instead impose a non-income tax on those taxpayers in place of the inapplicable income tax.

Governments may, of course, exclude categories of taxpayers from the normative scope of their general income tax for many reasons. These may include administrative convenience, tax policy, economic policy, or other policy goals. And in connection with a decision to exclude those taxpayers from the income tax, a foreign government may design an alternative tax to be imposed on them, calculated on some basis other than net income. In such a case, it may well be clear that the excepted classes of taxpayers were not, are not, and never will be subject to the general income tax, for whatever administrative or policy reasons the government found persuasive. But it is likewise clear that the in-lieu tax is still imposed in substitution for the income tax that the government chose not to impose. Such a tax system clearly falls within the plain language of section 903, yet would be excluded by the proposed regulations’ requirement of a showing that the foreign income tax “would be imposed” on the taxpayer that is subject to an in-lieu tax.

Moreover, even if it could be squared with the language of the statute, the proposed “would be imposed” test raises an impossible question of proof by requiring taxpayers to show what the purely hypothetical income tax result would be if foreign law were rewritten to repeal the in-lieu tax. The creditability of a foreign tax cannot logically depend on such a speculative inquiry entirely untethered from reality. Unproductive battles of foreign experts engaged in those untethered speculations would be the likely outcome if the proposed rule were actually adopted.

All of this is particularly puzzling given that the proposed “would be imposed” test is again inconsistent with the Metropolitan Life decision that the November 2020 preamble cites as authority for the proposed regulations’ interpretation of the statute. Metropolitan Life imposes no requirement that a taxpayer show that it would have paid an income tax absent being subject to an in-lieu tax. The court found as a factual matter that there was a “close connection” between the income tax and the in-lieu tax under foreign law, and it found that the foreign government chose to apply its in-lieu tax “because” for policy reasons it would not apply its income tax to the insurance businesses at issue.112

But the court also recognized that the foreign government, for tax policy reasons of its own, would under no circumstances impose its income tax on insurance companies.113 By requiring proof that the foreign income tax “would be imposed” on the taxpayer absent the in-lieu tax, the proposed regulation thus contradicts the court’s finding in Metropolitan Life that a tax imposed on insurance companies was a creditable in-lieu tax, even though it was clear that those companies would under no circumstances be subject to the foreign income tax.

In sum, because in-lieu taxes may be imposed in substitution for an income tax because the foreign government decided not to impose its income tax on the relevant class of taxpayers, the substitution requirement cannot logically be read to require a showing that the foreign income tax “would be imposed” on the taxpayer absent the in-lieu tax, and such a reading is unsupported by relevant precedent.

D. Concluding Observations on the Surpassing Strangeness of Treating the IRC as the Platonic Ideal of an Income Tax

The cumulative impact of the many changes summarized above would be to treat the IRC as the definitive income tax, so that if a foreign tax diverged in any substantial way from code rules, the foreign tax would be foreclosed from treatment as an income tax. From the elimination of all references to the predominant character of a tax, through the requirement of deduction rules that substantially match code deduction rules, to the addition of required fealty to code sourcing and accounting rules, etc., the November 2020 NPRM would come close to requiring that foreign countries mirror the IRC.

While I’m as big a fan of the code as anyone (admittedly, a modest claim), its deification in the proposed regulations is both inexplicable and a profound misreading of the statute, which, of course, requires only that the foreign tax be an income tax. Even with the later gloss clarifying that when the code refers to an income tax, it means an income tax “in the U.S. sense,” a tax may plainly be an income tax in the U.S. sense without being the U.S. income tax. That is, a foreign tax may be a tax imposed on “income” as we understand the term, without having to replicate all the highly idiosyncratic details of our ever-changing computation of taxable income under the code. Our code’s computation of income reflects the gradually accreted synthesis of thousands of discrete congressional decisions to pursue a myriad of goals and judgments that have often had nothing to do with a strict economic measurement of income.114

For this self-evident reason, Congress, Treasury, the IRS, and courts have never previously suggested that the code represents the only reliable manner of computing an income tax. The code may indeed tax some reasonable approximation of realized net income, most of the time, but its patchwork of variously intentioned rules results in very different taxable bases for different classes of taxpayers in different contexts. Given these extensive differences across taxpayers and contexts even within the single set of rules embodied in the code, it is remarkably bold for the preamble to suggest that there exist any generally observed international tax “norms reflected in the Internal Revenue Code that define an income tax in the U.S. sense.”115

Constitutionally, the 16th Amendment has been found to impose few limitations on congressional flexibility in the drafting of the federal income tax. For example, Congress may constitutionally choose to tax gross income rather than net income,116 and the code in fact taxes gross income in some contexts.117 Thus, an income tax in the U.S. sense may include gross-basis taxation — although that is certainly not the predominant character of the U.S. income tax (to coin a phrase). Deductions, because they are a matter of legislative grace, are allowed only when, if, and to the extent Congress determines118 — and Congress routinely determines to deny deductions for reasons not necessarily related to the economic measurement of income. The code’s rules sometimes — maybe even often — reflect an intent to accurately measure income, but just as often they implement other goals entirely, such as encouraging taxpayers to make charitable contributions (but not too many), encouraging taxpayers to buy houses, or simply ensuring that taxpayers pay some minimum amount of tax.119 Accordingly, what constitutes gross income and deductions in the U.S. system generally reflects the evolving political priorities of Congress over time rather than a determination of what constitutes the Platonic ideal of an income tax based on normative economic and tax policy principles.

Further, to whatever extent international norms for determining and taxing income exist and become reflected in the code, these norms change over time, sometimes with startling suddenness.120 And in some cases the United States pushes the norms forward; for example, the U.S. interest limitations under former section 163(j) were introduced before those of most other jurisdictions. In other cases, the United States trails behind the norms, for example regarding the imposition of broad anti-hybrid limitations.

Requiring strict adherence to U.S. norms for each taxpayer and context at a given moment will in no way “simplify and clarify the application of the rules.”121 Rather than determine whether a foreign levy applies to income in the normal instance, the proposed regulations require separate determinations for each class of taxpayers for which the application of the foreign levy results in a significantly different tax base. Each of those levies must then be tested against a similarly situated class under the U.S. rules. The foreign levy is considered an income tax only if the rules conform in detail for that taxpayer class. For this reason, providing a broader definition of a separate levy to allow the separate evaluation of a tax for a class of taxpayers does not mitigate the harm done by the elimination of the more flexible predominant character approach.

The result of analyzing foreign taxes as multiple separate levies for multiple classes of taxpayers will be to eliminate consistent treatment of particular national taxes, requiring a separate analysis of the tax’s creditability for the various classes of taxpayers to which it applies. Thus, rather than being able to rely on a general understanding that a country’s corporate income tax is an income tax in the U.S. sense, each taxpayer will be required to test any foreign levies to which it is subject to take into account the aspects of the foreign system that may apply to its industry.

Further, these required determinations would be fact intensive because all deviations from the putatively pure income tax system of the code will have to be identified and weighed. Some deviations will create a separate class of taxpayers (and therefore a separate levy), while other deviations would have to be weighed for significance. And because neither the U.S. system nor any foreign system is static, these assessments would have to be made periodically (or even annually?) to determine whether even small changes to either the foreign system or the code render a creditable tax non-creditable (or vice versa). Thus, rather than achieving simplicity and clarity for taxpayers, the proposed approach would exacerbate the existing complexity and confusion and require the government to address more often, rather than less often, the creditability of foreign income taxes.

The removal of the current final regulations’ more flexible approaches would also have the effect of prioritizing the form of a tax over its substance. It is, of course, a fundamental principle of federal income taxation that tax results must turn on substance rather than form.122 In fact, the Supreme Court has found that the substantive effect of a tax must be considered when evaluating whether a foreign tax constitutes an income tax.123 Thus, adopting rules that determine the creditability of a foreign tax based on its form rather than its substance would be inconsistent with the fundamental principles generally underlying the code and with Supreme Court guidance on the FTC.

In sum, under section 901, an FTC is provided for any foreign tax that is an income tax in the U.S. sense, but the November 2020 NPRM would instead limit the credit to foreign taxes that closely match the provisions of the U.S. income tax in force in the year the foreign tax is paid. By requiring an approach that eliminates the flexibility to recognize the essential characteristics of tax systems as they evolve over time, the proposed regulations would frequently lead to double taxation because a taxpayer suffers double taxation whenever its income is subject to two income taxes, not only when U.S. and foreign income taxes are fraternal twins. The proposed regulations do not express much concern about the extent to which foreign income taxes would become non-creditable under the proposed new tests, but that result would be difficult to square with the history and purpose of section 901 discussed in Section II above.

IV. Conclusions

A. The November 2020 NPRM Should Be Withdrawn

The history of the FTC summarized above reveals the tax policy foundations that underlie the credit. The November 2020 NPRM departs in fundamental ways from the lessons of that history and should therefore be withdrawn. The sequence of historical developments supporting this conclusion include the following:

  1. When Congress enacted the FTC in 1918:

    • It referred simply to an income tax. Although later interpretation has glossed that term to mean an income tax “in the U.S. sense,” no legislative or judicial authority has suggested that the term encompasses only a foreign income tax that closely resembles the IRC.

    • It sought to prevent international double taxation because such double taxation would unjustly impose disparate tax burdens on taxpayers with similar net incomes, depending on the source of their income and the amount of foreign tax they paid. Congress also sought to protect the competitiveness of U.S.-based international businesses.

  2. The 1918 statute referred to income taxes that a foreign country imposed “upon income derived from sources therein.” When Congress added the FTC limitation two years later, it deleted the reference to income derived from sources in the taxing country and replaced it with a limitation that restricts the size of the credit to the amount of U.S. tax imposed on the foreign income in the relevant income category, without inquiring into the basis for the foreign country’s assertion of taxing jurisdiction. The system was thereby fully protected from foreign taxes on U.S.-source income by the limitation.

  3. This basic structure of the rules has remained unchanged through all subsequent changes to the FTC limitation and related international tax rules, which have been frequent and voluminous.

Given this sequence, it is fair to conclude that if Congress had intended to authorize a jurisdictional requirement to be imposed as an additional filter on the creditability of foreign taxes, it would have said so at some point in the last 100 years, and that it is inappropriate for Treasury to do so now by administrative action.

Similarly, it is fair to conclude that the adoption of definitional provisions that would render many foreign taxes non-creditable, even though those taxes are factually imposed on the taxpayer’s net income, would be inconsistent with the statute’s straightforward reference to foreign income taxes. It would also be inconsistent with the intent of Congress to prevent double taxation to promote horizontal equity among taxpayers by ensuring similar total tax burdens regardless of their sources of income and to protect the competitiveness of U.S.-based international businesses.

These conclusions follow from the straightforward history of legislative activity summarized above and should not require further elucidation by any learned doctrines of statutory construction. Thus, my primary focus here is not on the interplay between legislative and administrative action but rather on section 901 itself. The statute is the primary authority to be construed, and it has remained unchanged through many reenactments. Thus, the scope of the FTC under section 901 should be governed first by the plain language of the unamended statute, and to the extent that interpretive assistance is required, Congress’s original legislative purpose in enacting the statute should remain the principal source of insight regarding the intended scope of the FTC.

But it is also worth noting that courts often ascribe interpretive significance to sequences of legislative, judicial, and administrative action and inaction. For example, the Supreme Court has treated the adoption of an authoritative interpretation of a statute, followed by reenactment of that statute without alteration, as implying a congressional endorsement of that interpretation. In Bragdon, for example, the Court stated that “when administrative and judicial interpretations have settled the meaning of an existing statutory provision, repetition of the same language in a new statute indicates, as a general matter, the intent to incorporate its administrative and judicial interpretations as well.”124

Of course, the common law’s endearing schizophrenia also gives us authorities pointing in a different direction:

We need not stop to inquire whether, in absence of the Treasury Regulations under the 1934 Act, the administrative construction of “acquisition” under the earlier Acts was of such a character (Higgins v. Commissioner, 312 U.S. 212), and the prior judicial decisions had such consistency and uniformity, that Congressional reenactment of the language in question was an adoption of its previous interpretation within the rule of such cases as United States v. Dakota-Montana Oil Co., 288 U.S. 459. That rule is no more than an aid in statutory construction. While it is useful at times in resolving statutory ambiguities, it does not mean that the prior construction has become so embedded in the law that only Congress can effect a change. Morrissey v. Commissioner, 296 U.S. 344, 296 U.S. 355. And see Murphy Oil Co. v. Burnet, 287 U.S. 299. It gives way before changes in the prior rule or practice through exercise by the administrative agency of its continuing rulemaking power. Helvering v. Wilshire Oil Co., 308 U.S. 90, 308 U.S. 100-101.125

Nevertheless, depending on the extent to which they are ultimately found to apply, canons such as the prior construction doctrine may provide further reason to question the soundness of the November 2020 NPRM, because there are no authoritative interpretations of the statute predating the November 2020 NPRM that would suggest that Congress endorsed any of the interpretive innovations reflected in those proposed rules. To the contrary, the proposed rules would seek to reverse long-established authorities interpreting the statute and to revive administrative positions previously rejected in the regulatory process and in litigation.126

Accordingly, the November 2020 NPRM’s attempts to impose a new jurisdictional nexus restriction and to restrict creditable foreign taxes to those that closely resemble those imposed under the IRC are inconsistent with the history and purposes of both the FTC under section 901 as enacted in 1921 and never substantively amended and the in-lieu credit enacted in 1942 and never substantively amended. The proposed regulatory changes would result in the double taxation of cross-border income in a manner inconsistent with the normative goal of the statute (achieving horizontal equity between taxpayers with similar amounts of income, wherever derived) and with its economic goal of enhancing the competitiveness of U.S.-based international businesses. Further, while these considerations suggest that Treasury and the IRS would lack the regulatory authority to promulgate rules so flatly inconsistent with the history and purposes of the relevant statutes, that inconsistency means that the regulations should not be finalized even if Treasury and the IRS conclude that they have authority to do so. In sum, I respectfully suggest that the November 2020 NPRM be withdrawn.

B. Future Changes to Section 901 or Its Implementing Regulations

If Treasury and the IRS conclude that there are concerns that justify imposing a new nexus requirement in addition to the statute’s existing sole requirement that a foreign tax be an income tax, that revision to the statute’s creditability standards should be proposed to Congress and legislatively enacted in the ordinary course. Regulations are not the proper forum to impose additional restrictions on creditability not set forth in the statute itself — and particularly not a restriction that was deleted from the statute by Congress a century ago.

If Treasury and the IRS nevertheless determine that regulatory, rather than legislative, change should be pursued, any changes that would alter long-standing standards governing the creditability of foreign taxes should include transition rules recognizing that the existing regulations under section 901 have been in place and relied on by taxpayers for nearly 40 years. A sudden lurch in regulatory guidance such as that proposed in November 2020, particularly in light of the proposals’ inconsistency with decades of precedent under an unchanged statute, would be deeply disruptive to many U.S. businesses with foreign operations. The financial projections that companies relied on to make those investments over the past several decades would be disrupted by a sudden change in the creditability of foreign income taxes that were routinely and appropriately treated as creditable under the existing regulations, rulings, and case law. Planning for the acquisition or expansion of any international business has been thrown into profound uncertainty by the November 2020 proposals, which would seemingly deny FTCs for many long-creditable foreign income taxes, including conventional corporate income taxes imposed by major U.S. trading partners that deviate in some way from the operation of the IRC.

In light of these considerations, I respectfully suggest that, in addition to withdrawing the November 2020 proposals, Treasury and the IRS seek to minimize business disruptions by (1) giving due weight to the history and purpose of sections 901 and 903, including both the legislative and the regulatory history summarized above; (2) deploying the rarely used advance NPRM device to solicit early stakeholder input regarding the potential direction of future changes to the rules in a manner less disruptive than a full-blown NPRM that could result in nearly immediate finalization after a brief comment period; and (3) ensuring that future changes will take effect only prospectively, after an appropriate transition period that enables taxpayers to make necessary changes in business and investment structures (to the extent that it is possible for them to do so).

I conclude with Sir Edmund Burke’s famous observation that “to tax and to please, no more than to love and to be wise, is not given to men.”127 Maybe so, but I hope that Treasury and the IRS won’t be discouraged by that wry adage from taking a fresh look at the November 2020 NPRM. They could please at least one observer (and probably many) by giving more weight to the long history of section 901 as they decide where to take it in the future.

V. Appendix

Informal, Unofficial Redline of Existing Reg. Section 1.901-2(b)(4) Against Prop. Reg. Section 1.901-2(b)(4)128

Reg. section 1.901-2(b)

(4) Net income.

(4) Cost recovery requirement.

(i) In general.

(A) Requirement. A foreign tax satisfies the net income cost recovery requirement if, judged on the basis of its predominant character, the base of the tax is computed by reducing gross receipts (including gross receipts as computed under described in paragraph (b)(3)(i)(B) of this section) to permit

(B) Recovery recovery of the significant costs and expenses (including significant capital expenditures) attributable, under reasonable principles, to such gross receipts; or. In addition, a foreign tax satisfies the cost recovery requirement if the foreign tax law permits recovery of an amount that by its terms may be greater, but can never be less, than the actual amounts

(B) Recovery of such significant costs and expenses computed under a method that is likely to produce an amount that approximates, or is greater than, recovery of such significant(for example, under a provision identical to percentage depletion allowed under section 613). A foreign tax whose base is gross receipts or gross income for which no reduction is allowed under foreign tax law for costs and expenses. does not satisfy the cost recovery requirement, even if in practice there are few costs and expenses attributable to all or particular types of gross receipts included in the foreign tax base. See paragraph (b)(4)(iv) of this section (Example 3).

(B) Significant costs and expenses.

(1) Timing of recovery. A foreign tax law permits recovery of significant costs and expenses even if such costs and expenses are recovered at a different timeearlier or later than they would be if are recovered under the Internal Revenue Codeapplied, unless the time of recovery is such that under the circumstances thereso much later (for example, after the property becomes worthless or is disposed of) as effectively to constitute a denial of such recovery. For example, unless the time of recovery is such thatThe amount of costs and expenses that are considered to be recovered under the circumstances there is effectively a denial of suchforeign tax law is neither discounted nor augmented by taking into account the time value of money attributable to any acceleration or deferral of a tax benefit resulting from the foreign law cost recovery method compared to when tax would be paid under the Internal Revenue Code. Therefore, the cost recovery, the net income requirement is satisfied where items deductible under the Internal Revenue Code are capitalized under the foreign tax systemlaw and recovered either immediately, on a recurring basis over time, or upon the occurrence of some future event, or where the recovery of items capitalized under the Internal Revenue Code occurs more or less rapidly than under the foreign tax system.law.

(2) Amounts that must be recovered. Whether a cost or expense is significant for purposes of this paragraph (b)(4)(i) is determined based on whether, for all taxpayers in the aggregate to which the foreign tax applies, the item of cost or expense constitutes a significant portion of the taxpayers’ total costs and expenses. However, costs and expenses related to capital expenditures, interest, rents, royalties, services, or research and experimentation are always treated as significant costs or expenses for purposes of this paragraph (b)(4)(i). Foreign tax law is considered to permit recovery of significant costs and expenses even if recovery of all or a portion of certain costs or expenses is disallowed, if such disallowance is consistent with the types of disallowances required under the Internal Revenue Code. For example, foreign tax law is considered to permit recovery of significant costs and expenses if such law disallows interest deductions equal to a certain percentage of adjusted taxable income similar to the limitation under section 163(j), disallows interest and royalty deductions in connection with hybrid transactions similar to those described in section 267A, or disallows certain expenses based on public policy considerations similar to those disallowances contained in section 162. A foreign tax law that does not permit recovery of one or more significant costs or expenses, but that does not meet the cost recovery requirement, even if it provides alternative allowances that effectively compensate for nonrecovery of such significant costs or expenses, is considered to permit recovery of such costs or expensesin practice equal or exceed the amount of nonrecovered costs or expenses. However, in determining whether a foreign tax (the “tested foreign tax”) meets the cost recovery requirement, it is immaterial whether the tested foreign tax allows a deduction for other taxes that would qualify as foreign income taxes (determined without regard to whether such other tax allows a deduction for the tested foreign tax). See paragraph (b)(4)(iv) of this section (Example 5).

(3) Attribution of costs and expenses to gross receipts. Principles used in the foreign tax law to attribute costs and expenses to gross receipts may be reasonable even if they differ from principles that apply under the Internal Revenue Code (e.g., principles that apply under section 265, 465 or 861(b) of the Internal Revenue Code). A foreign tax whose base, judged on the basis of its predominant character, is computed by reducing gross receipts by items described in paragraph (b)(4)(i)(A) or (B) of this section satisfied the net income requirement even if gross receipts are not reduced by some such items. A foreign tax whose base is gross receipts or gross income does not satisfy the net income requirement except in the rare situation where that tax is almost certain to reach some net gain in the normal circumstances in which it applies because costs and expenses will almost never be so high as to offset gross receipts or gross income, respectively, and the rate of the tax is such that after the tax is paid persons subject to the tax are almost certain to have net gain. Thus, a tax on the gross receipts or gross income of businesses can satisfy the net income requirement only if businesses subject to the tax are almost certain never to incur a loss (after payment of the tax). In determining whether a foreign tax satisfied the net income requirement, it is immaterial whether gross receipts are reduced, in the base of the tax, by another tax, provided that other tax satisfies the realization, gross receipts and net income requirements.for example, principles that apply under section 265, 465 or 861(b) of the Internal Revenue Code).

(ii) Consolidation of profits and losses. [Unchanged.]

(iii) Carryovers. [Unchanged.]

FOOTNOTES

1 Adams, “International and Interstate Aspects of Double Taxation,” 22 Nat’l Tax Ass’n Proc. 193, 196 (1929) (remarks to the National Tax Association). Prosaic? Not dramatic or thrilling? Fictitious interest? We’ll see about all that.

2 Quoted in 1916 Chicago Tribune interview, as reported in K. Kris Hirst, “Did Henry Ford Really Say ‘History Is Bunk’?” ThoughtCo., Sept. 4, 2019. See also Roger Butterfield, “Henry Ford, the Wayside Inn, and the Problem of ‘History Is Bunk,’” 77 Proc. Mass. Hist. Soc. 53 (1965).

3 Compare, for example, the events reported in Jonah E. Bromwich, “Whatever It Is, It’s Probably Not Hair Dye,” The New York Times, Nov. 19, 2020.

4 See REG-101657-20, 85 F.R. 72078, 72087 (Nov. 12, 2020).

5 The TCJA made other significant changes to the FTC rules but did not amend the relevant language of section 901(b). See P.L. 115-97, “An Act to Provide for Reconciliation Pursuant to Titles II and V of the Concurrent Resolution on the Budget for Fiscal Year 2018.” Although sadly that legislation was passed without the benefit of a short title, I adopt the universal practice of referring to it as the TCJA.

6 A tip of the cap to the drafters of the U.S. Council for International Business (USCIB) comments on the November 2020 NPRM for providing this redline presentation, which highlights the two words that Congress added and the one that it deleted. See USCIB comments on REG-101657-20, at 5 (Feb. 8, 2021). In light of there being no substantive difference between the 1921 version of the statute and any subsequent version as it relates to the type of foreign tax that is creditable, for convenience I generally refer to both current section 901 and its predecessors under prior IRCs and revenue acts as “section 901” unless the context otherwise requires. The history of section 901 is discussed in more detail in Section II.A, below.

7 See id. at 3-12.

8 Preamble to REG-101657-20, 85 F.R. 72078-72116. The proposed regulations also seek to reinstate several IRS litigating positions rejected by courts. While this too is a form of response to history, the preamble does not highlight this aspect of the proposals. See the discussion below in sections III.B and III.C.

9 Fortunately, as discussed in Section II below, some excellent historical work by others has facilitated exploration of the early legislative history and other materials underlying the enactment of the FTC.

10 I have tried to stick with the following taxonomy throughout this report: (1) the term “foreign income tax” means a foreign tax that is creditable under section 901 because it is a foreign income, war profits, or excess profits tax within the meaning of that section; (2) the term “in-lieu tax” means a foreign non-income tax that is creditable under section 901 via section 903 because it is imposed in lieu of a generally imposed foreign income tax within the meaning of the latter section; and (3) the term “creditable foreign tax” means a foreign tax that is creditable because it is either a foreign income tax or an in-lieu tax.

11 Reg. section 1.901-2(a)(1)(ii) (referring to an income tax “in the U.S. sense”); Rev. Rul. 56-51, 1956-1 C.B. 320 (referring to foreign tax imposed on amounts that constitute income items “by United States standards”). See also Lanman & Kemp-Barclay & Co. of Colombia v. Commissioner, 26 T.C. 582, 587 (1956) (stating that “it is well settled that the determination of whether or not a foreign levy qualifies as an income tax within the meaning of section [901] is to be made not upon the characterization of the foreign law, but under the criteria established by the internal revenue laws of the United States,” and citing Biddle v. Commissioner, 302 U.S. 573 (1938); Keasbey & Mattison Co. v. Rothensies, 133 F.2d 894 (3d Cir. 1943); Wilson v. Commissioner, 7 T.C. 1469 (1946); and Commissioner v. American Metal Co., 221 F.2d 134 (2d Cir. 1955)).

12 A similar diagram in Section III.C below illustrates the comparable series of hurdles that the proposed regulations would impose before permitting a foreign non-income tax to be treated as a creditable in-lieu tax under section 903.

13 As noted above, the early history of the FTC and related provisions has been helpfully analyzed by others. The discussion that follows is deeply indebted to the detailed 1997 article by Michael J. Graetz and Michael M. O’Hear, “The ‘Original Intent’ of U.S. International Taxation,” 46 Duke L.J. 1021 (1997). Other secondary sources that have usefully reviewed the early history of these rules include a 2001 volume by the National Foreign Trade Council (NFTC), The NFTC Foreign Income Project: International Tax Policy for the 21st Century; Part Two — Relief of International Double Taxation, and the recent USCIB comments, supra note 6.

14 My focus here is on the language of the statute defining the foreign taxes eligible to be credited and on the remarkable absence of change to that language. To be sure, other aspects of the design and drafting of the credit have been modified over the years, such as the integration of separate provisions addressing credits for corporate and individual taxpayers into a single provision addressing both and the separation of various rules into distinct statutory sections addressing direct credits (section 901), indirect credits (section 902 (before its 2017 demise) and section 960), the FTC limitation (section 904), etc. As discussed below, this ongoing legislative tinkering with many aspects of the FTC make it all the more remarkable that time and tide have not availed to change the language defining creditable foreign taxes.

15 Revenue Act of 1918, section 238. A similar credit was provided to individual taxpayers under section 222 of the same act. Although the Revenue Act of 1918 was not enacted until February 1919, I generally refer to it as 1918 legislation, based on its gestation period, short title, and effective dates (and despite the pedantic appeal of confusingly referring to the Revenue Act of 1918 as 1919 legislation).

16 Revenue Act of 1921, section 238. A redline between the 1918 and 1921 versions of the language highlights one important substantive change to the rule, which continued to provide a credit for “the amount of any income, war-profits and excess-profits taxes paid during the same taxable year to any foreign country, upon income derived from sources therein, or to any possession of the United States.” The importance of the 1921 deletion of the 1918 language requiring a creditable foreign tax to be imposed on income sourced in the country imposing the tax is discussed in Section II.B.2 below.

17 Section 901(b)(1).

18 This redline does not highlight minor punctuation changes, but those changes included dropping the hyphens from the references to war profits and excess profits taxes and eliminating the comma that separated the references to a foreign country and a U.S. possession. It may also be worth noting that the statute’s now somewhat mysterious references to war profits and excess profits taxes reflect the fact that when the credit was being drafted in 1918 and 1921, the United States imposed war profits and excess profits taxes of its own (in the nature of alternatively applicable surtaxes) and apparently felt constrained to grant a credit for similar taxes imposed by other countries. See, e.g., Revenue Act of 1921, section 335.

19 Graetz and O’Hear, supra note 13, at 1041; NFTC, supra note 13, at 149.

20 See, e.g., Edwin R.A. Seligman, Double Taxation and International Fiscal Cooperation 32-37 (1928); and Adams, “Interstate and International Double Taxation,” in Lectures on Taxation: Columbia University Symposium 101 (1932). Adams summarizes Seligman’s work as showing that serious discussion of the problem of double taxation “began as early as the thirteenth century and continued for centuries, apparently with no very important results beyond revealing the stubborn conflict of interest between the jurisdiction of residence and the jurisdiction of situs, and the absence of any simple or customary rule of jurisdiction as regards personal property capable of resolving the conflict satisfactorily in accordance with the dictates of ‘natural justice’ and common sense.” Id.

21 See Seligman, supra note 20, at 37-57.

22 Graetz and O’Hear, supra note 13, provide a detailed summary of Adams’s role as a Treasury official in forming the “original intent” of the U.S. international tax regime, as well as his central role in the early development of an international consensus regarding the taxation of cross-border income at the League of Nations and in other international forums.

23 If both countries impose tax at a 50 percent rate and neither country allows a deduction, the total tax paid will equal the total income earned, which is not sustainable. Even if one country permits a deduction for the other country’s tax, the total tax paid will equal $75, a level of taxation that many would regard as excessive and likewise unlikely to be sustainable.

24 Graetz and O’Hear, supra note 13, at 1033-1035; Seligman, supra note 20, at 137-138 (criticizing the U.S. FTC approach and recommending instead “the assignment of the income tax to the country of domicile”). This recommendation is based on a somewhat abstruse analysis developed at length in chapters V and VI of the Seligman book, which was based on (and quotes extensively from) a similar analysis adopted in a report issued by a committee of experts (notably including Seligman) under the auspices of the International Chamber of Commerce. See International Chamber of Commerce, First Congress, London, 1921, Double Taxation, Part I — Report of the Select Committee of the Chamber (1921), as cited and quoted throughout Chapter VI of the Seligman book.

25 Seligman later sniffed that “the United States is making a present of the revenue to other countries” and that the U.S. approach “is an over-generous and one-sided arrangement.” He proffered what may amount to an early accusation that the credit amounts to corporate welfare: “This may indeed be endurable or even desirable for a country which is anxious to favor the business enterprises carried on abroad by its own nationals; and which is ready to make sacrifices for that end. But in the ordinary run of cases the sacrifice would be too great.” Seligman, supra note 20, at 135.

26 “Every state insists upon taxing the non-resident alien who derives income from sources within that country, and rightly so, [or] at least inevitably so.” Adams, supra note 1, at 197.

27 “Prevention of double taxation, in short, calls for a self-denying ordinance in the home state — rather than concessions from a foreign state. The state of domicile must protect its own residents.” Adams, supra note 20, at 121.

28 Id. at 112-113.

29 See H.R. Rep. No. 65-767, at 11 (1918).

30 56 Cong. Rec. 677 (1918). In addition to chairing the Ways and Means Committee between 1915 and 1919, Kitchin was House majority leader during the same period. He was elected to represent North Carolina’s second congressional district 12 times, serving in the House from 1900 until his death in 1923.

31 Id. Kitchin also mentioned two other concerns about taxpayer responses to unrelieved double taxation: First, it would encourage U.S. taxpayers to form foreign subsidiaries whose earnings would not be subject to U.S. tax until repatriated, and second, it would encourage U.S. citizens residing abroad to give up their citizenship “in order to escape the large and double taxation imposed.” Id.

32 56 Cong. Rec. 661-702 (the length of the debate is noted at 678 and 702). Although some commentators have referred to these remarks as the views of “some Congressmen,” the fact that they were presented by the chair of the Ways and Means Committee (and majority leader) should entitle them to more weight than the views of a random member of Congress. Cf. Roswell Magill and William Schaab, “American Taxation of Income Earned Abroad,” 13 Tax L. Rev. 115, 118 (1958) (stating that “some Congressmen saw it as a method to encourage foreign trade and to prevent revenue loss through incorporation of foreign subsidiaries or expatriation”).

33 Graetz and O’Hear, supra note 13, at 1051-1054.

34 Id.

35 Tennessee v. Whitworth, 117 U.S. 129, 137 (1886).

36 Adams, supra note 1, at 197-198.

37 Id. at 198.

38 Id.

39 Id. at 197.

40 A mechanical approach would also be at odds with much of the judicial precedent under section 901, which has tended to read the rule in a manner consonant with its purpose, rather than merely mechanically. See, e.g., PPL Corp. v. Commissioner, 569 U.S. 329 (2013), rev’g 665 F.3d 60 (3d Cir. 2011), rev’g 135 T.C. 304 (2010) (finding that the IRS’s “rigid construction” of section 901 was “unwarranted” and “cannot be squared with the black-letter principle that ‘tax law deals in economic realities, not legal abstractions’” (PPL, 569 U.S. at 340, citation omitted)).

41 Revenue Act of 1921, section 238. For readers’ convenience, I reproduce here the redline between the 1918 and 1921 versions of the language providing a credit for “the amount of any income, war-profits and excess-profits taxes paid during the same taxable year to any foreign country, upon income derived from sources therein, or to any possession of the United States.”

42 For readers’ further convenience, I reproduce here the redline between the 1921 and 2021 versions of the language providing a credit for “the amount of any income, war profits, and excess profits taxes paid or accrued during the same taxable year to any foreign country or to any possession of the United States.” I know this is the third time I’ve used this redline, but (1) would you rather go digging back through these endless footnotes to find it again, and (2) it continues to be worth emphasizing that a century-old, but always vital, tax statute has been subjected to so little change.

43 Revenue Act of 1918, sections 222(a)(1) and 238(a).

44 Revenue Act of 1921, sections 222(a)(5) and 238(a).

45 The 1921 version of the corporate FTC under section 238(a) added a new proviso that immediately followed the operative language granting a credit for foreign income, war profits, and excess profits taxes: “Provided, That the amount of credit taken under this subdivision shall in no event exceed the same proportion of the taxes, against which such credit is taken, which the taxpayer’s net income (computed without deduction for any income, war-profits, and excess-profits taxes imposed by any foreign country or possession of the United States) from sources without the United States bears to its entire net income (computed without such deduction) for the same taxable year.” The individual FTC was similarly limited by new language added in section 222(a)(5).

46 The continuing lack of congruence in national approaches to sourcing rules is usefully addressed in several of the public comments responding to the November 2020 NPRM. See, e.g., Silicon Valley Tax Directors Group comments (Feb. 9, 2021); and Software Coalition comments (Feb. 9, 2021).

47 For a useful discussion of this history, see the recent USCIB comments, supra note 6, at 5-7.

48 For a concise summary of the FTC limitation’s checkered history, see Graetz and O’Hear, supra note 13, at n.141; and the recent USCIB comments, supra note 6, at 8-9.

49 This congressional focus on the FTC is understandable in light of its revenue significance. For example, IRS Statistics of Income data indicate that in 2016 the credit ceded more than one-quarter of the corporate tax base to foreign sovereigns, because the total pre-credit corporate income tax liability of roughly $321 billion was offset by FTCs of roughly $89 billion. SOI, “Corporate Foreign Tax Credit Table 1, 2016.”

50 In appropriate circumstances, a court may consider the import of legislative silence. The Supreme Court has accepted the “common sense” implications of such silence, by analogy to Sherlock Holmes solving the “Silver Blaze” mystery in part based on a dog’s failure to bark. “All in all, we think this is a case where common sense suggests, by analogy to Sir Arthur Conan Doyle’s ‘dog that didn’t bark,’ that an amendment having the effect petitioner ascribes to it would have been differently described by its sponsor, and not nearly as readily accepted by the floor manager of the bill.” Church of Scientology of California v. IRS, 484 U.S. 9, 17-18 (1987).

51 S. Rep. No. 77-1631, at 131 (1942).

52 In the decades after the enactment of section 903, the IRS somewhat curiously vacillated in its views on whether foreign gross-basis withholding taxes were creditable only as in-lieu taxes under section 903 or could be viewed as income taxes in their own right. See D. Kevin Dolan, “General Standards of Creditability Under the Sections 901 and 903 Final Regulations — New Words, Old Concepts,” 13 Tax Mgmt. Int’l J. 167, 174-176 (1984). Apparently, the IRS was at least partially motivated by discomfort with the implication that if foreign withholding taxes are creditable only under section 903, U.S. gross-basis taxes under sections 881, 882, 1441, and 1442 must not be income taxes either. Regardless of that historical discomfort, however, final regulations issued in 1983 definitively resolved the issue on the side of treating gross-basis withholding taxes as creditable only under section 903.

53 Indeed, the IRS sometimes argued that section 903 should be interpreted to require the amount of an in-lieu tax to be comparable to the amount of the underlying income tax. See id. at 180. But the IRS abandoned that view in the existing final regulations and has not proposed to revive it in the November 2020 NPRM.

54 Readers are asked to take judicial notice of this well-known history; please don’t make me write yet another stultifying footnote proving what you already know.

55 I guess this is one boring footnote I can’t skip: See 44 F.R. 36071 (June 20, 1979) (proposed regulations); 45 F.R. 75692 (Nov. 17, 1980) (reproposed regulations); T.D. 7739, 45 F.R. 75647 (Nov. 17, 1980) (temporary regulations); 48 F.R. 14641 (Apr. 5, 1983) (re-reproposed regulations); and T.D. 7918, 48 F.R. 46272 (Oct. 12, 1983) (final regulations).

56 Based on the final regulations’ consistency with prior authorities, one commentator concluded that “the regulations under section 901 do not significantly modify the definitional standards of what is an income tax as set forth in prior cases and rulings.” The same commentator further stated that while the creditability rules under section 903 were more significantly altered by the final regulations, “these changes have far more academic interest than practical importance,” concluding that the final rules “will have very little practical effect in making noncreditable taxes which would have been considered creditable under prior law.” Dolan, supra note 52, at 168. Unfortunately the same could not be said of the November 2020 NPRM.

57 Having once been an IRS reg drafter myself (in the previous century), I know that it can be tempting to discount the work of prior generations of reg drafters. But having now joined a prior generation of reg drafters, I can see the error of my youthful hubris in this regard.

58 The rules summarized here are set forth in prop. reg. section 1.901-2(c).

59 See, e.g., OECD, “Transfer Pricing in Brazil: Towards Convergence With the OECD Standard” (undated) (analyzing “significant gaps and divergences” between Brazilian transfer pricing rules and OECD guidance on the arm’s-length principle and concluding that “Brazil’s transfer pricing regime is not fully aligned with the international standard, the ‘arm’s length principle,’ embodied in Article 9 of the OECD Model Tax Convention”).

60 Prop. reg. section 1.903-1(c)(1)(iv).

61 Preamble to REG-101657, 85 F.R. at 72088.

62 See generally Graetz and O’Hear, supra note 13, passim.

63 Id.

64 Preamble to REG-101657, 85 F.R. at 72088. However, this attempt to link the definition of an income tax to traditional notions of taxing jurisdiction ignores the history of section 901 and in particular the 1921 deletion of the source rule contained in the 1918 version of the statute.

65 Indeed, extraterritorial assertions of taxing jurisdiction, and U.S. legislative responses to them, have been occurring for many decades without previously being seen to implicate the definition of a foreign income tax. For example, since at least 1939, section 891 has allowed the rates of U.S. tax imposed on citizens and corporations of a foreign country to be doubled, based on a presidential finding that the country imposes discriminatory or extraterritorial taxes on U.S. citizens or corporations.

66 As a more specific example, the rule would seem to throw into further question the creditability of the excise tax (likewise referred to as “novel” by Treasury and the IRS) that Puerto Rico imposes on the acquisition from related persons of property manufactured in Puerto Rico and services performed there. See Notice 2011-29, 2011-16 IRB 663. See also the comments on the November 2020 NPRM submitted by the Puerto Rico Treasury secretary (Sept. 30, 2020).

67 Preamble to REG-101657-20, 85 F.R. at 72089.

68 The trade would be beneficial for the United States only if countries imposing DSTs agree to a consensus approach that requires them to impose narrower DSTs than those they would otherwise impose. But that may be difficult to achieve given the modest leverage that Treasury has in these discussions, even with the threat of non-creditability under the jurisdictional nexus rule. (The amount of leverage the United States derives by threatening to double tax its own taxpayers could be interestingly debated — but not here.)

69 Preamble to REG-101657-20, 85 F.R. at 72089.

70 See sections 894 and 7852(d).

71 For example, SOI data for 2016 indicate that nearly 70 percent of FTCs claimed in that year were “deemed paid” credits for taxes paid by foreign affiliates. SOI, supra note 49 (showing deemed paid credits of roughly $61.3 billion versus direct credits of roughly $27.2 billion). SOI data for post-TCJA years are not yet available, but given the enactment of the section 951A tax nominally imposed on GILTI (but in fact imposed on most CFC earnings not otherwise taxed under subpart F), CFC taxes seem likely to continue representing a substantial portion of potentially creditable foreign taxes. The inapplicability of treaties to those taxes would substantially limit the utility of treaty confirmations of the creditability of covered taxes.

72 Reg. section 1.901-2(a)(3)(i).

73 Reg. section 1.901-2(b)(1).

74 Id.

75 This principle was addressed in the decisions of courts at all levels, including the Supreme Court, in cases addressing the creditability of the windfall profits tax imposed by the United Kingdom after the privatization of some nationalized industries. See PPL, 569 U.S. at 329; and Entergy Corp. v. Commissioner, T.C. Memo. 2010-197, aff’d, 683 F.3d 233 (5th Cir. 2012). Although these cases were decided long after the current regulations were finalized, they rest on the same earlier authorities that are reflected in the substance-based approach adopted by the regulations.

76 Reg. section 1.901-2(b)(2)(i).

77 Reg. section 1.901-2(b)(3)(i).

78 Preamble to REG-101657-20, 85 F.R. at 72090.

79 Prop. reg. section 1.901-2(b)(3)(ii)(A), Example 1. On the same facts, current reg. section 1.901-2(b)(3)(ii)(A), Example 1 reaches the opposite conclusion.

80 Id.

81 For further discussion, see infra notes 89, 122, and 123, and accompanying text.

82 45 F.R. 75692 (Nov. 17, 1980) (reproposed regulations); T.D. 7739, 45 F.R. 75647 (Nov. 17, 1980) (temporary regulations).

83 48 F.R. 14641 (Apr. 5, 1983) (re-reproposed regulations).

84 T.D. 7918, 48 F.R. 46272 (Oct. 12, 1983) (final regulations).

85 Phillips Petroleum Co. v. Commissioner, 104 T.C. 256 (1995).

86 PPL, 569 U.S. at 329; Entergy, T.C. Memo. 2010-197, aff’d, 683 F.3d 233.

87 Realized cash receipts gave way to the accrual method, which itself has given way to financial statement income. Compare section 451 (2018), with section 451 (2012).

88 See, e.g., sections 475 (mark-to-market accounting method for securities dealers); 1256 (mark-to-market for some financial contracts); and 1296 (mark-to-market for interests in passive foreign investment companies).

89 See discussion, supra note 81, and infra notes 122 and 123, and accompanying text.

90 The format of the November 2020 NPRM, with its amending instructions adding and removing text from various places in the final regulations, makes it hard to get a clear sense of the scope of the proposed changes. So I cobbled together a redline between the existing final regulations and the regulations as they would stand if the NPRM were finalized as proposed. The appendix is excerpted from that redline. It compares reg. section 1.901-2(b)(4) with prop. reg. section 1.901-2(b)(4). But caveat lector: I found the redline useful but can offer no warranties, express or implied, as to its accuracy. One potential source of inaccuracy is that I had to manually create the full text of the revised regulation by following the amending instructions set forth in the NPRM, requiring a challenging level of attention to detail.

91 Prop. reg. section 1.901-2(b)(4)(i).

92 But in the regulations’ defense, similar examples can be found elsewhere. For example, under the section 954(h) active financing exception to subpart F, the statute helpfully defines the term “customer” as “any person which has a customer relationship with [the CFC] and which is acting in its capacity as such.” In other words, a customer is a customer.

93 Prop. reg. section 1.901-2(b)(4)(i)(B)(2).

94 Id.

95 Id.

96 See H.R. Rep. No. 115-466, at 645-654 (2017) (Conf. Rep.).

97 The proposed regulations’ embrace of section 163(j) as the apotheosis of interest deduction limitations seems particularly strange given how frequently, recently, and substantially that code provision has been rewritten by Congress, as discussed in the text accompanying note 98, infra.

98 See further discussion at Section II.E, below.

99 Reg. section 1.901-2(b)(4)(i)(B).

100 Prop. reg. section 1.902-1(b)(4)(i)(A).

101 Prop. reg. section 1.901-2(b)(4)(i)(B)(2).

102 For example, several provisions in the code and regulations permit inclusions or deductions to be based on per diem amounts or standard mileage costs rather than actual costs.

103 As a corporate tax example of the ways in which the varied purposes pursued by Congress can modify a rule over time, consider section 163(j). It was first enacted to address base erosion concerns in relation to foreign-owned corporations but was then revised as a part of the TCJA to limit incentives to use debt in any business’s capital structure. H.R. Rep. No. 115-409, at 247 (2017) (Ways and Means Committee report on the TCJA).

104 See OECD, “Limiting Base Erosion Involving Interest Deductions and Other Financial Payments — Action 4: 2015 Final Report” (2015).

105 For relevant background on section 903, see Section II.C.2, above.

106 See, e.g., definitions of in lieu of from the Cambridge Dictionary and Merriam-Webster.

107 Preamble to REG-101657-20, 85 F.R. at 72095.

108 Id. (citing Metropolitan Life Insurance Co. v. United States, 375 F.2d 835 (Ct. Cl. 1967)).

109 Metropolitan Life, 375 F.2d at 837-840.

110 Prop. reg. section 1.903-1(c)(1)(ii).

111 Compare, for example, the largely unedifying brawls of foreign law experts in such cases as Riggs National Corp. v. Commissioner, 107 T.C. 301 (1996); and Amoco Corp. v. Commissioner, T.C. Memo. 1996-159.

112 Metropolitan Life, 375 F.2d at 840.

113 Id.

114 Which is not even to mention that any two economists would likely generate at least three opinions regarding the correct economic measurement of income.

115 Preamble to REG-101657-20, 85 F.R. at 72078 and 72087.

116 See Patients Mutual Assistance Collective Corp. v. Commissioner, 151 T.C. 176, 208 (2018).

117 See, e.g., section 280E (denial of deductions for those trafficking in controlled substances); and section 882 and reg. section 1.874-1(a) (denial of deductions for foreign persons that fail to timely file returns).

118 See White v. United States, 305 U.S. 281, 292 (1938).

119 See, e.g., sections 163(j) (providing limitations on the deductibility of interest); 170(b)(2) (limitation for charitable deductions); and 55 (individual alternative minimum tax).

120 See, e.g., the U.S. system’s 2017 lurch from generally deferring U.S. taxation of the earnings of CFCs to a system in which most of those earnings are currently taxed at a reduced rate but some are exempt.

121 Preamble to REG-101657-20, 85 F.R. at 72087.

122 See, e.g., section 7701(o) (codification of the economic substance doctrine); Minnesota Tea Co. v. Helvering, 302 U.S. 609, 613 (1938) (stating that “a given result at the end of a straight path is not made a different result because reached by following a devious path”); Gregory v. Helvering, 293 U.S. 465, 470 (1935) (“To hold otherwise would be to exalt artifice above reality and to deprive the statutory provision in question of all serious purpose.”); and True v. United States, 190 F.3d 1165, 1174 (10th Cir. 1999) (stating that substance over form is a “fundamental tax principle” and applies to “look beyond the taxpayers’ characterization” of the challenged business transactions).

123 See PPL, 569 U.S. at 331 (“Consistent with precedent and the Tax Court’s analysis below, we apply the predominant character test using a commonsense approach that considers the substantive effect of the tax.”).

124 Bragdon v. Abbott, 524 U.S. 624, 645 (1998); see also Antonin Scalia and Bryan A. Garner, Reading Law 322 (2012) (stating that “if a statute uses words or phrases that have already received authoritative construction by the jurisdiction’s court of last resort, or even uniform construction by inferior courts or a responsible administrative agency, they are to be understood according to that construction”); accord Caleb Nelson, Statutory Interpretation 483-484 (2011).

125 Helvering v. Reynolds, 313 U.S. 428, 432 (1948).

126 The discussion in this paragraph merely skims the surface of some deep waters, noting the potential relevance of the canons of construction without attempting to describe their precise scope or the scope of the government’s “continuing rulemaking power.” As noted in the text, the primary focus here should be on the language and history of the statute, which are more than enough to show that the November 2020 NPRM is headed in the wrong direction.

127 The quoted language is cited in dozens of places, notably including the IRS website. See IRS, “Tax Quotes” (updated May 25, 2021). The comment was made in a 1774 parliamentary debate on the issues of American taxation that soon led to the Revolution. See Eighteenth Century Collections Online, “Speech of Edmund Burke, Esq. on American Taxation, April 19, 1774.” Burke describes the history of British taxation of the American colonies, including the stamp duties that were disastrously reinstated, in part, under the leadership of Chancellor of the Exchequer Charles Townshend. The quotation appears in Burke’s discussion of Townshend’s role: “Here this extraordinary man, then Chancellor of the Exchequer, found himself in great straits. To please universally was the object of his life; but to tax and to please, no more than to love and to be wise, is not given to men. However he attempted it.”

128 Please see the caveat lector in note 90 of the main text regarding the highly unofficial redline from which this material was excerpted.

END FOOTNOTES

Copy RID