Menu
Tax Notes logo

Is GILTI Constitutional?

Posted on Jan. 11, 2021
Reuven S. Avi-Yonah
Reuven S. Avi-Yonah

Reuven S. Avi-Yonah is the Irwin I. Cohn Professor of Law at the University of Michigan. He thanks Philip Baker, Steve Bank, Allison Christians, Kim Clausing, Tsilly Dagan, Mike Durst, Nir Fishbien, Cliff Fleming, Ed Fox, Jeff Kadet, Ajay Mehrotra, David Miller, Susan Morse, Steve Ratner, Mathias Reimann, H. David Rosenbloom, Luís Eduardo Schoueri, Fadi Shaheen, and Stephen E. Shay for helpful comments.

In this article, the author argues that the global intangible low-taxed income regime may be an unconstitutional attempt to tax the foreign-source income of foreign entities, and he offers an alternative.

Copyright 2021 Reuven S. Avi-Yonah. All rights reserved.

At first glance, the question in the headline may appear to be frivolous. The global intangible low-taxed income regime must be constitutional — GILTI builds on subpart F, which built on the foreign personal holding company (FPHC) regime that was declared to be constitutional in Eder back in 1943.1

But perhaps it isn’t that easy. I consider this issue not to argue the United States should overturn GILTI, but rather to suggest GILTI should be replaced with something better — something that is constitutional — because it may not be. GILTI depends on the deemed dividend concept, a concept invented in 1937 to tax U.S. shareholders on the passive income of FPHCs. Three years earlier, Congress enacted the personal holding company (PHC) provision, which was identical to the FPHC regime except that the PHC was a domestic corporation and was therefore taxed at the corporate level — or if it was foreign, the PHC was only taxed on U.S.-source income.2 The PHC provisions did not address foreign corporations earning foreign-source income, because Congress believed it lacked jurisdiction to tax those entities on foreign-source income.3 Therefore, when it enacted the FPHC provision to close this loophole, Congress invented the concept of deemed dividends, perhaps reasoning that because the U.S. individual shareholders controlled the FPHC, they could make it pay a dividend any time (under the constructive receipt doctrine developed by the Supreme Court in 1930).4 Congress also provided that an FPHC could not be a PHC. In 2004 the FPHC provision was repealed — it was redundant after the enactment of passive foreign investment company regime in 1986 — and foreign corporations were excluded from the scope of the PHC provisions.5

Congress’ belief that it could not tax foreign corporations on foreign-source income was well grounded. In Cook, the Supreme Court explained that jurisdiction to tax must rest on one of two bases: nationality/residence or territoriality.6 The first (in personam) basis justified taxing U.S. residents on worldwide income, and it also justified taxation of U.S. citizens on worldwide income even if they permanently lived outside the United States (the case itself involved a U.S. citizen living in Mexico who derived income from real estate located in Mexico). The second (in rem) basis justified taxing nonresident aliens on U.S.-source income. But there was no basis for taxing foreigners on foreign-source income.7

The FPHC deemed dividend rule was challenged as unconstitutional in Eder. The taxpayer in Eder was a U.S. shareholder in a Colombian FPHC that was forbidden by Colombian law from distributing more than $1,000 per year to its shareholders. In rejecting the challenge, the Second Circuit Court of Appeals (with Judge Jerome Frank writing for a unanimous panel that included Learned Hand) justified the deemed dividend concept based on the following analysis:

That the result under the statute here before us may be harsh is no answer to the government’s position; the purpose of Congress was to deal harshly with “incorporated pocketbooks,” and the motive of a particular taxpayer who has such a “pocketbook” we have held to be irrelevant. . . . Interpreting the statute to bring about such a consequence does not render the statute unconstitutional; the Congressional purpose was valid and the method of taxation was a reasonable means to achieve the desired ends. [Internal citations removed.]

This is rather conclusory: There is no real analysis regarding why Congress needed to deal harshly with incorporated pocketbooks or whether another method of taxation — for example, a tax on dividends with an interest charge, as adopted later in the PFIC regime — would have sufficed and prevented the harshness of taxation when dividends were prohibited under foreign law.8 Nor is there any discussion of the limits that international law places on tax jurisdiction or of Cook. This is probably because the taxpayer argued the case based on Macomber, not Cook, and in that context Frank had it right — Macomber does not bar Congress from imposing a tax on deemed dividends, and it had already been limited to its facts in 1936 when the Supreme Court found that a distribution of common stock on preferred stock was taxable income.9 Frank, a great supporter of the New Deal who served as chair of the SEC before being elevated to the federal bench, simply brushed all these issues aside.

When subpart F was enacted in 1962, Congress lifted the deemed dividend concept from the FPHC provisions (and FPHC income is one of the major components of subpart F income to this day).10 This was challenged on constitutional grounds in Garlock, a case involving a Panamanian controlled foreign corporation, and rejected by the Second Circuit in the following words:

The argument that 951, which requires an American shareholder to include in income his pro rata share of a CFC’s profits, is unconstitutional we think borders on the frivolous in the light of this court’s decision in Eder v. Commissioner, 138 F.2d 27, 28 (1943). That case held constitutional the foreign personal holding provisions of the income tax laws upon which subpart F was patterned, permitting taxation of United States shareholders on the undistributed net income of Colombian corporations even though Colombian law made the taxpayer unable to receive such income in the United States in excess of $1,000 per month.11

There is not much analysis here either. Further, the other cases that the Garlock case relies on (that is, besides Eder) don’t truly support the outcome. Alvord held for the taxpayer, finding that the U.S. shareholder was not required to include disputed amounts in income when, under the specific circumstances, the IRS forbade distributions out of a FPHC, without addressing constitutional issues.12 In Marsman, the Fourth Circuit considered whether income earned by a FPHC before its sole shareholder became a U.S. resident could be taxed.13 The court resolved that issue in the taxpayer’s favor, but the constitutional issue was not raised.

The basic problem therefore remains: Is it clear that the deemed dividend mechanism is constitutional under Cook, which remains valid authority?14 This issue was not addressed in Eder or Garlock. The argument that it is unconstitutional would rest on the assertion that deemed dividends are economically equivalent to taxing a CFC on foreign-source income, which is unconstitutional under the jurisdictional limits of Cook and is also a violation of international law.15

Interestingly, when Cook was decided, foreign subsidiaries of U.S. parent corporations were covered by a mandatory consolidation regime adopted in 1917. The War Revenue Act of 1917 (specifically regulation 41, articles 77 and 78) gave the commissioner the authority to require related corporations to file consolidated returns “whenever necessary to more equitably determine the invested capital or taxable income.” In 1921 a statute authorized the commissioner to consolidate the accounts of affiliated corporations “for the purpose of making an accurate distribution or apportionment of gains, profits, income, deductions, or capital between or among such related trades or business.” That legislation was enacted, in part, because of the tax avoidance opportunities afforded by possessions corporations, which were not eligible to file consolidated returns with their domestic affiliates. The mandatory consolidation regime was repealed in 1928 and replaced by the first version of IRC section 482.16 This might have occurred because Congress realized that mandatory consolidation violated the jurisdictional limits set forth in Cook.

If the deemed dividend mechanism is unconstitutional, it raises serious issues for subpart F, which relies on deemed dividends17 — and, in turn, it undermines GILTI, which relies on subpart F.

The issue could have been raised in the recent Moore case, which attacked the constitutionality of the transition tax imposed by IRC section 965 in 2017 as a violation of the realization requirement and as a tax on wealth, not income.18 In Moore the district court correctly held that realization is no longer a prerequisite, since later cases have narrowed the application of Macomber to its facts, and that IRC section 965 applies to income, not wealth. As the court explains:

Given the cabining of Macomber by the Supreme Court and the clear departure from it by other courts, there is no reason for this court to conclude that Macomber currently controls whether the [mandatory repatriation tax] is an income tax. Accordingly, the [mandatory repatriation tax] does not violate the Apportionment Clause, as it is a tax on income rather than a direct tax.19

But the Moore plaintiffs did not raise the jurisdictional limits issue. The question is whether some other plaintiff might. While CFCs may lack standing to sue because they are not subject to tax, their U.S. parents can sue on their behalf because they are subject to tax under GILTI — a tax that raises more than $10 billion each year.20

If the Supreme Court declared GILTI (and subpart F) unconstitutional, what could Congress do?

Presumably, mandatory consolidation of the parent and the CFCs or a mandatory check-the-box election to treat CFCs as branches of the parent would be off-limits for the same reason that deemed dividends are: They would also amount to taxing the CFCs on foreign-source income. Perhaps a formal case could be made that the tax applies to the U.S. parent and not to the CFCs and that ignoring the separateness of subsidiaries has become the norm in other areas of law (for example, bankruptcy or environmental law). Another alternative that would presumably be constitutional is to tax the parent on the change in value of the CFCs’ stock resulting from the CFCs’ earnings. Like the PFIC mark-to-market regime, a tax on the unrealized appreciation of the parent’s stock is not the same as a tax on the CFCs. But the tax base would be different — the value of the stock is not always perfectly correlated with the underlying earnings — and, because the stock is not publicly traded, valuation may lead to disputes.

Congress could mandate that the value of a CFC’s shares for mark-to-market purposes is equivalent to the CFC’s earnings, but that would once again run into the constitutional issue because it is economically equivalent to taxing the CFC on foreign-source income.

The best solution would be to redefine the residency of CFCs, that is, to change the definition of a domestic person in IRC section 7701(a)(4) to include CFCs (as defined in IRC section 957). Nothing in the Constitution or in Cook precludes Congress from doing that. Article 4(1) of U.S. tax treaties defines resident as “any person who, under the laws of that Contracting State, is liable to tax therein by reason of his domicile, residence, citizenship, place of management, place of incorporation, or any other criterion of a similar nature.” Article 4(4) further states that when “by reason of the provisions of paragraph 1 of this Article a company is a resident of both Contracting States, such company shall not be treated as a resident of either Contracting State for purposes of its claiming the benefits provided by this Convention.” To the extent this language precludes redefining CFCs as U.S. residents, Congress can override the treaties.

There have been many proposals to redefine corporate residency, such as including a managed and controlled standard (like most of our trading partners do), although that would raise issues regarding where a CFC is managed. IRC section 7874 (2004) already redefines the residency of some inverted corporations as domestic based on U.S. share ownership; this rule has not been challenged, just avoided.21

Redefining residency would subject all CFCs to current tax in full, and it would also permit them to join in a consolidated return and use their losses to offset U.S. domestic income.22 But that is in fact the right outcome, especially in a world in which the taxpayer controls whether a foreign subsidiary is treated as such or as a branch, and in which GILTI applies to all CFCs as a group.23

FOOTNOTES

1 Eder v. Commissioner, 138 F.2d 27 (2d Cir. 1943).

2 IRC section 542 (1934). See Jerome B. Libin, “Personal Holding Companies and the Revenue Act of 1964,” 63 Mich. L. Rev. (1965).

3 John H. Alexander, “Foreign Personal Holding Companies and Foreign Corporations That Are Personal Holding Companies,” 67 Yale L.J. 1173 (1958).

4 The Joint Committee on Taxation, “Report on Tax Evasion and Avoidance” (Comm. Print 1937), explained the FPHC proposal as follows:

This proposal recommends a method of taxation which is a departure from any previously used with respect to corporate income. The committee feels, however, that this innovation is necessary to protect the revenue and prevent further use of one of the most glaring loopholes now existing. The proposal would affect only foreign corporations which are owned 50 percent or more by five American citizens or residents (including members of their families) and which have the same type of investment income which makes a domestic corporation subject to tax as a personal holding company. Real foreign operating companies or widely held holding companies are not included. . . . The committee believes that the recommendation is not any more drastic than the situation requires.

Previous congressional hearings focused on examples like Col. Jacob Schick, inventor of the electric razor, who transferred his patent to a tax haven and then expatriated, living off the accumulated royalties. See the discussion of the intent of FPHC provision in Alvord v. Commissioner, 277 F.2d 713 (4th Cir. 1960), emphasizing that the FPHC provision only applies when the taxpayer can require a distribution. Under U.S. Treas. reg. section 1.451-2, “dividends on corporate stock are constructively received when unqualifiedly made subject to the demand of the shareholder.” See Corliss v. Bowers, 281 U.S. 376 (1930).

5 IRC section 542(c)(5). The IRS believes the accumulated earnings tax applies to foreign corporations’ foreign-source income, but this does not seem to have been tested in court. See IRC section 532(b) (not excluding foreign corporations other than PFICs).

6 Cook v. Tait, 265 U.S. 47 (1924) (“Plaintiff assigns against the power, not only his rights under the Constitution of the United States, but under international law, and in support of the assignments cites many cases. It will be observed that the foundation of the assignments is the fact that the citizen receiving the income and the property of which it is the product are outside of the territorial limits of the United States. These two facts, the contention is, exclude the existence of the power to tax. Or to put the contention another way, to the existence of the power and its exercise, the person receiving the income and the property from which he receives it must both be within the territorial limits of the United States to be within the taxing power of the United States. The contention is not justified . . . the basis of the power to tax was not and cannot be made dependent upon the situs of the property in all cases, it being in or out of the United States, nor was not and cannot be made dependent upon the domicile of the citizen, that being in or out of the United States, but upon his relation as citizen to the United States and the relation of the latter to him as citizen. The consequence of the relations is that the native citizen who is taxed may have domicile, and the property from which his income is derived may have situs, in a foreign country and the tax be legal — the government having power to impose the tax.”).

7 The American Law Institute Restatement (Fourth) of Foreign Law divides jurisdiction into jurisdiction to prescribe, adjudicate, and enforce. For jurisdiction to prescribe, the restatement specifies that a state has jurisdiction if it falls under one of six headings: territory (the conduct occurs within the state’s territory), effects (it has significant effects within its territory), active personality (there is a personal connection between the object of the rule and the state (nationality or residence)), passive personality (the conduct affects nationals or residents), the protective principle (involving acts that threaten the sovereignty and security of the United States), and universal jurisdiction (applies to some crimes that do not require a personal or territorial connection). None of these bases for jurisdiction appear to justify imposing a U.S. tax on nonresidents’ foreign-source income. For case law rejecting attempts by the United States to require controlled foreign corporations to abide by U.S. sanctions as exceeding U.S. jurisdiction to prescribe, see Reuven S. Avi-Yonah, “National Regulation of Multinational Enterprises: An Essay on Comity, Extraterritoriality, and Harmonization,” SSRN (Aug. 2002) (citing, e.g., Fruehauf Corp. v. Massardy, Court of Appeals of Paris, 14th Chamber (May 22, 1965); and Compagnie Europeenne des Petroles SA v. Sensor Nederland BV, District Court of the Hague (Sept. 17, 1982)). Of course, some U.S. courts may take a different view of the matter. See also Avi-Yonah, “Does Customary International Tax Law Exist?” in Research Handbook on International Taxation 2 (2020) (arguing that the jurisdictional limits are customary international law). Nationality jurisdiction as approved in Cook v. Tait is a well-established aspect of international law, despite being unusual in the tax context.

8 Other courts sought to mitigate the harshness. See, e.g., Alvord, 277 F.2d 713.

9 Eisner v. Macomber, 252 U.S. 189, 207 (1920). See also Koshland v. Helvering, 298 U.S. 441, 443 (1936) (noting that Macomber “affected only the taxation of dividends declared in the same stock as that presently held by the taxpayer”); and Moore v. United States, No. C19-1539-JCC (W.D. Wa. 2020) (for a discussion of Moore, see infra notes 18-19 and related text).

10 The 1961 JCT report on the Kennedy administration proposal concluded that the tax was unconstitutional under Macomber, distinguishing the FPHC as an antiavoidance rule. Joint Committee on Internal Revenue Taxation, “Constitutional Power to Tax Shareholders on the Undistributed Income of a Corporation” (1961).

11 Garlock Inc. v. Commissioner, 489 F.2d 197 (2d Cir. 1973). The main issue in Garlock was not the constitutional challenge but the taxpayer’s attempt to avoid CFC status by selling 51 percent of voting preferred shares to a friendly foreign investor, which the court also rejected. Congress then changed the law to “by vote or value.” See also Whitlock v. Commissioner, 59 T.C. 490, 507 (1972) (upholding subpart F against a Macomber-based challenge because Congress could ignore the separateness of the CFC (“If Congress has legislatively declared that it will bypass the corporate entity in taxing certain foreign-source income, the mere circumstance of petitioners’ interposition of a foreign corporate framework between themselves and income over which they had complete control would certainly be no constitutional barrier to the taxation of that income to petitioners.”)); and Dougherty v. Commissioner, 60 T.C. 917, 928-929 (1973) (“the doctrine of Eisner v. Macomber, supra, does not prevent Congress from bypassing the corporate entity in determining the incidence of Federal income taxation. This principle has special application in the case of controlled foreign corporations, which Congress has found to be frequently abused by their U.S. shareholders as vehicles for the avoidance of tax. S. Rept. No. 1881, supra, 1962-3 C.B. at 784-785. In subpart F, Congress has singled out a particular class of taxpayers, U.S. shareholders, whose degree of control over their foreign corporation allows them to treat the corporation’s undistributed earnings as they see fit. Compare Corliss v. Bowers, 281 U.S. 376, 378 (1930), where the Supreme Court said, in connection with the taxability of a grantor on the income of a revocable trust, ‘The income that is subject to a man’s unfettered command and that he is free to enjoy at his own option may be taxed to him as his income, whether he sees fit to enjoy it or not.’”). Both of these cases were decided on the basis of constructive receipt, but it is not clear that constructive receipt can apply in the case of a U.S. shareholder like Moore who only has slightly more than 10 percent of the vote and does not control dividend decisions. Nor does “bypassing the corporate entity” apply to subpart F because some of the income of the CFC is included and some is not, while treating the CFC as a branch would entail including all of its income as well as its losses on the U.S. parent’s return.

12 Alvord, 277 F.2d 713.

13 Marsman v. Commissioner, 205 F.2d 335 (4th Cir. 1953), aff’d, 216 F.2d 77 (4th Cir. 1954).

14 Cook is a crucial decision. It underpins the U.S. system of worldwide taxation for citizens, which has been part of U.S. tax law since 1861. The Supreme Court would be reluctant to overrule it for this reason, however misguided it may be. See Avi-Yonah, “The Case Against Taxing Citizens,” University of Michigan Law and Economics Empirical Legal Studies Center Paper No. 10-009 (Mar. 25, 2010).

15 Some argue that GILTI violates U.S. tax treaties by imposing tax on the business profits of CFCs resident in treaty partner countries that do not have a U.S. permanent establishment, but I agree with H. David Rosenbloom and Fadi Shaheen that GILTI is acceptable for treaty purposes as a tax on the U.S. parent and covered by the saving clause in article 1(4). See Rosenbloom and Shaheen, “The TCJA and the Treaties,” Tax Notes Int’l, Sept. 9, 2019, p. 1057.

16 T.D. 2694 (1918). See also War Revenue Act of 1917, ch. 63, 40 Stat. 300 (1917); Revenue Act of 1921, ch. 136, section 240(d), 42 Stat. 260 (1921) (reenacted in Revenue Act of 1924, ch. 234, section 240(d), 43 Stat. 288 (1924); and Revenue Act of 1926, ch. 27, section 240(f), 44 Stat. 46 (1926)). See generally Avi-Yonah, “The Rise and Fall of Arm’s Length: A Study in the Evolution of U.S. International Taxation,” 15 Va. Tax Rev. 89 (1995) (updated version available at 9 Fin. and Tax L. Rev. 310 (2006)).

17 This is not a problem for “incorporated pocketbooks,” which are taxed under the PFIC rules; the problem that led Congress to invent deemed dividends in 1937 does not arise.

18 Interestingly, Moore was a 13 percent minority shareholder; arguably, he could not have forced a distribution.

19 Moore v. United States, No. C19-1539-JCC. See also Annagabriella Colón, “U.S. District Court Dismisses Repatriation Tax Case,” Tax Notes Int’l, Nov. 23, 2020, p. 1108.

20 See Thomas Horst, “Preliminary Estimates of the Likely Actual Revenue Effects of the TCJA’s Provisions,” Tax Notes Int’l, Sept. 16, 2019, p. 1153.

21 See also J. Clifton Fleming Jr., Robert J. Peroni, and Stephen E. Shay, “Defending Worldwide Taxation With a Shareholder-Based Definition Of Corporate Residence,” 2016 BYU L. Rev. 1681 (Dec. 2016).

22 This assumes, as would generally be the case, that the CFC is at least 80 percent owned by the U.S. parent by vote or value (most CFCs are 100 percent owned by a single U.S. parent). I would support limiting that redefinition of residency to CFCs eligible for consolidation, because CFCs with significant foreign minority shareholders raise difficult questions. An 80 percent threshold is also consistent with IRC section 7874.

23 See, e.g., Fleming, Peroni, and Shay, “Expanded Worldwide Versus Territorial Taxation After the TCJA,” Tax Notes, Dec. 3, 2018, p. 1173; and Avi-Yonah, “Hanging Together: A Multilateral Approach to Taxing Multinationals,” University of Michigan Public Law Research Paper No. 364 (Aug. 4, 2015).

END FOOTNOTES

Copy RID