Expanded Worldwide Versus Territorial Taxation After the TCJA

Stephen E. Shay

Robert J. Peroni

J. Clifton Fleming Jr.

J. Clifton Fleming Jr. is the Ernest L. Wilkinson Chair and professor of law at the Brigham Young University J. Reuben Clark Law School. Robert J. Peroni is the Fondren Foundation Centennial Chair for Faculty Excellence and professor of law at the University of Texas School of Law. Stephen E. Shay is a senior lecturer at Harvard Law School. The authors thank those who commented on earlier drafts of this report: Yariv Brauner, Patrick Driessen, Rebecca Kysar, Omri Marian, Gladriel Shobe, participants in a tax panel at the Southeastern Association of Law Schools 2018 annual meeting, and participants in an October workshop at the Vienna University of Economics and Business.

In this report, the authors analyze whether the Tax Cuts and Jobs Act, through its purely outbound international provisions, has caused any movement toward either territorial or expanded worldwide taxation of U.S. multinationals’ foreign-source active business income. The report also recommends a path for getting from the TCJA to expanded worldwide taxation.

The views expressed in this report are those of the authors and do not reflect those of their universities, any organization for which any of the authors serves as an officer or renders pro bono services, or, in the case of Shay, any client. Shay discloses on his Harvard website faculty page activities not connected with his position at Harvard Law School, one or more of which may relate to the subject matter of this report.

Copyright 2018 J. Clifton Fleming Jr., Robert J. Peroni, and Stephen E. Shay.
All rights reserved.

I. Introduction

In the run-up to enactment of the Tax Cuts and Jobs Act (P.L. 115-97),¹ one of the principal U.S. tax policy issues was how foreign-source active business income of U.S. multinational enterprises² should be taxed by the United States if the system of deferring U.S. tax on active foreign income of a foreign subsidiary was ended. Should active foreign income be taxed under a territorial or exemption system — that is, bear no residual U.S. tax³ — or should it be subjected to expanded worldwide taxation⁴ — that is, current taxation at regular U.S. rates coupled with a credit for foreign income taxes paid or accrued, but limited to the U.S. tax on the foreign-source income as measured for U.S. tax purposes?⁵

The opposing sides were not without common ground. Both agreed that the existing U.S. system for taxing the foreign-source active business income of U.S. MNEs needed to be changed because it generally did not impose U.S. tax until the active income of foreign subsidiaries was repatriated, either through dividends or by sale of subsidiary stock at a price reflecting accumulated foreign-source income. This was problematic for the advocates of territoriality because it required payment of a U.S. tax before controlled foreign corporation earnings could be directly accessed by U.S. parent corporations when no home country tax would have to be paid by MNEs from many competitor countries. It was unacceptable to worldwide taxation advocates because the resulting deferral of U.S. tax effectively created a preferential tax rate for CFC income that encouraged U.S. MNEs to locate operations in and engage in income shifting to low-tax foreign countries. The two sides were, however, at loggerheads over whether the foreign-source active business income of U.S. MNEs should bear a current U.S. tax at regular rates, subject to a limited foreign tax credit, or should bear no U.S. tax at all. Neither view prevailed in the TCJA.

One major purpose of this report is to take a preliminary look at the extent to which the TCJA caused a degree of movement in either direction. The substantive analysis will begin by examining the Joint Committee on Taxation’s revenue scores for the relevant TCJA provisions. This report argues that although those scores are the best available estimates of the effects of these provisions on U.S. tax revenues before considering dynamic macroeconomic effects, the manner in which revenue estimates are constructed tends to conceal important elements of what has happened. Accordingly, this revenue-effects approach requires taking additional considerations into account. The report also explains why the post-TCJA U.S. outbound international income tax regime remains far removed from an ideal expanded worldwide taxation regime and suggests steps that would move it toward that end.

This choice of focus means that the base erosion and antiabuse tax⁶ is not discussed. Although the BEAT is a prominent feature of the TCJA international tax provisions and although it has an impact on the foreign-source income of U.S. MNEs in some circumstances,⁷ its primary effect is on the U.S.-source income of comparatively large foreign corporations.⁸ Moreover, the JCT’s score for the BEAT is not broken down into its inbound and outbound components. Thus, there is only a lump-sum score skewed toward the BEAT’s impact on the U.S.-source income of foreign taxpayers. That makes the score unhelpful in analyzing the TCJA’s movement toward territorial or expanded worldwide taxation of the foreign-source income of U.S. MNEs. For these reasons, this report does not include discussion of the BEAT and instead focuses on the most significant purely outbound TCJA international elements.

The task undertaken here is neither a simple nor a precise exercise because the relevant TCJA provisions have components cutting in offsetting ways that require the JCT’s revenue scores to be used with caution. Moreover, reaching a conclusion requires comparing the pre-TCJA U.S. regime for taxing the foreign-source active business income of U.S. MNEs with the system as revised by the TCJA and then attempting to evaluate the effects of the differences. Thus, Section II delineates that comparison. Section III summarizes the JCT’s scoring of the TCJA changes and concludes that according to that scoring, the TCJA created a more complex system that appears to tilt in the direction of expanded worldwide taxation. From that standpoint, the major impact of the TCJA is change that raises new issues and increases complexity,⁹ thus providing an employment bonanza for lawyers and accountants. The changes manage to include elements of territoriality, expanded worldwide taxation, and even deferral and consequently do not produce clear movement toward either territoriality or expanded worldwide taxation. In Section IV, we argue that although the JCT score reflects a reasonable approach, it needs adjustment to take account of additional considerations. When it is expanded to do so, the TCJA international tax provisions appear to be revenue losers that move the U.S. international tax system toward territoriality, although not nearly as much as territoriality advocates had hoped.

Section V concludes that the TCJA has left us with an extraordinarily complicated outbound international tax regime that falls somewhere between expanded worldwide taxation and territoriality. Consequently, the direction of U.S. policy for the taxation of foreign business income of U.S. MNEs remains ambiguous, unresolved, and unsatisfactory. Thus, to move that policy toward what we have concluded is a normatively preferred position of expanded worldwide taxation, Section V identifies how the new global intangible low-taxed income regime may serve as the platform to shift the U.S. international tax regime to expanded worldwide taxation and identifies steps that would accomplish that objective.

II. Pre- and Post-TCJA Comparison

A. What We Had Before the TCJA

Before the TCJA, most active foreign business income earned by U.S. corporations through CFCs was taxed on a deferred basis. That is, U.S. tax was not imposed until the CFC distributed dividends or the U.S. parent corporation sold CFC shares at a price reflecting income accumulated within the CFC. This effectively created a U.S. tax rate on CFC income that was less than the regular corporate rate (often significantly so), thus encouraging U.S. MNEs to locate business operations in low-tax foreign countries.¹⁰ Moreover, because U.S. MNEs faced a U.S. tax liability when their CFCs distributed dividends, U.S. MNEs were said to face a lockout effect that encouraged them to avoid financing their U.S. businesses with dividends from their CFCs.¹¹

Before the TCJA, the foregoing effects were intensified because of the following:

1. Subtstantially unlimited cross-crediting,¹² effectively expanded with look-through rules,¹³ reduced the deferred U.S. tax on dividends when they were ultimately distributed by CFCs. Although this reduced the lockout effect, it encouraged U.S. MNEs to earn income in and shift income to low-tax foreign countries.

2. The amount of foreign-source income earned in low-tax foreign countries could be inflated by aggressive transfer pricing,¹⁴ and the amount of income treated as foreign-source could be inflated by the export sales source rule.¹⁵

3. The check-the-box rules¹⁶ allowed U.S. MNEs to reduce the foreign tax burden on CFC income by allowing profits to be shifted from high-tax foreign countries to tax havens while avoiding current U.S. tax under the subpart F rules.¹⁷ This decreased the foreign tax cost of earning income through CFCs and enhanced the attractiveness of foreign locations.¹⁸

4. By operating new foreign enterprises through branches, U.S. MNEs could deduct foreign start-up losses against U.S.-source income currently and avoid having those losses trapped in a foreign separate entity. When a foreign activity became profitable, it could then be brought within a CFC so that U.S. tax on the profits would be deferred prospectively. Even when the branch loss recapture rule would restore a prior loss to U.S. income, there was no interest charge. Thus, a time value of money benefit was obtained.¹⁹

Before the TCJA, some current U.S. taxation of CFC income was achieved through subpart F,²⁰ which generally taxes the passive income and base company income of CFCs as it is earned. Subpart F, however, does not apply to the manufacturing income of CFCs, including manufacturing by third parties under a highly questionable regulatory interpretation.²¹ Subpart F is (and remains) highly avoidable with careful planning.

In some circumstances, subpart F also is capable of being planned into where it would be helpful under post-TCJA law. For example, it would be possible to cause GILTI that would be high-taxed non-subpart F income (meaning that it bears a foreign effective rate of 18.9 percent or higher) to be subpart F income instead. At foreign effective rates between 18.9 and 21 percent, one could elect section 954(b)(4) deferral so that the income qualifies for section 245A exemption when distributed and bears only the foreign tax (and saves the difference between that tax and the U.S. 21 percent rate). For income subject to a foreign effective rate of 21 percent or more, section 954(b)(4) deferral should not be elected, because the foreign tax should offset the U.S. tax up to 21 percent and the foreign taxes exceeding 21 percent would be preserved as general category FTCs with FTC cross-crediting in the general category and full carrybacks and carryovers. The effective rates referred to in this example are in relation to the U.S. tax base after allocations of expenses, including allocable U.S. expenses. It hardly needs comment that optimizing under the complexity of these rules will be costly for taxpayers, and if pursued by taxpayers, will be difficult for the government to audit.

B. What the TCJA Delivered

The TCJA gave us:

Dividend exemption.

1. New section 245A provides a 100 percent deduction for the foreign-source portion²² of any dividend received from a specified 10 percent-owned foreign corporation²³ by a U.S. domestic corporation that is a U.S. shareholder.²⁴ This new deduction obviously frees the dividend recipient from residual U.S. tax on that portion (while losing use of any associated FTCs). So to the extent that shareholders are subject to the section 245A deduction, there is no point in pursuing the deferral strategy by accumulating foreign-source income in CFCs.

a. The TCJA disallowed FTCs and deductions for foreign taxes with respect to any portion of a distribution that qualifies for the section 245A deduction.²⁵ Section 245A made the section 902 indirect FTC unnecessary, so it was repealed. The section 960 indirect FTC was preserved, however, and applied on a current-year basis.²⁶

b. The following shareholders, however, are ineligible for the deduction and therefore continue to face taxation on dividends received from a foreign corporation:

i. individual shareholders;

ii. S corporations;

iii. corporate shareholders that do not qualify as U.S. shareholders under the TCJA’s expanded definition thereof²⁷ — in other words, the deduction is unavailable to foreign corporations or to U.S. domestic corporations that do not own at least 10 percent of the voting power or value of the distributing foreign corporation (directly, indirectly through a foreign entity, or constructively under the attribution rules in section 958(b));

iv. all shareholders of any distributing corporation that is not a specified 10 percent-owned foreign corporation — that is, all shareholders of a distributing corporation that is not a foreign corporation having at least one U.S. domestic corporate shareholder that qualifies as a U.S. shareholder;²⁸ and

v. all shareholders of passive foreign investment companies that are not also CFCs.²⁹

c. The JCT estimated that the foregoing changes would have a revenue cost/tax-saving effect of $223.6 billion in the 2018-2027 period.³⁰

2. The TCJA did not change the passthrough status of foreign branches. Moreover, the TCJA did not extend the section 245A deduction to foreign branch income. Thus, foreign branch losses continue to be immediately deductible by U.S. owners, and branch net income continues to be immediately taxable at the U.S. owners’ respective regular rates. The TCJA did, however, create a new FTC basket limitation for foreign branch income.³¹ Thus, excess FTCs on foreign branch income cannot be offset against the U.S. residual tax on other types of foreign-source income, but unlimited cross-crediting is allowed within this new basket.³² The JCT did not give a separate score for the net effect of these FTC changes. Instead, the effect was included in the $112.4 billion revenue-raising/tax increase estimate, infra, for the GILTI provisions.³³

3. It is well known that a deduction for dividends received from CFCs constitutes a powerful incentive to shift business operations and profits to low-tax foreign countries. The TCJA tries to restrain this effect for the new section 245A deduction by providing a carrot (the new foreign-derived intangible income (FDII) regime) and two sticks (continuation of the subpart F regime, with some expansion, and adoption of the new GILTI regime).

An export subsidy.

1. New section 250 establishes an FDII regime that is effectively an export subsidy.³⁴ This new regime operates by giving each U.S. domestic corporation an income tax deduction equal to 37.5 percent of the corporation’s FDII.³⁵ The deduction has a limitation based on taxable income,³⁶ but when the full deduction is available, it reduces the effective U.S. corporate income tax rate on FDII from 21 percent to 13.125 percent.³⁷

a. In simplified terms, FDII is a U.S. domestic corporation’s income from export sales of goods and services to the extent that this income exceeds a 10 percent deemed return on the corporation’s tangible assets.³⁸ This definition is supposed to ensure that deemed tangible income is excluded from FDII and taxed at the regular 21 percent rate so that the FDII tax benefit is reserved for intangible income. There is no assurance that this is the result, however.³⁹

b. Foreign branch income is excluded from FDII.⁴⁰ This is supposed to prevent a U.S. domestic corporation from earning FDII through a directly owned, foreign-located factory.

c. The FDII regime may have a short life because it may be held to violate WTO law.⁴¹ Consequently, this report does not parse the details. For present purposes, the FDII regime’s important feature is that the JCT estimated the regime’s revenue cost/tax-saving effect to be $63.8 billion during the 2018-2027 period.⁴²

d. There is, however, a small countervailing provision in the TCJA’s repeal of the export sales source rule.⁴³ That rule gave a foreign-source classification to 100 percent of the profits from export sales of goods in some cases, and 50 percent in other cases, when right, title, and interest passed to the foreign buyer outside the United States. Because export sales profits frequently bear no foreign tax, the effect of the export sales source rule was often to create zero-taxed foreign-source income with a full U.S. residual tax that could be used to absorb excess credits resulting from high foreign taxes. The TCJA eliminated this opportunity,⁴⁴ and the JCT scored this change as a $0.5 billion revenue-raising/tax increase measure for the 2018-2027 period.⁴⁵ Unfortunately, the method chosen by Congress to repeal the export sales source rule — namely, allocating and apportioning the source of income from sales of inventory property produced in the United States and sold abroad, or produced abroad and sold in the United States, “solely on the basis of production activities with respect to the property”⁴⁶ — probably provides a distortive incentive to locate production assets in foreign countries.⁴⁷

Subpart F.

1. Subpart F imposes an immediate U.S. tax at regular rates on each U.S. taxpayer’s pro rata share of the passive income and active base company income (referred to collectively as “subpart F income”) of each CFC for which the taxpayer is a U.S. shareholder.⁴⁸ Although this regime was mostly untouched by the TCJA, the following amendments were made:

a. For corporations that are U.S. shareholders, subpart F income is taxed at the regular U.S. corporate rate. Thus, the reduction in the regular rate lessens the bite of residual U.S. tax under the subpart F regime for those corporate taxpayers.

b. Before the TCJA, the top 35 percent federal corporate income tax rate plus the effect of subnational corporate taxes yielded a top average U.S. corporate income tax rate of 39.1 percent on all income (domestic and foreign) in 2013.⁴⁹ In comparison, the GDP weighted average of top corporate rates for all OECD countries, other than the United States, was 28.4 percent in 2013.⁵⁰ Although the effective combined tax rate of numerous MNEs was significantly below 39.1 percent,⁵¹ there was substantial support for a sufficient reduction in the top federal corporate rate to move the combined U.S. corporate rate toward the OECD average, assuming that the amount of the reduction gave appropriate attention to deficit concerns. The Alliance for Competitive Taxation, a large group of American MNEs,⁵² pushed for reducing the U.S. corporate income tax rate to no more than 20 percent.⁵³ President Trump went further and advocated a 15 percent rate.⁵⁴

c. The TCJA provides a 21 percent flat federal corporate income tax rate.⁵⁵ This is a 40 percent reduction from the former 35 percent top U.S. federal income tax rate and is substantially similar to the 20 percent rate promoted by U.S. MNEs. Although the resulting $1.349 trillion revenue loss⁵⁶ applies to corporate income generally, part of it substantially reduced the tax burden on subpart F income and should be taken into account when evaluating the effects of the TCJA on subpart F. However, the JCT gave an overall score for the corporate tax rate reduction and did not break out the subpart F impact.⁵⁷ Of course, doing so was unnecessary for the JCT’s purpose of stating the overall revenue effect of the TCJA, and the JCT has resource constraints that prevent it from satisfying all possible avenues of curiosity. Nevertheless, an allocation to subpart F of the appropriate part of the $1.349 trillion revenue loss would have given a clearer picture of the TCJA’s international effects.

d. The TCJA expanded the definition of U.S. shareholder to include a U.S. person who owns less than 10 percent of the voting power of a particular corporation’s shares but who does own at least 10 percent of the total value of that corporation’s stock.⁵⁸ This increases the extent of the subpart F regime, and the JCT scored it as having a $1.3 billion revenue-raising/tax increase effect for the 2018-2027 period.⁵⁹

e. The TCJA removed from section 951(a) the stipulation that a foreign corporation is not a CFC for any tax year unless the CFC satisfied the CFC definitional requirements for at least 30 straight days during the year. Thus, subpart F now applies to foreign corporations if they qualify as CFCs for only one day in the tax year.⁶⁰ This expands the coverage of subpart F somewhat, and the JCT scored it as having a $600 million revenue-raising/tax increase effect for the 2018-2027 period.⁶¹

f. Further, the TCJA expanded the section 958(b) constructive stock ownership attribution rules by repealing section 958(b)(4), which, in the context of the owner-to-entity attribution rules in section 318(a)(3), had prevented attribution of stock ownership to a U.S. person from a person who is not a U.S. person. This enlarges the reach of subpart F somewhat and makes it much easier for a foreign corporation with only a minority of U.S. persons or even no U.S. persons directly owning its stock or owning its stock through a foriegn entity to constitute a CFC.⁶² The JCT did not provide a separate score for this item but instead included it in the $223.6 billion revenue cost/tax-saving estimate, supra, for the dividends received deduction (DRD) in section 245A.⁶³

g. In a move that reduced the scope of the subpart F regime, although not in a major way, the TCJA removed foreign base company oil-related income from subpart F income. The JCT scored this as having a $4 billion revenue cost/tax-saving effect for 2018-2027.⁶⁴

h. Because the JCT did not provide a score for the overall effect of these three expansions and this one contraction of the subpart F regime, the precise effect is not known.

Check-the-box.

1. The TCJA did not change the law giving rise to the check-the-box regulations. Thus, U.S. MNEs continue to be able to shift profits from high-tax foreign countries to low-tax foreign countries through deductible payments that do not create subpart F income. To the extent that foreign-source income can now be repatriated tax free by means of the 100 percent section 245A DRD, this shifting no longer has importance as a device for enhancing deferral of U.S. tax on foreign income. When the GILTI regime (discussed later) applies, however, the check-the-box rules are a powerful tool for manipulating entities to achieve tax benefits under post-TCJA rules — for example, by combining loss CFCs with profitable CFCs to obtain section 245A exemption instead of GILTI taxation under section 951A. Retention of the check-the-box rules without modification continues a flawed policy upheld by the U.S. MNE community, which has argued that reducing the tax bases of high-tax countries, most of which are important U.S. allies, is in the national interest.⁶⁵ Thus, the failure of the TCJA to restrict the check-the-box rules to a domestic context and to align the U.S. entity classification rules with foreign classification to the extent feasible is a disappointment, although mitigated by the TCJA’s adoption of anti-hybrid rules.⁶⁶

The GILTI regime.

1. The TCJA adopted new section 951A. This imposes a subpart F-like regime on the GILTI of CFCs.

a. The GILTI regime uses the CFC definition in section 957(a) with the enlarging amendments mentioned earlier. Thus, a foreign corporation is a CFC for GILTI purposes if more than 50 percent of the voting power or value of its shares is owned — directly, indirectly through foreign entities, or by attribution — by one or more U.S. shareholders on at least one day in the tax year.⁶⁷

b. The GILTI regime applies to the GILTI of each U.S. taxpayer who is a U.S. shareholder of at least one CFC during a particular year.⁶⁸

c. The taxpayer first calculates net CFC tested income for the year. In simplified terms, net CFC tested income is the sum of the taxpayer’s pro rata shares of the gross income of each CFC for which the taxpayer is a U.S. shareholder, minus each such CFC’s subpart F income and minus allocable deductions other than cost of goods sold.⁶⁹ If this computation yields a loss for a particular CFC, the loss is offset against the net positive amount, if any, for the taxpayer’s other CFCs.⁷⁰

d. An overall loss from the preceding computation cannot be deducted against non-GILTI income and cannot be carried backward or forward.

e. If the preceding computation yields an overall positive amount for a particular year, the taxpayer calculates 10 percent of the taxpayer’s pro rata share of the qualified business asset investment (QBAI) for all profitable CFCs for which the taxpayer is a U.S. shareholder and subtracts the taxpayer’s pro rata share of specified CFC interest expense from that 10 percent amount.⁷¹ The resulting sum, which is called the net deemed tangible income return (NDTIR), is then subtracted from the overall positive amount.⁷² If this deduction for 10 percent of QBAI, less specified interest expense, yields a negative amount, the negative amount cannot be used for any purpose.

f. Generally speaking, QBAI is the taxpayer’s pro rata share of the quarterly average of the total basis in tangible foreign business property of all profitable CFCs for which the taxpayer is a U.S. shareholder.⁷³

g. Any net positive amount that emerges from the preceding steps is immediately included in the taxpayer’s gross income under section 951A(a) and, for U.S. shareholders that are C corporations, for tax years starting before 2026, 50 percent of that amount is deductible under section 250(a). The deduction is reduced to 37.5 percent for tax years starting after 2025.⁷⁴

h. The preceding means that foreign-source active business income is exempt from the GILTI regime to the extent it does not exceed 10 percent of QBAI net of specified allocable interest expense. However, for tax years starting before 2026, the section 250 deduction means that before taking account of FTCs, GILTI exceeding the 10 percent benchmark is generally taxed at the reduced effective rate of 10.5 percent⁷⁵ for a domestic C corporation and 37 percent for an individual in the highest marginal rate bracket.⁷⁶ Thus, the immediate tax on GILTI is a movement in the direction of expanded worldwide taxation, but the bargain rate for domestic C corporations tilts in the direction of territoriality.⁷⁷

i. Eighty percent (not 100 percent) of the foreign tax on GILTI is creditable against the U.S. tax thereon.⁷⁸ FTCs for GILTI are quarantined in a separate basket limitation, but there is unlimited cross-crediting within that basket (except for passive category income that is placed in the separate FTC basket for that income in section 904(d)(1)(C)).⁷⁹ However, excess FTCs in the GILTI basket may not be carried back or carried forward to any other tax year,⁸⁰ and no FTC is available for foreign taxes paid or accrued on the NDTIR because the NDTIR is effectively tax exempt under section 245A.⁸¹ The net effect is that the separate GILTI FTC basket is only a weak restraint on cross-crediting given that GILTI includes most of a CFC’s foreign-source business income. Unlimited cross-crediting conflicts with expanded worldwide taxation but is consistent with territoriality because it reduces or eliminates the residual U.S. tax that would be applied under an expanded worldwide system.

j. Subpart F income is not eligible for the GILTI bargain tax rate applicable to domestic C corporations.⁸² Instead, it is taxed at the relevant taxpayer’s normal rate net of FTCs (without the 20 percent haircut on allowable credits that applies to GILTI). On the other hand, manufacturing income exceeding the QBAI exemption, which is excluded from the application of subpart F and which was eligible for deferral before the TCJA, is now subject to an immediate tax, but at the GILTI bargain rate for domestic C corporations and net of FTCs (with the 20 percent haircut on allowable credits that applies to GILTI).

k. If the exemption for 10 percent of QBAI exceeding interest expense were eliminated, if the section 250 deduction for a domestic C corporation’s GILTI were eliminated so the regular U.S. rate were substituted for the GILTI bargain effective rate, and if cross-crediting were substantially eliminated and the FTC haircut removed, then this reformed GILTI regime together with the subpart F regime would cause most foreign-source net corporate income to be subject to an immediate tax at no less than the regular U.S. rate. Expanded worldwide taxation would be achieved. Simplification could then be accomplished by repealing the subpart F regime and having a single, simpler worldwide taxation regime under a revised GILTI.

2. The JCT scored the new GILTI regime as having an overall $112.4 billion revenue-raising/tax increase effect for the 2018-2027 period.⁸³

The transition tax.

1. The TCJA was required to address the issue of the taxation (or not) of the large amounts of foreign-source income earned by U.S. residents before the TCJA and for which U.S. tax had never been paid. Generally speaking, the solution was a one-off transition tax on previously untaxed amounts earned after 1986.⁸⁴ The tax was imposed on U.S. shareholders of affected corporations at a rate of 15.5 percent for income held in cash or cash equivalents and 8 percent for other income.⁸⁵ The tax is payable in eight interest-free annual installments.⁸⁶ The two rates are higher than the 8.75 percent and 3.5 percent rates on earnings accumulated in cash or in other forms, respectively, previously proposed by then-House Ways and Means Committee Chair Dave Camp.⁸⁷ The two TCJA rates are also higher than the 5.25 percent rate on income repatriated under the temporary repatriation window enacted in 2004.⁸⁸ Thus, the TCJA transition tax rates are a disappointment to the U.S. MNE community. On the other hand, those rates are substantially less than either the 35 percent top corporate income tax rate that applied before the TCJA transition tax or the 21 percent corporate income tax rate that now applies. Thus, these TCJA transition tax rates are also a disappointment to some expanded worldwide taxation advocates.⁸⁹ Nevertheless, the JCT scored this one-off transition tax as having a $338.8 billion revenue-raising/tax increase effect for the 2018-2027 period.⁹⁰

III. The JCT Scorecard

The post-TCJA U.S. outbound international income tax system is clearly neither purely territorial nor purely expanded worldwide.⁹¹ A rough cut at assessing the degree to which the TCJA moved the existing regime in either direction involves simply totaling the JCT’s revenue gain and loss scores indicated in Section II (without separately taking account of the corporate tax rate reduction).⁹²

Of course, those scores lack scientific precision. They involve assessing the net effects of the offsetting aspects of the TCJA provisions described in Section II, which in turn involve inherently uncertain macroeconomic and behavioral predictions extended over a 10-year period. Nevertheless, there are no better scores accessible to the public, and even their roughness can shed some light on movement directions. Accordingly, the following table is presented to summarize the JCT’s 2018-2027 revenue gain or loss (tax savings increase or decrease) estimates for the outbound TCJA international tax provisions discussed in Section II:

IV. Expanding the Scorecard

Thus, the JCT scored the provisions in Section II of this report as actually increasing the tax burden on the U.S. MNE community. This is consistent with the reality that the pre-TCJA U.S. international tax provisions raised little revenue. Harry Grubert and Rosanne Altshuler found that the United States collected tax of only $32 billion on all 2006 foreign-source income (implying less than a 4 percent effective rate),⁹³ and Jane G. Gravelle found that U.S. effective rates on all foreign income were only 7 percent in 2007 and 5 percent in 2008.⁹⁴ Consequently, even a slight tightening of the U.S. international tax regime would be expected to yield a revenue gain.

More importantly, the JCT’s projected revenue increase of $157.2 billion over the 2018-2027 period is minuscule in terms of the federal fisc. In fact, it represents only 0.37 percent of the $41.98 trillion in federal revenue receipts projected by the Congressional Budget Office for the same period.⁹⁵ This small revenue score is not surprising given that the minimum tax feature of the TCJA, the GILTI regime, includes an 80 percent FTC. Thus, the GILTI tax shrinks as foreign tax increases and generally switches off entirely when the foreign tax is at least 13.125 percent (for tax years starting before 2026). Many of the major trading partners of the United States have effective corporate tax rates that are close to or greater than this.⁹⁶ Moreover, shrinkage of the GILTI tax is exacerbated by the unlimited cross-crediting permitted under the FTC component of the GILTI regime (except for GILTI that is passive category income and placed in the passive category income FTC basket limitation).

The bottom line is that the TCJA has created substantial additional complexity but, according to the JCT score, has caused hardly any significant movement toward either territorial taxation or expanded worldwide taxation.

The JCT score was, however, prepared only a short time after the TCJA became law. Thus, the JCT did not have the benefit of observing the legislation’s operation after tax advisers have learned to fully exploit loopholes. As shown in Section II, the TCJA international tax provisions are complex, and some tilt in favor of expanded worldwide taxation while others tilt toward territoriality. The obvious difficulty of netting these conflicting features against each other and projecting the results over a 10-year period means that the JCT’s $157.2 billion revenue gain score must be regarded as significantly speculative. Moreover, there are several specific reasons for believing that the JCT scorecard understates the degree to which the TCJA actually moved the needle in the direction of territorial taxation.

First, we must remember that without the $333.8 billion revenue-raising/tax increase effect of the one-off transition tax, the JCT’s score would show a $176.6 billion revenue loss/tax-saving effect for the 2018-2027 period. Thus, the TCJA’s apparent tilt toward expanded worldwide taxation is temporary, and when the impact of the transition tax vanishes at the end of the eight-year installment payment period, the TCJA international tax provisions — as scored by the JCT — will swing toward territoriality. Moreover, it would have been correct policy to apply the transition tax by using the pre-TCJA rate table that topped out at 35 percent instead of the 15.5 percent and 8 percent rates that were actually adopted.⁹⁷ The pre-TCJA rates would have materially increased the cost of the transition tax, and the avoidance of that cost should be considered a major win for territoriality advocates.

Moreover, Congress could have gone in another direction and enacted an expanded worldwide system that set the tax rate on GILTI for domestic C corporations at the 21 percent regular corporate rate and that more substantially eliminated cross-crediting.⁹⁸ The JCT did not score this alternative, but its first report on tax expenditures under the TCJA scored GILTI’s use of a 10.5 percent rate, instead of the regular 21 percent corporate rate, as a $404.2 billion cumulative tax expenditure for the 2017-2021 period.⁹⁹ This estimate cannot be neatly plugged into the JCT’s scoring regime for two reasons. First, the scoring regime uses a 10-year calculation period (2018-2027) while the tax expenditure estimate uses a five-year period (2017-2021). Second, the scoring regime incorporates predicted behavioral changes while the tax expenditure estimate does not. Nevertheless, the five-year tax expenditure estimate clearly suggests a meaningful 10-year revenue loss/tax saving from Congress’s decision to use a bargain rate for domestic C corporations in the GILTI regime. Adoption of an expanded worldwide system would have eliminated this item and materially tilted international income tax reform toward expanded worldwide taxation. Avoidance of this outcome must be taken into account in considering the direction of movement under the TCJA outbound international income tax provisions.

Finally, the TCJA outbound international tax provisions will attract the ingenuity of tax planners. When their tax minimization skills are fully directed at those provisions, it seems likely that the JCT will have overestimated revenue gains and underestimated revenue losses.¹⁰⁰

Modifying the JCT scorecard to reflect the one-off and bargain-corporate-rate character of the transition tax, the revenue loss from the GILTI regime’s low rate, and the likelihood that the JCT has not fully accounted for the effects of tax planning, suggests that the TCJA outbound international tax provisions move the system in the direction of the zero U.S. tax rate sought by territoriality advocates even though the TCJA creates a mixed regime that is only partly territorial.¹⁰¹

V. Toward Expanded Worldwide Taxation

Thus, although the TCJA preserved subpart F and replaced deferral for U.S. shareholders of CFCs with the GILTI regime’s current U.S. tax, the U.S. outbound international income tax system remains defective for several reasons. The first is that the TCJA has preserved the location distortion aspect of prior law, but it now is more directed at real investment. Under the GILTI regime, a zero foreign tax rate generally yields a U.S. tax on GILTI income of only 10.5 percent even though there is no FTC. When this result is compared with the 21 percent rate that generally applies to U.S. domestic income of C corporations, it is clear that the GILTI regime has moved the system in a territorial direction and can provide a strong incentive for U.S. MNEs to shift profits to, and locate business activity in, CFCs formed in low-tax foreign countries.¹⁰² Example 1 illustrates this point.

Example 1: Assume a year in which the TCJA is in effect, but before 2026. M Corp. is a U.S. C corporation that has $1 million with which to build a factory that will manufacture simple consumer goods for sale to foreign customers outside both the United States and New Landia. (New Landia is a low-tax country that is encouraging M to build the factory within its borders.) If M builds the factory in the United States, M can earn $80,000 of annual net profit, calculated before the imposition of U.S. tax (that is, after all allocable deductions). If M builds the factory in New Landia, M can earn $75,000 of annual net profit, calculated before the imposition of U.S. and New Landia taxes. (Assume that this lower profit is because of New Landia’s poor infrastructure and lower-cost but less efficient workforce.) The United States taxes corporate business profits at a rate of 21 percent while New Landia’s rate is 5 percent. Assume that New Landia does not impose either a withholding tax on dividends paid to nonresident parent companies by New Landia corporations or a branch profits tax and that it allows a local law deduction for all expenses such that its tax base is the same as the U.S. tax base. If M uses a wholly owned New Landia CFC to build, own, and operate the factory in New Landia, and if the GILTI tax is zero because of the 10 percent QBAI deduction, the after-tax results will be:

If the U.S. GILTI regime does apply, the after-tax results will be:

If the factory is built in the United States, the after-tax results will be:

Thus, in both cases (when the section 245A DRD applies or when it does not but GILTI and section 959(a) apply), the factory will be built in New Landia even though the New Landia factory will produce a lower pretax return than a U.S. factory.

Of course, the GILTI regime and the TCJA’s reduction of the top U.S. corporate income tax rate from 35 percent to 21 percent have a combined effect that narrows the range within which results such as those in Example 1 can occur.¹⁰⁴ For instance, if the New Landia factory’s annual pretax profit were only $70,000 and the GILTI tax applied, the New Landia factory’s after-tax profit would be only $61,950¹⁰⁵ compared with the $63,200 U.S. profit, and the new factory would be built in the United States. Nevertheless, Example 1 shows that the comparatively low GILTI tax rate can distort the location decision and cause a foreign location that is inferior to a U.S. location on a pretax basis to become superior on a post-tax basis, thus inefficiently drawing business operations to a low-tax foreign location. Moreover, the 10 percent QBAI deduction from the GILTI tax base creates its own incentive to locate tangible business assets, like factories, in low-tax foreign countries.

This dynamic means that the TCJA international system gives U.S. MNEs an incentive to inflate the income of their low-taxed CFCs by engaging in aggressive transfer pricing — a tactic to which the TCJA international changes do not respond other than to make it potentially more expensive to shift assets out of the United States.¹⁰⁶ Thus, the outbound transfer pricing problems of pre-TCJA law largely have been carried forward. Moreover, the TCJA international changes give U.S. MNEs a comparative advantage in the U.S. market over U.S. businesses that operate and sell only in that market and therefore cannot use profits from favorable tax treatment of foreign activities to support U.S. activities.¹⁰⁷ The existence of the FDII regime does not alter these conclusions. That regime’s nonapplicability to an amount of income equal to 10 percent of QBAI,¹⁰⁸ plus the fact that its maximum benefit is a 13.125 percent effective rate on covered income (for tax years before 2026),¹⁰⁹ makes it unattractive compared with the option of carrying on business through a CFC that is subject to the GILTI regime assuming a sufficiently low foreign effective tax rate.¹¹⁰

A related piece of this picture is that the post-TCJA international tax regime continues to provide an incentive for a U.S. corporation to use a foreign subsidiary in a low-tax country to produce products that it will resell to customers in the United States. This is illustrated in Example 2.

Example 2: USCo sells widgets in the United States. USCo purchases the widgets from ForSub, a wholly owned foreign subsidiary of USCo. ForSub makes the widgets at its factory in Lowtaxia, a low-tax foreign country, and delivers them by common carrier to USCo at a foreign port where right, title, and interest passes to USCo outside the United States.¹¹¹ One hundred percent of ForSub’s profits from sales to USCo are foreign-source income subject to the advantages of the GILTI regime.¹¹² As previously noted, ForSub’s profits can be increased by aggressive transfer pricing.

The dynamics illustrated in Example 2 provide an additional incentive for U.S. MNEs to locate manufacturing operations in CFCs formed in low-tax foreign countries or countries that have favorable incentive regimes (such as the Mexican IMMEX/maquiladora regime). If expanded worldwide taxation were adopted, this incentive would disappear.

The defects identified may be mitigated through reforms that limit the distortive effects of the current regime by eliminating or at least decreasing the section 250 GILTI deduction and thereby increasing the GILTI effective tax rate and by curtailing the unlimited cross-crediting that is available in the GILTI FTC basket limitation.

The platform provided by the GILTI regime suggests that it may offer a bridge to expanded worldwide taxation. There is a clear path to accomplishing expanded worldwide taxation by increasing the GILTI bargain rate to 21 percent and curtailing cross-crediting by applying the GILTI FTC basket limitation on a per-country basis.¹¹³

VI. Conclusion

This report has shown that the TCJA outbound international income tax provisions cut in favor of territoriality while also embracing expanded worldwide taxation elements, albeit at a low effective rate. The principal positive for expanded worldwide taxation advocates is that the TCJA changes are not as bad as they would have been if there were no transition tax and no GILTI minimum tax on foreign-source business income. But the most important observation for these advocates regarding the TCJA is that even though deferral is gone for most U.S. corporate taxpayers, the distortive effects of taxing a significant portion of U.S. corporations’ foreign-source income earned through CFCs at a low effective rate continue under the GILTI regime. Those effects call for increasing the GILTI tax rate to 21 percent for a U.S. corporation, rolling subpart F and the GILTI regime into a single simplified construct, more substantially curtailing cross-crediting with a per-country FTC limitation, and turning attention to crafting stronger restraints on aggressive transfer pricing.