GILTI of Putting All of Your Taxes in One Basket

We are often warned not to put all of our eggs in one basket. What is true of eggs is also true of taxes, and if certain parts of President Biden’s “Made in America Tax Plan” or Senators Ron Wyden’s (D-Ore.), Sherrod Brown’s (D-Ohio), and Mark Warner’s proposed (D-Va.) “Overhauling International Taxation” proposal (the “Senate white paper”) are adopted, there will be more—and correspondingly smaller—baskets in which U.S. shareholders of controlled foreign corporations (CFCs) may put them.

Tax code Section 904(d)(1) sets forth four general categories, or baskets, for foreign tax credit (FTC) purposes. One is for non-passive category income that is includible in gross income under tax code Section 951A (the “GILTI basket”). Both the Made in America Tax Plan (first described as part of President Biden’s “American Jobs Plan” infrastructure proposal released March 31, 2021) and the Senate white paper (released April 5, 2021) argue that the use of a single basket for all global intangible low-taxed income (GILTI) inappropriately permits U.S. taxpayers to earn income (through CFCs) in low-taxed jurisdictions without incurring residual U.S. tax cost. This is due to their ability to use foreign tax credits resulting from higher taxed income from other jurisdictions to offset the lower-taxed income. Both proposals seek to stop this ability to blend, or “cross-credit,” high-taxed income and low-taxed income in the GILTI basket.

Although the evils of cross-crediting can be explained simply (i.e., refer frequently and vehemently to “tax havens”), proposals actually to limit cross-crediting tend to be much more complicated. To ensure that a taxpayer cannot blend high-tax income from one jurisdiction with low-tax income from another jurisdiction, one has to wall-off the income in the low-taxed jurisdiction, the taxes in the high-taxed jurisdiction, or both. This requires not just preventing the taxes from being mixed: the gross income, deductions associated with that gross income, and taxes imposed on the gross income or net income all have to be identified, determinations as to whether income is high-taxed or low-taxed must be made, and the relevant items must be segregated into the desired separate baskets.

Per-Country ‘Tis of Thee

The initial materials for the Made in America Tax Plan provide little detail regarding this proposed FTC change. According to the U.S. Department of Treasury description of the plan, released on April 7, 2021, one of the changes made by the Made in America Tax Plan would be to “calculate the GILTI minimum tax on a per-country basis.” The Senate white paper provides a bit more detail on what the authors have in mind for per-country reporting: “One country-by-country option is to expand the existing system for foreign tax credits—with the use of foreign tax credit ‘baskets’—essentially applying the current GILTI rules separately for each country in which a corporation operates. For example, if a corporation operates in nine countries, it would have nine GILTI ‘country baskets,’ with no aggregation among them.” In fairness, the goal of these documents is to argue for policy changes rather than to set forth technical detail. The “Fact Sheet: The American Jobs Plan,” released March 31, 2021; the Treasury explanation of the Made in America Tax Plan, released April 7, 2021; and the Senate white paper are more focused on why the change is being proposed and justifying their goal of ensuring that profits in “tax havens” are subject to the minimum tax. Nonetheless, unless one defines “tax haven” as “any jurisdiction with an effective tax rate below whatever the GILTI effective tax rate winds up being,” the Made in America Tax Plan proposal will apply much more broadly to individuals and companies with foreign-source income, regardless of where the income is earned.

Still, if one wants to adopt a per-country limitation, there are several models available. Many countries, including some of our largest trading partners, apply a type of per-country limitation in applying their foreign tax credit. Pillar Two of the OECD/G-20 Inclusive Framework on base erosion and profit shifting (BEPS) describes approaches for applying a per-country limitation in the context of its minimum tax and the issues that arise. See Section 3.4.2 of the Pillar Two Blueprint. Yet, as much as the Made in America Tax Plan cites Pillar Two, I suspect that the U.S. is going to use its general rules for categorizing, allocating, and apportioning expenses and taxes to gross income, rather than starting with the OECD country-by-country reporting approach or another country’s rules.

It is also important to note that both the Made in America Tax Plan and the Senate white paper do not suggest replacing the existing baskets, based on type of income, with baskets based on source. Rather, they suggest layering a per-country limitation into the existing FTC rules. This distinction is significant because, historically, the FTC basket approach (such as we use now) and per-country limitation were seen as alternative—perhaps even rival—means of limiting cross-crediting. Indeed, during its century-plus of federal income tax, the U.S. has tried per-country limitations, separate categories, and overall limitations. To add a per-country limitation to our existing separate category approach, however, will introduce a mix-and-match element to our FTC rules that will result in complexity beyond what would occur in either a per-country or separate category approach by itself. So, even if the per-country limitation rules of old Treasury Regulations Section 1.904-1 had not been repealed by Treasury Decision 9882 as “deadwood,” they would not provide much of a model in this case.

Granted, there is something close to a per-country limitation that exists in the separate category approach of Section 904(d). This is Treas. Reg. Section 1.904-4(k) which implements the separate basketing rules of Section 904(d)(6) and Section 865(h). Those tax code sections deal with income and taxes re-sourced due to application of a U.S. income tax treaty. They effectively require taxpayers to follow the “normal” basketing rules for items of income and taxes but to place them in a special treaty-re-sourced basket. Thus, if U.S.-source income is re-sourced pursuant to a U.S. income tax treaty and that income would be treated as income in the passive category, the re-sourced income is treated as income in the passive category as re-sourced under the U.S. income tax treaty, with the high-tax kick-out rules and such applying separately to this particular category. See Treas. Reg. Section 1.904-4(k)(1)(ii) and -4(k)(2). Indeed, the suggested approach in the Senate white paper could be seen as applying Section 1.904-4(k)(2), only substituting the particular taxing jurisdiction for the particular income tax treaty.

Further, the per-country limitation is only one feature of the changes to GILTI the Made in America Tax Plan and Senate white paper would make. There are reductions in the Section 250 deduction and elimination of the net deemed tangible income return in determination of the amount of GILTI too, and so one cannot evaluate the per-country limitation in isolation of those additional changes. However, one thing that is clear is that creating more and smaller GILTI baskets will exacerbate the effect of existing FTC limits in GILTI. One example is the effect of Section 960(d), which generally limits the amount of creditable taxes to 80% of the allocable share of the aggregated tested foreign income taxes paid or accrued by the CFC. The smaller the groupings of income and taxes, the more likely the 80% limitation will apply in particular instances, even though the overall tax burden in a country may be well above the U.S. tax rate. Perhaps more importantly, smaller baskets make the lack of carryback and carryforward of foreign taxes in GILTI even more arbitrary. Sometimes, income (measured under U.S. tax principles) is high-taxed or low-taxed solely due to timing issues. In baskets with more blending of income and taxes, timing issues regarding payment and accrual of foreign income taxes can offset each other; in baskets with limited items of income, timing differences can effectively mean denial of foreign tax credits. With reduced ability to cross-credit, fairness will require a relaxation of the strictness of the current GILTI FTC rules.

Of course, the call for a per-country limitation is not based on a philosophical view that income and taxes from two different jurisdictions have no business fraternizing with each other. Given the strident references to preventing companies from shifting income to tax havens, the Made in America Tax Plan and Senate white paper imply a simple world in which there are high-tax countries where companies have to operate and low-tax countries where companies choose to book income through intangibles or other unseen ways. The presentation of this simplistic worldview makes it hard to tell whether the concern about cross-crediting is attributable to taxpayers’ ability to reduce U.S. residual tax by blending high- and low-taxed income or the way it can distort investment and activity. At least the Senate white paper implies that the goal is to prevent blending of high- and low-taxed income, and if there is a simpler way to do that than per-country limitation, then that is fine with the author.

Gotta Keep ‘Em Separated

The Senate white paper accordingly sets forth an alternative: “A second option that achieves the same country-by-country objective, potentially in a much simpler fashion, is to divide global income into two groups—low-tax and high-tax. Rather than applying the foreign tax credit system to every single country separately, GILTI would only be applied to income from low-tax jurisdictions. This would allow a significant amount of global income to be aggregated, but without any of the abuses present in the current GILTI system.”

The Senate white paper argues that this would be a simpler approach. In particular, it suggests converting the GILTI high-tax exception of Treas. Reg. Section 1.951A-2(c)(7) from an election to a mandatory rule: “Income from high-tax countries would be excluded from GILTI through the use of a mandatory high-tax exclusion—if a corporation paid a foreign country a tax rate that was above the GILTI rate, it would be excluded from GILTI altogether. All the income that remains in the system is, by definition, from countries where the foreign tax rate was below the GILTI rate—in tax parlance, there would be no “excess credits” in the system to cover for low-tax income. These earnings would be aggregated and subject to the current GILTI rules. This achieves the goals of a country-by-country system in a simpler way, making it easier for the IRS to enforce.”

I do not know whether a high-tax, low-tax distinction is simpler than a per-country limitation. I do know that nothing about GILTI or the high-tax exception is simple. First, although many companies and practitioners called for regulations to create a high-tax exception for GILTI, the actual exception in the regulations stops far short of the requests in the comments. Indeed, the preamble to the final regulations for the GILTI high-tax exclusion clearly expresses the IRS’s and Treasury’s concern about allowing low-taxed income to escape GILTI because it was blended with higher-taxed income. See the discussions in T.D. 9904 rejecting the approach of applying the high-tax exclusion on a CFC-wide basis.

Further, because the factual scenarios in which these rules will have to apply are a lot more complicated than the “real activity” in high-tax countries and “profit shifting” in low-tax countries dichotomy, the new FTC rules will be much more complicated than might appear from their advocates. For example, to apply a per-country limitation, one has to design rules for identifying specifically where income is earned and taxed. There are some existing rules for doing this, e.g., the treaty re-sourcing rules of Sections 865(h), 904(d)(6), and 904(h)(10); Section 901(j); and the international boycott rules. But the limited scope of those rules prevents them from providing much of a model for the more broadly-applicable per-country limitation anticipated by the Made in America Tax Plan or Senate white paper.

The high-tax exception approach avoids having to deal with this sourcing rule, but it has its own issues of determining how one concludes that income is high taxed. Section 1.951A-2(c)(7) considers “high-taxed” to be 90% of the federal corporate tax rate and to be measured on a “tested unit” basis, with lots of special rules for disregarded payments. The level at which the limitation applies, whether at the CFC or a lower level, has to be determined since CFCs can have branches and own interests in pass-through entities. In that regard, the GILTI high-taxed exception rules adopt a “tested unit” approach for determining which income, expenses, and taxes are grouped together to determine whether the income is high-taxed. How to do this is not obvious, and the painful twists and turns the IRS and Treasury took to get to the tested unit concept is laid bare in the preambles to the various sets of proposed and final foreign tax credit regulations. Further, payments that are disregarded for one country’s tax purposes but regarded for another’s can distort the tax base used to measure effective tax rate. It may take only a couple of sentences to explain these concepts but many pages of regulations to implement them.

Because the Section 954(b)(4) high-tax exception, as set forth in proposed regulations, would apply both to GILTI and Subpart F, the question naturally arises as to how the new GILTI FTC rule would affect GILTI’s interaction with Subpart F. GILTI and Subpart F already have different FTC rules. Unlike GILTI, Subpart F income does not have its own basket. Rather, income is allocated to the general or passive category (or other Section 1.904-4(m) basket, if applicable). Further, foreign taxes cannot be carried back or forward under the GILTI regime and are subject to the 80% haircut. So, even if one (like the IRS and Treasury) thinks that the high-tax exception should apply uniformly to Subpart F and GILTI, applying the regulatory elective rule as a mandatory rule, as proposed in the Senate white paper, would result in a very significant change for Subpart F without—until now—articulating why such a change to Subpart F is appropriate.

Of course, there is another model for separating high-taxed and low-taxed income, and that is the high-tax kick-out rules of Treas. Reg. Section 1.904-4(c). Those rules are not simple but they do provide clear grouping rules and a detailed methodology.

Whatever model one adopts, one has to identify where the income and taxes go if they are kicked out of GILTI. In the high-taxed income kick-out rule of Treas. Reg. Section 1.904-4(c), the high-taxed income and associated taxes go to the general basket, foreign branch income basket, GILTI basket, or other specified separate category, based on where the FTC rules would otherwise assign it. In this case, income and associated taxes kicked out of the GILTI basket would go to the residual category, at least based on my reading of Treas. Reg. Section 1.951A-2(c)(7)(ii)(A). That would appear to be the general basket (since the foreign branch income basket is not applicable to CFCs and the high-taxed income should also be kicked out of the passive basket, at least in most instances) or a specified separate category under Section 1.904-4(m). This raises the question of how high-taxed income thrown out of the GILTI basket would interact with Subpart F income of the CFC (which would presumably be the only other source of income in the general basket for CFCs under Section 960). The FTC category to which the high-taxed income goes is not just a theoretical question; it affects whether the high-taxed income is included in current income (through Subpart F) or not included in income until distributed (or not taxed at all if Section 245A applies) and the credits thereby lost. Which approach a taxpayer prefers could depend on what the previously taxed earnings and profits “PTEP” rules eventually say.

Conclusion

The anti-cross-crediting proposals engender a lot of to-be-answered questions, but still come conclusions may be drawn.

The proposed changes to GILTI aggravate the basic problem that Subpart F and GILTI are two separate regimes with similar but not identical goals. As long as we have two different such regimes and there is some ability to structure transactions so that taxpayers can choose whether they want Subpart F or GILTI to apply, the well-advised will compare treatment and choose accordingly. Policymakers have no ground to complain when that happens when they intentionally make design choices that encourage rational taxpayers to act that way.

Treating regarded and disregarded entities differently in the new GILTI FTC rules will lead to distortions and tax planning. If a multinational group concludes that it gets a better GILTI FTC result by filing check-the-box elections to convert wholly-owned CFCs to disregarded entities (or vice versa), then it will. The five-year limitation on change in entity classification will limit switching back-and-forth, but such elections would still be sensible to fix structural problems.

Domestic corporations and U.S. individuals are taxed differently on income earned through CFCs. Designing rules with domestic corporations in mind, and then subjecting U.S. individuals to some of those rules (e.g., GILTI and Subpart F) and not to others (e.g., Section 245A) does not make sense. If these new changes are adopted, individuals who are U.S. shareholders should not be an afterthought as they were in the Tax Cuts and Jobs Act.

If the goal is to encourage U.S. companies to act in the ways described in the Made in America Tax Plan or the Senate white paper, the rules need to be clear and coordinated. Subjecting foreign-source income to high taxes in some cases and not in others will inevitably lead taxpayers to explore ways to achieve a less harsh result. Simpler and coordinated rules may eliminate some of that response, but it is even more important to be reasonable and realistic in what policymakers can accomplish with these changes.

This column does not necessarily reflect the opinion of The Bureau of National Affairs, Inc. or its owners.

Author Information

John L. Harrington is one of the co-leaders of Dentons U.S. tax practice.

The views expressed in this article are solely those of the author’s.

Bloomberg Tax Insights articles are written by experienced practitioners, academics, and policy experts discussing developments and current issues in taxation. To contribute, please contact us at TaxInsights@bloombergindustry.com.