Menu
Tax Notes logo

Should CFC Regimes Grant a Tax Credit for Qualified Domestic Minimum Top-Up Taxes?

Posted on June 27, 2022
Heydon Wardell-Burrus
Heydon Wardell-Burrus

Heydon Wardell-Burrus is completing his doctorate of philosophy in law at the University of Oxford and is a researcher at the Oxford Centre for Business Taxation. He previously worked for the Australian Taxation Office and contributed to its project on the OECD pillars. He thanks Michael Devereux, John Vella, Paul Oosterhuis, Steve Shay, Casey Plunket, Noam Noked, and Jason Yen for their helpful comments.

In this article, Wardell-Burrus considers whether the qualified domestic minimum top-up tax in the OECD’s pillar 2 model rules should be treated as creditable from a policy perspective.

On December 20, 2021, the OECD released its pillar 2 model rules, which introduced the concept of a qualified domestic minimum top-up tax (QDMTT), an idea that did not appear in any of the inclusive framework’s public announcements or the OECD’s pillar 2 blueprint or October 2021 statement.

In essence, a QDMTT allows a source country to impose a top-up tax, which would otherwise be imposed by the income inclusion rule or undertaxed profits rule.1 While its operation appears simple, it raises some difficult questions regarding the operation of the global anti-base-erosion (GLOBE) rules and their interaction with today’s international tax system.

At a high level, the QDMTT concept appears to respond to the criticism that pillar 2 would give new taxing rights to the ultimate parent entity’s jurisdiction and would not advantage source countries. That criticism has always been a bit misleading because even without a QDMTT, the GLOBE rules would still have given priority to the source country’s regular corporate income tax. If the rules were adopted without a QDMTT, source countries could still have imposed domestic minimum taxes, which would have ensured that multinational enterprises had effective tax rates (ETRs) equal to the minimum rate under the GLOBE rules. In other words, even without a QDMTT design, source countries could have prevented a foreign top-up tax by adopting domestic minimum taxes.

How to Interpret the QDMTT Concept

This article assumes that a QDMTT imposes the amount of top-up tax that would generally be payable under an IIR imposed on an ultimate parent entity that wholly owns the subsidiary. More precisely, it considers a QDMTT a domestic minimum tax equal to a jurisdiction’s top-up tax percentage (15 percent less the ETR) multiplied by the amount of excess profit (total accounting profit less the amount of the substance-based income exclusion (SBIE)). Based on that interpretation,2 a QDMTT is designed to generally equal the jurisdictional top-up tax.3

Under an alternative interpretation of the definition of a QDMTT, the source country must raise its domestic tax liability to 15 percent for excess profits in the jurisdiction.4 It does not calculate the ETR under the GLOBE rules, so it does not take cross-jurisdictional taxes into account. The source country must itself impose the 15 percent tax liability.

While that interpretation might be initially attractive based on the wording of the definition of a QDMTT, it is not preferred because it produces absurd results.

First, the model rules’ definition of a QDMTT refers to increasing the “domestic tax liability with respect to domestic excess profits to the Minimum Rate.” That wording does not rely on the concept of covered taxes at all but instead on a new concept — domestic tax liability — not used elsewhere in the model rules or accompanying commentary. It would be strange in the context of the QDMTT for the GLOBE rules to abandon the infrastructure built into the definition of covered taxes. For instance, the GLOBE rules address both timing differences and GLOBE losses by increasing the amount of covered taxes. That operates to increase the ETR and therefore decrease the top-up tax liability in the jurisdiction. If a QDMTT did not take into account the ETR and top-up tax percentage, it could impose an additional tax liability because of timing differences and GLOBE losses when the IIR and UTPR would not. There is no sensible reason for that outcome.

Second, there are cases when it would be odd not to take into account foreign taxes. Consider an entity that has $100 of excess profit that was subject to withholding of 10 percent in the payer’s country (so the company receives $90). The recipient company’s domestic tax system has a tax rate of 20 percent and provides foreign tax credits for withholding taxes paid. The company must therefore pay an additional $10 in domestic taxes under its regular corporate income tax ($20 domestic tax liability ($100 * 20 percent), minus a tax credit of $10 for the withholding tax). In that case, the entity has paid $20 of tax on $100 of income ($10 of foreign withholding tax and $10 additional domestic tax liability). If a QDMTT doesn’t take into account foreign taxes, it must result in the imposition of an additional $5 of tax to reach the minimum 15 percent domestic tax liability. This would produce an overall tax rate of 25 percent. Again, there is no reasonable justification for that outcome.

Both outcomes appear too far-fetched to have been intended under the GLOBE rules. There is a sensible policy rationale for the first interpretation — the source country should get to apply the top-up tax that would generally otherwise arise under the GLOBE rules. The second interpretation would produce arbitrary results and make a QDMTT extremely risky for any country that sought to adopt it. In many cases, a QDMTT would impose significantly greater taxation than that under the GLOBE rules.

For those reasons, the first interpretation — that a QDMTT considers foreign cross-jurisdictional taxes in determining the amount of top-up tax due — should be adopted.5

Two Types of Domestic Minimum Taxes

There is a subtle but important difference between a QDMTT and a domestic minimum tax that would raise a jurisdiction’s ETR to the minimum rate.

The GLOBE rules operate by calculating the ETR for a jurisdiction’s total profits (adjusted GLOBE income). Once the ETR is calculated, the top-up tax percentage (15 percent minus the ETR) is applied to excess profits only (adjusted GLOBE income less the SBIE amount). The top-up tax liability is the top-up tax percentage multiplied by excess profits. The SBIE is considered only for determining the amount of excess profit; it is irrelevant for determining a jurisdiction’s ETR.

The innovation of the QDMTT is that it allows the source country to impose the top-up tax liability itself; otherwise, that top-up amount would be imposed in a different country (or countries) under the IIR or UTPR. A QDMTT shifts the incentives for source countries adopting domestic minimum taxes: Without it, a source country would need to raise its ETR to 15 percent to prevent the IIR or UTPR from applying. Adopting a QDMTT means the source country does not need to raise its ETR to the minimum under its regular corporate income tax regime. Instead, it can simply impose the top-up tax liability that would otherwise arise under the GLOBE rules. In practice, that means the source jurisdiction can retain a tax rate below 15 percent on profits that are covered by the SBIE without triggering the IIR or UTPR. Based on a recent consultation document, that appears to be Ireland’s proposed strategy.

For a domestic minimum top-up tax to be qualified (and therefore a QDMTT), it must operate to increase the rate on excess profits to the minimum rate consistently with the GLOBE rules. Setting aside that a QDMTT can apply a different accounting standard than that generally required under the GLOBE rules, a QDMTT should generally match the liability that would otherwise arise under an applicable IIR in a wholly owned group.

Interaction With CFC Rules

There is an interaction problem between the QDMTT concept in the model rules and preexisting controlled foreign corporation rules.6 A QDMTT must increase the domestic tax liability for domestic excess profits to the minimum rate. To do so, it must determine the current ETR under the GLOBE rules, which require in article 4.3 that cross-jurisdictional taxes (including CFC taxes) be counted in determining the CFC jurisdiction’s ETR. Accordingly, for a tax to be a QDMTT, it must calculate that jurisdiction’s ETR, having taken into account CFC taxes on income attributed to that jurisdiction. Put differently, a QDMTT applies the top-up tax on a post-CFC tax basis.

The question then becomes how QDMTT amounts are to be treated under the relevant CFC rules. An HM Revenue & Customs official raised that question during an International Fiscal Association event in May,7 and on June 8 a U.S. Treasury official said “the ordering priority between a CFC tax regime and a QDMTT is still being debated, with no clear answer yet.”8

A reasonable starting point for that question is considering a QDMTT simply to be an alternative minimum tax imposed by another country on the corporate income of a domestic entity (or permanent establishment). It is difficult to see immediately why that kind of tax would not be a creditable tax under both domestic and foreign tax rules (as applicable under the CFC regime) and under any applicable tax treaty.

One observer has considered whether a QDMTT should be non-creditable under U.S. FTC regulations (T.D. 9959) by reason of being a soak-up tax.9 U.S. Treasury official Isaac Wood advised taxpayers and practitioners not to spend too much time reading the FTC regs or divining an answer therefrom because they will have to be updated if pillars 1 and 2 are adopted.10

Insofar as a QDMTT is a soak-up tax, that is the clear intention of the GLOBE rules and the inclusive framework agreement — source states are entitled to soak up the amount of tax that would otherwise be imposed under the rules. The rationale behind the non-creditability of soak-up taxes is a policy objection to other countries adopting them. It would be strange to accept a QDMTT as a key design element of the GLOBE rules only to then claim there is a policy objection to soak-up taxes. Accordingly, if CFC regimes deny a credit for a QDMTT, that should be justified by reasons other than that it might be a soak-up tax.

It is also unclear from the pillar 2 model rules or their commentary whether a country is expected to credit a QDMTT — even though the commentary considers the closely related question of crediting the IIR and UTPR:

To preserve the intended rule order, domestic tax regimes should not provide a foreign tax credit for any tax imposed under a Qualified UTPR or IIR which is implemented in a foreign jurisdiction, otherwise the application of that domestic tax regime would create circularity issues since those Taxes have already been determined prior to applying the Qualified UTPR or IIR.11

Considering the importance of a QDMTT, the commentary’s silence on crediting it is noteworthy. However, that may simply reflect the time pressure on the negotiation process.

QDMTT Creditable Under CFC Rules

If a QDMTT is a creditable tax, there is a feedback loop between the imposition of CFC taxes and a QDMTT itself.

The essence of the problem is that both rule sets are intended to operate as provisions of last (or at least later) resort. A QDMTT calculation takes into account the CFC taxes applicable to the relevant income that feeds into the ETR calculation. While the mechanism is slightly convoluted, the effect is that a QDMTT credits the CFC tax. Similarly, the amount of CFC tax would be giving a credit for applicable QDMTT tax.

Example

Consider a simplified example in which the GLOBE tax base perfectly matches the CFC tax base. The source jurisdiction (which is also the CFC’s jurisdiction) imposes a regular corporate income tax of 5 percent (assume $100 of profit) and has adopted a QDMTT (for simplicity, there is no SBIE).

The parent jurisdiction imposes a CFC tax of 12.5 percent with a full credit for source-jurisdiction taxation (including any QDMTT). The relevant income is not passive income, so the rule limiting the push-down of CFC taxes on passive income does not apply.

Before a QDMTT is applied, the source country has imposed regular corporate income tax of $5. The parent jurisdiction includes all $100 of profit in the parent’s income for a tax liability of $12.50 ($100 * 12.5 percent) and, as noted, fully credits source taxation; accordingly, the parent is liable for $7.50 in CFC tax ($12.50 - $5).

The source jurisdiction then applies a QDMTT. The GLOBE base recognizes $100 of GLOBE income and $12.50 of covered taxes ($5 of source-state corporate tax and $7.50 of parent-state CFC tax). That produces an ETR of 12.5 percent, so there will be a QDMTT liability of $2.50 (because there is no SBIE).

So far so good: There has been total tax of $15 imposed on $100 of excess profit. However, if the parent jurisdiction is required to credit the QDMTT, then the parent entity has paid too much CFC tax. It amends its return, noting that it paid $7.50 of creditable tax in the source jurisdiction ($5 of regular corporate tax and $2.50 of QDMTT) but should have been liable for only $5 (the $12.50 liability reduced by the $7.50 credit). Thus, it requests a refund of $2.50 for excess CFC taxes paid.

However, the MNE has now paid only $10 total tax on its GLOBE income ($5 corporate tax and $5 CFC tax), so the QDMTT must be recalculated. The top-up tax should be $5 rather than $2.50, so the parent would have to pay an additional $2.50 of QDMTT, which must be credited against the original CFC tax.

While that pattern might strike the reader as circular, it is in fact a spiral. The amount of CFC tax is consistently reduced by increasing QDMTT credits. On the fourth iteration, there is no CFC tax at all — the $5 source-country corporate tax and $7.50 QDMTT completely offset it, resulting in a final result of $5 of source-country corporate tax and $10 of QDMTT.

Those outcomes are shown in the table, which considers each iteration as a recalculation of the CFC tax (then QDMTT), taking into account a tax credit for the QDMTT.

CFC Taxes and the QDMTT

 

First Iteration

Second Iteration

Third Iteration

Fourth Iteration

Profit

$100

$100

$100

$100

Domestic Tax

 

 

 

 

Regular Corporate Income Tax (5%)

$5

$5

$5

$5

CFC Tax (12.5%)

 

 

 

 

Liability Inclusion

$12.50

$12.50

$12.50

$12.50

Credit

$5
($5 corporate tax)

$7.50
($5 corporate tax + $2.50 QDMTT)

$10
($5 corporate tax + $5 QDMTT)

$12.50
($5 corporate tax + $7.50 QDMTT)

CFC Tax Liability

$7.50
($12.50 - $5)

$5
($12.50 - $7.50)

$2.50
($12.50 - $10)

$0
($12.50 - $12.50)

QDMTT

 

 

 

 

Covered Taxes

$12.50
($5 corporate tax + $7.50 CFC Tax)

$10
($5 corporate tax + $5 CFC)

$7.50
($5 corporate tax + $2.50 CFC)

$5
($5 corporate tax)

ETR

12.5%

10%

7.5%

5%

Top-Up Tax Percentage

2.5%

5%

7.5%

10%

Excess Profit

$100

$100

$100

$100

QDMTT Liability

$2.50
(2.5% * $100)

$5
(5% * $100)

$7.50
(7.5% * $100)

$10
(10% * $100)

Total Tax

 

 

 

 

 

$15
($5 corporate tax + $7.50 CFC tax + $2.50 QDMTT)

$15
($5 corporate tax + $5 CFC tax + $5 QDMTT)

$15
($5 corporate tax + $2.50 CFC tax + $7.50 QDMTT)

$15
($5 corporate tax + $10 QDMTT)

That structure produces a cliff effect for the amount of CFC tax. If the CFC tax liability creates an ETR of 15 percent, then there is no QDMTT at all. All the top-up tax goes to the parent jurisdiction under its CFC rules. However, as soon as the CFC tax falls below 15 percent, the spiral effect comes into play: The amount of CFC tax declines, and the QDMTT amount increases to claim all the revenue. The number of iterations required to reach that result depends on how close the CFC liability is to the QDMTT liability.

The example ignores three important sources of complexity: (1) the CFC tax base is unlikely to match the GLOBE tax base; (2) the difficulty of cross-crediting (when there is a pool of FTCs under the relevant CFC rule), as well as potentially applicable FTC limitations; and (3) the SBIE. The feedback loop becomes more complicated when the QDMTT amount is reduced in each iteration by being inapplicable to SBIE amounts.

All three of those complicating factors apply in the context of the U.S. global intangible low-taxed income regime. While a full analysis of that regime is beyond the scope of this article, various differences between the GILTI and GLOBE bases make it unclear how a GILTI liability would interact with a QDMTT liability in any particular case. While one might initially assume that the GILTI liability would have to be below any cliff (at least before 2026) as a result of a total tax rate of between 10.5 and 13.125 percent, the rate is not the end of the story. Differences between the relevant tax bases and covered taxes, as well as differences between the GLOBE SBIE concept and the GILTI concept of net deemed tangible income return mean there is no simple answer.

QDMTT Not Creditable Under CFC Rules

The above outcomes follow if a QDMTT is treated as a creditable tax under the FTC rules that apply to a CFC regime. If a QDMTT is not treated as a creditable tax, the spiral problem and related cliff effect are avoided. However, that approach has several problems of its own.

First, if a QDMTT would be creditable under current law, the law would need to be changed. Further complications could arise if a tax treaty requires the country imposing CFC taxation to grant a tax credit for income taxes paid in the other country. While domestic law may be changed, an applicable treaty, which is more difficult to amend, might require a tax credit for the QDMTT.

Second, failing to credit a QDMTT might both inappropriately grant the parent jurisdiction priority over the source country’s domestic minimum tax and create unintended incentives to expand CFC regimes. It appears that a QDMTT is meant to allow the source country to have first bite at top-up tax revenue — that is, the source country can claim the top-up taxing rights itself and does not have to raise its ETR to the minimum to protect itself from top-up taxes applied by other jurisdictions. By refusing to grant a credit for a QDMTT, parent jurisdictions effectively jump ahead of the source country’s domestic minimum tax. If an applicable CFC regime would raise the ETR to the minimum rate, the top-up tax will go to the parent jurisdiction applying the CFC rule instead of the source country. That undermines the source country’s ability to protect its position by simply adopting a QDMTT. To avoid that outcome, source states would need to adopt an alternative minimum tax that raises the ETR to 15 percent — which, as noted above, is different from a QDMTT.12

That could also encourage jurisdictions to adopt CFC rules that seek to include approximately 15 percent of GLOBE income.13 By expanding their CFC regimes to include approximately 15 percent of GLOBE income, parent jurisdictions could jump the queue and collect what is effectively top-up tax revenue ahead of source jurisdictions that impose a QDMTT. In many, if not most, cases, source countries will have imposed a QDMTT to avoid other jurisdictions imposing top-up tax on income sourced in their jurisdiction.

Importantly, that scenario would not be costless to an MNE subject to the CFC tax, which does not replace a QDMTT at a 1-1 ratio unless there is no SBIE amount in the jurisdiction. The CFC tax is counted toward the ETR (with a denominator of adjusted GLOBE income), while a QDMTT counts toward the top-up tax liability (which is applied to excess profits). That means a greater amount of CFC tax is required to displace any QDMTT amount; accordingly, there will be a higher total tax burden on the MNE if the CFC rate rises to replace a QDMTT. Whether there is a meaningful difference in those amounts will be a function of the SBIE amount in the jurisdiction.

There is also an interesting question of whether the common approach of pillar 2 implicitly prevents a country from expanding its CFC regime in that way. In explaining the common approach in the October 2021 statement, the OECD said:

This means that [inclusive framework] members . . . are not required to adopt the GloBE rules, but, if they choose to do so, they will implement and administer the rules in a way that is consistent with the outcomes provided for under Pillar Two, including in light of model rules and guidance agreed to by the [inclusive framework].

The GLOBE rules arguably operate on the basis that a QDMTT gives the source country the first opportunity to apply the top-up tax, followed by the IIR jurisdiction and, finally, UTPR jurisdictions. If a country were to amend its CFC rules such that the amount of GLOBE income were included in the parent entity’s taxable income and subject to a 15 percent rate, would that be consistent with the common approach? What if the parent country adopted a CFC rule that applied a 15 percent tax rate calculated on its own tax base and applied to active income? Would that be a legitimate exercise of state sovereignty or a violation of the common approach? Where is the line between an inconsistent application of the GLOBE rules and merely expanding applicable CFC rules?

It seems plausible that at least some jurisdictions may have agreed to the GLOBE rules on the assumption that they could avoid the application of top-up taxes to profits sourced within their borders by adopting a QDMTT. If those countries were then to discover that that proved practically ineffective because other countries expanded their CFC rules to take precedence over a QDMTT, they might argue that their legitimate expectations had been disappointed. A technical response to the criticism that the rules always gave priority to CFC rules — including those that are not common outside the GILTI regime — might not be politically acceptable.

Refusing to grant a tax credit for a QDMTT discourages source states from relying on a QDMTT in lieu of their existing corporate income taxes. Some commentators have written about incentives created under the GLOBE rules to reduce the regular corporate income tax under the existing corporate tax base and rely on a QDMTT.14 If FTC regimes do not grant credits for QDMTTs, source countries could face a major downside in relying on QDMTTs in lieu of their regular corporate taxes because countries applying a CFC tax would be able to impose taxation ahead of them.

That outcome may itself lead source countries to adopt domestic minimum taxes that apply an alternate minimum tax equal to 15 percent of GLOBE income — that is, raising the jurisdictional ETRs for MNEs operating in their jurisdiction to 15 percent. That would be an alternate minimum tax that does not take into account the SBIE (and therefore differs from a QDMTT). The more that source countries take that approach, the less incentive there is for countries to expand their CFC regimes to jump the queue ahead of a QDMTT. Those interactions create what one observer has called “an iterative cross-border strategic game.”15

Implementation

An important question is how a mutually agreed answer — that is, to credit or not to credit — can be achieved consistently, given the structure of the GLOBE rules and that they will be subject to a common approach. The best approach will depend on the answer.

If there is a multilateral decision that a QDMTT should not be creditable, there will need to be treaty amendments to ensure that changes can be made to domestic law without treaty obstacles. That would best be achieved through a multilateral instrument, which is already envisaged for the subject-to-tax rule. That multilateral instrument would update tax treaties to ensure a QDMTT is not a creditable tax under them (domestic law would then also need to be amended).

If there is a multilateral decision that a QDMTT should be creditable, a multilateral instrument could also be used — but a simpler mechanism might be better. The GLOBE rules could restrict the definition of a CFC tax by not treating a tax as a CFC tax unless it credited a QDMTT. In that case, there would be no incentive to expand the CFC rules to the minimum rate to jump the queue. If the CFC regime credited a QDMTT, it would be imposed after, rather than ahead of, a QDMTT. If the CFC regime did not credit a QDMTT, it would be imposed in addition to a QDMTT, IIR, and UTPR — that is, in addition to any GLOBE top-up tax liability.

GILTI

The interaction issues addressed above apply to CFC taxes on both active and passive income.16 However, they are far more salient for active CFC rules — and therefore the U.S. GILTI regime,17 which clearly should not be treated as a qualified IIR unless it gives an FTC for a QDMTT. It would be an unfair advantage for the United States if all other countries adopted IIRs giving priority to source-country QDMTTs, but the United States could apply GILTI ahead of any applicable QDMTT.

If the GILTI regime is treated as a CFC tax and is not required to grant a credit for a QDMTT, the United States again may have an advantage over the rest of the world. GILTI would apply a broad CFC rule to active income with priority over any source QDMTTs (unlike an equivalent IIR). Supporters of that position may argue that other countries can adopt similar CFC rules. However, countries that do so will arguably invert the agreed rule order under the common approach (which grants priority to a source-country QDMTT over GLOBE top-up taxes) and undermine a QDMTT’s role as a provision designed to apply ahead of pillar 2 top-up taxing provisions. As noted above, it is unclear whether source jurisdictions (and developing countries in particular) would consider that a breach of the inclusive framework’s implied agreement. While that outcome can be avoided by CFC rules granting a credit for a QDMTT, it gives rise to the spiral problem and cliff effect described above and would add major complexity to GILTI’s interaction with GLOBE rules.

The best outcome is likely to be modifying the GILTI regime so that it is treated as a qualifying IIR (and therefore is not a CFC tax under the GLOBE rules). That would significantly reduce the pressure for CFC regimes to grant credits for a QDMTT because there would not be an interaction problem for CFC taxes on active income. While that outcome would not remove the incentive for countries to expand their CFC regimes to active income, it would remove the largest source of pressure to do so. It could be accompanied by an agreed position that expanding CFC rules to active income (at roughly the GLOBE rate) would be treated as an inconsistent implementation of the GLOBE rules (and a breach of the agreement). Of course, it would also raise difficult questions regarding the limits on expanding CFC taxation in a pillar 2 world.

If GILTI is not modified, the question of rule order needs to be answered — should a QDMTT credit GILTI, or should GILTI credit a QDMTT? The undesirable spiral and cliff effects arise when the answer is both. If a QDMTT credits GILTI, then no FTC should be granted in the United States for a QDMTT. This outcome would give the United States an advantage over other countries in that its CFC rule on active income would apply ahead of a QDMTT, while the IIR would apply after it. Other countries might try to rectify that imbalance by expanding their CFC rules to active income (at approximately the GLOBE rate), which, in turn, could encourage source countries to abandon a QDMTT in lieu of alternative minimum taxes that raise the ETR to 15 percent. That course of events would undermine a QDMTT as a part of the bargain struck by countries in the GLOBE rules.

If GILTI credits a QDMTT, then a QDMTT should not take into account cross-jurisdictional CFC taxes imposed under GILTI when calculating the ETR. That does not appear to be the agreed design of the GLOBE rules (under which CFC taxes are taken into account in determining the ETR and top-up tax amount). Of course, a special arrangement could be reached regarding GILTI to alter that outcome.

In the U.S. green book for fiscal 2022, the Biden administration effectively proposed that the GILTI regime credit any IIR taxes paid in the context of a U.S. intermediate parent of a multinational group with a foreign parent. That proposal would have accepted that GILTI would be applied after GLOBE top-up taxes on foreign-headquartered MNEs.

It is difficult to see why GILTI should not credit a QDMTT if the Biden administration was willing to credit the IIR in those circumstances. Granted, the fiscal 2022 green book was released in May 2021, and the GLOBE rules have developed significantly since then, so it is of limited use in considering the administration’s current position on the interaction between GILTI and a QDMTT.

An Administrative Solution?

The GLOBE rules offer a choice: either CFC rules do not credit a QDMTT, or a QDMTT and CFC rules try to credit each other (leading to the spiral effect). An alternative design would be for a QDMTT not to credit applicable CFC taxes, but that is inconsistent with the way a QDMTT is calculated under the GLOBE rules.

An alternative solution could be to create a concession under which a domestic minimum tax could still be treated as a QDMTT even if it does not take CFC taxes into account in calculating the ETR.

As noted, for a domestic minimum tax to be a QDMTT, it must calculate the ETR consistently with the GLOBE rules — and that includes taking into account cross-jurisdictional taxes. However, jurisdictions could agree to concede as part of implementation that even if a domestic minimum top-up tax did not include foreign CFC taxes in its ETR calculation, it would still be treated as a QDMTT.

That would give a QDMTT priority over CFC taxes and solve the spiral problem and related cliff effect. The benefit of that approach is that the source country would collect the relevant revenue ahead of both CFC rules and the IIR/UTPR. It also would be consistent with the general premise of both the GLOBE rules and current international tax architecture, which gives source countries priority in imposing tax.

There are several downsides to that approach. First, it would create a divergence in the calculation of a QDMTT and otherwise applicable amounts of top-up tax under the IIR or UTPR. There would be two calculations: a QDMTT calculation (without taking into account cross-jurisdictional taxes) and an IIR/UTPR calculation (that would take relevant CFC taxes into account). However, the amount of a QDMTT calculation should generally be higher than what would otherwise arise under the IIR or UTPR.18 That is because the ETR calculated under a QDMTT would generally be lower (because it did not take into account cross-jurisdictional taxes) than that under the IIR/UTPR (which will be increased by cross-jurisdictional taxes). The result generally would be that the amount of QDMTT would exceed the amount that would otherwise arise under the IIR/UTPR. Because the IIR/UTPR provides a full credit for any QDMTT paid, there would generally be no additional IIR/UTPR.

Those two calculations involve additional compliance and administrative complexity, which could be mitigated via a simplification measure under which an IIR calculation is not required when there is an applicable QDMTT (although there could be independent reasons not to accept this).

Another downside is that adopting a QDMTT with that design would no longer apply the minimum level of total tax — that is, it would generally impose greater taxation than would otherwise arise under the GLOBE rules. For countries seeking to attract foreign investment, that might be a disadvantage. However, by adopting this option as a concession (rather than fundamentally changing the design of a QDMTT), countries would effectively have a choice. They could adopt a QDMTT of this design (not taking into account CFC taxes) or a regular QDMTT. Under a regular QDMTT, they would impose the minimum level of top-up tax but would potentially cede revenue to other jurisdictions imposing CFC taxes. Under the concession, they would retain their taxing priority at the potential cost of imposing additional taxation.

An Aside

The concession discussed in this section is premised on the notion that a QDMTT, by definition, must be calculated in the same manner as the top-up tax liability that arises under the GLOBE rules for an IIR or UTPR. The commentary to the pillar 2 model rules states that a QDMTT must be implemented and administered consistently with the outcomes provided for under the model rules and commentary. It is unclear what the rules consider an outcome, but if the inclusive framework were to adopt a wide interpretation, it could arguably conclude that a QDMTT that does not credit CFC taxes meets the commentary’s requirement despite the difference in calculation.

Using a concession rather than changing the QDMTT requirements would allow source states to choose which type of QDMTT better suits their goals. There would also be reduced criticism that the GLOBE rules were being rewritten, which could be a serious concern — if one element of the GLOBE rules is being reconsidered, there could be pressure to reopen others. On the other hand, offering alternatives to jurisdictions adopting QDMTTs would add complexity for states considering how best to respond to the wider adoption of the GLOBE rules.

Further, countries with CFC rules would still need to address how to treat QDMTTs of each type. Under existing laws, it could be difficult to justify drawing a legal distinction between the two types of QDMTTs to provide a credit for one (a concessionary QDMTT that does not take CFC taxes into account when calculating the ETR) and deny a credit for the other (a regular QDMTT). If legislation were to address that, it would be unlikely to pose a large problem. However, if the creditability of the tax is to be determined under existing treaty provisions, the outcome will likely be difficult to support. Again, a consistent outcome could be best achieved via a multilateral instrument to update treaties.

Conclusion

The pillar 2 model rules as drafted offer a choice: either CFC rules do not credit a QDMTT, or a QDMTT and CFC rules both try to credit each other. The first introduces incentives to expand CFC regimes and arguably undermines a QDMTT as a source-country response to pillar 2. The second leads to spiral and cliff effects and adds complexity to already complex rules.

The administrative solution outlined here could be adopted to address some of those concerns. It would allow jurisdictions to choose to adopt QDMTTs that do not take into account CFC taxes, which would avoid the spiral problem at the cost of increasing other types of complexity. It would not be a panacea, and the issues outlined above would still need to be addressed for non-concessionary QDMTTs.

This article has analyzed interaction concerns between CFC taxes and QDMTTs, but CFC taxes are not the only cross-jurisdictional taxes for which crediting will be an issue. Further complexity will arise as interaction issues are considered for those other taxes (for example, taxes on PEs and transparent or hybrid entities). Also, crediting QDMTTs could produce interaction concerns with other components of domestic tax systems (for example, how an amount of income subject to a QDMTT should be treated under anti-hybrid rules or the subject-to-tax rule).

Further, there will be allocation questions when a QDMTT has been paid by one entity in a jurisdiction but based on the tax characteristics of other group entities (which may have different ownership).19

Some jurisdictions might not want to treat a QDMTT as creditable under their domestic tax systems simply to avoid the complexity that will arise from those issues — and that outcome is not without costs. It will raise questions of whether a jurisdiction can refuse to treat a QDMTT as a creditable tax under applicable treaties. It could also encourage countries to expand their CFC rules to active income at approximately the GLOBE rate. That might trigger the need to draw boundary lines between legitimately expanding a jurisdiction’s CFC rules and breaching the requirements of the common approach under pillar 2.

Expanded CFC regimes could fairly be regarded as undermining a QDMTT as a vehicle for source countries to address pillar 2 without being subject to foreign top-up taxes. Source countries could foreseeably argue that their support for pillar 2 was predicated on the assumption that adopting a QDMTT would be effective. Accordingly, resolving those problems should not be considered a mere technicality but an important element of both the design of and political support for pillar 2.

FOOTNOTES

1 For an outline of the QDMTT’s role in pillar 2, see Mindy Herzfeld, “How Does the Qualified Domestic Minimum Top-Up Tax Fit Into Pillar 2?Tax Notes Int’l, Apr. 18, 2022, p. 315; and Michael Devereux, John Vella, and Heydon Wardell-Burrus, “Pillar 2: Rule Order, Incentives, and Tax Competition,” Oxford University Centre for Business Taxation Policy Brief 2022 (Mar. 11, 2022).

2 As adopted by Herzfeld in “More on GLOBE Ordering: CFC Rules,” Tax Notes Int’l, May 2, 2022, p. 603.

3 Even so, differences can arise because of the application of different accounting standards under the applicable IIR and the QDMTT.

4 Both interpretations were raised at the U.S. Branch of the International Fiscal Association’s annual conference (June 2-3, 2022).

5 This article assumes that interpretation to be the correct one.

6 For a broader discussion of CFC regimes under the GLOBE rules, see Herzfeld, supra note 2.

7 See Stephanie Soong Johnston, “Pillar 2 and CFC Tax Credit Issue Needs Work, U.K. Official Says,” Tax Notes Int’l, May 16, 2022, p. 947.

8 See Andrew Velarde, “Treasury Outlines Viable Approaches for GILTI Pillar 2 Allocation,” Tax Notes Int’l, June 13, 2022, p. 1474.

9 See Noam Noked, “Defense of Primary Taxing Rights,” 40(2) Va. Tax Rev. 341, 358 (2021) (arguing that a domestic minimum tax should not be a soak-up tax under the current definition). However, whether a QDMTT would be considered a soak-up tax is likely to be an academic question only.

10 See Velarde, “Treasury Defends GILTI as Good Pillar 2 Regime,” Tax Notes Int’l, May 23, 2022, p. 1094.

11 Commentary to the model rules, article 4, para. 45.

Noked has suggested that the rationale supporting that passage could be extended to a QDMTT but concludes that “further guidance on the question of the creditability of QDMTTs is needed”; see “Designing Domestic Minimum Taxes and Subsidies in Response to the Global Minimum Tax,” Intertax (forthcoming 2022).

12 For discussion of the trade-off for source countries in adopting a QDMTT versus a domestic minimum tax that raises the ETR to 15 percent, see Noked, supra note 11.

13 That point is raised in Noked, supra note 11, citing Steven Towers, “Will the GloBE Rules Cause the Expansion of CFC Regimes?” International Tax Bytes blog, Apr. 1, 2022.

14 See Devereux, Vella, and Wardell-Burrus, supra note 1.

To clarify, a QDMTT is still a profit-based tax on corporate income; other observers may prefer to describe the relevant incentive as one for states to shift from their current corporate income tax base to a new base that is calculated under the GLOBE rules and qualifies as a QDMTT. For simplicity, this article refers to a QDMTT as separate from the corporate income tax. However, countries will need to consider whether to administer QDMTTs as part of their corporate tax regimes — a question raised by Ireland in its recent consultation paper, discussed infra.

15 Allison Christians, “Let the GILTI/GLOBE Games Begin,” Tax Notes Int’l, May 16, 2022, p. 913.

16 The interaction issues are limited to the extent the limitation on pushing down passive income applies. However, many CFC regimes apply to other income (such as “tainted” sales income) which would not be subject to the passive income limitation under the GLOBE rules; those regimes are reasonably described as CFC regimes on passive income despite applying to income other than passive income within the narrow definition in the GLOBE rules.

17 For further discussion, see Herzfeld, supra note 2.

18 That will not necessarily be the case for (at least) two reasons. First, a QDMTT can be calculated using a different accounting standard than that for the IIR/UTPR calculation. Differences in the recognition of income or expenses under different standards can lead to a divergence in those calculations in both directions. Second, the IIR/UTPR calculation can take into account additional top-up tax as required under article 5.4 of the pillar 2 model rules, which might not be taken into account under a QDMTT calculation.

19 See Herzfeld, supra note 1, at 318.

END FOOTNOTES

Copy RID