Reuven S. Avi-Yonah is the Irwin I. Cohn Professor of Law at the University of Michigan Law School. He would like to thank Lukas Hrdlicka and Martin Sullivan for helpful comments.
In this installment of Reflections With Reuven Avi-Yonah, Avi-Yonah reflects on the relationship between investment incentives and the possible neutralization of pillar 2 effects.
In a recent column, Tax Notes’ Martin Sullivan asked whether a country that wishes to neutralize the effect of pillar 2 on its investment incentives can get around the OECD prohibition on a multinational enterprise receiving what amounts to a refund of the pillar 2 tax it pays to that country. He writes that:
It would make a mockery of the pillar 2 taxation system if an investment hub imposed a 15 percent minimum tax on a company — thereby shielding profit in that hub from other jurisdictions’ pillar 2 tax — and then, through a separate mechanism, unconditionally returned the new revenue dollar-for-dollar to the taxpaying company. Pillar 2 model rules prevent that. But what if the benefits offsetting the tax are not dollar-for-dollar but instead merely approximate the revenue raised from the new tax?
And what if it’s easy for a company to satisfy the conditions for receiving the benefit — perhaps by doing business the same way it did before? Should the collateral benefits — whether deliberately or by coincidence offsetting the burden of the new tax — prevent other countries’ imposition of pillar 2 tax on investment-hub profits?1
Sullivan points out that under the OECD rules, a country that imposes a qualified domestic minimum top-up tax (QDMTT) may not “provide any benefits that are related to” the QDMTT (emphasis added). The question is what “related to” means. Sullivan argues that if this rule is interpreted too broadly, it “would seem to preclude all the benefits — the cash grants, accelerated depreciation, and qualified refundable tax credits — that were clearly intended.” He concludes that “it is difficult — perhaps impossible — to say where we should draw the boundary between ‘no benefit’ qualifying and disqualifying incentives.”2
Sullivan then gives a specific example: Singapore. To offset the effect of its new QDMTT on investment, Singapore is proposing to adopt a refundable investment credit. MNEs can receive a credit of up to 50 percent of qualifying expenditures on a broad set of costs, including capital expenditures and wages. Because these expenditures are higher than profits, “it will be commonplace for the new refundable investment credit to potentially offset the entire burden of the new DMTT on a taxpayer.”3
How will Singapore prevent those credits from costing it more than the revenue it collects from the QDMTT?
The key is that the new tax credit “will be awarded on an approval basis,” through the Singapore Economic Development Board and Enterprise Singapore, with the total quantum of credits that a company is eligible for being determined by those authorities.
The Singapore Economic Development Board and Enterprise Singapore are government agencies charged with enhancing:
Singapore’s position as a global center for business, innovation, and talent, championing the development of Singaporean businesses [and supporting] the growth of Singapore as a hub for global trading and startups.4
Sullivan argues that negotiations will enable Singapore to precisely calibrate the refundable credits to an MNE’s QDMTT liability:
Let’s assume for discussion’s sake that Singapore — despite the perceived necessity of adopting a 15 percent DMTT — wants to maintain the level of tax benefits it now provides to U.S.-headquartered multinationals. (And why wouldn’t that be a goal, given its overwhelming success in attracting foreign investment?)
Singapore could observe a company’s recent tax history in the country and then estimate the new DMTT liability it expects to collect from that company. For example, it might be that a subsidiary of a U.S. corporation in Singapore is booking $100 million of profit in Singapore and paying Singaporean tax at an effective rate of 5 percent. Singapore’s corporate tax collections from that company would rise from $5 million to $15 million in 2025 when the DMTT is in effect.
Now here’s the crux of the matter. Singaporean authorities may also observe, for example, that the company has annual capital expenditures of about $60 million and annual employee compensation of $80 million. Singaporean authorities could then, as part of the discretionary approval process, set a refundable investment credit rate about equal to 7 percent on the $140 million of qualified expenditures to generate about $10 million of the credits. That would negate the additional burden of the new DMTT.
I doubt this subterfuge will be effective in avoiding the pillar 2 rules.5
The problem is that if a country wants to prevent a refundable investment incentive from costing it more revenue than it raises through the QDMTT, it must have a mechanism to adjust the credit to the investor’s QDMTT liability. But in that case, the OECD could adopt a specificity rule like the U.S. foreign tax credit rules and the EU state aid rules that would render the DMTT unqualified and therefore result in the application of top-up taxes by other countries in which the MNE operates. Even without adopting such a rule, the OECD could apply it in practice through the pillar 2 peer review process.
The United States confronted the refundability problem in the context of the FTC years ago when countries including Saudi Arabia and Indonesia imposed a “tax” in exchange for a specific benefit such as the right to extract oil or gas there.6 The result was the adoption of reg. section 1.901-2:
Notwithstanding any assertion of a foreign country to the contrary, a foreign levy is not pursuant to a foreign country’s authority to levy taxes, and thus is not a tax, to the extent a person subject to the levy receives (or will receive), directly or indirectly, a specific economic benefit (as defined in paragraph (a)(2)(ii)(B) of this section) from the foreign country in exchange for payment pursuant to the levy . . .
(B) Specific economic benefit. For purposes of this section and sections 1.901-2A and 1.903-1, the term “specific economic benefit” means an economic benefit that is not made available on substantially the same terms to substantially all persons who are subject to the income tax that is generally imposed by the foreign country, or, if there is no such generally imposed income tax, an economic benefit that is not made available on substantially the same terms to the population of the country in general. Thus, a concession to extract government-owned petroleum is a specific economic benefit, but the right to travel or to ship freight on a government-owned airline is not, because the latter, but not the former, is made generally available on substantially the same terms. An economic benefit includes property; a service; a fee or other payment; a right to use, acquire or extract resources, patents or other property that a foreign country owns or controls (within the meaning of paragraph (a)(2)(ii)(D) of this section); or a reduction or discharge of a contractual obligation. It does not include the right or privilege merely to engage in business generally or to engage in business in a particular form . . . [Emphasis added.]
Similarly, the EU state aid rules prohibit a country from granting a “specific” subsidy to a taxpayer in the form of a tax benefit that is not generally available to other taxpayers. To constitute prohibited state aid, a tax measure:
must be specific or selective in that it favors some undertakings or the production of some goods.
Selectivity may derive from a legislative, regulatory, or administrative provision, or from a discretionary practice on the part of the tax authorities. . . . The fiscal aid notice accordingly considers that the main criterion in applying article 87(1) of the EC Treaty to a tax measure is to prove that the measure provides in favor of some undertakings in the member state an exception to the application of the tax system. The tax system of reference should thus first be determined to decide whether a derogatory advantage has been granted. However, to determine if state aid is involved, it should then be proven that the advantage is specific or justified by the nature or general scheme of the tax system.7
Under both the U.S. and the EU rules, the Singaporean scheme (and similar schemes applied by other countries) would presumably fail because the negotiation process with specific MNCs would mean that the tax incentive is “a specific economic benefit.”
The basic problem is that the whole qualified refundable credit mechanism of pillar 2 is a mistake, because it is only vaguely related to the real dividing line between permitted and prohibited tax expenditures. The key question under pillar 2 is whether the purpose of the tax expenditure (whether given as a credit or a deduction or a lower tax rate) is primarily to draw investment away from other locations or to address an externality. If the former, it should be banned; if the latter, it should be permitted.8
Most tax incentives are primarily designed to attract investment from other locations. For example, the foreign-derived intangible income regime in the United States provides a lower tax rate of 13.125 percent for income derived from exporting goods, services, or intangibles out of the United States, and it is designed to encourage both U.S. and foreign MNCs to shift those activities into the United States. This is the type of incentive that should be prohibited under pillar 2, and a top-up tax should apply to raise the rate to 15 percent.9 But a credit for research and experimentation in the United States should be permitted because R&E creates positive externalities, and the green credits in the Inflation Reduction Act should also be permitted because they address the negative externality of greenhouse gas emissions.
The OECD should allow for these types of credits to not reduce the effective tax rate for pillar 2 purposes whether or not they are refundable, because refundability has nothing to do with whether the credit represents harmful tax competition, and invites game playing like what Singapore proposes.
FOOTNOTES
1 Martin A. Sullivan, “Will Singapore’s Refundable Investment Credit Trigger Pillar 2 Tax?” Tax Notes Int’l, May 13, 2024, p. 995.
2 Id.
3 Id.
4 Id. Sullivan adds that “Singapore is hardly unique in its use of negotiations with foreign investors to determine tax benefits. In a 2022 survey of more than 100 investment promotion agencies, the United Nations Conference on Trade and Development found that in only a minority of cases are investors automatically eligible for incentives based on measurable criteria.”
5 Sullivan is also doubtful, concluding that “if Singaporean authorities overtly make the calculation described above in the preliminary stages of their approval process, this refundable, expenditure-based credit would seem to violate the spirit of the no-benefit requirement: It effectively would be a rate reduction dressed up as a substance-based incentive. Given the difficulty of demonstrating that the amount of the credit isn’t overtly determined by reference to the amount of profit in Singapore — especially when a written approval agreement made possible under statutory law describes it as equal to a percentage of qualifying expenditures — it is hard to see how any practical pillar 2 rule can be developed to prevent these ‘related to’ benefits from being identified if incentives are negotiated.”
6 For the history, see the classic article by Joseph Isenbergh, “The Foreign Tax Credit: Royalties, Subsidies, and Creditable Taxes,” 39 Tax L. Rev. 227 (1983).
7 Pierpaolo Rossi-Maccanico, “A Review of State Aid in Multinational Tax Regimes,” Tax Notes Int’l, May 28, 2007, p. 941.
8 See Reuven S. Avi-Yonah, “The Case Against Expensing R&E,” Tax Notes Int’l, Feb. 5, 2024, p. 743.
9 That would require applying the undertaxed profits rule to the U.S.-source income of U.S.-based MNCs benefiting from FDII, which would be controversial, hence the current safe harbor from the UTPR for countries with a statutory rate of 20 percent or higher. But the FDII rate is scheduled to rise above 15 percent in 2025. In my opinion, FDII has fulfilled its purpose and should be allowed to expire before it is subject to a WTO challenge, because it clearly violates the WTO prohibition against export subsidies. See Avi-Yonah, “The Elephant Always Forgets: Tax Reform and the WTO,” University of Michigan Public Law Research Paper No. 585 (Jan. 1, 2018).
END FOOTNOTES