The Ambition and Limits of the Global Minimum Tax

In a previous article,¹ I laid out a political case for book minimum taxes and how they help address the challenge of coordinating political actors inside the United States and, most importantly, around the world. This article focuses on the global anti-base-erosion (GLOBE) agreement on a global (book) minimum tax to highlight exactly what that deal, as it now stands, would coordinate.

The agreement’s ambition is more limited than many proponents and critics have suggested. The minimum tax doesn’t end competition for many types of business activity or even limit the ability of governments to provide net subsidies to significant elements of their business sectors. Races to the bottom (or to the top in terms of size of the subsidy) almost certainly would continue even if the global agreement were fully implemented.² But that also means a key criticism of the minimum tax framework in the United States — that it would sharply limit Congress’s (and other governments’) ability to offer incentives for specific activities — doesn’t bear out.³ Such subsidies, including ones recently enacted in the CHIPS and Science Act (P.L. 117-637) and Inflation Reduction Act (IRA, P.L. 117-169) and many others, would face little limitation by the minimum tax as it has been negotiated.

So if the minimum tax does little to restrict these kinds of subsidies and competition, what does it do? The minimum tax targets the taxation of the largest returns to capital investment and efforts to shift profits to low-tax jurisdictions (which would result in the appearance of large returns to the actual activity in those jurisdictions). If implemented, it would restrict the related forms of tax competition: competition for paper profits and competition for activity associated with very large returns. That is a significant and important step forward, even though other competition for business activity would live on.

This article reviews how the agreement tries to achieve that targeting, with a focus on the line between what is treated as a spending-side grant (subject to minimal limitation) versus a tax cut (more sharply limited), and how the agreement tries to differentiate those two categories and not rely simply on governments’ labeling. Several of those line-drawing exercises have received considerable attention,⁴ including in a new report from the OECD on tax incentives and the GLOBE agreement.⁵ This article puts the line-drawing in broader context, making clear what is and isn’t the fundamental ambition of the global minimum tax as it stands.

Would more ambitious limits on tax and subsidy competition have been desirable? That question takes us back to the topic of my previous article on why to pursue book minimum taxes to begin with and the wisdom of seriously considering politics when evaluating these kinds of policy changes.⁶ The book minimum tax is responsive to the foundational political and institutional challenges of a world in which many governments understandably want to retain considerable fiscal flexibility. And while a broader ambit for the minimum tax would have come with real upsides of addressing those competitions for business activity, it would have come at a cost of each government’s flexibility — a step too far for those governments at present.

The hope now is that governments can implement the global minimum tax as it has been negotiated — and recognize its focus on the taxation of the largest profits and profit shifting. The agreement no doubt faces a rocky path toward implementation, given the lack of adoption so far in the United States and a lack of unanimity in Europe. The political and institutional barriers that the minimum tax attempts to solve are very real ones after all, but efforts so far have kept open a path, however rocky.⁷

I. Minimum Tax and Its Many Limits

To restate the basics: The GLOBE agreement applies top-up taxes to bring applicable multinational groups up to a 15 percent minimum effective tax rate in each country in which they operate. The details of this agreement have been described in considerable detail elsewhere,⁸ and I focus instead on the main limits to its ambit and what those limits mean practically for the targeting of the tax.

The GLOBE rules’ reach is limited by the agreement in several key ways:

• Tax cuts versus grants. The regime applies to tax only — and not how governments treat businesses on the spending side of the ledger.⁹ However, as has long been recognized (generating many an article), “tax” and “spending” are formalistic concepts and don’t in themselves capture much of anything substantively meaningful. Benefits to businesses can be delivered via both the tax and spending codes, and while the administering agency or relevant title of the law may differ, dollars (or euros or name your currency) are dollars, and economically speaking, the same effect can be had whether a benefit is denoted a tax cut or a spending grant.¹⁰

  Under the GLOBE rules, what is considered to be a spending-side grant would get picked up as income to a relevant business — like any other receipt to the company and consistent with book accounting principles.¹¹ Thus, it wouldn’t be counted as a tax cut. The difference is that, for a company with an effective tax rate that is below 15 percent and thus subject to the minimum tax, a spending-side grant would be subject to a minimum tax rate of 15 percent, but a tax cut would be reduced by 100 percent under the minimum tax (to bring the corporation up to the 15 percent minimum rate floor). To offset the 15 percent haircut, countries could also make the spending-side subsidy more generous if desired (by “grossing up” the subsidy). The same isn’t true of a tax cut, which is simply eliminated for a company facing the minimum tax.

  At first blush, that can present a puzzle. If tax and spending are interchangeable, would the agreement actually accomplish much of anything? It still would or, at least, should. As discussed below, governments should be meaningfully limited in terms of the kinds of benefits they can deliver — and how much they can shield high-return income from resulting in net transfers to governments — through what are characterized as grants. This is because the GLOBE rules effectively impose substantive guardrails distinguishing grants from tax cuts.

• Qualified refundable tax credits. The spending-side treatment under the agreement isn’t limited to benefits given through spending codes (and meeting the relevant criteria); it also applies to “qualified” refundable tax credits.¹² And, even if not refundable by the government, transferable tax credits — credits that the company sells to others to monetize the benefit — may get spending-side treatment.¹³ In those cases, the credit is then treated as income and, like a grant, subject only to a 15 percent haircut if the business is subject to the minimum tax.

• Depreciation. The minimum tax regime is intended to reduce additional liability resulting from timing differences between the book tax system underlying the minimum tax and tax depreciation schedules.¹⁴ And, under the GLOBE agreement, tax rules that “carve out” income through faster than book depreciation for tangible assets and research and development expenses — up to and including expensing of those investments — wouldn’t systematically generate additional liability under the minimum tax. That too gives governments significant flexibility when it comes to treatment of smaller returns to these investments. The immediate write-off of those investments as they are made, rather than as they economically depreciate, can effectively shield the “normal return” to investment — the return generally available in the market — from taxation.¹⁵ Supernormal returns — the excess returns — are what would remain for the minimum tax. In other words, countries can go as far as setting a zero rate on the normal return, and without constraint, under the minimum tax regime.

• Substance-based carveout. There is also the overt “substance-based income exclusion,” which effectively reduces the profits subject to the minimum tax by at least 5 percent of tangible assets and payroll — although starting at 10 percent of tangible assets and 8 percent of payroll before phasing down over 10 years.¹⁶ That is another way of effectively exempting relatively low return activity from the minimum tax. However, the emphasis here should probably be on relatively. After all, expensing can already effectively eliminate taxation on the normal return to investment; the substance-based carveout can be layered on top — and, initially, at relatively high rates of expenditure.

• Size threshold and equity method of accounting. Only large companies — groups with revenue exceeding €750 million annually — are subject to the minimum tax.¹⁷ Further, these groups must have a large stake in underlying operations for the minimum tax regime to even capture the relevant income from those operations. Operations in which the large multinational holds a minority but still significant stake are excluded from the minimum tax regime, even if income from those operations shows up on the books of the parent company (booked using what is known as the equity method of accounting).¹⁸ Allowing total discretion to domestic authorities when it comes to the treatment of smaller companies apparently also includes discretion regarding tax benefits delivered to subsidiaries in which large companies have a minority stake, and those tax benefits are then monetized by offsetting the large company’s tax liability. This includes credits like the low-income housing tax credit and the new markets tax credit.¹⁹

II. What Is Targeted

The list of carveouts could leave the impression that there is little left to tax — but there is or can be as long as the carveouts are sufficiently policed. What should be left are very large returns from the largest companies or profits shifted into a country from activity occurring elsewhere (and so appearing as large profits), with governments retaining substantial flexibility for all the rest.

The minimum tax should still have scope: There is considerable evidence of profit shifting around the world,²⁰ and there is considerable evidence that large returns to capital constitute a large and growing share of corporate profits subject to tax, at least in the United States. As William G. Gale and Samuel I. Thorpe have summarized, much of the U.S. corporate income tax base may have been excess returns in recent years given the combination of expensing rules protecting the normal return from taxation, the falling normal return to capital, and changes in the U.S. economy, like increasing concentration.²¹

Still, how are large returns and profit shifting targeted while leaving governments such flexibility beyond that? Here, I revisit several of the distinctions discussed above with a focus on “grants” versus “tax cuts” and how that line-drawing can distinguish some subsidies from others.

A. ‘Grant’ Versus ‘Tax Cut’ and Large Returns

To reiterate: In theory, a grant could reduce the effective tax rate applied even to large returns to investment and shifted profits by any amount — and, in fact, zero out the effective tax rate or more. Take two very simple (maybe glaringly simple) approaches to doing so:

• Key to taxable profits. The grant could be based on the amount of taxable profit reported. For example, if a grant were set equal to 15 percent of taxable profit, that would be the equivalent of a rate cut of 15 percentage points.²² The tax rate in the “tax system” would be 15 percent, but the grant would reverse it, achieving a 0 percent effective tax rate.

• Key to tax liability. To take the most direct approach for this: The government could give a grant equal to 100 percent of any tax liability. The effective tax rate is of course 0 percent then. Another approach is less direct but can achieve similar results. The grant could be contingent on economic activity but be very generous and then limited to the size of the company’s tax liability. That, too, keys to the amount of tax liability and limits the size of the credit to the taxes that would otherwise be due. The tax rate can be zeroed out — as long as the company engages in enough of the economic activity (and the credit is sufficiently generous). That is, of course, recognizable as the equivalent of a “nonrefundable” tax credit.

So the GLOBE rules must give some substance to the difference between grants and tax cuts or be rendered effectively meaningless, and they do — or certainly try to.

To determine whether a benefit is considered a grant, the GLOBE system doesn’t simply look to whether the government calls it a grant. Instead, the distinction is rooted in the accounting principles that govern what is treated as a grant versus a tax cut for financial accounting purposes. And those principles impose guardrails — though, as some commentators have noted, there is room for further definition, and the guardrails should probably be made even more explicit.²³

First, the grants cannot directly key off taxable income (or tax liability) without being recharacterized as a tax cut. That is the likely — and certainly desired — result of the accounting principles underlying the GLOBE system.²⁴ So the first route should be cut off.

Second, governments cannot limit the benefit they deliver through a grant under the GLOBE rules to the amount of taxes that the company at issue would otherwise pay.²⁵ That too would likely be recharacterized as a reduction in tax under the accounting rules and pillar 2 commentary. Additional restrictions apply for a tax credit to be treated like a grant. The credit must be “designed in a way such that it must be paid as cash or available as cash equivalents within four years” of the activity that establishes eligibility for the credit.²⁶ So it must be refundable — and within a limited period. Alternatively, it appears that tax credits that are transferrable and sold to other companies would also be treated as income and not as a reduction in tax — again, as long as they aren’t directly keyed to the amount of tax liability that the selling company would otherwise pay.²⁷

Those restrictions should have real bite. Governments cannot reduce the effective tax rate on large returns so easily. Direct approaches of keying off profits or tax liability shouldn’t work. Governments can provide unlimited subsidies for specific business activities, but they must be willing to either (1) actually send checks out for the excess over the amount of a company’s tax liability; or (2) allow the tax credits to be sold and reduce the tax liability of other corporations or individuals. There is no effective way of mimicking a rate reduction on large returns with those guardrails in place — the subsidies must be based on activity and not the profits or liability of the company.

B. Other Carveouts

The other carveouts have the same basic flavor, and large returns of the largest companies or profits shifted into the jurisdiction are still likely to face tax. The allowance for depreciation and the substance-based carveouts do that mechanically — allowing low tax rates on lower returns but maintaining the minimum tax for the larger returns. After all, the system — even before the substance-based carveouts are taken into account — allows for expensing of tangible investment and R&D expenses, which would eliminate taxation of the “normal” rate of return.

The size limit is also mechanical, up to a point. By their very nature, small companies are excluded from a minimum tax imposed only on large companies, and governments are then unconstrained in their treatment of small companies. The flexibility when it comes to the treatment of the large companies — to a point that could be problematic — comes in the approach to subsidiaries. The rules entirely exclude profits (and taxes) paid by subsidiaries in which the multinational parent has a minority stake — even if, under the accounting rules, net income is counted on the books through the equity method of accounting.²⁸

That presents a danger to the ambition of the minimum tax: Large corporations could push high-return activity into these smaller entities in which they retain only minority ownership. It’s a version of the broader challenge with any minimum tax of this kind and with a size threshold: Companies could try to stay small to avoid the tax. The practical limits on that will be the transaction costs associated with doing so: Many large companies seem likely to put a high value on retaining control of high-return relevant activity involving specialized know-how. Here, the targeting relies on what is, one hopes, a relatively “inelastic” margin for corporate restructuring.

C. Incentives in the United States

The issue of subsidies has spawned controversy in the United States — with some people arguing that important incentives that Congress wants to deliver would be too limited by the regime.²⁹

Congress, in fact, has many paths to provide incentives consistent with the minimum tax regime and without those subsidies effectively being reversed by other countries enforcing the minimum tax (or by the United States itself imposing a top-up tax). To reiterate, they can be delivered: (1) as a grant; (2) as a refundable tax credit; (3) as a transferrable tax credit in all likelihood; (4) as faster than economic depreciation; and (5) to smaller companies, even if they can be used only through monetization by larger companies, if those smaller companies aren’t majority-owned by those larger companies.

Those options cover almost all major tax incentives that Congress has provided to businesses with two notable exceptions described further below. For example, the new markets tax credit and low-income housing tax credit apparently wouldn’t face significant limitation because they are generally delivered to smaller companies and then monetized by investment of larger companies in a way the agreement probably doesn’t significantly limit. The same is true of faster-than-economic depreciation, which is favored in the United States (in the form of expensing and bonus depreciation and its many iterations). It is also the case when it comes to new incentives in the CHIPS and Science Act and the IRA — they have been delivered as a combination of grants, refundable tax credits, and transferrable credits, all apparently consistent with the GLOBE guardrails and so, at most, subject to an only 15 percent haircut if a company is otherwise subject to the minimum tax.

The two exceptions — which could be more significantly affected by the GLOBE rules — are the research and experimentation tax credit and the foreign-derived intangible income deduction. Still, there are readily available routes to reform when it comes to the R&E credit, and effects could be modest when it comes to FDII.

Starting with the R&E tax credit—this credit goes directly to a number of large companies and isn’t refundable. So, it could be caught in the GLOBE net. However, the credit could be reformed so that it works similarly to new refundable or transferable credits in the recent CHIPS and Science Act and IRA. Reformed in this way, the R&E credit would not face significant limitation by the GLOBE rules just as the new tax credits in those laws would not.

FDII offers a deduction — and thus, a lower effective tax rate —for specific profits that are supposed to be associated with intellectual property located in the United States. The effective tax rate is 13.125 percent through 2025 and 16.406 percent after and, if combined with other low-tax income, could trigger a minimum tax for a business. Notably, the minimum tax applies only if other company profits aren’t taxed at a sufficiently high rate. Many companies could probably maintain unrestricted access to the FDII incentive for better or worse — and even at the 13.125 percent rate — if that low-tax income for the company were also combined with higher tax income. Because of this and the fact that the FDII rate is not currently that far below 15 percent and is scheduled to increase, the effects on FDII could be modest.

The United States would be more sharply restricted in its ability to offer an even more generous FDII-like deduction, dropping the rate well below 15 percent, if that lower rate applied to a significant proportion of companies’ incomes. But that reflects the targeting of the minimum tax regime — tax incentives focused on profits like this rather than incentives for activity are restricted in their generosity.

The experience of the United States is illustrative more broadly. There is significant flexibility within the minimum tax framework to subsidize business activity, whether it be through grants or appropriately structured tax credits. The one route that runs squarely into the minimum tax is offering a very low tax rate on a significant share of large profits.

III. A Minimum Minimum Tax

The minimum tax that has emerged from the recent years of negotiation should be seen for what it is — and praised for what it can accomplish, rightly criticized for some of the limits on its ambition, and not painted as sharply restricting how governments treat business activity.

This regime won’t stop countries from delivering subsidies toward large businesses and even racing each other to do so. Competition among countries of this kind would live on. As long as they do so with grants, tax credits consistent with guardrails, faster than economic depreciation, or tax subsidies for activity undertaken by smaller companies, the minimum tax does little to arrest that. Today’s global race involving semiconductor manufacturing and clean energy technology is one example of that — and the United States is now undertaking a response. Perhaps one day, governments around the world will come to the view that the cost of such races is no longer worth bearing and would be willing to give up flexibility to stop it. But that isn’t the world we have now.

The agreement on the table does stop a competition: the race for paper profits through profit shifting and for activity associated with the largest returns. That shouldn’t be mistaken for a small goal. But it is a much more limited goal than significantly constraining government fiscal policy when it comes to business taxation and subsidies. So what has emerged is a minimum minimum tax in that important sense. And the question now is whether the world can move forward in cooperation.