Should Congress pay for a lower effective tax rate on global intangible low-taxed income by increasing the 20 percent haircut on foreign creditable taxes? Would it be better to increase the haircut on foreign tax credits in 2026 instead of reducing the section 250 deduction, as now scheduled under present law? If, as some Democrats have proposed, Congress hikes taxes on GILTI, should it reduce foreign tax creditability instead of raising the rate? The answer to all these questions is probably yes.
Mechanics
Since 2010, when President Obama proposed current taxation of excess returns associated with transfer of intangibles “in circumstances that evidence excessive income shifting,” we have seen more than a few minimum tax proposals. For the moment we will put aside the complex issue of what foreign income should be subject to minimum tax (income from high profits? income from intangibles? income from sales back to the United States?) and focus on the more mechanical aspects of a minimum tax: the difference (if any) of the minimum tax rate from the regular tax rate and the offset (if any) to U.S. tax for foreign income taxes paid. Once we put aside issues of the tax base, the underlying load-bearing structure of all the major minimum tax proposals may be summarized with the following simple formula:
MinTax = (A * tUS – B * tF) * MinInc
MinTax is minimum tax revenue. MinInc is the minimum tax base. The U.S. and foreign tax rates are tUS and tF, respectively. A is the ratio (always less than or equal to 1) of the effective minimum tax rate to the headline U.S. rate. B is the ratio (between 0 and 1, inclusive) of the U.S. tax reduction for any increase in foreign taxes paid on minimum taxable income.
Now let’s move away from the abstract into some real-world examples. Under prior law’s worldwide system with no deferral or in the case of subpart F income, the constant A is equal to 1. In other words, there is no separate minimum tax rate on foreign-source income. And because the FTC provides 100 percent reimbursement, B is also equal to 1, except when FTCs are limited and then B is equal to 0.
Under current law’s taxation of GILTI, the constant A generally is equal to 0.5 (because of the 50 percent of income deduction provided by section 250). The effective minimum tax rate is 10.5 percent. The variable A will decline to 0.675 and the effective rate will rise to 13.125 percent if Congress allows the section 250 deduction to decline on schedule to 37.5 percent beginning in 2026. The constant B is equal to 0.8 because of the 20 percent haircut in section 960(d) applied to creditable foreign taxes on GILTI.
In November 2013 then-Senate Finance Committee Chair Max Baucus floated a minimum tax proposal called Option Z. Under this proposal, 40 percent of minimum taxable income was exempt from U.S. income tax. That left 60 percent subject to U.S. tax and 60 percent of foreign income tax levied on minimum taxable income creditable against U.S. tax. In this case, A and B are both equal to 0.6.
For those readers who find pictures helpful, we can dust off our high school math and graph minimum tax lines. Let’s put the effective minimum rate of U.S. tax on the vertical axis and the foreign tax rate on the horizontal axis. The term AtUS is the effective rate of U.S. tax when the foreign tax rate is 0 (the vertical intercept). The variable B is the absolute value of the slope of the minimum tax line. The smaller B is, the closer our minimum tax line is to being horizontal.
In Figure 1A we graph minimum tax lines for all three variations of minimum tax described above. Subpart F income is fully taxed and foreign taxes are fully creditable (up to the foreign tax limit). The minimum tax line intersects the vertical axis at 21 percent. The slope is -1. So the minimum tax line also intersects the horizontal axis at 21 percent.
GILTI is taxed generally at half the U.S. statutory rate, and only 80 percent of foreign taxes are creditable. The minimum tax line intersects the vertical axis at 10.5 percent. The slope is -0.8. So the minimum tax line intersects the horizontal axis at 13.125 percent.
Option Z minimum taxable income is taxed at 60 percent of the U.S. statutory rate. The minimum tax line intersects the vertical axis at 12.6 percent. The slope is -0.6. So the minimum tax line intersects the horizontal axis at 21 percent.
Figure 1B shows the same three proposals but with global (U.S. plus foreign) tax on the vertical axis. In this case, the intercept of the vertical axis is the same as in Figure 1A. But now the slope of the line is 0 or positive and equals 1 plus the slope of the line in Figure 1A.
The adjacent table provides the values of A and B for current law, various proposals from U.S. lawmakers, and a few alternatives we have made up here. (More details about others’ proposals can be found in the references at the end of this article.) In an effort to organize this jumble of options, the table divides the proposals into three categories corresponding to the three proposals described above. In Figure 1B, all the total tax lines for Category 1 proposals are first horizontal, then have a positive slope of 1. All the total tax lines for Category 2 proposals are first slightly upward-sloping, then have a positive slope of 1. All the total tax lines for Category 3 proposals are upward-sloping with no break.
Full Offset of Foreign Tax (Category 1) | Limited Offset of Foreign Tax (Category 2) | Exemption or Partial Exemption (Category 3) |
|---|---|---|
|
|
|
aThese taxes, as proposed, may differ significantly in their nonstructural aspects. Their definitions of taxable income vary — from the very broad (for example, almost all non-subpart F income subject to low tax) to the very narrow (for example, excess income from intangibles transferred out of the United States). Some may be levied where foreign tax rates and limitations are levied on a per-country or an overall basis. Most of these taxes are drafted as new categories of subpart F income, but the tax on GILTI is separate. (Notably, the otherwise limited discussion of minimum taxation in the OECD’s “Public Consultation Document on Digital Taxation” states that a minimum tax “would supplement rather than replace a jurisdiction’s CFC rules” para. 96.) | ||
Going Old-School
Prominent commentators have made it unfashionable to discuss international policy objectives in terms of capital export neutrality (CEN) (which can be effectuated with an unlimited FTC) and capital import neutrality (CIN) (which is complete exemption of (active) foreign income from domestic tax). (See “Further Reading” at the end of this article.) One has even disparaged the relevance of thinking in terms of a compromise between CEN and CIN. Yes, there are other dimensions to international tax policy than worldwide efficient allocation of capital (assuming there must be a corporate tax or a tax on capital at all), such as domestic politics and international relations. Where precisely these considerations will lead, however, is far from clear. Although tax policy folks are surely fatigued by “the battle of the acronyms,” CEN and CIN are still useful points of reference, especially in a policymaking environment where providing unschooled lawmakers some guidance is better than none at all.
Suppose you are starting with a territorial system and you are a member of the dying breed of believers in CEN. Now also suppose your government has enlisted you to raise money from a minimum tax on foreign profits. Given that CEN is optimal (by your assumption), and given the broadly applicable rule in public finance economics that economic inefficiency (also known as deadweight loss) is proportional to the square of the difference between the actual and optimal tax rate, you will want your minimum tax line to minimize the sum of the squared differences between the optimal tax rate (which for a CEN person is 21 percent) and your minimum tax line. In this case the optimal minimum tax is just a downward-sloping line parallel to the “Subpart F” line in Figure 1A set at whatever rate is necessary to meet your revenue goal.
Now suppose instead you are a CIN purist, but political pressure and revenue needs require you to impose a minimum tax. If CIN is optimal, the same economic reasoning used above will tell us the most efficient outcome is a minimum tax line that on average is as close as possible to zero taxation. In this case, the optimal minimum tax is just a horizontal line set at whatever rate is necessary to meet your revenue goal.
In Figure 2, the downward-sloping line represents a CEN minimum tax with a 14 percent rate and a 100 percent creditability of foreign tax. The slope is -1, and both vertical and horizontal intercepts are at 14 percent. (For our equation, A would equal 2/3 and B would equal 1.) The total tax line is horizontal until the foreign rate equals 14 percent when FTC relief is shut off. Notice that these lines are the same shapes as the subpart F lines in figures 1A and 1B. So all those minimum taxes in Table 1 that resemble subpart F — including Obama’s 2012 proposal and Baucus’s 2013 Option Y — would in effect be superimposing worldwide taxation (CEN) onto a territorial system (CIN).
A CIN minimum tax is just a constant, flat tax with no adjustment for foreign taxes. In Figure 2, the rate is set at 7 percent. The intuition (such as it is) is that the foreign rate is the optimal rate, and given our revenue constraints, we want as much as possible to stay close to it. (For our equation, A would equal 1/3 and B would equal 0.)
In sum, the higher the intercept (reflected in the value of A) and the steeper the slope (reflected in the value of B), the more any minimum tax is pushing the system toward CEN.
Balance and Trade-Offs
Before and even now after the passage of the Tax Cuts and Jobs Act, the U.S. system of taxing international income is described as a hybrid territorial-worldwide system. Before the TCJA, it was impossible to be precise about this because the benefits of deferral (which gave the system its CIN flavor) varied and were often uncertain for each taxpayer. But now that territorial taxation is the foundation of our tax system, we can more concretely assess the system along the CEN-CIN spectrum.
Suppose, for example, it is decided to give equal weight to CEN and CIN in our tax policy. In words a bit more precise, it is determined that deviations from CEN and CIN are equally harmful, so we want to minimize the square of the distances of our minimum tax line from both the CEN and CIN ideals. (In math-speak, we are minimizing the sum of squared residuals from both the CEN and CIN line — but don’t repeat that at parties.)
Such a minimum tax is shown in Figure 3. It is similar to Option Z as portrayed in figures 1A and 1B. In fact, if we reset the parameters of Option Z from a 40/60 exemption/taxation split to a 50/50 split (so B = 0.5) and set the rate equal to half the statutory rate (so A = 0.5), Option Z and our compromise CEN-CIN minimum tax are exactly the same. Notice that the optimization that determines the compromise minimum tax gives us the rate of tax (as indicated by the vertical intercept) as well as the slope. Therefore, it also sets revenue raised by the tax. In Figure 3, the compromise minimum tax exactly bisects the triangle formed by the vertical axis, the CEN line, and the CIN line. The 10.5 percent U.S. minimum tax is reduced by 0.5 percent for every percentage point increase in the foreign tax rate. (Below we will explore optimization when revenue targets must be met. In that case, the compromise will no longer necessarily bisect the triangle.)
OK, so what did we learn from all the rigmarole? Some interesting tidbits emerge if we compare our compromise tax to current-law GILTI, as shown in Figure 4. First, wonder of wonders, in its design of GILTI, Congress actually selected the effective minimum tax rate — 10.5 percent — that coincides with our CEN-CIN compromise minimum tax. Second, if we use equal compromise between CEN and CIN as our standard, the current-law 20 percent haircut on FTCs is too small. It should be increased from 20 percent to 50 percent, and this would be a revenue raiser. All taxpayers with foreign tax rates below 21 percent would be losers. The biggest losers would be those with a foreign effective tax rate of 13.125 percent. Tax increases would be smaller as the foreign tax rate moves in either direction from this amount. Democrats who want to raise taxes on multinationals may want to consider increasing the FTC haircut instead of reducing the section 250 deduction. House Ways and Means Committee member Lloyd Doggett, D-Texas, and Senate Finance Committee member Sheldon Whitehouse, D-R.I., have introduced legislation that would, contrary to our suggestion, eliminate the haircut while raising the effective minimum tax rate. (See section 2(e) of S. 780.)
Of course, taxpayers don’t like tax increases and it is not our purpose here to advocate for one. So what if Congress was looking for a way to improve GILTI in a revenue-neutral manner? Some complicated calculus shows that if Congress is willing to adopt the standard of equal concern about CEN and CIN, it could lower the GILTI rate below 10.5 (which moves the tax unambiguously toward CIN) and offset that change with a reduction in the haircut from below 50 percent but still well above 20 percent. (More precise adjustments will depend on the distribution of foreign tax rates paid by U.S. multinationals out there in the real world.)
Skin in the Game
So far, we have been discussing the design of minimum taxes using CEN and CIN as guiding principles. And if we agree to a compromise in the war between these two extremes, the analysis so far could be used to justify to some that the haircut on the FTC should be increased. But there is another, entirely separate reason to consider this route.
For years professor Daniel N. Shaviro of New York University School of Law has pointed out that CEN and CIN provide guidance only on one margin, on the marginal incentive to invest abroad. He emphasizes that a second margin is also critical — the margin concerning a multinational’s sensitivity or cost-consciousness to foreign taxes. Without deferral, a 100 percent credit for foreign taxes leaves U.S. multinationals that have not reached the credit limit with zero incentive to reduce foreign taxes, and as such, 100 percent creditability is not in the interest of the United States (unless the United States is expecting reciprocity or trying to implement foreign policy that includes Marshall Plan-like foreign aid). Also, in some circumstances 100 percent creditability provides incentives for foreign governments to raise their taxes (or not reduce them).
In the not-too-distant past, the idea of an across-the-board cut in the creditability of foreign tax would have met with responses ranging from silent puzzlement to vociferous outrage. Sacrilege. But the logic of at least moving in that direction is inescapable, and somehow Shaviro’s revenue-raising observation managed to make it through the enigmatic tax-making process that ends with a presidential signature. By 2016 a revised minimum tax proposal from Treasury included a 15 percent haircut justified with stilted language: “It may be difficult for the IRS to verify that a taxpayer has exhausted practical remedies under foreign law to reduce its reasonably expected foreign tax liability over time in a manner consistent with a reasonable interpretation of foreign law.” And, of course, in 2017 the TCJA’s GILTI included a 20 percent haircut to the FTC.
But how large should the reduction in the FTCs be? Who came up with 20 percent anyway? Shaviro, in fact, does more than suggest the haircut. He provides a quantitative answer. From a national (as opposed to a worldwide) welfare-maximizing perspective, it is optimal for the United States, if a U.S. taxpayer is doing business abroad, to minimize foreign income tax just like any other cost and to maximize after-foreign-tax profit. To expend the correct amount of effort on foreign tax minimization — whether that be through tax planning, tax lobbying, or relocation of physical assets and activities to lower-tax jurisdictions — foreign taxes should be deducted instead of credited at the same rate of tax the United States is imposing on foreign-source income.
If you are in the majority of those who viscerally reject the idea of downsized FTCs, one reason you may be overreacting is that you are automatically associating the deduction-for-credit switch with a tax increase. In the context of international taxation, that’s the totally disparaged concept of national neutrality. Yes, if foreign profits are taxed at the U.S. statutory rate and deductions replaced credits, that would result in an impossibly large tax increase on U.S. multinationals. But as mentioned above, and as carefully explained by Shaviro, replacing credits with deduction of foreign taxes achieves optimal foreign tax cost-consciousness at any rate the U.S. imposes on foreign profits. That need not be the same as the statutory rate. It can be the effective 10.5 percent imposed by GILTI (which would result in a more horizontal line than any of those in Figure 1B). It can be a rate lower than 10.5 percent that results in revenue-neutral replacement of GILTI (which would result in a more horizontal line than the proposal in the prior sentence). It can even be a rate of zero — that is, a territorial tax that would be a horizontal line in Figure 1B and a large tax cut from current law.
So, Shaviro’s work gives us another good reason to be unafraid of significantly cutting the foreign tax creditability across the board in exchange for a lower rate of U.S. tax on foreign profits. In terms of our simple equation, where under GILTI the constant A is equal to 0.5 and B is equal to 0.8, Congress with the help of revenue estimators may want to explore values of A less than 0.5 and values of B less than 0.8. In a half-dozen years, the thinking on minimum taxation of foreign profits has progressed from what we call Category 1- to Category 2-type proposals. Perhaps we are ready to move to Category 3.
Further Reading
Kimberly A. Clausing and Daniel N. Shaviro, “A Burden-Neutral Shift From Foreign Tax Creditability to Deductibility,” Tax L. Rev. 431 (2011).
Doggett and Whitehouse press release, “Whitehouse, Doggett Author Bills to End Trump Tax Breaks for Exporting Jobs & Profits” (Mar. 13, 2019).
Michael J. Graetz, “Taxing International Income: Inadequate Principles, Outdated Concepts, and Unsatisfactory Policies,” Tax L. Rev. 261 (2000).
James R. Hines, “The Case Against Deferral: A Deferential Reconsideration,” Nat’l Tax J. 385 (Sept. 1999).
Thomas Horst, “A Note on the Optimal Taxation of International Investment Income,” Q. J. of Econ. 793 (1980).
Shaviro, “The Case Against Foreign Tax Credits,” J. of Legal Analysis 65 (2011).
Shaviro, “Rethinking Foreign Tax Creditability,” Nat’l Tax J. 709 (2010).
Shaviro, Fixing International Taxation (2014).
David A. Weisbach, “The Use of Neutralities in International Tax Policy,” Coase-Sandor Institute for Law and Economics Working Paper No. 697 (2014).
White House and Treasury, “The President’s Framework for Business Tax Reform” (Feb. 22, 2012).
White House and Treasury, “The President’s Framework for Business Tax Reform: An Update” (Apr. 2016).