The OECD is committed to finding reasonable solutions for addressing digital taxation this year. As previously reported, the organization has been considering a three-pronged approach involving a proposal for a minimum tax, favored by the Germans and the French; a proposal for a shift to destination-based source rules, favored by the United States; and a proposal for allocating taxing rights based on user location, favored by the United Kingdom.
The United States obviously supports minimum tax proposals, given that it adopted a variation — new section 951A, the global intangible low-taxed income regime — as part of the Tax Cuts and Jobs Act. With France, Germany, and the United States all seeming open to a minimum tax, it’s likely that any comprehensive solution will include that kind of proposal.
Other countries are showing interest in a minimum tax outside the digital tax work. Speaking last month at the annual Institute on Current Issues in International Taxation, cosponsored by George Washington University Law School and the IRS, Brian Jenn, Treasury deputy international tax counsel, noted the attention other countries are giving TCJA reforms, especially GILTI.
Given the heightened interest, it’s worth exploring what an OECD-proposed version of a GILTI-type minimum tax might look like. A primary imperative for any widely adopted minimum tax is that it must be far simpler than the U.S. version: Officials working on the OECD version will need to consider the compliance challenges GILTI presents — together with its complexities and flaws — and refrain from imposing them on its members.
Before trying to design a less complex version of GILTI, it’s helpful to understand just why the U.S. system is so complex. The answer lies in both the pre-TCJA U.S. international tax regime and specific congressional policy choices made in drafting the TCJA.
GILTI + Subpart F
One reason the new U.S. international tax regime is so complicated is that rather than repealing any of the prior system’s complexities, Congress simply layered GILTI on top of the controlled foreign corporation rules. However, GILTI is just different enough from subpart F so as to require that taxpayers perform two very distinct sets of calculations in computing their U.S. tax on foreign subsidiaries’ earnings.
Among the more important differences between GILTI and CFC rules is that subpart F income is calculated solely on a CFC-by-CFC basis — and separately for each U.S. shareholder of a CFC — while GILTI requires that the calculation be performed on a net basis at the U.S. shareholder level, with amounts included from each CFC then reallocated back to the CFC for tracking previously taxed income (PTI). As a result, U.S. multinationals must perform two computations each tax year, a requirement traceable to numerous policy choices made by TCJA drafters.
Netting Gains and Losses
The U.S. tax system generally allows the netting of gains and losses in computing a taxpayer’s tax liability for the year. The subpart F regime doesn’t allow for that, at least among the profits and losses generated by different CFCs. Thus, if one CFC has subpart F income and net positive earnings and profits in a given tax year, its U.S. shareholder will have an inclusion of taxable income, regardless of whether all the other CFCs in the group have E&P deficits. Any subpart F deficit in a loss-making CFC can possibly be used in later years by the same CFC (or by other CFCs in limited circumstances).
Congress abandoned that approach for GILTI, instead giving taxpayers greater flexibility in using deficits generated by loss-making CFCs to offset the net GILTI inclusion from profitable CFCs with positive tested income. It’s the allowance for that offset that mandates the shift away from the simpler, separate-entity calculation under subpart F toward one that requires that calculations be performed at the U.S. shareholder level: To offset a profit-making entity against a loss-making entity, one must necessarily perform a type of consolidated calculation at the shareholder level.
The decision to calculate the net deemed tangible income return (the GILTI-exempt portion of a CFC’s foreign earnings) at the U.S. shareholder level also creates complexity. (The TCJA didn’t invent that type of exemption: Former House Ways and Means Committee Chair Dave Camp included a version in his 2014 tax bill.) Calculating net deemed tangible income return at the shareholder level necessarily follows from the decision to allow for blended foreign tax credits and the netting of income of profit- and loss-making CFCs. If qualified business asset investment and net deemed tangible income return remained CFC-level attributes, the benefits of a blended FTC and the QBAI exemption would be much less beneficial for taxpayers (QBAI mostly being associated with high-tax jurisdictions).
The GILTI Deduction
The TCJA drafters also chose to tax CFCs’ foreign earnings at a lower rate than their domestic earnings, and to provide that lower rate via a deduction at the shareholder level (new section 250 allows taxpayers a 50 percent deduction on GILTI inclusions). While one can debate the wisdom of that policy — the corporate community and many economists were adamant that taxing foreign earnings at the same rate as U.S. earnings would encourage U.S. businesses to reincorporate overseas, and Democrats have generally favored higher rates on foreign earnings, if not complete parity with the rate on domestic earnings — the mechanism for achieving the lower rate is surely more complicated than need be. For example, an easier way to achieve a reduced rate would be to assess a differential (or top-off tax) on any foreign earnings taxed below a specific rate. In the April 2016 update to its framework for business tax reform, the Obama administration proposed subjecting foreign earnings to current U.S. tax at a 19 percent rate less an FTC equal to 85 percent of the per-country, average foreign effective tax rate.
The shareholder-level deduction for GILTI inclusions creates substantial confusion because of the way it interacts with other provisions, such as the new limitation on interest expense (section 163(j)) and the calculations and deductibility of net operating losses.
All the above calculations would be time-consuming and challenging enough for taxpayers to comply with even if performed only annually. But despite theoretically moving the United States to a territorial regime — under which distributions of dividends from foreign companies shouldn’t be subject to tax in the hands of their U.S. shareholders — the TCJA retains the concept of PTI (or previously taxed E&P). As a result, the difficulties inherent in continuously maintaining and tracking PTI accounts haven’t been eliminated, but instead multiplied exponentially. Now taxpayers must keep track of not only subpart F PTI (a small part of their foreign E&P), but also GILTI PTI, which will have been taxed at a different rate. And section 956 PTI, while mostly irrelevant, still remains a theoretical possibility and thus must also be separately tracked.
Tracking PTI remains necessary in the new regime — even with a participation exemption for foreign dividends — partly because Congress decided that time of repatriation is an appropriate moment to recognize gain or loss on foreign currency fluctuations that occur between when amounts are included in income and when they’re repatriated. It’s also the result of other legislative policy choices, including retaining capital gains taxation on sales of foreign entities and maintaining separate FTC baskets.
FTCs: 80 Percent and 100 Percent
To minimize incentives for other countries to raise their foreign tax rates to exactly offset the FTC against GILTI inclusions, the TCJA limits the FTC on GILTI to 80 percent of foreign taxes paid (the Obama proposals contained a similar limitation). That less-than-100-percent FTC increases the problems with the system — especially when considering that taxes associated with subpart F inclusions remain 100 percent creditable, thus requiring a careful tracking of taxes associated with different items of income.
Allocation of Deductions
Much discussion of GILTI has involved the proper allocation of shareholder-level deductions (see, for example, a report from the New York State Bar Association). Allocating shareholder-level deductions to GILTI often increases the combined U.S.-foreign tax rate on foreign non-subpart-F earnings to over 13.125 percent, even when the foreign tax rate is less than that.
Allocating U.S. shareholder-level expenses to foreign assets and earnings has been a feature of the U.S. tax system for decades. But it mostly made a difference in a taxpayer’s ability to claim FTCs, which are allowed only to offset U.S. taxes on foreign-source income. Foreign taxes attributable to GILTI can’t be carried forward or back, which exacerbates the drawbacks of excess allocation of deductions to GILTI. Practitioners had hoped Treasury would take a more relaxed approach to allocating shareholder-level deductions, but in proposed regulations (REG-105600-18) released in December, the government largely adhered to the old method.
The United States has long used separate FTC baskets to prevent cross-crediting of taxes earned on highly mobile passive income against real operating income. At various times, the FTCs operated on a per-country basis. After the Tax Reform Act of 1986, there were nine FTC baskets, but by the time the TCJA was passed, there were only two — the passive basket and the general limitation basket. The TCJA increased the FTC baskets to four, adding a GILTI basket and a foreign branch basket. Leaving the other two baskets in place, however, created confusion and odd results — such as when foreign branch income might also be passive income.
Having to allocate U.S. shareholder-level expenses to earnings of foreign subsidiaries was always complex, and the costs of doing it incorrectly are now even higher.
Did GILTI Get Anything Right?
The drafters of the TCJA didn’t set out to make GILTI overly complicated; to the contrary, they envisioned that in many respects, it would simplify the CFC rules. For example, in some respects, GILTI removed the necessity of analyzing and differentiating types of CFC earnings by essentially including all foreign earnings in its scope.
Being universal, the regime ignores the foreign country’s tax rate as a threshold matter in determining whether it applies. It also relieves the taxpayer — and the government — from having to perform any type of subjective analysis regarding the taxpayer’s intent in generating earnings in a particular jurisdiction. It adopts a bright-line approach that is quantitatively, rather than qualitatively, driven, which theoretically should make it easier to apply on a broader scale.
A Simplified OECD Version?
Understanding the policy choices behind the provisions that create GILTI’s complexities could make it easier for the OECD and governments to adopt a minimum tax that prevents the types of problems occurring in the United States. There are several options for designing a less complicated regime.
Eliminate the QBAI
Because much of the complexity of the GILTI calculation results from the exemption for a return on QBAI and calculating that exemption at the shareholder level, eliminating the exemption could simplify the GILTI regime.
Plus, it’s unclear whether the exemption produces much benefit for the government or taxpayers. Democrats have already begun pushing for its elimination, claiming that it encourages taxpayers to move assets overseas. (Prior analysis: Tax Notes Int’l, Dec. 10, 2018, p. 1049.) And as some economists have shown, the return on tangible assets probably represents only a very small portion of CFC income. (Prior analysis: Tax Notes, May 7, 2018, p. 773.)
Mandating an inclusion from profit-making CFCs while disallowing current losses from unprofitable subsidiaries makes calculating a minimum tax on foreign profits much easier. But it also undermines a fundamental bedrock of any tax system — allowing loss-making activities to offset profitable ones. Allowing netting would move the international tax system closer to one that treats multinational enterprises as a single entity.
While U.S. taxpayers could in theory plan around a system that disallows netting of profits and losses of different CFCs through check-the-box elections, prohibiting the netting of profit- and loss-making entities is potentially much more onerous for non-U.S. multinationals that can’t avail themselves of similar rules.
No Separate Rates
Taxing GILTI at a lower rate than U.S. and subpart F earnings creates many challenges — but it’s not a requirement for a minimum tax regime. U.S. lawmakers, however, considered a lower rate necessary: If those earnings were taxed at the statutory U.S. corporate rate, and only the United States enacted a minimum tax, U.S. companies would have felt pressure to move overseas.
But that kind of pressure would be substantially reduced if the OECD’s inclusive framework commits to adopting a comprehensive proposal for a minimum tax. If all countries agree to tax foreign and domestic earnings the same, the pressure to reincorporate overseas is lessened — unless the U.S. corporate rate remains much higher than those of other countries.
Here again, Democrats could lead the way, having offered proposals to tax foreign earnings at the same rate as domestic earnings (H.R. 5108, for example). If that’s the direction the United States is headed, U.S. multinationals might have reason to support an OECD proposal that includes rate parity.
A system without the U.S. requirement to track PTI of foreign subsidiaries would be much simpler. In the U.S. regime, PTI is needed (and produces taxpayer-favorable results) because it allows for recognition of gain (or loss) on currency fluctuations associated with PTI inclusions. It also allows foreign taxes associated with distributions of PTI to be creditable — but maintaining PTI accounts is not necessarily the only way to give taxpayers that benefit. Further, the imperative for capturing gain or loss associated with currency movements is unclear, especially from the fisc’s perspective.
No Allocation of Deductions
While most jurisdictions aren’t foolish enough to try to allocate a portion of shareholder-level expenses to foreign income, the United States long ago decided that such a regime was needed to prevent taxpayers from incurring expenses there and then using them to generate income that would be taxed at a lower rate overseas. But if all foreign earnings are currently taxed at a rate that approaches that on domestic earnings, the need to allocate expenses between foreign and domestic earnings is greatly reduced. And most jurisdictions address concerns about domestic deductions being used to create foreign income by simply not granting a 100 percent participation exemption.
Not many countries use separate FTC baskets, so omitting a basketing concept in a minimum tax regime shouldn’t pose a large hardship. Allowing an exemption for foreign income earned by branches, which several countries do, would also alleviate the need for baskets. Even without baskets, however, a decision must be made regarding how to address highly mobile passive income easily shifted to jurisdictions with lower rates.
The Joint Committee on Taxation’s explanation of the TCJA and a Republican-sponsored Ways and Means discussion draft of a bill proposing technical corrections to the TCJA both recognize the act’s numerous errors, particularly in the GILTI regime. But even if all the proposed technical corrections were made, the U.S. version of a minimum tax would still be enormously complex — and likely unworkable for OECD members, let alone the broader group of countries that make up the inclusive framework.
If a minimum tax is to be part of OECD proposals to address the taxation of the digital economy, it must be a scaled-down and less complicated version of the GILTI regime — and designing that simplified system will itself be a challenge.
Mindy Herzfeld is professor of tax practice at University of Florida Levin College of Law, of counsel at Ivins, Phillips & Barker Chtd., and a contributor to Tax Notes International.
Follow Mindy Herzfeld (@InternationlTax) on Twitter.