Next year is the 85th anniversary of the adoption of the arm’s-length standard (ALS) by the U.S. government as the norm to be used for pricing transactions among related parties for the purpose of calculating their U.S. corporate income tax (CIT).
The 1935 U.S. Treasury Regulations in tax code Section 45-1(b) (later renumbered as Section 482) defined the ALS as: “The purpose of section 45 to place a controlled taxpayer on a tax parity with an uncontrolled taxpayer, by determining, according to the standard of an uncontrolled taxpayer, the true net income from the property and business of a controlled taxpayer.” In other words, the purpose was to determine what the taxpayer’s net income from its property or business would have been if all of its transactions had been conducted with unrelated parties rather than with other affiliates in the multinational enterprise (MNE). The U.S. Congress feared that MNEs would engage in income shifting to lower tax jurisdictions outside of the U.S. through mispricing of related party transactions. The ALS was designed to prevent this by ensuring that transfer prices would clearly reflect MNE income and deter tax evasion.
Over the past 84 years, our understanding of the ALS has been clarified and codified through multiple U.S. Treasury Regulations and U.S. Tax Court decisions. Transfer pricing professionals now view the ALS as requiring related party transactions to be priced as if they had been negotiated between unrelated (arm’s-length) parties engaged in the same or similar transactions under the same or similar facts and circumstances. In other words, “What would independent enterprises have done?” Answering this hypothetical question therefore requires a detailed comparability analysis and adjustments to ensure that sufficiently similar transactions took place under sufficiently similar facts and circumstances.
The simple sentence adding the ALS to U.S. tax law in 1935 has now spread and diffused around the world. Canada was an early adopter; Section 23B was added to the Income Tax Act in 1939 and the first Information Circular was published in 1987. Mexico followed later. The ALS was added to the Mexican Income Tax Law in 1976; enforcement was added in Article 64A in 1992.
Worldwide diffusion of the ALS has been primarily through the efforts of the Organization for Economic Cooperation and Development (OECD), which included the ALS in Article 9 (associated enterprises) of the 1964 draft (finalized in 1977) Model Tax Convention and issued its first Transfer Pricing Guidelines in 1979. The United Nations included the ALS in its 1977 draft Code of Conduct on Transnational Corporations and its 1978 Model Tax Convention. With the active encouragement of the OECD and UN, the number of countries with detailed transfer pricing laws, regulations, and rulings based on the ALS has tripled since 2003 and doubled since 2010, from 40 countries in 2003 to 59 countries in 2010 to 124 countries in 2019 (EY 2003: 3, 2010: 2, 2019a: 2).
With an 85th anniversary just around the corner, one might be expecting transfer pricing professionals to be getting ready to celebrate. However, that is not the case. The ALS now appears to be in retreat. Many transfer pricing experts argue the ALS is “dead.” Shepherd (2012: 467), for example argued that, “The transfer pricing guidelines are a sorry vestige of a system that will be gone in 10 years.”
Criticisms of the ALS have been longstanding, going back at least to 1986 (Bird and Brean, 1986; Langbein, 1986; 1992). The criticisms fall into two broad categories. First there are concerns, most prominently argued by non-governmental organizations such as the Tax Justice Network and Christian Aid, that MNEs have been deliberately engaging in abusive transfer pricing that is extensive, unfair and draining development. The second criticism is that the transfer pricing rules are too difficult to implement for various reasons, of which the two most important reasons are that arm’s-length comparables often do not exist (e.g., hard-to-value intangibles) and that MNEs benefit from synergies not available to unrelated parties.
Elsewhere, in Eden (2016), I have addressed the second criticism: the ALS is unworkable due to lack of comparables and MNE synergy benefits. I assessed the criticism from both theoretical and practical bases, concluded that the transfer pricing rules can be salvaged, and made several recommendations for fine-tuning the ALS. Interested readers are directed to that article. In this piece, I want to address the first criticism: the ALS should be abandoned because MNEs are misusing the ALS to engage in extensive income shifting. My goal is to show that the ALS is not the problem. Moreover, I believe that many of the OECD’s BEPS (Base Erosion and Profit Shifting) reforms go a long way to remedying the worst income-shifting excesses of the past 20 years. On the eve of the 85th anniversary of the ALS, there is much to celebrate—although more remains to be done.
Quick Review of International Income Tax Principles
The current international tax system as set out in the OECD and UN Model Income Tax Conventions and bilateral income tax treaties is based on the residence and source principles whereby different types of income are allocated to either to the residence country (where the owner resides) or the source country (where the income is earned). Typically the source (or host) country is given first crack at the corporate income tax base and the country of residence—assuming it taxes its MNEs’ foreign source income—provides tax room for the source country through a foreign tax credit or deduction of foreign income and withholding taxes up to the level of the residence country tax rate. Foreign subsidiaries are treated as separate entities for tax purposes, that is, as standalone legal entities in the host country using separate accounting. Foreign branches may also be treated as de facto separate entities if they meet the permanent establishment test.
Because foreign subsidiaries and permanent establishments (branches) are treated as separate from their parent firm, their profits are determined on a country-by-country basis. Transactions among the MNE parent and its foreign affiliates can therefore affect where the MNE’s group profits are declared and taxes paid. Over-invoicing (under-invoicing) of exports shifts profits into (out of) the exporting affiliate; the reverse is true for the importing affiliate. The ALS is therefore necessary as a backstop to the international tax system. The ALS prevents MNEs from engaging in tax avoidance and evasion strategies designed to shift income to lower taxed activities and jurisdictions by requiring that transfer prices clearly reflect the income earned by the separate entities within the MNE.
The Tax Design Problem
I believe the first critique made by professionals and academics who would like to scrap the ALS is misplaced. The excessive income shifting among tax jurisdictions that we have witnessed over the past 20 years is not a transfer pricing problem but rather an income tax design problem. I believe the criticism is a good example of “shooting the messenger” rather than facing up to and addressing the underlying problem. Abusive transfer pricing is caused by perverse incentives—set in place by governments—that encourage MNEs to manipulate transfer prices to take advantage of differences in tax rates across jurisdictions. This is not a transfer pricing problem but an international tax regime “design” problem. It is best handled by fixing the gaps in the international tax system rules. If governments, NGOs, and the general public do not like the way that multinational firms are allocating their taxable income among countries and the amounts of tax (or lack of tax) they are paying, the problem should be laid at the feet of governments, not the MNEs. The prescription should be: Physician heal thyself!
Arguing that the OECD, UN, and national tax authorities should scrap the ALS is misplaced and misguided. The international tax system has gaping holes that provide many legal opportunities for MNEs to engage in regulatory arbitrage. Moreover, the rules have become ever more complicated and complex, both in terms of the length and variety of regulations and the number of countries with these rules, making it ever more difficult for MNEs to keep up with national regulations.
The best solution would be to re-establish the principles of international equity and efficiency that underpin the international tax system by tightening and eliminating the tax loopholes that create the incentives for transfer mispricing. While there will always be firms that will push the envelope in terms of tax aggressiveness—moving across the “bright line” from tax avoidance into tax abuse and possibly tax evasion—the majority of MNEs pay their taxes and follow the rules that have been laid down for them by national governments.
To end abusive transfer pricing, the first step must be to reduce the incentives that governments have put in place to encourage and enable MNEs to engage in these income shifting activities. Here I discuss three possible alternatives. The first is a simpler international tax system—one based on the classical system of residence taxation of foreign source income (FSI) with foreign tax credits for source-based taxation on an accrual basis. If large OECD home countries were to adopt such a system, it would eliminate most of the incentives for abusive behavior that riddle the current international tax regime. The second alternative—closing the tax loopholes—is well underway with the OECD’s BEPS project and the new Multilateral Instrument. The third alternative—replacing the current system with source-based taxation using global formulary apportionment (GFA)—would, I believe, be a disastrous mistake.
Alternative 1: Return to First Principles—International Tax Equity and Neutrality
Alternative 1 is probably the least likely to happen, particularly since the 2017 U.S. Tax Cuts and Jobs Act (TCJA) has moved in exactly the opposite direction. Still, it is useful to explore this option.
To reduce the incentives for MNEs to engage in abusive transfer pricing, residence countries should tax foreign source income (FSI) on a worldwide basis as earned, with no deferral provided for income kept offshore. Common residency definitions should be adopted so that MNEs cannot exploit differences in definition across countries so as to become stateless and tax free (for example, as Ireland did for many years with its two definitions of MNE residency).
I am not alone in this view. For example, Avi-Yonah (2016) argues that OECD governments should return to the traditional way of taxing MNE profits: home country taxation of the MNE’s worldwide income on an accrual basis coupled with foreign tax credits for foreign source income taxes. He recommends that “Each OECD member country impose tax at its normal corporate rate on the entire profit of multinationals that are managed and controlled from headquarters within it, with a foreign tax credit for taxes imposed by other jurisdictions “ (Avi-Yonah 2016: 114).
Developing countries would benefit from this policy change because it would enable their governments to re-establish corporate income and withholding taxes, creating a revenue stream that could be used for social and other domestic purposes. Developing countries would not have to worry about the potential loss of inward foreign direct investment (FDI) due to competition from tax havens and other lower taxed jurisdictions. In effect, the OECD countries would provide an “umbrella” level of corporate income taxation under which developing countries could “shelter” their own corporate income taxes, protected from evisceration by mutually destructive rate cutting. The “first crack” principle would enable developing countries to raise their taxes on inward FDI income, knowing that these taxes were creditable up to the level of the home country tax.
Figure 1 illustrates the impact of the umbrella effect and first crack principle. Assume country A (the home or residence country) has a CIT rate of 30%. Multinationals resident in A have foreign affiliates in countries B, C, and D where CIT rates range from 10% to 25%. Assume country A taxes the worldwide income of its MNEs on an accrual basis (no deferral), with a foreign tax credit up to the level of A’s CIT rate. In this situation, A’s MNEs pay a 30% worldwide CIT rate regardless of where their foreign affiliates are located. As a result, countries B, C, and D are free to set their own CIT and withholding tax rates up to the creditable level in country A without fear of a negative impact on inward FDI from country A. In effect, country A provides a “tax umbrella” or ceiling of 30% for its host country partners, under which these governments can shelter their own rates.
Figure 1: How to End Abusive Transfer Pricing—First Crack and the Umbrella Effect
For many years the U.S. has one of the world’s highest statutory corporate income tax (CIT) rates at 35%. The U.S. taxed worldwide income of its MNEs; however, the U.S. also practiced tax deferral whereby FSI was not taxed until it was repatriated. As a result, most U.S. MNEs kept their FSI offshore to avoid the additional U.S. tax that would have been due on repatriation.
Still, the situation pictured in Figure 1 did apply to offshore income earned by controlled foreign corporations (CFCs) if that income qualified as passive income under Subpart F of the U.S. corporate income tax. That income was taxed at the U.S. rate, with a foreign tax credit for the host income tax. The umbrella effect of a high U.S. CIT together with the Subpart F provision discouraged MNEs from shifting passive income to low-tax jurisdictions. This situation remained in place until the “check the box” provision permitted MNEs to avoid Subpart F through creating hybrid structures.
The December 2017 TCJA lowered the U.S. statutory federal corporate income tax (CIT) rate from 35% to 21%, effective in 2018. The TCJA also eliminated taxation of repatriated dividends; in effect, moving the U.S. from a worldwide to a territorial system, exempting FSI (whether repatriated or kept offshore) from U.S. taxation. Thus, not only did the U.S. reduce its CIT rate, shifting the umbrella downwards; in addition, the U.S. exempted all FSI from taxation, in effect, removing the umbrella.
We can see this in Figure 1. Assume country A does not tax FSI (i.e., country A follows a territorial system of exempting FSI from taxation). Then the three host countries B, C, and D, which are competing for inward FDI from country A, will offer A’s multinationals very different tax rates. While CIT rates are typically not the most important factor affecting FDI location, they do matter. Tax competition could precipitate a CIT war, leading to a “race to the bottom.”
As Figure 1 suggests, differences in host country CIT and withholding rates should now have more impact on FDI location; they could also encourage abusive transfer pricing between the host countries. Heinemann et al. (2018) supports these conclusions. The authors find that the TCJA lowered the effective tax burden on U.S. inward FDI from the EU as well as on U.S. outward FDI to the EU, with the benefits of U.S. tax reform being highest for low-tax EU members (e.g., Ireland) and the costs highest for high-tax EU members (e.g. Germany). One likely impact is that high-tax jurisdictions will be forced to follow the U.S. lead, lowering their CIT and withholding taxes in order to retain inward FDI.
As Gravelle and Marples (2019) argue, however, lowering the CIT rate and exempting FSI from U.S. taxation were not the only international tax changes introduced in the TCJA. The U.S. also added a worldwide tax on global intangible low-taxed income (GILTI), a tax incentive program for U.S. intangibles (FDII), and a base erosion and anti-abuse tax (BEAT) to curb profit shifting out of the U.S.. GILTI effectively imposes a global minimum tax on foreign affiliates (branches and subsidiaries) that have U.S. parents. FDII encourages U.S. exports of intangibles and services by taxing their profits at a lower rate. BEAT is an alternative minimum CIT that can disallow U.S. tax deduction for “above normal” payments to foreign related parties. The legislation and (coming) regulations implementing TCJA are highly complex and their long-term impact is unclear. Gravelle and Marples (2019) suggest that the combination of GILTI, FDII, and BEAT should work to discourage profit shifting, possibly offsetting the impact of lowering the CIT rate and moving to a territorial system. What is clear, however, is that alternative 1—moving to worldwide taxation on an accrual basis—may now be a non-starter in the U.S. after the 2017 TCJA.
Alternative 2: Closing the Most Egregious Loopholes (BEPS)
A second alternative is to close up some of the most egregious of the income tax loopholes. The OECD’s Base Erosion and Profit Shifting (BEPS) initiative is built on this approach with 15 action items designed to reduce the incentives for MNEs to engage in income shifting.
A particularly useful policy invention is the new Multilateral Instrument (MLI), which is designed to update thousands of bilateral tax treaties to include the BEPS Action Items as they are adopted at the country level (Oguttu 2018). Another useful—and practical—innovation is the new Tax Inspectors Without Borders initiative, which has already significantly increased tax revenues for developing countries (OECD/UNDP 2019).
Stronger anti-avoidance rules for activities with no business purpose would also provide a backstop against the most egregious activities. Some examples of proposals that I support are summarized in Box 1 below. I am not alone in making these policy recommendations, and many of them are included in the BEPS Action Items and now being introduced into the MLI. Many practitioners and academics have also supported these proposals. For example, Durst (2016) argues for strengthening controlled foreign corporation (CFC) rules that prevent income shifting by MNEs, such as the U.S. Subpart F rules. He provides a nice history of CFC rules around the world, and how they have been eroded through legislation and MNE ingenuity. Durst argues for a “strengthened international net of CFC rules” (which I also support), for “self-protective measures” by developing countries such as limitation of deductions and reinstatement of withholding taxes (which I also support), and for a move from the ALS to formulary apportionment (which I do not support). Mehta and Siu (2016: 339) also provide a variety of suggestions for developing countries that would help ensure they receive a “fair deal on tax justice”.
Box 1: Solving the Tax Design Problem—Eliminate Income Shifting Opportunities
Country-by-country reporting (CbCR) is already providing additional information would may be helpful for governments; the greater visibility should also discourage base shifting activities by MNEs. MNEs must provide much better information about their activities, both by country and by line of business, in their public reports. Greater transparency in MNE operations on a worldwide basis would also go a long way to reducing opportunities for income shifting. One concern, however, is that CbCR may be misused to “back door” into a formulary apportionment approach to allocating the MNE’s tax base among countries, as it appears to be doing in the current BEPS 2 proposals for taxing the digital economy (Eden, Srinivasan, and Lalapet, 2019).
Alternative 3: Global Formulary Apportionment Is Not the Answer
Most of the critics of the ALS have recommended that it should be replaced by global formulary apportionment (GFA) where the worldwide profits of the MNE are allocated among its home and host countries using a formula based on some combination of capital, labor, and sales.
I am not in favor of GFA for allocating the MNE’s global profits for tax purposes among countries on a worldwide basis for a variety of reasons that I spell out in Eden (1998: 561-583). Twenty years later, I still believe these are key impediments to implementing GFA on a global—or even on a regional—basis. One of the key reasons is the simple adage that “an old tax is a good tax.” The ALS has been the global norm for multiple decades. Achieving international consensus—or even regional consensus—around another other international tax standard seems highly unlikely. A mixed system with both the ALS and GFA in place would impose huge compliance costs on MNEs and significantly raise the risk of double taxation.
Second, GFA is simply too blunt a tool to achieve the ultimate purpose for which the ALS was designed, that is, to clearly reflect the true taxable income of the MNE in each of the countries where the MNE has activities, and to prevent (or at least reduce) the incentives for income shifting among countries. In addition, the three-factor formula most likely to be used (sales, labor, and capital (tangible assets)) does not include the “sina qua non” of the multinational enterprise—its intangible assets, which are the core source of MNE long run competitive advantage (Eden, Levitas, and Martinez 1997).
A third reason is both political and practical. The problematic history of the multi-jurisdictional compact among the U.S. state governments with sharing of U.S. state-level CITs provides a sobering example that should raise a cautionary red flag to even the most enthusiastic proponent of GFA. Altshuler and Grubert (2010) provide us with good examples of how shifting from assets (origin based) to sales (destination based) factors can affect the location of MNE income. Tangible assets (i.e., property, plant, equipment and inventories) are a “minority of total assets even in nonfinancial businesses” (Altshuler and Grubert, 2010: 1179)—and that percentage is expected to fall in the digital economy. It is perhaps not surprising that most of the U.S. states have switched from assets to sales as the most important factor for allocating state CIT receipts. The situation is likely to be much worse when national governments—not sub-federal ones—can use GFA to play tax competition games. Moreover, many of the income gaming techniques currently used by MNEs could also be used to shift income among jurisdictions under GFA.
There may be certain areas where the related-party transactions are so interwoven that they cannot be disentangled, much less priced such as 24-hour global trading. In these situations, GFA may be least inappropriate way to determine true taxable income. A partial introduction of formulas, perhaps as safe harbors or as guaranteed taxable income floors for the least developed countries, may also be needed. See, for example, the proposals in Avi-Yonah and Benshalom (2011).
However, I do not believe GFA is the best way forward for the international tax regime, even in a world dominated by digital MNEs. The OECD/G20 (2019) is currently considering fractional apportionment as one of three possible methods for attributing income to countries under the Pillar 1 proposal (reallocation of taxing rights involving profit allocation and nexus rules) for taxing the digital economy. I believe the problems outlined above would also apply if GFA were used for Pillar 1 and am therefore not in favor of this proposal.
The historical approach to taxing intrafirm transactions of multinational enterprises—the ALS—has been criticized as unworkable, out-of-date and on death’s door. Many academics and policy makers now advocate getting rid of the ALS and shifting to GFA as an alternative. Even the OECD, long a supporter of the ALS and opponent of GFA, now includes fractional apportionment as a possible method for attributing income among countries under its Pillar 1 proposals for taxing the digital economy.
My views are different. My preferred policy response to the income shifting problem is two-fold. First, I believe that many of the criticisms of the ALS in terms of abusive transfer pricing are misdirected; the criticisms should be more appropriately aimed at weak international tax rules that need to be fixed. The OECD’s BEPS project goes a long way to correcting many of the inconsistencies and loopholes in the international tax system. As the BEPS Action Items are adopted at the country level and diffused across countries through the Multilateral Instrument, many of the most egregious incentives for income shifting should disappear.
Second, assuming the loopholes in the international tax regime can be fixed, I believe that the ALS remains the appropriate international norm for taxing MNEs. New thinking, particularly for the digital business models that are now starting to dominate international production, is probably needed (Srinivasan, Lalapet, and Eden 2019). It is even possible that new technologies such as blockchain could make it easier for MNEs to comply with the ALS (Sim, et al. 2019). Fine-tuning the ALS for the 21st century is the appropriate solution.
My conclusion: The arm’s-length standard is not the problem. Don’t shoot the messenger!
This column does not necessarily reflect the opinion of The Bureau of National Affairs, Inc. or its owners.
Lorraine Eden is Professor of Management and Emerita Professor of Business at Texas A&M University. Please address any correspondence regarding this article to firstname.lastname@example.org.