Hallmarks of tax treaty policy
Potential shape of reform

Direction of BEPS project
Implications for US treaty policy

Some candidates for rethinking


US tax treaty policy can be affected by pressures for tax reform from within the United States and by major developments in international taxation from without. Talk of US tax reform has been widespread for years, although it is sometimes hard to gauge how much of the talk is serious. Similarly, the Organisation for Economic Cooperation and Development (OECD) base erosion and profit shifting (BEPS) project represents an unprecedented worldwide challenge to received wisdom regarding the taxation of cross-border investment, although it is difficult to see how the ambitious goals of this project can be reconciled with the interests of individual countries.

If there is any eventual reality in either of these movements, they are likely to cast a shadow over the policies that have long underlain US tax treaties and tax treaty policy. Neither US tax reform as it is likely to progress nor BEPS as it may unfold is accommodating to that policy. The question on the table is how that policy might change.

Hallmarks of tax treaty policy

The US Model Income Tax Convention dates from November 15 2006. The fifth in a series of model conventions first issued in 1976, it exhibits three features that sharply distinguish it not only from the OECD Model Tax Convention on Income and on Capital, but also from all other model conventions produced by international organisations or other countries. The US model is the document in which the United States stakes out a position for negotiations with prospective treaty partners. Therefore, its distinguishing features can fairly be characterised as the pillars of US tax treaty policy.(1)

One unique feature of US tax treaty policy – which is unlikely to be affected by either tax reform or BEPS – is the limited grant of treaty benefits to US persons. For more than 55 years the United States has been consistently reluctant to grant those benefits to its citizens and residents. Flowing directly from a hearing of the Senate Foreign Relations Committee in 1957, when Harvard Law School Professor Stanley Surrey spoke out forcefully against a 'tax sparing' provision in a proposed treaty with Pakistan, US policy has remained remarkably firm on this issue regardless of the political map or the persons inhabiting the US Treasury.(2) US treaties are not used to encourage foreign investment or other foreign activities by US persons.(3) However, the United States cannot withhold all treaty benefits from US persons because the basic 'deal' inherent in a tax treaty involves a reduction or elimination of tax in the country of source and a commitment by the country of residence to avoid double international taxation. This commitment is essential to any tax treaty. If it were not accorded to US persons, such treaties could not be concluded.

The mechanism devised to accommodate a policy of using treaties primarily to make concessions to foreign persons investing in the United States, and not to US persons investing abroad, while nevertheless abiding by commitments to residents that are indispensable to any treaty is the 'saving' clause. Found in Paragraphs 4 and 5 of Article 1 of the US model, the article dealing with the general scope of the treaty, the saving clause declares that the treaty has no application – does not exist – for US persons, including both US citizens and US residents (after application of the dual resident provisions of Article 4 of the model). Having thus removed US persons from any entitlement to treaty benefits, Article 1(5) carefully restores a limited number of benefits to all US persons and a second group of narrow benefits to US persons who are neither US citizens nor US permanent residents. Into the first group fall the articles on relief from double taxation, non-discrimination and the mutual agreement procedure, as well as the commitment in Article 9 to make correlative adjustments in transfer pricing cases(4) and rules for pensions, social security payments, child support(5) and the treatment of pension funds.(6) In the latter category are other rules for pension funds, the treatment of income from government service(7) and income of students, trainees(8) and diplomats.(9) The saving clause can sometimes be overlooked or its potency underestimated, but it appears in every US tax treaty and US treaty policy cannot be understood without it. Nothing in the tax reform debate or in the discussions on BEPS suggests a need for any change.

A second unique feature of the US model lies in its extraordinary tilt towards the country of residence. The most common species of international double taxation in cross-border operations is the taxation of income by both the country of source (where income is earned) and the country of residence (where the taxpayer resides). All countries are countries of both source and residence, but the relative degree to which they fall in those categories differs greatly from country to country – as do perceptions of that degree (which may be a different matter). The OECD model is relatively favourable to the country of residence, which is hardly surprising. The roots of the OECD are in Europe and until relatively recently its membership has been limited to capital-exporting nations. However, the US model goes much further than the OECD model in favouring the country of residence. This can be clearly seen in Article 11 on interest, in which the US model calls for a reciprocal maximum rate of 0% in the source country, while the OECD model proposes 10%. Interest probably represents the largest of cross-border income flows.

Article 11 is not the only evidence of the US approach. Article 6 deals with income from real property, which allows taxpayers to elect to be taxed at source on that income on a net, rather than gross, basis.(10) In Article 16, dealing with "entertainers and sportsmen", the US model provides that relatively low-paid taxpayers are excepted from the article's provisions calling for taxation of the first dollar of income earned in the source country.(11) In Article 20, concerning students and trainees, the US model provides for exemption at source of a modest amount of income from personal services.(12) Other directionally similar provisions could also be cited.(13)

The US model reflects the view of the United States that it is overwhelmingly a capital-exporting country and, in its treaty relations with other developed countries, it should approach the world seeking maximum, reciprocal, source-country concessions.(14)

Arguably, there are two reasons for this. As a capital exporter, the United States would stand to benefit economically from reductions in taxation in source countries. The tax burden on US resident taxpayers would be lower and thus more residence-based tax would be collected by the US government. Given expected cross-border income flows, with much larger amounts coming back into the United States than going out, the cost in terms of US source-based taxation forgone is well worth the trade.(15)

Another explanation – and a more altruistic one – is that a residence country is inevitably better positioned to address the threat of international double taxation than a source country, because the residence country sees a taxpayer's entire world while any given source country sees only the part of the taxpayer's operations that crosses its borders.(16) For some industries – such as shipping – deferring to a greater extent to the source country would invite not merely double but multiple taxation.

So for either or both of these reasons, the United States has staked out a position that is more favourable to the residence country than even the OECD model. This stands in marked contrast to the position of the United Nations, whose own model convention – developed with the active participation of many capital-importing countries – is much more oriented towards source-based taxation than the OECD model.(17)

A third unique feature of US tax treaty policy flows from the US view that treaties should aim for substantial source-country reductions of tax. The United States has long taken a dim view of treaty shopping, in which persons having no or only tenuous relations with a US treaty partner form an entity in that country for the purpose of taking advantage of the reduction in source taxation that all US treaties – especially those with developed countries – provide.(18) This policy against treaty shopping is a logical extension of the US favouritism of residence-based taxation, because the US effort to seek maximum reductions of tax at source is likely to produce disparate results in actual negotiations. Some countries (eg, the United Kingdom and the Netherlands) will be pleased to accede to the US proposal to reduce source-based taxation as much as possible. Their own views of the world coincide with those expressed in the US model and it will not be difficult to achieve the kinds of reciprocal reduction called for by that model. However, other countries are less enthusiastic about wholesale reductions of tax at source and, in negotiations with the United States, will not willingly go so far as the US model would. If the United States remains steadfast in seeking reciprocal reductions at source, it will wish to maintain a degree of pressure on these countries (eg, Canada, Japan and Italy), with the far-ranging goal that ultimately they can be persuaded to accept ever-greater reductions of tax at source.

It is obviously a hindrance to this goal if taxpayers from countries that balk at accepting reductions in source taxation can achieve reductions on their own by investing in the United States through entities formed in third countries whose treaties with the United States are closer to the US model. If an Italian or Canadian investor can obtain a better deal than is available under the treaty between the United States and Italy or Canada by forming an entity in the Netherlands and having that entity claim benefits under the Netherlands-US treaty, the investor and management will have little interest in the level of source-based taxation in the US treaty with their home countries.(19)

Hence the negative view of treaty shopping taken by the United States – or, more specifically, the US interest in maintaining the integrity of each bilateral treaty that it negotiates. The United States wishes to discourage third-country investors from taking advantage of its treaties, because that advantage would dilute the pressure to reduce source-based tax that the United States seeks to maintain on all countries, at least in the developed world.(20)

There have long been judicial doctrines serving to inhibit tax treaty abuse, but these have always been somewhat imprecise and unpredictable in their application to real-world transactions.(21) Thus, the concept of 'limitation on benefits' was born. The idea – by now a standard feature of every modern US income tax convention – is that within the body of the treaty, there should be a separate article requiring a purported claimant of US treaty benefits to establish, according to mostly objective criteria, that it is a genuine resident of the treaty partner for reasons other than merely seeking those benefits. The limitation on benefits article has changed over the years, but the original concept remains intact. If the United States succeeds in restricting treaty benefits with, for example, the Netherlands to persons with a real, non-tax nexus with the Netherlands, that may have some effect on persons from Canada or Italy who are deterred from using the Netherlands as a conduit for investment into the United States. Those persons, in turn, may approach their own governments with a request that they negotiate treaties with the United States providing directly (and reciprocally) for low taxes at source. Although the process takes years, this has been the story of Article 22 (the limitation on benefits article) of the US model.

Potential shape of reform

The call for tax reform has been so omnipresent in the United States for so many years that the meaning of reform has been thoroughly blurred. An actual attempt at a comprehensive change of the US tax laws produced by the House Ways and Means Committee(22) was rejected by the House of Representatives. There has been little actual progress or activity. It is not even clear whether the debate, such as it is, encompasses all income tax, only corporate taxation or even only corporate international taxation.

Wherever the debate goes, it is safe to predict that any changes will be to the bedrock. For example, although the details of a viable reform are far from settled, it is reasonable to assume that it would entail a reduction of the 35% maximum corporate tax rate, since the fact that 35% is the highest statutory rate in the world figures in most critiques of the existing US rules.

If there is an attempt to devise a reform that collects as much revenue as the present US tax system, preferences studded throughout the Internal Revenue Code would have to be reduced. There would be fewer special rates and benefits, and permissible reductions of the tax base would be limited. This is entirely appropriate, since the arguments for not taxing all income in the same manner and to the same extent are strained and theoretical.(23) Moreover, the fewer categories that are established in the law, the easier it will be to understand and administer that law and the more difficult it will be to play games with it.

In the best case, reform would place a premium on simplification. The Internal Revenue Code is a peculiar document.(24) It is nearly impossible for taxpayers to decipher, or for the Internal Revenue Service (IRS) to apply uniformly and consistently. The result is that the United States has developed two completely different tax laws – one embodied in two large volumes (and six accompanying books of regulations), the other found in the practices and policies of individual IRS offices around the country, which labour with greater or lesser consistency to collect the revenue.

Of course, simplicity by itself cannot be an engine of change.(25) Too many constituencies have a vested interest in complexity. To achieve change, a sharp pruning of the existing rules should be the priority. As a necessary complement, the rules for taxing foreign income will change. At least to some extent, the foreign credit will yield to a territorial regime.(26) US multinational companies have been lobbying for this change for years now – a reversal of their prior opposition to a change that might encourage calls for distributions to shareholders. Although a territorial system for the United States would require some safeguards, and these might not be so readily accepted by the international community, the existing foreign tax credit regime has become too cumbersome and expensive to retain.

Finally, there should also be a much greater focus on income earned in the United States. The will is not there to end deferral, to pierce the corporate veil for taxation purposes, to tax all US persons currently on their worldwide income or to run down all the schemes and strategies that the existing byzantine rules have engendered. In the modern world, with its communications possibilities, the limited resources accorded to tax administrations and armies of sophisticated advisers, a complete worldwide tax reach is hard to defend.

Moreover, the United States can ill afford to act as if it is still 1945. It exports considerable capital, but for decades it has also imported considerable capital. The US market is not going anywhere, regardless of how entities or instruments are characterised or the vanishing acts that have been devised to shunt income into black holes. The domestic rules on inbound taxation have remained substantially unchanged since the Foreign Investors Tax Act of 1966. As the United States begins the overdue process of revamping its tax system, it should think more, if not exclusively, about the home tax base and homegrown income; that focus should be maintained insistently and vigorously. For too long it has been common to equate international taxation with outbound taxation, giving near-exclusive attention to the concerns and issues of US multinational companies.

Taxation is moving away from a world in which the source country reduces or eliminates its tax to attract investment and the residence country assumes the responsibility for providing relief from double taxation. Arguably, these developments were inevitable with a shrinking world, a growing population and increased understanding (by everyone) of what is going on beyond the borders. Residence countries have abetted the movement by abdicating their role as tax imposers in favour of subtle and sophisticated enabling of tax avoidance. And there will always be jurisdictions seeking to turn a quick (and often transitory) profit through 'beggar thy neighbour' policies designed to reduce the tax income of other countries.(27)

Perhaps the United States should take a leaf from the books of India, Brazil and every source country that finds it expedient and politically advantageous to enhance its taxation of income earned within its markets. As recent developments in France and the United Kingdom have shown, it is not only developing countries that are thinking about greater taxation at source.(28) Would such a development, if adopted by the United States, have a dampening effect on US investments by foreign persons? At the margin, perhaps. However, there are reasons for doubt.

The lobbyists will make every effort to persuade Congress that serious taxation of income earned in the US market would have a deleterious effect. After all, the notion that an ounce of taxation will give rise to a ton of deterrence is one of the core theories that brought the Internal Revenue Code to its present unhappy state. Taxes definitely deter. However, the question is not the abstract disincentive effect of taxation, but the practical one of how much taxation it takes to deter prospective investors from coming to the United States.

The answer is almost surely a great deal more than at present. It is imperative to undertake a thorough review and revision of the rules regarding inbound investment, with the United States thinking of itself more as a country of source (as it is) and less as the country of near-exclusive residence taxation (as it was following World War Two).

Direction of BEPS project

The BEPS project, which began in 2013, consists of 15 actions:(29)

  • Address the tax challenges of the digital economy;
  • Neutralise the effects of hybrid mismatch arrangements;
  • Strengthen controlled foreign corporation rules;
  • Limit base erosion via interest deductions and other financial payments;
  • Counter harmful tax practices more effectively, taking into account transparency and substance;
  • Prevent treaty abuse;
  • Prevent the artificial avoidance of permanent establishment status;
  • Assure that transfer pricing outcomes are in line with:
    • value creation;
    • intangibles, risks and capital; and
    • other high-risk transactions;
  • Establish methodologies to collect and analyse data on BEPS and the actions to address it;
  • Require taxpayers to disclose their aggressive tax planning arrangements;
  • Re-examine transfer pricing documentation;
  • Make dispute resolution mechanisms more effective; and
  • Develop a multilateral instrument.

Although some of these actions are directed to the residence country (eg, the effort to alter the design of the rules regarding controlled foreign companies) and several others are neutral as between the country of residence and the country of source, many of the actions point clearly in the direction of greater source-basis taxation. Thus, the actions addressed to the digital economy, base erosion, neutralising the effects of hybrid mismatch arrangements, preventing the artificial avoidance of permanent establishment status and ensuring that transfer pricing outcomes are in line with value creation either explicitly or implicitly invite more robust taxation in the country where income is earned. Several other actions (eg, requiring taxpayers to disclose their aggressive tax planning arrangements) can easily be interpreted to produce that result.

The stated goal of BEPS is a coordinated effort to reduce corporate tax avoidance. However, the real-world effect is more likely to be seen in the efforts of individual countries to impose a greater tax burden on inbound investment.(30) Coordination of the BEPS actions seems unlikely in a world with disparate views of the function of an income tax. The more likely result is a cacophony of new rules, predicated on BEPS and tempered only by the views of individual countries regarding adverse impacts on the inflow of capital. For countries convinced that their markets will remain attractive regardless of how far they push their income tax, there are no limitations. The implications for the volume of cross-border disputes, the role of tax treaties and the difficulties of satisfactory resolution are apt to be dramatic.

BEPS is likely to have a far-reaching and long-lasting impact. Developing countries, increasingly aware that taxation is a critical element in their growth strategies, are engaged in the debate. They are focused on the implications of the BEPS action plans for source-basis taxation.(31) If BEPS is curtailed, there should be greater collection of revenue in the source country. The perception is that large amounts of income are going untaxed and there is a growing realisation that investments will not be deterred if greater taxation is imposed by a jurisdiction having sufficient economic appeal.(32)

In the classic view of international taxation, the source country's claim to tax is superior to that of the residence country.(33) The strength of the source country claim varies with the type of income in question; so the claim to tax income from real property, which is immobile, is stronger than the claim to tax interest and other mobile items of income. The BEPS project flows in large part from the perception that source-country claims are being siphoned off by base erosion, while residence countries have not been doing a good job of levying the tax to which they claim entitlement. This situation fuels the lament of the developing countries, but it is hardly limited to them. Prominent taxpayers have succeeded in reducing their worldwide obligations to zero or nearly zero. Politicians who normally would not pay much attention to the subject of international taxation seem to be paying attention now. They have enlisted international organisations – primarily the OECD, but also the United Nations – in an intensive search for new rules and approaches.

Implications for US treaty policy

In these circumstances, tax treaties are being rethought, revoked or avoided altogether. Some previously lonely voices expressing scepticism about the desirability of treaties are finding new audiences.(34) A movement throughout the world towards territorial regimes arguably reduces the need for treaties in any case.

Another major element in any rational effort to reform the US rules regarding taxation of cross-border operations would be the intelligent use of tax treaties. The treaties are increasingly important in a world with large and growing investments in and from other countries.(35) The treaties provide all that, regardless of whether they tilt towards source or residence. The treaties are a useful, even indispensable, part of a cross-border strategy and any US reform should include a much greater integration of the treaties into national policy.

The US tax treaties and tax laws should be better coordinated. The problem is not that the negotiators of treaties are flying solo and in the face of statutory rules, but rather that US treaty policy is set in a vacuum with little contribution from outside the Treasury Department and without ongoing attention from those who make the laws. This is a missed opportunity.

Even without the greater focus on inbound taxation that seems appropriate, US tax treaty policy is not synchronised with the facts of modern life. Foreign investment in the United States is large and growing.(36) If Congress declines to address the phenomenon of US companies 'inverting' to claim to foreign status (ie, becoming foreign companies), the importance of inbound investment will only increase.(37)

There are profound implications for US tax treaty policy in any greater US assertion of source-basis taxation, as a result of BEPS, US tax reform or otherwise. In its treaties the United States cannot realistically cling to its extraordinary emphasis on residence taxation in a world that is moving in the opposite direction. The result will simply be a migration of the tax base out of the country. The pressures on US policy regarding treaty shopping will increase exponentially. This is the direction in which the BEPS project is likely to go, although developments may be uneven. Countries will adopt the parts of the BEPS agenda – whether it is an approach to hybrid mismatches or creative assertions of the permanent establishment concept – which serve their individual interests. US companies will find themselves footing the bill – to their own detriment and that of the US Treasury.

Some candidates for rethinking

If the US tax system does begin to focus more seriously on source-basis taxation, it would seem that US treaty policy will have to be reconsidered. In particular, the insistent favouritism of residence taxation will be a hindrance to new policies expressed in the Internal Revenue Code. Further, the limitation on benefits concept will require strengthening.

A natural starting point would be Article 24 on non-discrimination. This "baffling"(38) provision represents a significant obstacle to a more vigorous US inbound regime. The United States might consider either downplaying it or removing it altogether from the US model.

In general, Article 24 forbids a source country from taxing residents (or nationals or capital) of the treaty partner country more heavily than it taxes its own. Commentary to the OECD model notes that the principle did not originate in tax law, but instead was borrowed from commerce, navigation and shipping conventions.(39) Early tax treaties were often accompanied by explanations that the parties had included the non-discrimination article for customary, non-tax reasons.(40)

The current US model contains four paragraphs proscribing different varieties of discrimination. Article 24(1) bans taxation in the source country of "nationals" of the other country if that taxation is "more burdensome" than the source country's nationals would face "in the same circumstances". The 1996 US model broadened the term 'nationals' to include not only individuals, but also entities.(41)

Article 24(2) addresses permanent establishments. It prevents a source country from levying tax "less favourably" on a permanent establishment than it imposes on domestic firms "carrying on the same activities". Article 24(4) of the US model guards against indirect discrimination, preventing source countries from denying deductions for remittances to foreigners "under the same conditions" as those made to domestic recipients. Finally, Article 24(5) deals with capital, providing that foreign-owned capital may not be taxed in a manner that is "more burdensome" than that of domestic "similar enterprises".

These rules hinder the United States' ability to tax inbound investment. There are sound reasons to tax investment by foreign persons differently from investment by US persons. It is obviously more difficult to collect tax from foreigners,(42) and jurisdictional constraints limit the scope of US taxing power once the tax base has left the country.(43) Therefore, in the international tax realm, measured forms of discrimination may be both desirable and reasonable – desirable because of the potential loss of revenue for the government and reasonable because foreign businesses have a greater ability than domestic counterparts to escape the US tax system altogether. In short, the United States should discriminate against foreign investment, but the non-discrimination article stands in the way.

Forfeiting the ability to discriminate might be a fair trade-off if there were concomitant benefits. But justifications for the article wither under scrutiny. Some claim that the non-discrimination concept is intended to reduce impediments to cross-border trade and investment – a means of preventing "tax protectionism".(44) However, selective and thoughtful tax discrimination should not adversely affect cross-border trade; the sheer size of the US market and the stable and healthy returns enjoyed by foreign investors should allow for a fair and appropriate tax on foreign investment without fear of scaring it away. From an economic perspective, the typical rationale is that source-country non-discrimination promotes capital import neutrality,(45) an efficiency criterion in which all investments in a given country bear the same marginal rate of taxation regardless of the residence of the investor.(46) But this efficiency norm is not necessarily the most desirable of the available options. Capital export neutrality attempts to eliminate tax considerations for investors choosing between domestic and foreign investments. Commentators have observed that without harmonising global tax rates, it is impossible to achieve capital import and capital export neutrality simultaneously.(47) Further, there is some suggestion that capital export neutrality is the superior efficiency criterion because it results in fewer distortions as to the location of international investment.(48)

The non-discrimination provision allocates revenue between source and residence countries – it effectively limits the amount of tax that the source country can levy and thus the amount of double taxation that the residence country will be called on to relieve.(49) However, there is no obvious advantage in allocating taxing rights this way. An alternative norm, reciprocity, governs the taxation of portfolio investment – for example, treaty partners typically agree on a mutually acceptable rate of tax at source on royalties which is not tied to or limited by the taxes levied on purely domestic royalty arrangements.(50) In theory, at least, it is possible to negotiate for reciprocal rates of taxation on permanent establishments in the same manner. First principles do not necessarily determine which norm – reciprocity or non-discrimination – is fairer. Moreover, if the residence state is concerned with the rate of tax in the source country, it seems inapposite to choose a principle so often invoked for allegedly discriminatory taxation methods.(51)

Finally, although the United States would never endorse discrimination as a general principle, Congress often seeks to, and regularly does, enact discriminatory tax legislation. The qualifications in the text of the article's paragraphs – "similar enterprises", "same activities", "under the same conditions", "in the same circumstances" – reflect the tension between the rhetorical and sentimental appeal of the non-discrimination principle on the one hand and the hard realities of tax enforcement and revenue raising on the other.(52)

For at least several decades, Congress has leaned on these phrases in enacting laws that distinguish between domestic and foreign taxpayers. What has become clear is that the qualifying words are accommodating enough to justify increasingly distant deviations from the spirit, if not the letter, of non-discrimination. By manipulating the level of generality at which the "conditions" or "circumstances" are described, it has been possible to discriminate in substance while professing non-violation of US treaty commitments.

For example, the earnings stripping provisions of Section 163(j) deny or defer deductions for "disqualified interest" paid by thinly capitalised companies to related persons not fully subject to US tax.(53) Upon enactment, critics pointed out that this provision might violate the deductibility paragraph of the non-discrimination article since the overwhelming majority of denied deductions would be for payments to foreign lenders.(54) Congress defended the provision because it also applied, in theory, to domestic tax-exempt organisations (which are not ordinarily in the business of lending money).(55) The logic, apparently, was that taxable domestic lenders and treaty-eligible foreign lenders were not receiving interest payments under the same conditions of being fully subject to US tax.

The legislature has made similar rhetorical moves elsewhere. Section 367(e)(2) taxes the built-in gain on assets distributed in liquidation by a US subsidiary to its foreign parent, when the same tax would not apply in a purely domestic context. The Treasury technical explanation to the US model considers and dismisses the possibility that this provision violates the capital ownership non-discrimination paragraph.(56) According to the corresponding 1996 explanation, the two situations are insufficiently "similar" since the subsequent disposition of the asset would be outside the US taxing jurisdiction and therefore would escape US tax.(57)

In other instances, lawmakers and treaty negotiators have enacted questionable statutory fixes to address non-discrimination concerns. The branch profits regime, for example, subjects unincorporated US businesses of foreign corporations to a second level of US taxation intended to mimic a withholding tax on dividend distributions.(58) The tax is imposed annually and may apply regardless of whether funds are actually remitted to the foreign head office. As the American Law Institute has pointed out, this scheme likely violates the non-discrimination principle by virtue of subjecting foreign-owned US businesses to more onerous taxation than similarly situated US corporations.(59) In cognisance of this argument, Congress expressly requested that the Treasury Department renegotiate outstanding treaties to permit application of the tax; however, it also declined to concede that the tax violated the non-discrimination article in the first place.(60)

Similarly, Section 897(i) allows a foreign corporation to elect to be treated as a domestic corporation for purposes of the Foreign Investment in Real Property Tax Act 1980.(61) This provision was enacted out of concern that the act discriminated against foreign corporations by imposing withholding requirements and denying non-recognition treatment afforded to domestic corporations under the same circumstances.(62) Although Section 897(i) is a creative bit of statutory drafting, the treaty non-discrimination rule is expressed in mandatory, unequivocal terms.(63) Allowing a corporation to choose to be treated as a domestic entity is not the same thing as refraining from discriminating against it. Like the other examples above, the provision illustrates a hypocritical proclivity to discriminate against foreign investment while denying that discrimination is taking place. Even if there were some merit to the non-discrimination principle, the frequency of its defiance by US lawmakers would seem to limit its real-world efficacy.

Removing the non-discrimination article would be no easy task. Despite its flaws, the rule commands significant support among the legislature, treaty negotiators and international tax planners.(64) At present, it may be possibly only to confine the scope of the article rather than eliminate it. There is, however, relevant precedent. Unique among tax treaties currently in force, the 1982 Australia-US tax treaty contains a source-friendly version of the non-discrimination article. In contrast to the US model's private right of action, the non-discrimination provision in the convention creates only a government-to-government remedy. The article begins, uniquely, with the words "Each Contracting State in enacting tax measures, shall ensure that…" and Paragraph 4 provides:

"Where one of the Contracting States considers that the taxation measures of the other Contracting State infringe the principles set forth in this Article the Contracting States shall consult together in an endeavor to resolve the matter."(65)

A substantial net capital importer at the time of negotiation, Australia was adamantly opposed to the non-discrimination article, fearing that it would limit its ability to impose a branch profits tax, reallocate cross-border profits, implement thin capitalisation rules and regulate foreign firms for non-revenue-raising reasons.(66) US negotiators insisted on retaining some form of non-discrimination article in the text of the treaty. The result was a compromise between those two positions.

US treaty negotiators might consider the parallels between Australia's 1982 stance on non-discrimination and the current US position. Given these parallels, Paragraph 4 provides a rough-and-ready blueprint for negotiations with countries insistent on the non-discrimination principle. Even if the non-discrimination article stubbornly persists, casting it in the form of a public rather than private remedy removes much of its bite.

Such a move would clear one impediment to more effective taxation of inbound investment, but there are other targets in the US model. In recent decades, earnings stripping has significantly eroded the tax base of the United States and many other countries.(67) Section 163(j) purports to address the problem,(68) but it is largely ineffective and has the spectre of the non-discrimination article hanging over it. Article 11(1) of the US model exempts most types of interest from source-country taxation. The subsequent paragraph provides an exception for interest "determined with reference to receipts, sales, income, profits, or other cash flow of the debtor", which may be taxed at the maximum rate of 15%.(69) To support a more rigorous source-based tax regime, the United States might consider broadening the exception to include some of the base-eroding hybrid mismatch arrangements currently in the OECD's crosshairs – for example, by allowing withholding on interest when the payer is a disregarded entity from the point of view of the recipient's domestic law.(70)

Interest is probably the most important of the investment categories, owing both to the volume of cross-border interest flows and the arbitrage opportunities arising from the fungibility and tax deductibility of debt financing. However, the treatment of other kinds of investment may also merit rethinking. On dividends, the US model generally provides for preferential withholding rates of 5% (if the recipient is a corporation owning at least 10% of the voting stock of the corporation paying the dividend) or 15% in most other cases.(71) However, some categories of dividend, such as those paid by real estate investment trusts, are sometimes not entitled even to the higher 15% preferential rate.(72) As with interest, the exceptions might be broadened to include other categories of objectionable distribution. US model rules on royalties should receive similar scrutiny, since royalties are taxable only in the residence country under those rules.(73) This exemption dates from the era when the United States was a major net exporter of intellectual property. In light of its current status as a net IP importer, the United States might revisit the general zero withholding rate as well.

Perhaps the most direct impediment to robust inbound taxation is the flinty definition of the term 'permanent establishment' in the US model. Under Article 5, foreign companies may conduct significant business in the United States without fear of triggering US tax. The enumerated list of activities that do not constitute a permanent establishment is long. For example, Articles 5(4)(a) and (b) exempt the "use of facilities" and the "maintenance of a stock of goods" for the purpose of delivering goods to customers.

These exceptions are passing into obsolescence. Modern technology makes it possible to negotiate and conclude sales contracts without a physical presence in the market nation. As a result, businesses that operate predominantly online have found it increasingly easy to shoehorn their activities into once-narrow carve-outs. Online sellers of physical products, whose business models often depend on proximity to customers and fast delivery times, are major beneficiaries.(74)

Several BEPS reports propose updates to the definition of PE, with the goal of fixing the incongruity between the enumerated carve-outs and the modern economy. Those recommendations include specifying that all activities described in the OECD model's Article 5(4) must be of an "auxiliary or preparatory nature", as well as removing the word 'delivery' from paragraphs (a) and (b) of that article (which exempt "the use of facilities" and "the maintenance of goods", respectively, when undertaken "solely for the purpose of storage, display or delivery of goods or merchandise belonging to the enterprise") (emphasis added).(75) Other recommended changes include clamping down on so-called 'commissionnaire' arrangements by imputing a permanent establishment when firms regularly hire ostensibly independent intermediaries to act on their behalf.(76) Although US policymakers should not blindly adopt these recommendations, they are useful for spurring discussion on changes to the analogous article in the US model.

More ambitiously, the OECD has suggested supplementing the brick-and-mortar permanent establishment concept with a new 'digital presence' threshold.(77) This test would target companies whose products are largely dematerialised and intangible (eg, Facebook and Google). Factors to be potentially included in such a test include the volume of sales transactions conducted with customers in the source country, the number of active website accounts and the use of source-country banking or logistics services.(78) The prospects for implementing the concept are meagre at best; such a project would involve upending a longstanding and obdurate global consensus that permanent establishments are based on physical presence alone. But US negotiators should at least begin rethinking this consensus. In the coming decades, the treaty system will need to reckon with the fact that corporeality is no longer a good proxy for permanence.

Finally, the limitation on benefits article needs a thorough re-evaluation. Whether the United States continues to see itself as primarily a residence-basis country – and there is little doubt that will be the case – or adopts a more neutral position, the inconsistencies and discontinuities in Article 22 of the US model must be addressed. One option would be a much shorter and less objective set of tests, as is found in the limitation on benefits article in the Cyprus-US treaty.(79) If this proves to be too subjective, something of this nature should at least serve as a supplement to objective rules that can provide a roadmap for treaty shopping.


The Treasury is working on a new version of the US model. Informally, it has let it be known that all aspects of the 2006 model are candidates for review and revision. In the process of that review, the Treasury should pause to consider whether policies that have animated the US treaty negotiating position for decades are themselves worthy of a fresh look. If the concepts advanced in this article are correct, the process could produce dramatic changes to the model in the future.

For further information on this topic please contact H David Rosenbloom or Joseph Brothers at Caplin & Drysdale by telephone (+1 202 862 5000) or email ([email protected] or [email protected]). The Caplin & Drysdale, Chartered website can be accessed at


(1) The United States stands prepared to diverge significantly from the US model in negotiating with developing countries. See Eric J Smith, "The US-Mexico Tax Treaty" 8 Fla J Int'l L 97, 102 (1993) ("In order to take into account the importance of source-based taxation for [developing] countries, the United States often includes several provisions based on the U.N. Model in its treaties with [those] countries").

(2) See generally Stanley S Surrey, "The Pakistan Tax Treaty and 'Tax Sparing'", 11 Nat'l Tax J 156 (1958).

(3) Ernest R Larkins, "US Income Tax Treaties in Research and Planning: A Primer", 18 Va Tax Rev 133, 185 (1998) ("The intended treaty benefit to US citizens and US corporations is a decrease in their foreign taxes, not their US tax burdens").

(4) See US model, Article 9(2).

(5) US model, Article 17.

(6) US model, Article 18.

(7) US model, Article 19.

(8) US model, Article 20.

(9) US model, Article 27.

(10) See US model, Article 6(5).

(11) US model, Article 16(1).

(12) US model, Article 20(2).

(13) In recent US treaties, but not yet the US model, the United States has agreed to a 0% tax rate at source on specific dividends concerning direct investment. This provision speaks volumes in regard to the US embrace of residence-based taxation. See, for example, the Convention Between the United States of America and the Federal Republic of Germany for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion With Respect to Taxes on Income and Capital and With Respect to Certain Other Taxes, August 29 1989, Article 10(3), 1708 UNTS 3, incorporating Protocol 1, June 1 2006 (providing for a 0% rate on dividends paid from subsidiaries to parent companies owning at least 80% of subsidiaries' shares).

(14) Curiously, this perception does not seem to have squared with the facts for at least several decades. See, for example, Julie Roin, "Rethinking Tax Treaties in a Strategic World With Disparate Tax Systems", 81 Va L Rev 1753, 1758 n 17 (noting that there is "no question that the United States is a net importer of capital on a current basis").

(15) For converse reasons, developing nations are often reluctant to relinquish source taxation. See Charles I Kingson, "The Coherence of International Taxation", 81 Colum L Rev 1151, 1166-1167 (1981).

(16) See, for example, Michael J Graetz and Itai Grinberg, "Taxing International Portfolio Income", 56 Tax L Rev 537, 570 (2003) (claiming that source countries are poorly situated to gauge taxpayers' ability to pay).

(17) See Hugo Hurtado, "Is Latin American Taxation Policy Appropriate for Promoting Foreign Direct Investment in the Region?" 31 NW J Int'l L & Bus 313, 333 ("Basically, the UN Model increases the source country's right to tax income aimed at benefitting the developing country").

(18) See generally Simone M Haug, "The United States Policy of Stringent Anti-Treaty Shopping Provisions: A Comparative Analysis", 29 Vand J Transnat'l L 191 (1996).

(19) Indeed, some of the most popular avoidance structures have involved use of Netherlands tax treaties by shell entities. See Richard A Westin, The Tax Lexicon 158 (1989) (describing the "Dutch sandwich", involving the tax treaty networks of both the Netherlands and the Netherlands Antilles).

(20) See American Law Institute, International Aspects of United States Income Taxation II: Proposals on United States Income Tax Treaties, 152 (1992) (noting that if companies from non-treaty nations are permitted to form corporate entities in order to obtain treaty benefits, this reduces pressure on their home countries to enter into treaties).

(21) See, for example, Aiken Indus v Comm'r, 56 TC 925 (1971); Northern Ind Public Service Co v Comm'r, 105 TC 22 (1995); Del Commercial Prop v Comm'r, 78 TCM (CCH) 1183 (1999), aff'd, 251 F 3d 210 (DC Cir 2001).

(22) Jonathan Weisman, "A Republican's Tax Overhaul Envisions Big Changes", New York Times (February 26 2014), available at plan-would-bring-big-changes.html?_r=0.

(23) See, for example, Noel B Cunningham and Deborah H Schenk, "The Case for a Capital Gains Preference", 48 Tax L Rev 319, 320 (1993) (noting that others have made a "persuasive case that the arguments in favor of a [capital gains] preference are weak").

(24) Section 509(a) states that:

"For purposes of paragraph (3), an organization described in paragraph (2) shall be deemed to include an organization described in section 501(c)(4), (5), or (6) which would be described in paragraph (2) if it were an organization described in section 501(c)(3)."

(25) See Robert J Peroni, J Clifton Fleming Jr and Stephen E Shay, "Reform and Simplification of the US Foreign Tax Credit Rules", Tax Notes Int'l, September 29 2003, p 1177 (describing reduction in complexity of the code as a "worthwhile, although somewhat elusive, objective of tax reform efforts").

(26) See Fleming Jr, Peroni, and Shay, "Designing a US Exemption System for Foreign Income When the Treasury Is Empty", 13 Fla Tax Rev 397, 400 ("[T]here is a significant likelihood that Congress will sooner or later be considering legislation to create a US territorial or exemption system").

(27) The European Commission has recently taken the view that Ireland, Luxembourg and the Netherlands belong in this category. See European Commission, "State aid: Commission investigates transfer pricing arrangements on corporate taxation of Apple (Ireland) Starbucks (Netherlands) and Fiat Finance and Trade (Luxembourg)" (June 11 2014), available at http://

(28) UK Chancellor George Osborne has recently proposed a controversial diverted profits tax, which would impose a 25% levy on companies using "contrived arrangements" to shift profits outside the United Kingdom. See Vanessa Houlder, "Business Leaders Attack UK 'Google Tax'", Financial Times (December 10 2014), available at 00144feabdc0.html.

(29) See generally OECD Centre for Tax Policy and Administration, "BEPS 2014 Deliverables", available at

(30) BEPS Project Head Pascal Saint-Amans has stated that "source taxation is more on the table than in the past". PwC, "An interview with Pascal Saint-Amans" (February 2014), at 2, available at

(31) See, for example, Yariv Brauner, "What the BEPS?" 16 Fla Tax Rev 55, 74-75 (2014) (describing characterisation and sourcing rules as a "major bone of controversy between developing countries and the OECD").

(32) See generally Bret Wells and Cym Lowell, "Tax Base Erosion and Homeless Income: Collection at Source Is the Linchpin", 65 Tax L Rev 535 (2011).

(33) See Ruth Mason and Michael S Knoll, "What Is Tax Discrimination?" 121 Yale L J 1014, 1029 (2014) ("Under longstanding custom, the tax entitlement of the source state is superior to that of the residence state").

(34) See Antonio Hugo Figueroa, "Tax treaties to avoid international double taxation or to transfer resources from developing countries to developed ones?" 2:14 Voces en el Fenix 128 (May 2012); Tsilly Dagan, "The Tax Treaties Myth", 32 NYU J Int'l L & Politics 939 (2000).

(35) See Joseph Isenbergh, International Taxation (2014), Section 101.4 ("[T]he mere fact of exposure to the tax system of another country imposes significant transactional complexities on enterprises venturing beyond their national boundaries").

(36) According to Economy Watch, the value of direct inbound investment at the end of 2013 was $2.815 trillion; the value of direct outbound investment was $4.854 trillion. See "United States FDI Statistics", Economy Watch (March 17 2015), available at States/FDI_Statistics/.

(37) A large proportion of the tax savings generated by corporate inversions occurs through post-inversion earnings stripping transactions. See Jim A Seida and William F Wempe, "Effective Tax Rate Changes and Earnings Stripping Following Corporate Inversion", 57 Nat'l Tax J 805 (2004).

(38) See Mary Bennett, "The David R. Tillinghast Lecture: Nondiscrimination in International Tax Law: A Concept in Search of a Principle", 59 Tax L Rev 439, 439 (2006).

(39) See commentary to OECD model, Article 24(6) (2010).

(40) See, for example, US Treasury Department Memorandum Explaining Article XXI of the 1945 US-UK Treaty ("It will be observed that this article extends to all taxes, both Federal and local. Such extension, however, is in keeping with several commercial treaties (such as that with Norway, of 1928, and that with Germany, of 1923) to which the United States is now a party. It has no practical effect, since our domestic taxation does not discriminate as between United States citizens and British nationals residing in the United States").

(41) Compare Sanford H Goldberg and Peter A Glicklich, "Treaty-Based Nondiscrimination: Now You See It Now You Don't", 1 Fla Tax Rev 51, 82 (1992) (noting that 1981 US model limits this paragraph to individuals) with 1996 US model, available at (paragraph applies to both individuals and entities).

(42) For this reason, payers of passive income are generally required to withhold it at source. See generally Section 1441.

(43) This policy rationale underlies, among other provisions, Section 367(e).

(44) OECD Thin Capitalisation Report, 27 P 66 (1987); see also Kees van Raad, Nondiscrimination in International Tax Law 127 (1986) ("The purpose of the permanent establishment nondiscrimination provision is clear: a State should not offer enterprises operated by residents any competitive advantage over similar enterprises of nonresidents").

(45) Mason and Knoll, supra note 33, 1020 (2012) ("[A] capital export neutrality construction of nondiscrimination would impose nondiscrimination obligations only on residence states, whereas a capital import neutrality construction of nondiscrimination would impose nondiscrimination obligations only on source states").

(46) See Charles H Gustafson, Peroni and Richard Crawford Pugh, Taxation of International Transactions, 17 (2d ed 2001).

(47) See Graetz and Alvin C Warren Jr, Income Tax Discrimination and the Political and Economic Integration of Europe, 1212-1220 (2006).

(48) See, for example, Treasury Department, The Deferral of Income Earned Through US Controlled Foreign Corporations: A Policy Study 25-42 (2000), available at Documents/subpartf.pdf.

(49) Robert A Green, "The Troubled Rule of Nondiscrimination in Taxing Foreign Direct Investment", 26 Law & Pol'y Int'l Bus 113, 128-129 (1994).

(50) See generally US model, Articles 10-12 (specifying reciprocal withholding rates for interest, dividends and royalties).

(51) For example, a 1990 proposal to extend the typical three-year statute of limitations to six years for foreign corporations was criticised and ultimately shelved as violating the non-discrimination provision. See Task Force of American Bar Association (ABA) Comm on US Activities of Foreigners and Tax Treaties, ABA Comments on Foreign Tax Equity Act of 1990 (HR 4308 and S 2410).

(52) The technical explanation to the US model reflects this tension, at times implying that providing sound reasons for discriminatory policies is enough to render those policies non-discriminatory. See US model, technical explanation, Article 24 ("[T]he common underlying premise [of the qualifying phrases] rule commands significant support among legislators, is that if the difference in treatment is directly related to a tax-relevant difference in the situations of the domestic and foreign persons being compared, that difference is not to be treated as discriminatory").

(53) See Section 163(j)(1)(A).

(54) See, for example, American Law Institute, Federal Income Tax Project: International Aspects of United States Income Taxation II, "Proposals on United States Income Tax Treaties", 258-259 (1992).

(55) See HR Rep 101-386, at 568 (1989) (Conf Rep).

(56) See US model, technical explanation, Article 24(5) ("The taxation of a distributing corporation under section 367(e) on an applicable distribution to foreign shareholders does not violate paragraph 5 of the Article because a foreign-owned corporation is not similar to a domestically owned corporation that is accorded non-recognition treatment under sections 337 and 355"). If this statement is true, as it appears to be, one may question what it leaves intact in the non-discrimination principle.

(57) US model, technical explanation, Article 24(1); see also Notice 87-66, 1987-2 CB 376 (1987).

(58) See Section 884.

(59) American Law Institute, Federal Income Tax Project: International Aspects of United States Income Taxation 144-145 (1987).

(60) See HR Rep 99-426, 433 (1985), reprinted in 1986-3 CB (Vol 2).

(61) PL No 96-499 Sections 1121-1125, 94 Stat 2682 (codified in scattered Sections of 26 USC).

(62) See Boris Bittker and Lawrence Lokken, Federal Taxation of Income, Estates and Gifts (2015), Section 67.7.6 (explaining that although the pre-1986 corporate non-recognition rules regarding distributions by corporations to shareholders provided the primary impetus for the enactment of the Section 897(i) election, the election remains available after repeal of those rules).

(63) See, for example, US model Article 24(1) ("Nationals of a Contracting State shall not be subjected") (emphasis added).

(64) Mason and Knoll, supra note 33, at 1018 ("Lack of a clear definition has not prevented prohibitions of tax discrimination from appearing in (or being read into) statutes, constitutions, and international treaties").

(65) See Convention Between the United States of America and the Government of Australia for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion with Respect to Taxes on Income, August 6 1982, Article 23(4), 1 Tax Treaties (CCH) P 520.

(66) See Goldberg and Glicklich, supra note 41, at 58 ("Australia is apparently of the view that certain serious disadvantages may arise from including a broad nondiscrimination provision in its treaties. Such a provision would conflict with what Australia considers a proper division of taxing rights between the parties to the tax treaties").

(67) See generally Ilan Benshalom, "The Quest to Tax Interest Income in a Global Economy: Stages in the Development of International Income Taxation", 27 Va Tax Rev 631 (2008).

(68) See Treasury Department, "Report to the Congress on Earnings Stripping, Transfer Pricing and US Tax Treaties", 23 (November 2007), available at tax-policy/Documents/ajca2007.pdf.

(69) See US model, Article 11(2).

(70) See OECD, "BEPS Action 2: Hybrid Mismatch Arrangements" (September 2014).

(71) See US model, Article 10(2). Several US treaties now reduce the rate on some direct investment dividends to 0%.

(72) US model, Article 10(4).

(73) US model, Article 12(1).

(74) See OECD, "BEPS Action 1: The Digital Economy" (September 2014), at 127-129.

(75) OECD, "BEPS Action 7: Artificial Avoidance of PE Status" (discussion draft) (October 31 2014), at 15-16.

(76) Id at 12-14.

(77) OECD, "BEPS Action 1: The Digital Economy" (Sept. 2014), at 143-145.

(78) Id.

(79) Convention Between the Government of the United States and the Government of the Republic of Cyprus for the Avoidance of Double Taxation and the Prevention of Fiscal Evasion With Respect to Taxes on Income, with Related Notes, March 19 1984, Article 26, 35 UST 4737.

An earlier version of this update first appeared in Tax Notes International: H David Rosenbloom and Joseph P Brothers, "Reflections on the Intersection of U.S. Tax Treaty Policy, U.S. Tax Reform, and BEPS", Tax Notes International, May 25 2015, pp 759-769.