The US Model Income Tax Convention: emerging challenges and treaty evolution

October 2025  |  SPECIAL REPORT: CORPORATE TAX

Financier Worldwide Magazine

October 2025 Issue

The US periodically publishes a model double tax treaty (US Model). The first US Model was published in 1976 and the current iteration in 2016: United States Model Income Tax Convention 2016. The US Model represents the policy positions of the US and is intended to serve as a starting point, if not a template, for US treaty negotiations. Many of the US Model provisions are similar or identical to provisions in the Model Double Taxation Convention on Income and Capital, and its Commentaries, published by the Organisation for Economic Co-operation and Development (OECD Model). We focus here on several issues that arise, somewhat uniquely, under US treaties that follow the US Model: (i) the practical application of mandatory binding arbitration in cases arising under the US Model’s mutual agreement procedure (MAP); (ii) how certain recently-enacted provisions of US tax law affect interpretation of US Model treaties; (iii) why US state and local taxing jurisdictions are not covered by the US Model; and (iv) how the Internal Revenue Service (IRS) believes the US Model applies to the relatively new US net investment income tax (NIIT).

Mandatory binding arbitration

The US Model provides for mandatory binding arbitration in the event the competent authorities cannot reach agreement on the arm’s length outcome in a MAP or advance pricing agreement (APA). The US Model approach empowers an arbitration panel that is selected jointly by the two governments – each government appoints one arbitrator and those two select a third – to choose only from among the respective positions of the governments. The panel is not free to develop its own approach. The idea is that, if the arbitration panel must select only one of the two positions and cannot provide a negotiated settlement, both competent authorities will be incentivised to moderate their positions and come to the middle. If the governments moderate their positions sufficiently, they should be able to reach agreement without an arbitration process. If the process works optimally, the governments will achieve agreement in every case and actual arbitrations would become a rarity.

In our experience, the US Model’s arbitration approach, which has been implemented through memoranda of understanding negotiated by the US and relevant foreign competent authority, has worked almost exactly as intended, particularly for resolving US-Canada MAP and APA cases. Seldom does a US-Canada case actually go to arbitration. All seem to be resolved before arbitration deadlines – two years for a MAP and four years for an APA – are reached. Two areas could be improved, however: (i) the governments allow cases to linger, resolving most shortly before the arbitration deadline; and (ii) arbitration has not led to a material decline in US-Canada cases. Both shortcomings likely result from the general uncertainty that accompanies transfer pricing (TP) cases. Choosing a specific transfer price or TP method is an art, not a science, and only after extensive fact development can both countries be relatively confident in their positions. In addition, there are rules of comity for this intergovernmental activity and simple respect may require acceptance of all proffered cases.

The US Model’s inclusion of mandatory binding arbitration reflects the principled intentions of the US. No government would cede its decision-making authority to an arbitrator if it had mischief in mind. Further, the US Model provides minutely detailed procedures for how arbitrations are to be conducted. Those procedures have the effect of truly binding the US and foreign competent authorities to the process and to its outcome.

In addition, use of mandatory binding arbitration seems to have had positive knock-on effects for how the IRS addresses TP cases generally (at least where a foreign competent authority could be involved in a dispute). Specifically, the IRS exam function for TP (the Transfer Pricing Practice) must seek review by the US competent authority function, the Advance Pricing and Mutual Agreement Program (APMA) of any adjustment that could be subject to MAP. This allows APMA to bring its arbitration-enhanced understanding of likely outcomes to bear before an adjustment has been made and to rebuff adjustments that likely will not be sustained. This gatekeeping activity also exposes the IRS exam function to the pressure of mandatory binding arbitration and discourages proposed adjustments that might not survive the arbitration process. We do still see what seem likely arbitrary IRS adjustments when APMA is not watching. We would recommend that APMA be consulted on all TP adjustments since it is in a unique position to understand the likely outcomes that are produced when an effective arbitration process is in place.

Interaction of US Model with recent US legislative developments

In 2017, the US passed the Tax Cuts and Jobs Act, which, among other changes, introduced three new taxing regimes: the Global Intangible Low-Taxed Income (GILTI) regime, the Foreign Derived Intangible Income (FDII) regime and the Base Erosion and Anti-Abuse Tax (BEAT) regime. These regimes were retained and somewhat modified (including their names) by the 2025 One Big Beautiful Bill Act. The GILTI regime taxes, on a current basis, certain income of foreign subsidiaries of US multinationals. FDII allows for a reduced US income tax rate on income earned from foreign sales. The BEAT imposes an alternative minimum tax after adding back certain expenses paid to foreign persons.

A key issue for the US Model is whether these US taxing regimes are consistent with existing treaty limitations on the ability of the US to tax income earned in the treaty partner’s jurisdiction. The US has indicated that it is considering changes to the US Model to address these new regimes and that its official position is that any negotiated treaty will reserve US rights. A reservation to that effect was included in the US-Chile double tax treaty effective from 2023.

Non-coverage of US state and local taxing jurisdictions

Neither the US Model nor any existing US treaty limits the taxing authority of state or local taxing jurisdiction within the US, except in the non-discrimination article. A foreign company or business that sells into the US and lacks a US permanent establishment thus may nonetheless be subject to state or local income taxation. These outcomes may be unexpected to taxpayers doing business in the US for the first time.

Nor do the US Model or US treaties impose on foreign taxing jurisdictions any requirement to credit state and local income taxes against their domestic income taxes on the same income. Since many jurisdictions historically have maintained participation exemption systems that exempt income derived abroad, the lack of creditability of state and local income taxes may not have produced double taxation. This may change, however, with the introduction of Pillar Two taxes in non-US jurisdictions that seek to tax the US income of their domestic multinationals. Switzerland, whose cantons retain powers similar to US states, has incorporated them into and made them subject to its treaties. Perhaps the next iteration of the US Model will adopt this approach.

By contrast, foreign analogues to state and local income taxes often are covered, including taxes imposed by Swiss cantons and German lander. The US-Canada treaty includes an unusual provision that covers state and local taxes to the extent they are similar to federal taxes.

Application of the US Model to the US NIIT

The NIIT is a 3.8 percent tax on certain net investment income of individuals, estates and trusts. Where a US citizen is resident in a foreign country and that foreign country imposes tax on income that is subject to the NIIT, the treaty question is whether the foreign tax should be creditable against the NIIT.

The IRS takes the position that foreign taxes are never creditable against the NIIT because US treaties allow a credit only if US law, independent of treaties, allows for a credit. US law provides for credits against US income tax, but not against the NIIT.

Some taxpayers have disagreed and challenged the IRS view. For example, in the ongoing Christensen v. United States case, on appeal in the US Court of Appeals for the Federal Circuit, the taxpayer paid the NIIT and is seeking a refund on the basis that the US-France treaty provides for a credit and overrides US domestic law. The US Court of Federal Claims agreed with the taxpayer and the IRS appealed. The case is being heard together with a case involving creditability of the NIIT under the US-Canada treaty, Bruyea v. United States. We await a decision for the cases.

Conclusion

We have only scratched the surface of topics under the US Model. We expect a new iteration of the US Model in the near future. It hopefully will provide a clear US position on the implications of GILTI, FDII and BEAT on US treaty obligations. In an ideal (but unlikely) world, the IRS would retreat from its position in Christensen.

 

Clark Armitage is a member of Caplin & Drysdale. He can be contacted on +1 (202) 862 5078 or by email: carmitage@capdale.com.

© Financier Worldwide

BY

Clark Armitage

Caplin & Drysdale

©2001-2025 Financier Worldwide Ltd. All rights reserved. Any statements expressed on this website are understood to be general opinions and should not be relied upon as legal, financial or any other form of professional advice. Opinions expressed do not necessarily represent the views of the authors’ current or previous employers, or clients. The publisher, authors and authors' firms are not responsible for any loss third parties may suffer in connection with information or materials presented on this website, or use of any such information or materials by any third parties.