Bruyea’s Blunder
Sep 5, 2026
We finally have a decision from the Federal Circuit appeal of Bruyea v. United States,[1] and it did not go as I had hoped — nor is it, in my view, correct in law. The appellate court unequivocally rejected the contention that Article XXIV of the US-Canada tax treaty[2] allows a credit for Canadian tax to be claimed against the US net investment income tax (NIIT). Practically speaking, this reversal eliminates the stronger of two contested theories for how US citizens residing in Canada may avoid NIIT exposure.
Foreign tax credits are the main reason citizenship-based taxation is more of a paperwork headache than a net tax impact. At the end of the day, with proper planning, an American living in Canada may very often pay the same total tax as their similarly-situated Canadian neighbour. But the NIIT stands out as an exception to this backdrop: a special “extra” tax that threatens to attach a persistent financial burden to the retention of US citizenship.
The NIIT is a 3.8% tax on the investment income of high-earning US citizens and residents, enacted during the Obama administration[3] as Chapter 2A of the Internal Revenue Code.[4] The IRC on its own does not provide a foreign tax credit against the NIIT,[5] but it has long been recognized that tax treaties can expand access to foreign tax credits versus the IRC.[6] Despite this, the US Treasury Department initially indicated that typical treaty language foreclosed credit against NIIT;[7] this view was promptly criticized.[8] Bruyea is the first court case to address the question under the US-Canada treaty.
In this post I explain three critical errors that I believe the Federal Circuit made in its analysis of the treaty. I will also comment on the alternative theory, which is still unsettled, that the NIIT is avoidable under the social security totalization agreement.
The court overreads the clear grammatical scope of the US Law Limitation
Article XXIV contains more than one provision for credit against US tax. There is a general foreign tax credit rule in paragraph 1, which is “subject to the limitations of the law of the United States” (the “US Law Limitation”). And there is a special “three-bite rule” in paragraphs 4, 5, and 6 for the situation of a US citizen who is a resident of Canada,[9] which includes a credit against US tax in paragraph 4(b).
The court reasonably regards the absence of any foreign tax credit against NIIT in the Internal Revenue Code as a “limitation” of US law that is properly incorporated into the general rule of Article XXIV paragraph 1.[10] But the court goes on to say that the US Law Limitation also applies to the paragraph 4(b) credit,[11] which is an egregious atextual extrapolation.
Paragraph 1 is “subject to the provisions of paragraphs 4, 5 and 6” — that is, the three-bite rule takes precedence over the general rule. And paragraph 1 does not say that all of the treaty or even all of Article XXIV is subject to the limitations of US law. As written, the US Law Limitation is a subordinate clause that modifies only the general rule in paragraph 1.
To justify this departure from text, the court refers us to its reasoning in the companion case Christensen v. United States,[12] which invokes the distributive canon to read the US Law Limitation as an “overarching qualifier”:
“The distributive canon, for example, recognizes that sometimes where a sentence contains several antecedents and several consequents, courts should read them distributively and apply the words to the subjects which, by context, they seem most properly to relate.”[13]
But the distributive canon is inapropos here, because its function is to distribute subjects to objects within the same sentence, as in: “Men and women are eligible to become members of fraternities and sororities”.[14] It surely does not authorize us to “distribute” a modifying clause into an entirely different paragraph!
Instead, the court should have recognized meaningful variation[15] between the US Law Limitation’s presence in paragraph 1 and its omission from paragraph 4(b).
Differences that the court labels “anomalous results” are an intended outcome of the treaty’s negotiated balancing of jurisdictional interests
The court was guided by its aversion to the “anomalous result” of a hypothetical US citizen in Toronto enjoying a more generous credit than a US citizen in New York.[16]
By this logic, the entire three-bite rule is anomalous, and the treaty should not have had any provisions that begin as paragraph 4 does: “Where a United States citizen is a resident of Canada, …”
Fortunately the treaty negotiators were not so thoughtless. They understood that there is nothing anomalous about prioritizing the taxing rights of the country of residence. The policy underlying the three-bite rule “is that the United States, and not the treaty partner, should bear the burden of providing double tax relief in cases where double taxation results from the US imposition of tax solely on the basis of US citizenship.”[17] That policy quite sensibly discriminates between residents of New York and Toronto.
In harmonizing the treaty with the Code, the court should have recognized that explicit limitations deserve greater weight than implicit limitations
Having established that paragraph 4(b) does not subject itself to the US Law Limitation, we should have been left with the foundational rule that neither the treaty nor the Internal Revenue Code has preferential status;[18] rather they “must stand together in harmony” if possible.[19] Or, in the words of the Supreme Court, we try to interpret treaty and statute “so as to give effect to both, if that can be done without violating the language of either.”[20]
Under these principles, if the NIIT statute had contained an explicit limitation, such as “no credit for foreign taxes shall be allowed against the tax imposed by this section”, then we would easily apply that limitation as an overlay on the treaty. But because the NIIT statute is silent in regard to foreign tax credits, harmonization is best served by allowing the treaty credit.
In that way, Bruyea could have rationally distinguished itself from the prior cases of Kappus v. Commissioner[21] and Jamieson v. Commissioner,[22] which held that paragraphs 4, 5, and 6 did not override an explicit statutory limitation on foreign tax credits for purposes of the alternative minimum tax (AMT).[23]
This type of harmonization exercise could also have easily dispensed with the court’s anxieties about unfair double benefits under the foreign earned income exclusion.[24] A coordination rule explicitly disallows any credit for foreign taxes on excluded income.[25] With statute and treaty on equal footing, it does not seem difficult at all to say that the rule disallowing a double benefit applies to the treaty credit, while otherwise allowing treaty credit against NIIT.
Avoidance under the totalization agreement
All hope is not lost for US citizens abroad. Some have proposed a different theory based on social security totalization agreements (SSTAs):
“An alternate argument would contend that NII tax payments are akin to social insurance charges that might be covered by totalization agreements between the United States and other countries. … Totalization agreements lack the range of the tax treaty network, but going forward this could provide an avenue of relief for US taxpayers living abroad with NII tax liability.”[26]
This theory as applied to the US-Canada SSTA[27] was examined in more detail by Kevyn Nightingale, who wrote:
“Individuals employed primarily in Canada and self-employed individuals who reside in Canada are subject to the provisions of the CPP, not US Social Security. Consequently, if the net investment income tax is a social security tax, then an individual who lives in Canada and is subject to the CPP would be exempt from it.”[28]
The NIIT did not exist when the US-Canada SSTA was signed. Article II paragraph 1 defines “applicable laws” to include chapters 2 and 21 of the Internal Revenue Code, which govern social security and medicare contributions. But paragraph 3 adds that the SSTA “shall also apply to laws which amend, supplement, consolidate or supersede the laws specified in paragraph (1).”
The question whether the NIIT “supplements” the medicare tax for purposes of an SSTA was litigated inconclusively in Paul Young Kim v. United States.[29] The district court, ruling on a motion to dismiss, found that a textual analysis was supportive of this position:
“Plaintiff argues based on Chapter 2A’s title that Chapter 2A is intended to be a source of Medicare tax revenue, ‘supplementing’ the Medicare taxation provisions of Chapters 2 and 21. … This reading is supported by the legislative history of Chapter 2A.”[30]
However, the court went on to express skepticism that this conclusion was sufficiently “consistent with the shared expectations of the contracting parties”.[31] The case was subsequently settled without a decision on the merits.
I believe the SSTA theory in Canada has two significant weaknesses:
- There is probably no evidence that the governments of Canada and the US had any shared expectation of the SSTA applying to something like the NIIT, because nothing similar to the NIIT existed when the agreement was negotiated.
- Even if we grant that the SSTA “applies to” the NIIT under Article II, there is a second hurdle to clear: we must somehow read an exemption from NIIT into Article V, which says that an employee who works in Canada “shall, in respect of that work, be subject to the laws of only” Canada. This is troublesome because the NIIT is not imposed in respect of work compensation.
Nevertheless, it has been observed that “the Kim case provides enterprising taxpayers, who are not afraid of a dispute with the IRS, some support for the SSTA position.”[32]