Assuming the Treaty Works on Autopilot — What the US-Canada Tax Treaty Does and Doesn’t Do for You

Ask an American living in Canada why they’re not worried about US taxes, and you’ll usually hear some version of the same sentence: “There’s a tax treaty, so I’m covered.”

It’s the most reassuring sentence in cross-border tax, and it’s roughly one-third true. The treaty is real, it’s valuable, and it prevents a great deal of double taxation. But it has three properties almost nobody outside the profession knows about — and each one is a way the “I’m covered” assumption quietly fails.

Property one: most treaty benefits are claimed, not granted

The treaty is not a background process running on your finances. With limited exceptions, its protections are positions you take on a filed return — a credit you compute, a treaty-based position you disclose, a rate you assert. No return, no claim. The treaty can’t reduce the tax on a return that doesn’t exist, and it has no effect at all on whether you’re required to file in the first place.

This is the load-bearing flaw in “there’s a treaty, so I don’t need to file.” The treaty is a reason your bill is often zero. It is never a reason your return is optional.

Property two: the treaty mostly doesn’t apply to US citizens — by design

Buried in the treaty is a provision practitioners call the saving clause: the United States reserves the right to tax its own citizens as if most of the treaty didn’t exist. The majority of the treaty’s protections are written for residents of one country dealing with the other — and the saving clause carves US citizens out of most of them, no matter where they live.

A specific list of benefits survives the carve-out (relief from double taxation among them — which is why the system still mostly works). But the mental model of “the treaty shields me from the IRS because I live in Canada” has it backwards. For a US citizen, the treaty’s starting position is that the IRS’s claim on you survives intact, with enumerated exceptions. You live in the exceptions.

Property three: the treaty does nothing — literally nothing — about reporting

Here’s where the assumption gets expensive. The treaty is about tax: who taxes what income, and how double tax gets relieved. The disclosure system — the foreign bank account report, the foreign asset statements, the rest of the paperwork stack — exists in a separate legal universe the treaty doesn’t touch.

You can be in a position where the treaty legitimately reduces your US tax to zero, and your reporting obligations remain exactly what they would be with no treaty at all. The penalties on that side aren’t computed from tax owed, so “the treaty zeroed me out” is no defense to a missed disclosure. The two most expensive words to confuse in this field are taxation and reporting — the treaty governs the first and is silent on the second.

The TFSA: where “registered in Canada” means nothing in Washington

Canadians are trained to think of registered accounts as a family — RRSP, TFSA, RESP, all blessed by Ottawa, all tax-advantaged. The treaty does not see a family. It addresses certain retirement plans specifically, and the others simply aren’t there.

The TFSA is the famous casualty: tax-free in Canada, fully taxable to the US as an ordinary investment account — and depending on what’s inside it and how it’s structured, capable of triggering some of the heaviest reporting in the US system. “Tax-free” on a Canadian brochure is a statement about Canadian law. The treaty doesn’t extend it across the border, and assuming it does is one of the most common ways new arrivals walk into a paperwork problem.

The twist: the one place autopilot is real

Here’s the irony. The single spot where the treaty does now run automatically is the one where, for years, it didn’t.

US persons with an RRSP or RRIF once had to make an explicit election — on a dedicated form — to defer US tax on the growth inside the plan. Forget the election, and the IRS’s default position was that your retirement account’s internal growth was taxable annually. People missed it constantly; it was a classic trap.

In 2014, the IRS ended it: under current guidance, eligible individuals are treated as having made the election automatically, retroactively, with the old form eliminated. If a preparer or an old internet post tells you that you must elect RRSP deferral every year — or that your RRSP needs foreign-trust paperwork — they’re working from rules that have been dead for over a decade. (This exact stale claim still circulates widely; it’s a useful litmus test for whether a source is current.)

Two caveats keep this from being a free pass. The automatic deferral covers tax on internal growth — your RRSP still counts toward the account-reporting thresholds, so the disclosure side is unaffected. And it’s specific to the retirement plans the guidance covers; it doesn’t extend to the TFSA or anything else by analogy.

The takeaway

The treaty is a toolkit, not a force field. It works extremely well when applied — on a filed return, by someone who knows which of its provisions survive the saving clause, which accounts it covers, and where the reporting system runs on separate rails. It does approximately nothing for the person who invokes it from the couch as a reason not to file.

If your cross-border confidence rests on the sentence “there’s a treaty,” the right next step isn’t panic — it’s a return prepared by someone who can show you, line by line, where the treaty is actually doing its work for you.

This article is general education, not advice for your specific situation. How the treaty applies to your facts — and what it leaves untouched — is exactly the assessment a qualified cross-border professional exists to make.