Claims
What the journal concludes, and what each conclusion rests on. Outcome claims come first; claims about US intent are part of a labelled thought experiment and are listed separately.
Outcome (21)
What is being built, where it goes and who pays.
- On the current course, intended or not, both countries win from Canada's resource build-out. The US gets more supply, and the new southbound capacity arrives first, from 2027. Canada gets a bigger resource base and the capacity to sell elsewhere, which shifts bargaining power toward it.
- What Canada pays is set against what it gains: the capital, a lower tax take, the permitting and First Nations fights, a build-out that is almost all raw commodities, and the trade war itself. Counted together against 2024, Canada comes out ahead in 2034 unless about 57% of today's tariff damage is still in place, but the decade as a whole leaves it behind; how much it keeps after that depends on how long the tariffs last and whether it uses its new leverage to capture more value.
- We put 'faster, and more volume reaches the US' (this outcome or a stronger one) at about 85%, the figure from the journal's volume model, counting oil, gas, power, uranium, potash and metals: about 89% if Canada gets a full aluminum exemption, about 76% if the 50% aluminum tariff lasts. It is checked in 2034 against 2024, and counts only if the yearly value reaching the US, at fixed prices, is more than about US$4 billion above 2024, about 3% of that year's total, so that ordinary year-to-year swings don't count; electricity is compared at normal water levels in both years. The southbound volume is largely committed already: an oil pipeline expansion with a final investment decision and new uranium mines.
- Planned new oil pipeline capacity to the US adds up to about 730,000 barrels a day. Only 150,000 of it (Enbridge Mainline Phase 1, due in 2027) has a final investment decision. Most of the rest is South Bow's Prairie Connector (about 550,000 a day, 465,000 of it under 20-year shipper contracts, decision targeted for mid-2027, service around 2029). Enbridge postponed its Phase 2 (another 250,000 a day) on July 31, 2026, because producers would not commit.
- If everything planned is built, Canada's oil export pipeline capacity grows about 1.4 times (from 5.02 to 7.05 million barrels a day, our own sum of the capacities of existing and planned pipelines), and the share that reaches the coast rises from about 12.5% to about 27%. That is an upper bound, because some coast oil goes to US West Coast refineries. The biggest coast addition, the West Coast pipeline (1 million barrels a day), would come around 2032-34.
- If everything planned is built, LNG export capacity roughly triples (from 14 to about 45 million tonnes a year), and LNG would carry about 40% of Canada's gas exports overseas. LNG Canada's Phase 2 took its final investment decision on 29 September 2026: two more trains, doubling the plant from 14 to 28 million tonnes a year, about C$33 billion, with operations expected in the early 2030s. Gas sent to the US by pipeline may dip from 8.6 to 8.0 billion cubic feet a day by 2027, but the journal's model still expects more gas to reach the US by the mid-2030s, because it assumes most gas for new LNG comes from new production.
- Every signed LNG sales contract we found is for markets outside the US: BP (Woodfibre LNG), ExxonMobil's Asia-Pacific unit (Cedar LNG), and Shell, TotalEnergies and Uniper (Ksi Lisims, which has no final investment decision yet). SEFE's and Santos's deals with Ksi Lisims are non-binding.
- The long-term oil contracts are pipeline shipping commitments that Canadian producers pay for whether or not they use the space, not sales contracts with US buyers. Only Enbridge Mainline Phase 1 has a final investment decision; South Bow's Prairie Connector has 20-year shipper commitments but no decision yet, and producers would not commit to Enbridge's Phase 2.
- Canada is the largest single source of uranium for US reactors, but not by much, and its biggest producer's future sales are shifting overseas. Canadian uranium was 32% of what US reactor operators bought in 2025 (Kazakhstan: 28%) and 37% of what EU utilities received. Cameco's signed long-term sales were 38% to the Americas (the US, Canada and Latin America) at the end of 2025, down from 64% in 2022, against 40% to Europe and 22% to Asia, which includes a C$2.6 billion sale to India. Two new Saskatchewan mines, NexGen's Rook I and Denison's Phoenix, got construction licences in 2026 from the federal regulator under the ordinary process, not the new fast-track laws.
- New electricity export lines run south, not overseas. Two lines, to New England and New York, entered service in 2026 under contracts signed before the tariffs (about 20 TWh a year). But they mostly make up for sales Hydro-Québec is losing elsewhere: comparing a normal water year with a normal water year, Canada's energy regulator puts electricity exports at about 59 TWh in 2024 and about 57 TWh in 2034, so in the journal's model power adds about nothing to what reaches the US, like for like. The model doesn't forecast rain and snow. One caveat: 2024 was actually a drought year (Canada exported only about 36 TWh), so if 2034 is a normal year, the check in 2034 will make electricity look about US$2 billion a year better, for reasons that have nothing to do with building. The Churchill Falls deal adds no new line to the US; its power goes mainly to Quebec.
- About half of Canadian potash already goes overseas. Half of the new export terminal capacity (about 10 million tonnes a year) is at a US port, Longview in Washington state, which Nutrien chose on cost alone over Ottawa's objections.
- None of the new export routes is tied to a refinery, upgrader, smelter or steel plant; only two container terminals (Roberts Bank Terminal 2 and Contrecoeur) are built for manufactured goods. A few smaller critical-minerals processing projects get public support, such as Teck's Trail operations and a cobalt refinery in Ontario, but the build-out mostly fits the pattern of a raw-materials supplier.
- Canada's export diversification so far is more than half gold, and most of the rest is raw materials. Statistics Canada's headline figure, which includes gold, put the US share of goods exports at 69.8% in August 2026, up from 66.1% in July (balance-of-payments basis, seasonally adjusted). Leaving out gold sent to the UK, our calculation from customs data puts the US share at 79.1% in January-August 2024 and 74.6% in January-August 2026, and it has stayed around 74-75% in most months since April 2025 (monthly range about 72-78%). Of the rise in non-US exports between those two periods, about 52% is gold sent to the UK, and about two-thirds of the rest is raw materials. These are values, not volumes, and they say nothing about Washington's intent.
- Our count of comparable US deals, combined with the main forecast's odds (about 85%), puts more volume reaching the US plus locked-in preferential terms for it at about 23%, within a wide range (about 7% to 39%), using the middle of three bars for what counts. Nothing is locked in yet. New US equity in Canadian gas and LNG is limited to KKR and Apollo's indirect 29% of Enbridge's Westcoast pipeline and MidOcean's roughly 5% of LNG Canada, and Canada refused the US request for a right of first refusal on critical minerals in the August 2026 talks.
- All US ownership found in Canadian build-out assets is private, not government. US government stakes exist only in Vancouver-based companies whose mines are in the US (Trilogy Metals in Alaska, Lithium Americas in Nevada). Churchill Falls is closed to US capital.
- What the build-out costs the US is leverage, not supply. US leverage rested on Canada having few other buyers: in 2025 about 90% of Canadian crude exports by volume went to the US, the coast took 12.5% of oil export pipeline capacity, and LNG exports had barely started. The loss comes late: new US-bound oil capacity arrives first (2027-29), while most of the capacity to sell elsewhere comes around 2034.
- New resource regions, with no output today, open only through projects that depend on the new fast-track laws: the Ring of Fire roads in Ontario, Grays Bay in Nunavut, Sept-Iles iron ore in Quebec and the Mackenzie Valley Highway in the Northwest Territories.
- Canada's build-out was not obvious in advance. The US Energy Secretary said new Canadian pipelines were hard to build and did not expect the oil trade to change; conservative US analysts expected only a Conservative government to unleash oil and gas; and after Northern Gateway and Energy East fell through, 'Canada won't build' was a reasonable default. Some analysts foresaw Canada selling elsewhere, but in the US public record we found no one who predicted it would build fast and keep supplying the US. It is a surprising result that needs explaining.
- Europe is a real but small diversification channel, made of defence and resources. The binding items are Canada's agreement to join the EU's SAFE defence-procurement programme, the CPTPP trade pact now in force with the UK, Uniper's LNG contract and Germany's purchase of Lockheed Martin Canada's CMS 330 combat system; the rest (EU associate membership, the German submarine choice, SEFE's LNG deal) is non-binding. Canadian aluminum hit by US tariffs also went to the Netherlands and Italy in 2025. The EU took about 5.8% of Canada's goods exports in January-July 2026.
- On the current course, intended or not, Canada's resource build-out pays for both countries; the trade war doesn't. The US gets more supply, and the new southbound capacity arrives first, from 2027: it gets more supply either way; counted as income, the build-out pays it a little, but the trade war costs it more, so it comes out ahead in 2034 only if nearly all of today's tariff damage is gone. Canada gets a bigger resource base and the capacity to sell elsewhere, which shifts bargaining power toward it. By 2034, Canada comes out ahead unless most of today's tariffs are still in place; but the decade as a whole leaves it behind, and the longer the tariffs last, the less of it comes back.
- Counted as Americans' income, Canada's build-out adds about US$1.5 billion a year by 2034, mostly US owners' profits from Canadian oil, gas and LNG, but the trade war costs more. The US comes out ahead in 2034 only if about a tenth or less of today's tariff damage is still in place (Canada: unless about 57% is), which happens in about 14% of our runs if the tariffs fade the way past US tariffs did. Either way the amount is small for the US, about 0.006% of its economy in 2034. This counts income; the forecast that more Canadian supply reaches the US counts volume.
Mechanism (25)
How the pressure turns into building.
- US trade pressure since 2025 gave Ottawa the political cover to clear permitting and political blockages, producing permanent laws (fast-track permitting and investment tax cuts) that unlock a second wave of output.
- US tariffs spare raw resources and hit processed goods. Energy was taxed at 10% or exempt in every round, the potash rate was cut to 10%, and the 2026 Section 338 tariffs exempt energy, potash, critical minerals and fish, while steel, aluminum, copper, autos and lumber took the heaviest tariffs.
- Canada stands out not for how much it pays in US tariffs but for the tools used against it. Its effective tariff rate is at the low end of all trackers, and the Section 338 tariffs cover about 5% of its shipments. What is unusual is the first-ever use of Section 338, import bans, exclusion from US government purchasing, and trade talks cut off twice. Brazil shares part of that. Canada's retaliation and its deal with China coincide with the gap, but neither can be shown to cause it: countries that didn't retaliate, such as Brazil, India and Switzerland, were hit as hard or harder.
- Canadians' rally against the US shows in behaviour, not just polls: return trips from the US fell 25.4% in 2025 (all modes; car trips were still 27% below 2024 in August 2026), spending on US visits fell by C$3.3 billion, boycotts spread, and two premiers called snap elections citing the US threat. This shows the effect, not what caused it.
- The rally works unevenly: in Quebec it feeds sovereignty and 'control of our territory', not a blank cheque for fast-tracking, and in BC tariffs are a top issue for only 15% of voters.
- At home, new pipelines are sold as a way to reach buyers outside the US: in an Angus Reid poll, 55% picked 'sell oil outside the US' as a persuasive reason (several answers were allowed). So the link from the build-out to more output flowing south has to come from contracts and tariff design, not from the rally.
- Winning parties rarely spelled out their controversial fast-track tools before the vote: in our own count of 41 post-election actions by five tariff-period winners, checked against their platforms, 2 of 19 controversial actions were clearly promised, against 14 of 22 routine ones. The rally widened what could be done after the vote. This is a way to read platforms, not proof of concealment, and it supports 'Ottawa opportunistic' only weakly.
- C-5, the 2025 federal fast-track law, is durable, and this rests on recorded votes: no Conservative voted against the programming motion setting a fast timetable for it in the House (305-30), its second reading (304-29) or either part of its third reading (335-1 and 306-31), and during the 2025 campaign the Conservatives promised the same kind of system, faster ("One and Done").
- Nothing yet tests whether C-39 (the bill making the federal fast track permanent) and the Productivity Mega Deduction (a tax write-off for new investment) would survive a change of government: the Conservatives have not voted on either and have not said how they will vote on C-39, so claims that they are durable rest on statements.
- Cross-party support is for the projects, not the fast-track instruments: BC's Conservatives voted against both BC fast-track laws even though they promise more building.
- The weak point for durability is public money, not the permitting law. The Conservatives attack state capital (Trans Mountain helping finance the West Coast pipeline, the Canada Strong Fund, the Pathways deal with Alberta) as corporate welfare. So privately financed projects that send resources south should survive any government, while publicly financed projects to sell elsewhere are exposed after the 2029 election.
- The Pathways deal commits Alberta to support oil production growth for the West Coast pipeline but is non-binding, and that already cost private volume: producers cited it when Enbridge Mainline Phase 2 was postponed.
- A law passed is not a power switched on. BC's narrow Bill 14 was in force in 33 days and is being used for the North Coast Transmission Line, but only 1 of Bill 15's 48 sections was in force in late September 2026, 16 months after it passed, Quebec's Bill 5 lapsed, and Ontario has not yet designated any special economic zone. So the test of whether the build-out is locked in must ask whether powers are brought into force and used.
- The one-year review clock is soft (study time is excluded and extensions are allowed). What slows the build-out is First Nations lawsuits and the courts, whatever Washington intended; but in the record so far they delay projects and send approvals back more often than they stop them, and benefit agreements have ended several of the suits.
- Capital is waiting, not flowing. The summit's C$1 trillion target and the banks' and pension funds' pledges are statements with no money moved; producers held back Enbridge's Mainline Phase 2; Ksi Lisims has no final investment decision; and the UAE money hasn't been deployed. The clear exception is LNG Canada Phase 2, which took its final investment decision on 29 September 2026: two more trains, doubling the plant from 14 to 28 million tonnes a year, about C$33 billion, with operations expected in the early 2030s. The registered test 'Capital follows the law' is open.
- Public money carries the risk the market won't: the deficit-financed C$25 billion Canada Strong Fund, Trans Mountain as a Crown asset, Alberta's Pathways supports, C$10 billion of Indigenous loan guarantees and about C$6 billion for the trades. This fits 'the state is creating the volume', but plain hedging predicts it too, so it says nothing about Washington's intent.
- Official forecasts show no boom (about 1.1% growth in 2026, from Finance Canada and the Parliamentary Budget Officer), and much of the record foreign direct investment in 2025 was takeovers and reinvested earnings rather than new investment.
- US takeovers of Canadian producers and miners (Parkland, NuVista, New Gold) say nothing about Washington's intent: prices depressed by the tariffs invite buyers whatever the motive.
- The build-out only makes sense if the builders expect continued access to the US market: C-39's rail-competition clause for the Prairies takes effect only 'when there is a reliable trading relationship with the United States', and the firm new oil capacity runs south.
- The Productivity Mega Deduction would cut Canada's overall effective tax rate on new business investment from about 13% to 6.4% (US: 16.9%). It names mining property, pipelines and LNG, dropped the low-carbon condition for LNG, and applies to assets bought from September 15, 2026, though it is still a draft, not law. The tilt toward resources is clearest here, not in C-39.
- The idea of continental integration (Canada the resources, the US the demand, Mexico the assembly) has one anchor with legal force: Mexico's tariffs of up to 50% on Chinese imports from January 1, 2026. US pressure targets Chinese content coming through Mexico, which was named in a 2026 US trade investigation into excess capacity, while Canada was not.
- The Churchill Falls and Gull Island deal complicates the thesis. It began before the tariffs (negotiating mandates in 2023-24, a memorandum of understanding on December 12, 2024). In the tariff era it got bigger and politically safer (C$10 billion in federal financing, a referendum dropped) but not faster: every milestone slipped. The August 2026 agreement sets terms for binding contracts due by the end of 2026, and it adds no new power line to the US.
- Quebec overrode opposition to protect existing industry (Bill 93 for the Stablex waste site, passed under closure; Bill 11 with relief for the Horne smelter) but avoided a fight over new projects: its fast-track Bill 5 lapsed. The only party that pledged in writing to bring back a Bill 5-style law, the outgoing CAQ, won no seats in the 5 October 2026 election, which left the Parti Québécois short of a majority with 59 of 127 seats (preliminary results).
- Consultation gives First Nations no veto. Courts send approvals back for more consultation but rarely end projects: Trans Mountain's approval was quashed in 2018, granted again in 2019, and the pipeline was built. A nation that objects usually ends with an agreement or with an approval over its objection, so a nation that opposes and later signs reveals nothing about its motive. Provinces and towns holding the same lever end the same way.
- In the record so far, resource projects facing organized First Nations, Métis or Inuit opposition reached construction about as often as those without it (approved: 4 of 7 vs 1 of 2; under review: 2 of 4 vs 2 of 8); tiny, selected samples.
Ottawa (4)
How the Canadian government is playing it.
- Ottawa is opportunistic, following 'never waste a good crisis': it used the mandate and political cover of the tariff era to pass fast-track and tax tools. This judgment says nothing about Washington's intent.
- Ottawa is also in no hurry: it didn't ask for the crisis and isn't rushing to end it. It took a go-slow approach to the talks before July 2026, and walked away from the tentative deal announced on August 18, 2026, three days later.
- Some resource terms entered the talks from the Canadian side: at the White House on October 7, 2025, Carney pitched Keystone XL and more oil exports. In the foresight thought experiment, a blind grading found this leans slightly toward 'foreseen and wanted': under pressure, Canada offered more oil, and US officials welcomed it. The lean is slight, because plain self-interest would welcome the offer too.
- The Canada-BC Cooperative Prosperity Agreement is paid consent: federal money (C$3.9 billion toward the North Coast transmission line, C$500 million for the Red Chris mine expansion and up to C$3 billion for the Massey Tunnel replacement) and a guarantee of the tanker ban turned a province that opposed the pipeline into a conditional host. It needs no US involvement, so in the foresight thought experiment it fits plain self-interest and 'foreseen and wanted' equally.
Method (5)
How the evidence is handled.
- Evidence is ranked by type: actions with cost or legal force (tariffs, laws, votes, contracts, pipes, money spent, prices) outweigh formal documents, which outweigh statements; statements count only as a pattern or consistency check.
- Each item is graded against each reading: does it fit one reading better, or both equally? An item that 'foreseen and wanted' and plain self-interest predict equally is called non-discriminating: it can still matter for the outcome, but says nothing about intent.
- Output is split into projects that don't need the new fast-track laws and projects that depend on them. Only the second group can say anything about the effect of the tariffs, and neither says anything about Washington's intent.
- Output is measured as export capacity for now and as actual volumes later. The journal's volume model adds different commodities together at fixed prices, so a price rise never counts as growth; trade values must be split into price and volume before they are used for the outcome.
- What would count against the project's readings. Against 'both sides win': on the US side, capital never committing, or the second wave of output, the part that depends on the new laws, being sold mostly overseas while flows south stall; on Canada's side, most of today's tariff damage (about 57% or more) still in place in 2034. Against the foresight thought experiment: sustained escalation after the laws pass that damages the flow south, resources tariffed as hard as everything else, no resource terms ever appearing in a deal, or documented US support for lawsuits or provincial deals that slow the build-out.
Tested hypotheses: the thought experiment
Suppose Washington foresaw and wanted this outcome. What would we expect to see, and do we see it? These claims belong to that deliberately contrarian test, shown with the rival readings, not as the author’s belief. How the test works.
Tested hypothesis: intent (16)
- In the foresight thought experiment, the readings carry no odds: why Washington acted can't be observed, so no event could settle a number. What the record shows so far: no test has yet told 'wanted' apart from plain self-interest, and the events that do separate the readings lean toward self-interest, 54 to 7, in a blind grading.
- In the foresight thought experiment, nearly all statements about Washington's intent fit 'foreseen and wanted' and plain self-interest equally: Greer's 'still getting what we need', Hoekstra's 'economic fortress', Navarro noticing the rally, and Trump posting 'we don't need Canada' hours before saying the US needs Canadian aluminum 'badly'.
- In the foresight thought experiment, most hard US actions also fit 'foreseen and wanted' and plain self-interest equally: tariffs that spare raw resources (a pattern applied worldwide), the replacement tariffs after the US Supreme Court struck down the emergency tariffs in February 2026 (under laws that apply to all countries), the Bridger pipeline permit reusing Keystone XL assets (a private project US refiners want), and US takeovers and private-equity stakes.
- In the foresight thought experiment, a timing test found US tariff actions near Canada's 14 steps to lock in the build-out at chance level (9 against 8.3 expected), and US provocations rarer than chance. The test has little power. It rules out a scripted campaign, but says little about 'foreseen and wanted', which needs no timing.
- In the thought experiment's first foresight test ('Was it foreseen?'), we found nothing tied to administration circles predicting that pressure would push Canada to permit, build and export; the one insider asked directly, Energy Secretary Wright, dismissed the idea. Analysts outside the administration did warn that tariffs could push Canada to sell elsewhere. The result counted only as weak evidence for plain self-interest, because private foresight would not appear in public sources.
- In the foresight thought experiment: before Canada's response, what US voices wanted was a cooperative build-out framed as partnership (Project 2025's call for a hemispheric energy policy 'with Mexico, Canada', cross-border permitting bills, partnership resolutions in Congress), not one produced by pressure. None of it ties the build-out to tariffs.
- In the foresight thought experiment: the pitch to build more and send it south came from Canada (Alberta's 'Fortress North America' at the CERAWeek energy conference, March 12, 2025), and the only statement of the mechanism before C-5 that we found came from a Canadian, Jamie Tronnes, who said the pressure was possibly 'the kick in the butt that we need'.
- In the thought experiment's second foresight test ('Asymmetry'): US actions protect the flow south (raw goods spared, the Bridger pipeline permit, no objection to pipelines toward the US) and threaten Canada's efforts to sell elsewhere. But we have found no protection that costs the US something plain self-interest wouldn't pay, so the pattern fits 'foreseen and wanted' and plain self-interest equally.
- In the foresight thought experiment, US resistance to Canada selling elsewhere has stayed at threats and demands. Two threats (a 100% tariff over Canada's China deal, 'very serious tariffs' over closer EU ties) were not carried out, and in the August 2026 talks Canada refused US requests for a right of first refusal on critical minerals and for limits on its other trade deals. The US has taken no step against Asian or European buyers of Canadian crude, LNG or uranium.
- In the thought experiment's third foresight test ('Rerouting constraint'), no US action yet limits Canada's ability to sell elsewhere: no measure against its coast projects or named overseas buyers, and no clause on sales to other countries. The closest items are demands, not actions: a right of first refusal on critical minerals and limits on Canada's other trade deals, both refused in August 2026. If the US keeps tolerating the rerouting, that fits 'foreseen and wanted' slightly better than a plain self-interest focused on leverage.
- In the foresight thought experiment: the US request for a right of first refusal on Canadian critical minerals, first reported from unnamed sources in August 2026 and later cited by Carney among the demands Canada refused, claims existing output rather than asking Canada to build more, much like the 2025 US-Ukraine minerals deal. Reports of 'capped prices' come from an unverified viral list. The request fits 'foreseen and wanted' and plain self-interest equally.
- In the foresight thought experiment: the US Treasury lifted sanctions on Belarus's state potash producer in March 2026, and Trump said Belarus 'would like to sell it for a much lower price'. This complicates the idea that the US is securing Canadian supply, and fits plain self-interest: shop for the cheapest input.
- In the foresight thought experiment: the US has not used a tariff threat to extract specific resource commitments from Canada. The one resource demand Canada refused, a right of first refusal on critical minerals in August 2026, was not enforced. Other resource items in the talks, such as finishing the Keystone XL pipeline (which trade adviser Peter Navarro said the US wanted from the deal), were ones Canada had itself offered. By contrast, a tariff threat got Mexico to deliver treaty water within days in December 2025. This counts against "a worked-out plan with specific targets".
- In the foresight thought experiment: the official US reasons for the tariffs on Canada are lists of grievances with no resource request: fentanyl (the first tariff order), Canadian retaliation (the 35% tariff letter), Canada's auto surtax, alcohol boycotts and dairy quotas (the Section 338 proclamations), and the 2026 US trade-barriers report. The Canada-specific grounds shifted about three times. They show what the US chose to argue, not why it acted.
- In the foresight thought experiment: US government minerals money is not aimed at Canada. Pentagon grants for projects in Canada total about US$82 million since 2023; they began under Biden, were mostly co-funded by Ottawa and carry no US priority clauses, and the same tools went to US, Korean and Australian companies.
- In the foresight thought experiment, the deal-timing test ('No deal before the law') works in one direction only. A deal or easing within 120 days after the later of two dates, C-39 becoming law and the Mega Deduction becoming law, would support 'foreseen and wanted'. Continued pressure that still spares raw goods would not count against it; only sustained escalation that damages the flow south would.
Tested hypothesis: rival reading (6)
- Annexation as a working strategy is contradicted by its effect: support for joining the US fell from 13% to 9% (Léger, same question, December 2024 to August 2026) while the provocations ran, and the only annexation-flavoured acts with legal force are cheap and symbolic (the order renaming Lake Ontario 'Lake America' for US federal use). This counts against annexation as a strategy, not against annexation as a wish.
- Fragmentation (breaking Canada up through its provinces and separatists) and the thought experiment's 'foreseen and wanted' build-out need opposite things from Ottawa: a wanted build-out needs a strong Ottawa that passes C-39 and overrides holdouts, fragmentation a weak one. So the two can be told apart.
- Fragmentation is a live rival. One act counts toward it under our test, and none weakens Ottawa's ability to legislate. The act: three State Department meetings in 2025 with the Alberta Prosperity Project, a separatist group, which the US government has confirmed. It says they were staff-level, with no senior officials and no commitments, and that there will be no further meeting. Our test counts any confirmed US contact with separatists, so these count, although nothing in the record shows US support behind them. One approach is logged but not counted: in spring 2026 Quebec was approached from the US side for a separate deal on USMCA-linked sectors, which Radio-Canada described as an attempt to go around trade talks with Ottawa. Premier Christine Fréchette confirmed the approach in September 2026 and said she refused at once, without saying who made it. Québecor media, citing two Quebec government sources, said the US embassy in Ottawa made it, and CP24 reported that it touched tariffs including aluminum; neither detail is confirmed. The test counts such an offer only once we know who made it and on whose behalf.
- Against fragmentation as the working US plan: the US kept humiliating Canada as a whole (the 'Lake America' order, the burst of attacks in August and September 2026), which rallies Canadians behind Ottawa. A strategy of weakening Canada through its provinces would need those attacks to stop.
- No money linked to the US government or the US right was found behind any First Nations case we examined (not observed in public sources, not proven absent). US foundation money predates 2025 and opposes oil. Every US dollar we traced near a Canadian resource project since January 2025 goes into building it, including benefit agreements with Ksi Lisims LNG, a project led by Houston-based Western LNG, after which two nations dropped their lawsuits. This fits the thought experiment's 'foreseen and wanted' reading slightly better than fragmentation, but only as weak evidence.
- Fragmentation would work against itself through the courts: First Nations treaty concerns are why an Alberta court struck down the separatist referendum petition in May 2026 (Alberta is appealing), so funding treaty suits would strengthen the main legal barrier to Alberta separatism.
Nothing matches these filters.