Divide, then annex: could the U.S. take Canada province by province? A sourced simulation

Five key judgments

Our first analysis found that the U.S. is very unlikely to absorb Canada as a whole. This follow-up tests the strategy analysts single out as the least implausible alternative: encourage Quebec, Alberta and Saskatchewan to leave Canada first, then admit them one by one as separate states. We walk the route stage by stage. Every stage rests on a law, a poll, a peer-reviewed study, an official report or a named expert; the probabilities attached to each gate are our own judgments, labelled as such.

1. Splitting Canada first is the only annexation route with a coherent political logic. Admitting provinces separately avoids creating one giant Democratic-leaning state; the Lowy Institute calls separate admission “the more acceptable scenario” ^[16]. Alberta and Saskatchewan would likely send Republicans to Congress, which flips the partisan objection that sinks the whole-country version ^[5]. Confidence: Strong.
2. The first stage is already visible: senior U.S. figures have publicly courted Alberta, separatists have met State Department staff three times, and researchers have tracked foreign disinformation targeting the Alberta debate, about one fifth of it from U.S. sources ^{[18] [19] [29]}. What is missing is any sign that this pressure is working. Confidence: Strong.
3. Every later stage runs into a hard gate. No region has ever seceded from a well-established democracy by referendum ^[54]; a binding vote in Alberta cannot legally be held before First Nations are consulted ^[31]; the Clarity Act lets the House of Commons reject an unclear question or majority ^[49]; and Congress has refused statehood even to Puerto Rico after a 97% vote ^[70]. Confidence: Strong.
4. The strategy has a timing problem. The fastest legal path puts a binding Alberta vote in 2027 or 2028 ^[32], and the Parti Québécois has promised no referendum until after Trump leaves office in January 2029 ^[45]. The pieces could only be ready after the administration that wants them is gone. Confidence: Moderate.
5. Multiplying the gates, we put the chance that any of the three provinces becomes a U.S. state by 2036 at about 0.5%, and all three at well under 0.01%. Alberta independence alone is more plausible (about 1%), and Quebec independence more plausible still (about 5%), but in both cases independence is far more likely to end in a new country than in a new star on the U.S. flag. Confidence: Moderate.

Why “divide, then annex” is the route analysts take seriously

Admitting Canada as one state would create the most populous state in the Union and, on every published model, a reliably Democratic one ^{[3] [4] [5]}. That is the main reason our first analysis judged the whole-country version self-defeating. The Lowy Institute’s review of the practical obstacles makes the alternative explicit: joining the provinces as separate states is “the more acceptable scenario,” and it names Quebec, Alberta and Saskatchewan as the provinces where resistance to Ottawa runs deepest ^[16]. The Center for Strategic and International Studies (CSIS) in Washington goes further for Alberta: it judges that the “gravitational pull” toward U.S. annexation of an independent Alberta, as a state or as an unincorporated territory like Puerto Rico, “would be considerable,” mainly because of its energy ^[17].

The partisan arithmetic explains why. Using Statistics Canada’s July 1, 2026 estimates ^[72] and today’s U.S. ratio of about 761,000 residents per House seat (our calculation), the three provinces would carry the following weight.

Two senators per state is where the logic bites. Admitting Alberta and Saskatchewan would add four likely-Republican senators and about nine House seats. Partisan math has decided statehood fights before: in 2019 Senate Majority Leader Mitch McConnell pledged to stop statehood for Washington, D.C. and Puerto Rico because it would give Democrats four more senators ^[71]. The same logic would run in reverse for the Prairies.

History supplies the mechanics. The United States has acquired foreign territory by statute only twice, Texas in 1845 and Hawaii in 1898, and in both cases annexationists used a joint resolution after failing to win the two-thirds Senate vote a treaty needs ^[66]. Texas is the closest template for a breakaway province: an independent republic for nine years, admitted by a simple-majority resolution that let it keep its public lands and its own debt, and even allowed it to split into up to four more states ^[67]. Research on recognition adds the external accelerant: in Bridget Coggins’s dataset of secessionist movements from 1931 to 2000, two-thirds of great-power recognitions came before the parent state consented to independence ^[63].

So the hypothetical playbook has six stages: (0) pressure and influence, (1) a provincial mandate to start the process, (2) clearing the legal gates, (3) winning a binding vote, (4) negotiating the exit, (5) international recognition, and (6) admission by Congress. Each section below tests one stage.

Stage 0: pressure and influence (already under way)

A divide-first strategy starts by raising the cost of staying and lowering the perceived cost of leaving. Pieces of that are on the public record. In January 2026, Treasury Secretary Scott Bessent called Alberta “a natural partner for the U.S.” and said “I think we should let them come down into the U.S.” ^[18] Alberta Prosperity Project (APP) leaders met State Department staff three times in 2025 and discussed switching to U.S. currency and border security; a senior official said the meetings were staff-level and no commitments were made ^{[19] [20]}. One APP figure said proposals included fresh water, mining, land for a railway to Alaska and a pipeline to the Pacific, with oil royalties as compensation; another APP leader disputed that account ^[21]. The APP also claimed talks about a US$500-billion credit line, which the U.S. ambassador denied: “we’re not meeting with the separatists and strategizing this at all” ^{[22] [23]}.

The information space is contested too. A report titled National Unity Under Threat found Russian and pro-Trump U.S. actors amplifying Alberta separatist narratives ^[28]. Researchers tracking the debate counted 149 foreign disinformation items in two weeks of July 2026, about four in five linked to Russia, and 147 in early August, about 80% Russian and 20% U.S.-linked ^[29]. The CSIS director warned that the referendum is “rife for amplification” by foreign actors, while noting that open commentary by officials or influencers is not, legally, foreign interference ^[30]. On October 7, Elon Musk posted that Quebec “should separate” and that “Alberta and Saskatchewan should separate, too”; the premiers of both Prairie provinces publicly said no ^{[24] [25]}. In Quebec, the outgoing government said U.S. officials had approached it in the spring about a separate deal on trade sectors, which it refused before contacting CSIS ^[26].

Assessment: stage 0 exists, but it is diffuse, deniable and so far ineffective. Western alienation has actually fallen to a four-year low, and only 14% of Western Canadians would vote to separate ^[39]. A University of Saskatchewan longitudinal study finds that alienation in Alberta and Saskatchewan peaked after the 2019 federal election, not in 2026 ^[40].

Stages 1 and 2: Alberta’s mandate and the legal gates

Alberta is the keystone of the strategy, and its first test comes on October 19, 2026. Question 10 asks whether Alberta should remain a province or whether the government should “commence the legal process required under the Canadian Constitution to hold a binding provincial referendum” on separation ^[31]. It is non-binding by design ^[32]. The polls are consistent: Leger 70% remain and 22% for starting the process; Janet Brown 72–23; Ipsos 74–14; The Writ’s average 70–22 ^{[11] [9]}. Even on a stand-alone separation question, 30% say they would leave, and 41% of them say they mainly want to send “a message to Ottawa” ^[33].

Within the governing party the picture is different: UCP voters split 48% for starting the process and 43% for remaining (Leger), and a CBC-commissioned survey found 53% of UCP supporters would choose the separation option ^{[11] [35]}. A divide-first strategy would therefore need the movement to grow beyond one party’s base. Pollara found separatism to be “overwhelmingly” a Conservative phenomenon: 32% of federal Conservative voters in the West would vote to leave, against 3% of Liberals and 2% of New Democrats ^[39].

Suppose Option 2 wins anyway, now or in a later vote. The legal gates come next. First, the courts: a judge found that a binding separation referendum cannot be held before a constitutionally required consultation with First Nations on treaty rights; Alberta is appealing, but until then a binding vote “is not legally possible,” in the government’s own words ^{[31] [12]}. The treaty nations are not neutral parties. Treaty 8 Grand Chief Trevor Mercredi put it this way: “The province is a tenant on this land, not the landlord” ^[51]. A coalition of chiefs argued that “neither Alberta nor Albertans have the right to secede, unilaterally or otherwise,” and one chief warned that the petition risked “opening the door to Alberta becoming the 51st state” ^[52].

Second, the Clarity Act. Within 30 days of a province releasing its question, the House of Commons must decide whether it is clear; a question about a mandate to negotiate, or one that mixes in other arrangements with Canada, does not count ^[49]. The House must also weigh the views of Indigenous representatives, and later judge whether the majority is clear, looking at its size and the turnout ^[49]. Alberta Fact Check’s reading of the timeline: court challenges could push a binding vote into 2027, and possibly 2028 ^[32].

Stages 3 and 4: winning a binding vote and negotiating the exit

The base rate for stage 3 is close to zero. Research in the British Journal of Political Science finds that secessionists have never split a well-established democracy (at least ten consecutive years of universal suffrage) through a referendum or an election win ^[54]. A study of independence referendums in the Canadian Journal of Political Science finds that “leverage” referendums, held without the central government’s agreement, very rarely end in recognized statehood ^[55]. The Alberta government’s own technical paper notes that only one regional society, Western Australia in 1933, has voted to separate in a comparable way, and it did not succeed ^[56].

The committed core is also small. Abacus estimates that only about two-thirds of Option 2 voters would actually choose to leave, an 18% core ^[10]. Ipsos found that among Albertans who would vote to separate, 45% say “definitely” ^[34]. A clear majority would require roughly tripling that core.

If a clear majority did vote to leave, the Supreme Court’s Secession Reference would require all parties to negotiate in good faith; a referendum alone has no legal effect, and the Court left the talks to politicians rather than judges ^{[15] [50]}. The bill would be large. The Alberta government’s technical paper estimates that an independent Alberta would need to spend about $144 billion a year to replace federal transfers, services and programs, just over $60 billion more than the province spends today ^[57]. The APP’s draft plan claims the opposite, a surplus of $23.6–45.5 billion a year ^[58]; economists say that projection depends on assumptions about oil prices and federal services that are contested ^[60].

Geography is the hardest constraint. Most Alberta crude reaches the U.S. through pipelines that cross other provinces; the three lines that run straight from Alberta into the U.S. carry less than a tenth of Canada’s exports, and no gas pipeline runs directly from Alberta into Montana ^[59]. University of Calgary economist Trevor Tombe warns that Ottawa could charge fees on oil crossing Canada and that serious separation talk could push people, firms and capital out, as happened in Montreal ^[60]. On money, the C.D. Howe Institute lists three options: keep the Canadian dollar, create an Alberta dollar, or adopt the U.S. dollar. It notes that dollarization would leave Alberta without a lender of last resort, and that Czechoslovakia’s transitional currency union collapsed within weeks ^[61]. Here the divide-first strategy finds its real lever: a landlocked, newly independent Alberta would depend on Washington for market access, which is exactly the kind of dependence that makes a later request to join plausible ^[17].

Saskatchewan: a follower, not a leader

Saskatchewan matters to the strategy mainly as Alberta’s companion. Its separatists say so: Saskatchewan Prosperity Project president Brad Williams says some members want to unite with Alberta if Alberta leaves, after more than 40 town halls ^[42]. The legal path is weaker than Alberta’s: provincial law allows only a non-binding plebiscite, triggered by a petition signed by 15% of electors ^[41]. Premier Scott Moe says his party supports “a strong and growing province of Saskatchewan with the united nation of Canada,” though he would “follow the law” on a plebiscite, and he answered Musk with a one-word “No” ^{[41] [25]}.

Opinion is softer than Alberta’s, not harder. Pollara found 20% of Saskatchewan residents would vote to leave, the highest in the West but far from a majority ^[39]. In April 2025, 33% said they would vote to leave Canada if the Liberals won (they did), either for independence or to join the U.S.; 20% chose the U.S. option specifically ^{[38] [73]}. Among Saskatchewan Party voters, 42% supported holding a referendum and 52% opposed ^[43]. Assessment: Saskatchewan does not move first. Its odds are best read as conditional on Alberta.

Quebec: the strongest secession case, the weakest annexation case

Quebec is the province most likely to leave Canada and the least likely to join the United States. The Parti Québécois took office on October 5 with 59 of 127 seats, a minority that needs another party to adopt a referendum question ^[13]. Its leader has promised no vote until after Trump leaves office in January 2029, and polls show about 70% of Quebecers oppose holding one ^[45]. Support for sovereignty sits at 29% Yes and 59% No (Pallas Data, August 2026) ^[44]. For scale: in 1995 the No side won with 50.58% on a turnout of 93.52% ^[74].

Even a successful Quebec referendum would point away from Washington. The PQ’s plan is for a sovereign republic that joins NATO and NORAD, with its own army of about 20,000 and defence spending of 5% of GDP by 2035 ^[46]. When a PQ meeting with U.S. officials was mooted in early 2026, it was cancelled, and Washington has kept its distance from the election ^[27]. Only 24% of Quebecers told Research Co. their province would be better off as a U.S. state, the second-lowest share in the country ^[47]. And Quebec’s own borders would be contested: scholars note that the northern territories added in 1898 and 1912, about two-thirds of the province, were never agreed to by the Cree and Inuit, who assert a right to stay in Canada ^[53].

Quebec’s real role in a divide-first strategy is indirect. CSIS (Washington) notes that an Alberta exit could give new impetus to a Quebec referendum, and the reverse is also possible ^[17]. A sovereign Quebec would also physically separate Atlantic Canada from Ontario. In January 2025, 22% of Canadians said they would vote for their own province to join the U.S. if Alberta or Quebec became independent ^[48]. That is the domino the strategy would be betting on; it remains a minority position everywhere.

Stage 5: recognition, the step Washington controls

Recognition is where a great power can do the most with the least. Coggins’s research shows external recognition can start “a cascade of legitimacy,” with two-thirds of great-power recognitions coming before the parent state agreed ^[63]. Recognition is legally discretionary, and practice after Kosovo shows how one recognition invites another: scholars judge that Russia’s 2008 recognition of South Ossetia and Abkhazia was “extremely unlikely” without the earlier Kosovo precedent ^[64]. A U.S. administration that recognized an Alberta declaration before Ottawa agreed would be using exactly this lever.

The costs are equally clear. Crimea shows the other end of the scale: annexation after a controlled referendum met broad non-recognition and sanctions ^{[65] [8]}. Recognizing a breakaway region of a NATO ally would also damage U.S. security directly: CSIS (Washington) warns that a fractured Canada would split NORAD’s warning and control across several jurisdictions, stall basing agreements and could leave gaps in the approaches to North America ^[17]. The current U.S. ambassador’s stated position is that Alberta’s status is for Albertans and Canadians to decide ^[22].

Stage 6: admission by Congress

The last stage is entirely American. New states are admitted by Congress under Article IV ^[6]; there is no single path, and requirements have varied widely ^[69]. A joint resolution, as for Texas and Hawaii, needs only simple majorities in both chambers ^[7]. The Hawaii case is a warning as much as a template: more than half of Native Hawaiians signed a petition against annexation in 1897, and Congress went ahead anyway ^[68]. Treaty nations in Alberta have already said the province cannot leave without them ^[51].

Would Alberta even ask? Support for joining the U.S. is well below support for independence: 24% in Research Co.’s January 2026 poll (against 31% for independence), 17% for Leger in 2025, 16% for Probe Research in May 2026, and 6% in Leger’s June 2026 pulse check ^{[36] [37] [11]}. The APP itself pitches an independent country with its own dollar, not statehood ^[58]. And Congress is slow even when a territory begs: after Puerto Rico’s 2017 vote, in which 97% chose statehood, one territorial-law scholar put the odds of admission “between zero and negative-10 percent” ^[70]. The likelier U.S. offer, if any, would be a free-trade or compact relationship, or territorial status without votes in Congress ^[17].

The simulation: gates, odds and the fastest possible timeline

We now chain the stages. Each gate is a probability conditional on the previous one being passed, set by us from the evidence above. Multiplying them shows how a route that looks plausible step by step becomes very unlikely as a whole.

Consistency check: our first analysis put the chance of a breakaway region joining the U.S. by 2036 at about 2%, a top-down judgment that also counted looser forms of U.S. control. Building the estimate bottom-up, gate by gate, gives about 0.5%. We treat 0.5–2% as the honest range.

The timeline exposes the strategy’s central contradiction. The patron that would push hardest for a breakup leaves office before any province could legally leave, and Quebec’s sovereigntists have tied their own vote to that patron’s departure ^[45]. A divide-first plan would need sustained commitment across several U.S. administrations, which nothing in the record suggests.

What Washington would gain, and what it would lose

The prize looks smaller than the slogan suggests. Alberta already supplied 87.4% of the Canadian crude exported to the U.S. in 2023, and Saskatchewan another 8.9% ^[62]; nearly all of it already goes south ^[1]. Annexation would change who collects the royalties, not where the oil flows. Fiscally, the U.S. has no equalization program, and the University of Calgary’s School of Public Policy finds that federal money flows to U.S. states through many separate programs instead; how a resource-dependent state would fare is uncertain ^[14].

The losses are concrete: a broken NORAD command structure and stalled defence projects ^[17], a precedent that legitimizes recognition-first secession everywhere ^[64], and the non-recognition and sanctions that followed the last annexation by a permanent Security Council member ^[65]. Most Americans are not asking for it either: only 7% want Canada as a state and 63% want it to stay independent ^[2].

Signals that would change these odds

Limitations. Polls use different questions (start a process, separate, join the U.S.), so we do not merge them. Several key claims about U.S.–separatist talks come from separatists themselves and are disputed by U.S. officials ^{[21] [22]}. Gate probabilities are judgments: a reader who raises every gate by half still ends near 3% for Alberta statehood. Evidence is current to October 15, 2026; the Alberta vote on October 19 is the first test.

What it means if you are moving west

Alberta had the fastest population growth of any province in the year to July 1, 2026 ^[72], and in every scenario we rate above 1% it stays a Canadian province through 2036. If you are planning a move, the practical questions are cost, distance and timing: see the full breakdown for Montreal to Calgary, Montreal to Edmonton or Montreal to Saskatoon, or browse every route in our Canada moving cost guide. Moving for a new job? Check whether you can deduct your moving expenses, then get a written quote.