On October 18, 2021, Albertans cast ballots on a question that asked whether section 36(2) of the Constitution Act, 1867 — the provision that commits Parliament and the Government of Canada to the principle of equalization — should be removed from the Constitution, and 61.7 percent of those who voted marked yes, producing a result that reverberated through federal-provincial relations and set the stage for the far more ambitious referendum scheduled for October 19, 2026. That 2021 equalization vote was not the first time Albertans had been invited to express their will directly on a matter of constitutional or political significance, but it was the vote that demonstrated most clearly how a provincial referendum could be deployed as a tool of intergovernmental pressure even when everyone involved understood that the result could not, by itself, change the constitution or compel any federal action. The 2021 vote established a template: the Government of Alberta would frame a question, submit it to voters on a date aligned with municipal elections to maximize turnout at minimal cost, and then use the outcome as political ammunition in negotiations with the Government of Canada. Understanding what happened in 2021 — how the question was framed, what the yes vote legally accomplished, and what it did not accomplish — is indispensable to understanding the constitutional architecture of the 10 colour coded ballot questions that will appear on ballots across the Province of Alberta on October 19, 2026, including the 37-word separation question that sits at the centre of the litigation brought by the Athabasca Chipewyan First Nation and the earlier consultation-related challenges that shaped the December 5, 2025 Sylvestre decision of the Alberta Court of King's Bench.
Alberta's history of direct democracy by referendum stretches back further than most provinces, though the results have not always aligned with the expectations of the governments that called them. The province held its first liquor plebiscites in the years following the First World War, and for decades thereafter the referendum mechanism was associated primarily with morality-laden questions about alcohol regulation and, occasionally, with the Social Credit government's experiments in monetary policy that foundered on federal jurisdiction. The late twentieth century saw a different kind of referendum: the 1987 vote on whether Alberta should hold Senate nomination elections, a ballot that passed overwhelmingly and led to the system under which Albertans have since periodically elected "Senators-in-waiting" whose names are submitted to the Prime Minister as recommendations for Senate appointments when Alberta vacancies arise. That Senate election model demonstrated something important about provincial referendums on federal constitutional questions — the Prime Minister is not bound by the result, the constitution does not require the Prime Minister to appoint the person Albertans choose, and yet successive Prime Ministers have occasionally done so anyway because the political cost of ignoring a clearly expressed provincial preference can be high. The 2021 equalization referendum extended this logic from an appointment process to a fiscal transfer regime embedded in the constitution itself, and it did so in a context where the practical impossibility of amending the constitution meant that the yes vote would have to operate purely as an instrument of political suasion.
The equalization question put to Albertans in 2021 was narrow in its literal wording but expansive in its political implications. Voters were asked whether section 36(2) of the Constitution Act, 1867 should be removed, and the question did not explain that section 36(2) does not itself create the equalization program, does not determine the formula by which payments are calculated, and does not prescribe how much money flows to which provinces. The operational details of equalization are set out in federal legislation and can be changed by Parliament without any constitutional amendment at all; the constitutional provision simply commits Parliament and the Government of Canada to the "principle" of making equalization payments so that provincial governments have sufficient revenues to provide reasonably comparable levels of public services at reasonably comparable levels of taxation. Removing section 36(2) would not end equalization; it would remove the constitutional obligation to maintain some form of equalization while leaving Parliament free to continue the program or modify it in any way it chose. Critics argued that the question was misleading because it invited voters to believe they were voting to abolish the payments themselves, while supporters countered that Albertans understood full well they were expressing a view about whether the principle should enjoy constitutional protection. The debate over clarity echoed the federal Clarity Act, SC 2000, c 26, which requires the House of Commons to determine whether any referendum question on secession is clear before any federal obligation to negotiate would arise, but the Clarity Act applies specifically to questions about secession and did not formally govern the 2021 equalization vote.
What the 61.7 percent yes vote legally accomplished was nothing, in the strict sense that no constitutional provision was altered and no federal obligation was extinguished. The Constitution Act, 1982, Part V establishes the amending procedures for the Canadian constitution, and section 36(2) can only be amended by the general amending formula, which requires resolutions from the Senate and House of Commons and from the legislative assemblies of at least two-thirds of the provinces representing at least fifty percent of the population of all the provinces. A provincial referendum cannot substitute for those resolutions, and even a unanimous provincial referendum outcome cannot compel the Government of Canada to introduce the necessary federal resolutions or convince other provincial legislatures to pass theirs. The 2021 referendum was therefore advisory in the fullest sense: it told the federal government that a majority of voting Albertans wanted something to happen, but it did not and could not make that thing happen. The Government of Alberta under then-Premier Jason Kenney characterized the result as a mandate to press Ottawa for fiscal reform, and subsequent federal-provincial negotiations touched on questions of fiscal transfers and Alberta's share of federal spending, but section 36(2) remains in the constitution today unchanged, and no serious federal initiative to remove it ever materialized.
The 2021 experience illuminates the constitutional reality that the Supreme Court of Canada articulated in Reference re Secession of Quebec, [1998] 2 SCR 217: a clear expression of the democratic will of a province's population on a clear question creates political obligations, not legal ones, and those political obligations flow in the direction of good-faith negotiation rather than automatic implementation. The Secession Reference concerned the far graver question of whether a province could unilaterally secede from Canada, but its underlying logic about the interplay between democratic expression and constitutional change applies equally to lesser questions about the federation's fiscal architecture. A referendum result can shift the political terrain dramatically, can strengthen a provincial government's bargaining position, and can make it harder for federal politicians to dismiss provincial concerns, but it cannot rewrite the constitution or compel any amendment. The 2021 equalization vote proved precisely that: the yes side won decisively, the political conversation changed for a time, and yet the constitutional text remained untouched because no government initiated the amending process and no legislature passed the necessary resolutions.
This historical backdrop explains why the 2026 referendum poses questions that operate in layers. Some of the 10 colour coded ballot questions address matters that fall squarely within provincial legislative competence, and for those questions a majority yes vote could theoretically translate directly into provincial legislation without any need for federal cooperation or constitutional amendment, though political and practical constraints might still intervene. Other questions address matters that require federal action, intergovernmental agreement, or constitutional change, and for those questions the 2021 equalization precedent establishes that a yes vote is the beginning of a political process rather than the end of a legal one. The 37-word separation question is the starkest example: no provincial referendum can make Alberta cease to be a province of Canada, no matter how large the majority, because Reference re Secession of Quebec, [1998] 2 SCR 217 holds that secession requires a constitutional amendment under the Constitution Act, 1982, Part V and that such an amendment can only follow negotiations entered into in good faith by all relevant constitutional actors. The Government of Canada, the Government of Alberta, and the other provinces would all be parties to such negotiations, and the negotiations could take years and might fail, and failure would leave Alberta within the federation regardless of any referendum result.
The polling throughout 2026 has shown 60 to 65 percent support to remain in Canada, which suggests that the separation question, if voted on today, would fail on its own terms. But the Government of Alberta under Premier Danielle Smith of Alberta has framed the separation question as one option among many, and the broader referendum package is designed to give the provincial government negotiating leverage on subjects that range from taxation to policing to immigration. The 2021 equalization vote demonstrated that a provincial government can use a referendum loss on the ultimate question — in the sense that nothing legally changed — as a political victory that reshapes subsequent negotiations. If the 2026 separation question fails but several of the other questions pass with strong majorities, the Government of Alberta will likely argue that Albertans have mandated fundamental changes to the province's relationship with Canada short of independence, and the precedent of 2021 will be cited to show that such mandates carry political weight even when they lack legal force.
The litigation that preceded the 2026 vote has added a further dimension to the historical comparison. The December 5, 2025 Sylvestre decision of the Alberta Court of King's Bench addressed an earlier challenge to the consultation process that accompanied the 2026 referendum, and that decision established that the duty to consult articulated in Haida Nation v British Columbia (Minister of Forests), 2004 SCC 73 applies when the Crown contemplates conduct that may adversely affect established or asserted Aboriginal or treaty rights. No comparable duty-to-consult litigation arose in connection with the 2021 equalization referendum, likely because the removal of a constitutional commitment to fiscal transfers between provinces did not engage the same direct effects on treaty rights that a question about Alberta's future within or outside of Canada might. Treaties 6, 7, and 8 are treaties with the Crown in right of Canada, and they guarantee rights that depend on the federal Crown's continued existence and its constitutional relationship with the signatory First Nations. A referendum that asks whether Alberta should become an independent country, even if that referendum is advisory and cannot unilaterally accomplish separation, nonetheless contemplates a fundamental alteration of the constitutional order in which those treaty rights are embedded. The Sylvestre decision and the subsequent Athabasca Chipewyan First Nation v Alberta (Chief Electoral Officer), 2026 ABKB 375 ruling on May 13, 2026 both turned on this recognition: that the honour of the Crown and the duty to consult arise before irreversible steps are taken, and that a referendum question inviting Albertans to express their will on separation is a step with potential downstream consequences for treaty holders even though it does not and cannot accomplish separation on its own.
The 2021 equalization referendum did not engage those treaty considerations because its subject matter did not threaten to sever the federal-provincial relationship within which the treaties operate. Fiscal transfers between provinces, however important politically, do not implicate the treaty relationship between First Nations and the Crown in right of Canada in the same manner as a question about whether Alberta should cease to be part of Canada. The comparison reveals that the legal landscape for any given referendum depends heavily on the subject matter of the question and the identities of the parties whose constitutional or treaty rights may be affected. A referendum on liquor regulation, on Senate elections, or even on equalization does not trigger the same legal constraints as a referendum on separation, because separation alone would restructure the constitutional order so thoroughly that virtually every right guaranteed by the Crown could be called into question.
Alberta's Chief Electoral Officer has administered all of these votes under the province's Referendum Act and Citizen Initiative Act, and the mechanics of ballot preparation, voter registration, and result tabulation have remained largely consistent from one referendum to the next. What has changed is the constitutional scrutiny that courts have applied and the willingness of affected parties to challenge the process before the vote takes place. The Athabasca Chipewyan First Nation did not wait until after October 19, 2026 to argue that the separation question was constitutionally deficient; it sought relief in advance, and the Court of King's Bench in Athabasca Chipewyan First Nation v Alberta (Chief Electoral Officer), 2026 ABKB 375 agreed that the petition process underlying that question could not proceed without adequate consultation with First Nations holding rights under Treaties 6, 7, and 8. This pre-vote litigation represents an evolution in the legal treatment of referendums, one that recognizes the political force a referendum result can carry and that insists on constitutional compliance before the political momentum becomes irreversible.
The 2021 equalization vote and the 2026 vote thus occupy different positions in Alberta's referendum history, even though they share the same statutory machinery and draw on the same tradition of provincial direct democracy. The equalization vote was constitutional in its subject matter but not in its consequences; it asked Albertans whether a constitutional provision should be amended, and the answer had no legal effect because no amending process followed. The 2026 separation question is constitutional in both its subject matter and its potential consequences, because a clear yes vote on a clear question would, under Reference re Secession of Quebec, [1998] 2 SCR 217, trigger a duty on all parties to enter into negotiations whose outcome could include an amendment permitting separation. That trigger does not mean separation would happen, and it does not mean any amendment would pass, but it does mean that the referendum result would create legal obligations that did not exist before — specifically, the obligation to negotiate in good faith. The 2021 vote created no such obligation, partly because the Secession Reference's negotiation framework was developed in the context of separation rather than lesser constitutional changes, and partly because no one seriously expected the federal government to treat a provincial vote on equalization as the kind of democratic expression that compelled federal engagement.
The lesson for claims professionals, governance officers, and policy readers approaching the 2026 referendum is that referendums must be evaluated not only by their literal legal effect but also by the political and constitutional context in which they operate. The 61.7 percent yes vote in 2021 changed nothing in law and yet changed a great deal in politics, and the litigation over the 2026 questions has already changed the trajectory of those questions before a single ballot is cast. The December 5, 2025 Sylvestre decision and the May 13, 2026 ruling in Athabasca Chipewyan First Nation v Alberta (Chief Electoral Officer), 2026 ABKB 375 demonstrate that courts will intervene to enforce constitutional and treaty obligations even in the context of advisory votes, because those votes shape the political landscape in ways that can prejudice rights holders if proper processes are not followed. The 2021 precedent showed that a provincial government can use a referendum result as leverage; the 2026 litigation shows that other constitutional actors can use the courts to ensure that leverage is not obtained through processes that violate the duty to consult or the honour of the Crown.
The Constitution Act, 1982, Part V remains the lodestar for understanding what referendums can and cannot do. No provincial vote can substitute for the amending procedures set out in Part V, and no provincial government can achieve by referendum what it cannot achieve through the legislative and intergovernmental processes that Part V prescribes. But referendums can alter the political environment in which those processes operate, can strengthen or weaken bargaining positions, and can create or foreclose negotiating options that might otherwise have been available. The 2021 equalization referendum demonstrated these dynamics in a context where the constitutional stakes were relatively limited — equalization payments would continue regardless of how Albertans voted, and the worst-case scenario from the federal perspective was a period of difficult intergovernmental relations rather than the breakup of the country. The 2026 referendum operates in a context where the constitutional stakes are far higher, where the polling shows 60 to 65 percent support to remain in Canada and yet a well-organized campaign could shift those numbers before October 19, 2026, and where the litigation brought by the Athabasca Chipewyan First Nation and First Nations holding rights under Treaties 6, 7, and 8 has already reshaped what questions will appear on the ballot and under what conditions.
Understanding the 2021 vote is essential to understanding what the 2026 vote can accomplish, because the 2021 vote established that provincial referendums are instruments of political persuasion that operate within constitutional constraints rather than outside them. The Government of Alberta won the 2021 equalization question by a 61.7 percent margin and gained nothing in law but much in politics; the Government of Canada could have ignored the result entirely but chose instead to engage on fiscal matters in ways that acknowledged Alberta's concerns. Whether the same dynamic will play out after October 19, 2026 depends on factors that lie beyond constitutional law — on election results, on public opinion, on the willingness of Prime Minister Mark Carney and the Government of Canada to negotiate, and on the willingness of the Government of Alberta under Premier Danielle Smith of Alberta to accept negotiated outcomes rather than pressing for further referendums. What constitutional law establishes is the framework within which all of these political calculations must take place: referendums can ask questions, voters can provide answers, and those answers can carry political weight, but no referendum can change the constitution, end a treaty relationship, or accomplish separation without the full machinery of constitutional amendment that Part V prescribes and that Reference re Secession of Quebec, [1998] 2 SCR 217 explicates.