Published October 5, 2026 · Équipe Le Québec Vote
A party can ask voters to choose its platform. It can also ask them to choose it to prevent an outcome they dread. In 2026, a referendum offers precisely that second argument. But where does a legitimate warning end and exaggeration begin?
The technique changes the question voters ask themselves. Rather than comparing proposals for governing, they are invited to decide which party could protect them from a risk. The protective party’s platform may then receive less scrutiny.
This does not mean referendum concerns are invented. A change to Quebec’s political status deserves serious examination. Nor does it mean that someone opposed to a referendum is irrational or manipulated. The problem arises when the severity of a scenario replaces an explanation of its likelihood, conditions and consequences.
This analysis separates documented facts, research findings and interpretations of campaign messages. It does not measure how many votes these messages have shifted or claim knowledge of strategists’ private intentions.
A real commitment, presented in different ways
In its August 18, 2026 statement, the PQ maintains a consultation on independence during its first term but rules out a referendum before January 20, 2029. It also invokes cultural, linguistic and financial decline within Canada. That assessment is the party’s own.
The CAQ’s official website, consulted on October 3, carries October 2 messages contrasting public services with separation, linking prosperity to stability, and urging a vote against a referendum. This frames the campaign around risk and priorities.
On October 2, The Canadian Press reports that Milliard presents the Liberals as the federalist option, while Duhaime invokes a balance of power against a referendum.
These messages do not all say the same thing. Identifying a commitment, opposing it and predicting catastrophe are three different acts. Calling all of them fearmongering would itself prevent a proper assessment of their merits.
How does a fear appeal work?
The logic can be broken into four actions: identify a danger, associate it with an opponent, make it relevant to everyday life, and present one’s own party as protection. This is an analytical framework, not evidence of coordination between parties.
Making risk the main basis for a decision
The same commitment can be presented as a democratic opportunity, a distraction or a source of instability. The consequences selected for emphasis steer the discussion. The concept of *framing*, developed notably by Robert Entman, helps explain this selection of an issue’s features.
Consider this fictional example: “You like some of our opponent’s proposals, but are you willing to take this risk?” The question does not refute those proposals. It asks voters to give them less weight than the danger being described.
Making uncertainty personal
A constitutional future can seem abstract. Jobs, savings and family security are less so. Linking the first to the others can make a message more tangible. Each connection must nevertheless be demonstrated: announcing a referendum does not, by itself, establish a job loss or a reduction in benefits.
Kahneman and Tversky’s prospect theory shows that gains, losses and the presentation of risk influence decisions. It does not establish that every fear favours the status quo: some situations involving losses also encourage risk-taking. Its application here is interpretive, not a measurement of voting in Quebec.
Offering a way out of the threat
An electoral warning often includes a simple solution: give more power to whoever promises to prevent the danger. Voters are invited to turn concern into action.
A 2015 meta-analysis by Tannenbaum and colleagues finds an average positive effect of fear appeals on attitudes, intentions and behaviours; messages offering an effective response perform better. These studies, including health communication research, do not allow an electoral gain in Quebec to be calculated.
Encouraging a vote against rather than a vote for
A voter might prefer one platform but choose another candidate to block a scenario. That is a form of strategic voting. However, the party presenting itself as a province-wide safeguard is not necessarily best positioned in every riding. The claimed effectiveness of the protection also needs evidence.
What advantages are parties seeking?
The following are strategic possibilities, not established results for 2026.
| Potential advantage | What the message might accomplish | Limitation |
|---|---|---|
| Bring different voters together | Unite people around shared opposition despite disagreements over policy | A coalition against a project does not guarantee agreement on governing |
| Change the campaign’s focus | Put an opponent on the defensive about timing or risk | The record in office and other commitments remain relevant |
| Become the protector | Give voters an immediate reason to choose the party | Protection must be credible and achievable |
| Encourage strategic voting | Persuade some voters to leave their first choice | Several parties may divide the same electorate |
| Simplify a complex message | Turn constitutional debate into an easily remembered choice | Simplicity can conceal essential conditions |
For an incumbent government, the potential appeal is particularly clear: shift scrutiny from what it has done to what its opponent might do. This does not prove that every referendum discussion is an attempt to avoid scrutiny of a record. It describes a possibility voters can recognize and question.
Does it actually work?
Attention, anxiety and voting are different outcomes. A message can be memorable without persuading; it can persuade without changing a decision already made.
Ted Brader’s experiments, published in 2005, show that advertisements evoking fear can increase vigilance and facilitate persuasion. They do not establish the effect of Quebec’s current referendum messages.
The 2007 meta-analysis by Lau, Sigelman and Rovner does not confirm that negative campaigns generally win votes. It finds greater recall but small reductions in trust and political efficacy; it finds no reliable evidence of lower turnout. Negative campaigning and fear appeals overlap, but are not identical.
Saying “parties do it because it works” would therefore be too categorical. They may expect a benefit that the available evidence cannot establish here.
Disadvantages and risks
Credibility can weaken. A predicted catastrophe without a precise mechanism may seem exaggerated. A party dramatizing danger must explain its scenarios as rigorously as one promising an easy transition.
A rival may benefit from the protective argument. When several parties oppose the same project, anxiety alone does not determine which one gets the votes. A campaign can make an issue central without reaping the rewards.
Other issues may receive less scrutiny. Health care, housing, taxation and integrity require comparisons of their own. Their importance does not disappear because constitutional debate exists. Conversely, their urgency does not make that debate illegitimate.
The discussion can lock people into opposing camps. Someone may want a different government without wanting independence, support a consultation without supporting Yes, or reject a consultation at a particular time without rejecting all future discussion. A binary divide erases these distinctions.
Fear can be answered with another fear. Fear of separation can meet fear of decline if nothing changes. Repeated warnings may then replace comparisons of solutions with comparisons of threats.
These risks do not justify saying voters stop thinking. The research book *Anxious Politics* explains that anxiety can encourage information-seeking while directing attention towards threatening news. This research examines, among other subjects, immigration in the United States: its application to a referendum requires caution.
Why does this argument return in 2026?
The project is part of a current commitment. Discussing it does not require artificially reviving an issue that has disappeared.
Historical memory also supplies instantly recognizable symbols. According to Élections Québec, No won the October 30, 1995 referendum with 50.58% of valid votes; turnout was 93.52%. These figures document a close and highly mobilizing vote, without proving how today’s generations respond to its memory.
The constitutional question also offers a way to compete with the desire for change. In a Léger/Le Journal/TVA online poll of 1,001 eligible voters conducted September 25–27, 74% wanted a change of government. This is neither an election result nor an experiment establishing the effect of a fear campaign.
In response to that desire, the risk argument proposes another question: what kind of change are you prepared to accept? Its return can be understood through this competition between promised improvement and a demand for security. That is an interpretation of the context, not proof that fear determines the vote.
An election victory is not independence
Three decisions must remain separate: electing representatives, consulting the population and carrying out constitutional change.
Élections Québec describes Quebec referendums as consultative, governed by the Referendum Act. A referendum has its own question and date. An election victory is therefore not already an answer to that question.
The 1998 Reference re Secession of Quebec, explained on the Supreme Court’s website, distinguishes the absence of a right to unilateral secession from an obligation to negotiate following a clear majority on a clear question. It does not guarantee the outcome of negotiations.
The federal Clarity Act provides, among other things, for the House of Commons to assess the question and majority. Section 4 of Quebec’s Act respecting the exercise of the fundamental rights and prerogatives of the Québec people and the Québec State defines the winning option as the one receiving 50% of valid votes plus one vote. Identifying these provisions does not resolve constitutional disagreements.
This distinction does not eliminate transition risks. It rules out presenting the entire sequence as an immediate, automatic consequence of a provincial vote.
Is economic uncertainty necessarily a scare tactic?
No. Trade rules, currency, debt, rights and continuity of programs are serious issues to examine in a proposal to create a new state. Claiming all concerns are fabricated would unfairly excuse the project’s advocates from answering them.
But a concern is not a measured economic outcome. On October 2, according to Agence QMI, Christine Fréchette suggested that a period of uncertainty could freeze investment. This is a political forecast; the report provides no independent estimate establishing its scale.
The mechanism, period, assumptions and alternative scenarios would need to be specified. The same standard applies to a sovereigntist promise of easy gains or a trouble-free transition.
What about the referendum itself: benefits and costs?
A consultation can give citizens a separate opportunity to decide on their future, clarify preferences and require advocates to defend a proposal. Its democratic value depends on the question, the information available and the quality of debate.
It also uses public resources and political time; it can deepen divisions and leave questions unresolved when further steps require negotiation. This article establishes no total cost for a future referendum.
Opposing a referendum can be a legitimate democratic preference. Supporting one can be as well. The presence of emotion does not authorize anyone to make those choices for citizens.
Seven questions to test the strength of a warning
- Which stage is being discussed? An election victory, preparation for a consultation, its result or a transition?
- What is the actual timetable? Does the message accurately reflect the announced commitment?
- Which consequence is established? Is it a fact, forecast, possibility or opinion?
- What evidence connects the project to the predicted harm? A verifiable mechanism matters more than an alarming association.
- Which assumptions could change the conclusion? A serious analysis acknowledges the conditions of its scenario.
- Can the promised protection be delivered? What parliamentary power, procedure and commitments make it credible?
- What receives less scrutiny? Does the party also accept examination of its record and other proposals?
This standard works both ways. Referendum opponents must substantiate the dangers they predict; advocates must explain expected benefits, risks and possible safeguards.
Fear of a referendum remains useful as an argument because it can turn a political preference into a need for protection. But it does not release any party from demonstrating its claims. An informed debate leaves voters to assess risks while giving them the facts they need to do so.
*Analysis based on sources consulted October 3, 2026. Context photograph: Marc-Lautenbacher / Wikimedia Commons, CC BY-SA 4.0, June 7, 2009, without modifications.*