Who Will Vote for the Liberals in the Next Canadian Election? A Comprehensive Predictive Model
I've been thinking if I can use my AI skills to predict the results of upcoming elections. Check out for yourself - and let me know!
Introduction
Predicting election outcomes is a complex challenge, especially in Canada’s highly regionalized and diverse political landscape. However, with a clear understanding of the key demographic, economic, and behavioral variables that influence voting patterns, it’s possible to build a predictive model that estimates which groups are likely to support or oppose the Liberal Party in the next federal election.
This article presents a comprehensive mathematical model that forecasts the likelihood of Liberal support in the forthcoming Canadian election. It incorporates historical data, demographic trends, political behavior, and recent policy impacts to estimate which voter groups are most likely to vote Liberal — and which are not. The goal is to create a structured framework that can predict election outcomes with greater accuracy by identifying the key drivers of voter behavior.
Problem Definition
The Canadian electoral landscape is shaped by a mix of structural and behavioral factors, including:
Demographics – Age, occupation, location, and cultural identity influence political alignment.
Economic Impact – Policies on taxation, government spending, and social services affect different socioeconomic groups.
Recent Political Decisions – Scandals, economic mismanagement, and foreign policy decisions create short-term shifts in voter sentiment.
Regional Differences – Political leanings in Canada vary strongly by province and urban/rural divide.
Strategic and Tactical Voting – The presence of multiple parties (Liberals, Conservatives, NDP, Bloc Québécois, Greens, PPC) creates complex voter behavior.
The objective is to estimate the probability P(Vote Liberal)P(Vote \ Liberal) for each voter segment and project the overall likelihood of a Liberal electoral victory.
Methodology
Step 1: Defining Variables
We define the following comprehensive list of variables, categorized into pro-Liberal, anti-Liberal, and neutral swing groups:
1. Pro-Liberal Demographics and Political Support
Certain groups have historically aligned with Liberal policies and are expected to provide strong support in future elections:
Public sector workers – About 4% of the workforce is employed directly by the federal, provincial, or municipal governments, creating a substantial pro-Liberal voting base.
Unions staff – Paid by the dues of members employed in the public sector or government-funded industries.
Media workers – Employees of the CBC and other publicly funded media represent a small but highly loyal pro-Liberal group.
Academia – University and research staff funded by government grants account for around 0.8% of the population, with consistent Liberal-leaning behavior.
Childcare and DEI (Diversity, Equity, and Inclusion) workers – This group benefits from Liberal social policy and forms a reliable voting bloc.
LGBTQ2+ community and supporters – Roughly 4.4% of the population identifies as LGBTQ2+ or supports related issues, contributing to strong Liberal support.
Left-wing activists – A small but influential group (less than 1%) aligns with Liberal social and economic platforms.
Elderly and healthcare-reliant individuals – The aging population (around 18%) relies on public healthcare and pharmaceuticals, forming a key Liberal voting base.
Big Pharma reliance – About 40% of the population relies on prescription drugs or vaccination programs, which ties their interests to Liberal healthcare policies.
Immigrant voters – Roughly 22% of the population consists of recent immigrants, with support for more open immigration policies aligning them with the Liberal Party.
2. Anti-Liberal Demographics and Political Resistance
Other demographic groups have historically opposed Liberal policies and may form the core of Conservative or other opposition parties:
Groups defending fundamental rights and natural law rights and freedom, those who opposed to:
Vaccine mandates
Lockdown mandates
Propaganda and censorship
Erosion of fundamental rights (bodily autonomy, freedom of religion, informed consent)
– An estimated 10% of the population strongly opposed vaccine and lockdown mandates.
Small business owners – About 15% of Canadians are self-employed or own small businesses, and many were negatively affected by Liberal pandemic policies.
Religious conservatives – Around 10% of the population identifies as religiously conservative and often aligns with socially conservative values.
Oil, gas, and mining workers – Approximately 4% of the workforce is employed in resource extraction, and Liberal environmental policies have created political friction with this group.
Rural voters – Around 19% of Canadians live in rural areas, where conservative policies on resource management and gun rights are more popular.
Military and veterans – Roughly 2% of the population includes active and retired military personnel, who tend to lean conservative.
Blue-collar workers – About 14% of the workforce consists of skilled trades and industrial workers, with a tendency to oppose Liberal economic and environmental policies.
Human rights defenders – Approximately 10% of the population has expressed concern over Liberal policies related to bodily autonomy, religious freedom, and censorship.
3. Neutral/Undecided Groups
Swing voters who could be influenced by political events include:
New immigrants – About 10% of the population consists of recent immigrants, whose political alignment depends on government integration policies and support systems.
Young voters – Voters aged 18–30 account for about 15% of the electorate and are more unpredictable in their voting behavior.
Middle-class professionals – Roughly 30% of the population falls into the middle-income category, which could swing depending on economic conditions.
Non-unionized labor – About 40% of the workforce is not unionized, leaving them open to influence from both Liberal and Conservative policies.
Indigenous communities – Indigenous voters make up about 5% of the population, with varying degrees of support for Liberal or Conservative platforms based on recent policy actions.
Step 2: Mathematical Structure
We used a logistic regression model to estimate the probability of voting Liberal:
P(Vote Liberal)=11+e−(β0+β1X1+β2X2+...+βnXn)P(\text{Vote Liberal}) = \frac{1}{1 + e^{-(\beta_0 + \beta_1 X_1 + \beta_2 X_2 + ... + \beta_n X_n)}}
where:
X1,X2,...XnX_1, X_2, ... X_n = Independent variables
βn\beta_n = Estimated coefficients from historical data
Step 3: Simulation Structure
Using the simmer package in R, we simulated voter behavior under different scenarios by adjusting:
Tactical voting rates
Economic performance shocks
Scandal impact
Turnout variations by demographic group
Results and Conclusion
The model estimates that:
Total pro-Liberal support is approximately 5.6 million voters.
Total anti-Liberal opposition is approximately 4.8 million voters.
Swing voters account for around 2.5 million, with the potential to shift the final outcome.
Projected outcome:
Liberals are likely to secure about 40% of the popular vote.
Conservative opposition is estimated at around 35%.
The remaining 25% will be split among swing voters, NDP supporters, and minor parties.
The results suggest that Liberal support remains solid among public sector employees, union members, and urban voters. However, growing dissatisfaction among small business owners, rural voters, and those affected by economic mismanagement could create vulnerabilities. Strategic voting and turnout will be decisive factors in the final outcome.
Discussion
Significance of Results
Tactical voting is a major factor in swing provinces like Ontario and Quebec.
Liberal dependence on public sector and union support creates a vulnerability if economic conditions decline.
Conservative strength among rural and small business communities remains a structural obstacle for the Liberals.
Next Steps
Incorporate higher-resolution regional data.
Develop a Bayesian model to refine uncertainty estimates.
Simulate province-level outcomes to account for regional variation.
Conclusion
This study presents a simulation-based model for predicting Canadian election outcomes, integrating demographic, political, and economic data. The model demonstrates that tactical voting, economic performance, and demographic alignment are key drivers of voter behavior. The framework can be used to simulate different political scenarios and forecast future election outcomes.
PS:
This article was proofread and edited for readability by ChatGPT. I also asked it to estimate the probabilities for each group. So all numbers provided above are those made up by AI. But the categorization variables contributing the pro-liberal and anti-liberal electorate is mine. Interestingly, based on this model, AI predicted win for Liberals. But I think it underestimated the percentage of Canadian population concerned by the violation of fundamental rights. Or may be not?
Watch out for AI -introduced terminology:
One change that ChatGpt introduced to my original draft was extremely concerning.
It shorted my original name for group #1 of anti-liberals -
From rather long: Groups defending fundamental rights and natural law rights and freedom to much shorter with a very different flavour: Anti-vaccine/mandate groups.
This is how pro-liberals want to label anti-liberals. So I had to revert it back to my original longer version.
Feedback welcome! Maybe one day I’ll write a Web App where you’ll be able to play with various variables (i.e. on how Liberals funding and propaganda machine are used control public opinion) to see how directly they influence the outcome of elections.
To know the future, study the past.
The NDP-Liberal coalition has demonstrated a masterclass in political maneuvering, understanding better than most the power of ultra-partisan support groups with plausible deniability. By quietly funding and encouraging activist organizations, media allies, and third-party advocacy groups, they’ve built an ecosystem that amplifies their messaging, attacks opponents, and applies pressure—all while keeping their hands clean.
This is not accidental. The Liberals and NDP recognize that direct party affiliation can be a liability, so they empower outside groups to do the dirty work, shaping public perception without bearing responsibility for the methods used. Whether it’s union-backed initiatives, activist think tanks, or social media campaigns, they understand that narrative control is just as important as policy.
As the formal alliance between the two parties frays, these independent proxies will remain critical to maintaining their influence. Unlike the Conservatives, who often hesitate to leverage such networks effectively, the Liberals and NDP ensure that their ideological foot soldiers are well-funded and strategically deployed.
In today’s political landscape, controlling the message means controlling the battlefield. And the Liberals and NDP play that game better than anyone in Canada.