Place & Politics in Toronto - Part 7

The five Torontos

Aniket Kali

October 22, 2026

People's experiences of Toronto can vary dramatically from one neighbourhood to the next. Areas separated by just a few subway stops can differ sharply in income, life expectancy, access to transit and services, and housing – where rents can vary by over $1000 per month. Other posts in this occasional series, Place and Politics in Toronto, show that some parts of Toronto vote consistently, turn out in large numbers, and display a clear political orientation; elsewhere, voting is more divided or turnout is lower.

Those differences matter beyond who wins elections. Neighbourhoods with similar numbers of eligible voters can contribute very different numbers of votes, while governments make decisions about the transit people rely on, the homes they can afford, the services available nearby, and how public money is spent.

Using voting and turnout patterns from Toronto's three most recent elections (2023 mayoral by-election, 2025 provincial election, and 2025 federal election), we used a clustering algorithm to identify five broad political geographies – five different Torontos that participate and vote in distinct ways. We then used census data to describe the kinds of places those clusters represent; it does not determine which cluster a neighbourhood belongs to.

Five Torontos

Census tracts clustered by voting in the 2023 mayoral and 2025 provincial and federal elections

% of
population
Progressive Core 17.5% Soft-Left Belt 14.9% Civic Liberals 13.8% Suburban Conservatives 22.0% Low-turnout Suburbanites 31.9%

The five clusters run broadly from Toronto's most consistently progressive neighbourhoods toward its more conservative ones, but they are not simply five stops along a left-right line. Voter turnout matters separately from partisan preference, and similar voting patterns sometimes appear in neighbourhoods with very different histories, housing, and demographics. What emerges is not a map of five kinds of voters, but five recurring combinations of politics and place.

Progressive Core

109 of 585 census tracts, 17.5% of population (2021 Census)

Progressive Core City of Toronto average
Chow (2023 mayoral)
2023 mayoral turnout (citizens 18+)
Ontario NDP (2025 provincial)
Ontario PC (2025 provincial)
NDP (2025 federal)

Rooted in the twin hearts of Toronto's West End and East End, this is the city's clearest concentration of progressive voting. These neighbourhoods have given the NDP much higher shares than Toronto overall, even in historically difficult elections for the party, while giving considerably less support to Conservatives across levels of government. Olivia Chow – a former city councillor and NDP MP – also performed particularly strongly here in the non-partisan 2023 mayoral by-election. Voter turnout is above the city average across all three elections.

This is Toronto's densest cluster, with the highest renter share, the lowest reliance on cars for commuting, and relatively high university attainment. Much of it is made up of older, established central-city neighbourhoods. It has fewer newer high-rise districts that appear more prominently in the Soft-Left Belt cluster. This combination of educated, renter-heavy urban neighbourhoods and progressive voting is consistent with a broader shift documented across many Western democracies, including in research on Canada, in which the electoral base of the left has increasingly moved toward highly educated and economically downward urban voters.

% renter0%20%40%60%80%100%City of Toronto median: 43.0%Progressive Core median: 51.6%% commute by car0%20%40%60%80%100%City of Toronto median: 63.6%Progressive Core median: 44.0%

But this is not simply a recent realignment. Many of these same areas were centres of immigrant labour organization, left-wing politics, and later neighbourhood-based urban reform, including the fight against the Spadina Expressway. Toronto's reform movement was politically heterogeneous, but many of its fights over expressways, redevelopment, neighbourhood preservation, and local planning took place in and around the old urban core. Research on Toronto's political geography finds that neighbourhood-level patterns can persist across elections, even as the people living within them change. The electorate has shifted, but as a previous post in this series shows, the geography of progressive voting has proved remarkably durable.

Soft-Left Belt

85 of 585 census tracts, 14.9% of population (2021 Census)

Soft-Left Belt City of Toronto average
Chow (2023 mayoral)
Matlow (2023 mayoral)
2023 mayoral turnout (citizens 18+)
Ontario NDP (2025 provincial)
Conservative (2025 federal)

If the Progressive Core is Toronto's most concentrated geography of progressive voting, the Soft-Left Belt is what that politics looks like once it spreads into a more varied urban landscape. It stretches through newer condo districts and rental-heavy parts of the central city, established neighbourhoods farther west and east, and pockets of southwest Scarborough. These places do not share one built form or one history. What they share is electoral: higher voter turnout, broadly progressive voting, and a greater willingness to support Liberal or moderate alternatives than in the Progressive Core.

That difference is visible across elections. Chow and the provincial NDP both outperform their Toronto-wide results here, but by much smaller margins than in the Progressive Core. Josh Matlow also performs best in this cluster. At the same time, the Soft-Left Belt has the highest university attainment and income of the five clusters, while also having the highest prevalence of households spending 30% more on housing costs and the largest recent-migrant share. That combination fits parts of Toronto's changing central-city geography, where growth has increasingly taken the form of mid- and high-rise apartments and a large rented-condominium market; in 2021, more than half of downtown condominiums were rented.

The result is a political belt rather than a single neighbourhood type. They converge because they have similar political characteristics: above average voter turnout, left of Toronto overall, but less consistently aligned with the progressive left than central neighbourhoods.

Median pre-tax individual income$20k$40k$60k$80kCity of Toronto median: $38kSoft-Left Belt median: $52k% bachelor's degree+0%20%40%60%80%100%City of Toronto median: 39.1%Soft-Left Belt median: 60.1%

Civic Liberals

77 of 585 census tracts, 13.8% of population (2021 Census)

Civic Liberals City of Toronto average
Bailão (2023 mayoral)
Ontario Liberal (2025 provincial)
Ontario NDP (2025 provincial)
Conservative (2025 federal)
2025 federal turnout (citizens 18+)

Civic Liberal neighbourhoods are even harder to pin to one part of Toronto. They appear in parts of Midtown and Leaside, stretches of the Yonge corridor through Willowdale, pieces of The Beaches, Etobicoke-Lakeshore, and Scarborough along the lake.

This cluster has the highest turnout in the provincial and federal elections, and among the highest in the mayoral by-election. Liberal support is strong, and there is much less warmth toward the NDP and progressive mayoral candidates. This pattern is clearest provincially, where the Liberals receive just over half the vote and the NDP falls to only 8 per cent. Ana Bailão also leads Olivia Chow in the mayoral race, while federally the Liberals remain ahead but Conservatives perform considerably better than in either of the first two more progressive clusters.

There is some historical precedent for that kind of Toronto politics. The city's changing geography of wealth runs through parts of this cluster, especially along the Yonge corridor, where incomes have risen sharply over time. Toronto has long contained prosperous, high voter turnout districts that sit outside both the city's strongest progressive neighbourhoods and its most conservative suburbs. The Civic Liberals are a modern version of that middle ground: scattered across the map, but tied together by a politics that is engaged, higher income and higher education, Liberal-leaning, and notably less progressive than the clusters to their left.

Suburban Conservatives

129 of 585 census tracts, 22.0% of population (2021 Census)

Suburban Conservatives City of Toronto average
Bailão (2023 mayoral)
Ontario PC (2025 provincial)
Ontario NDP (2025 provincial)
Conservative (2025 federal)
Liberal (2025 federal)

Spanning much of Toronto's inner suburbs, the Suburban Conservatives overlap strongly with neighbourhoods built during the region's postwar suburban expansion: lower-density housing, high rates of homeownership, and greater automobile use. Today, they are the oldest of the five clusters as well. Their electoral pattern is just as clear: these neighbourhoods give the highest shares to the Ontario PCs and federal Conservatives, the lowest or near-lowest shares to the NDP, and favoured Ana Bailão over Olivia Chow in 2023. The federal Liberals narrowly led the Conservatives here in 2025, but this is where conservative parties and candidates perform most strongly overall.

That pattern echoes a much older divide between Toronto's core and its postwar suburbs. Etobicoke, North York, and Scarborough all grew rapidly through the postwar period before amalgamation in 1998, much of it around private homes and automobile travel. Zack Taylor has documented an enduring city-suburb divide in Toronto voting that predates Rob Ford's conservatism; in 2010, for example, the former Metropolitan Toronto suburbs voted heavily for Ford while the old City of Toronto went strongly for George Smitherman, as we described in a previous entry in this series.

But today's suburban politics no longer follow that older divide neatly. Much of Scarborough, despite sharing the same postwar and car-oriented built form, falls instead into the Low-turnout Suburbanites.

% renter0%20%40%60%80%100%City of Toronto median: 43.0%Suburban Conservatives median: 31.1%Average age30354045505560City of Toronto median: 41.5Suburban Conservatives median: 43.2% commute by car0%20%40%60%80%100%City of Toronto median: 63.6%Suburban Conservatives median: 73.4%

Low-turnout Suburbanites

185 of 585 census tracts, 31.9% of population (2021 Census)

Low-turnout Suburbanites City of Toronto average
Chow (2023 mayoral)
2023 mayoral turnout (citizens 18+)
Other mayoral candidates
Ontario PC (2025 provincial)
Liberal (2025 federal)

It is easy to look at Toronto's electoral maps and treat much of the inner suburbs as one conservative bloc, but this cluster complicates that picture. More than anything else, what distinguishes these neighbourhoods is low voter turnout. Turnout is lower than in every other cluster across the municipal, provincial, and federal elections, even though Low-turnout Suburbanites are Toronto's largest group, containing almost one-third of the city's population. These neighbourhoods also have the lowest income and university attainment, the highest racialized share, some of the largest households, most people with commutes over an hour.

Their voting is less ideologically fixed than that of the Suburban Conservatives beside them. Ontario PC support is relatively high, but Chow led the mayoral vote and the federal Liberals received a clear majority. Previous School of Cities work has documented a longer-term rise in Conservative support in immigrant- and minority-heavy GTA ridings, particularly at the provincial level, pointing toward disillusionment with an incomplete multiculturalism.

This political complexity sits on top of a broader geography of inequality. Research on Toronto's inner suburbs has documented lower incomes, weaker access to mass transit and city services, and longstanding underinvestment compared with more affluent parts of the city. Other reporting has traced how rising housing costs and gentrification have pushed lower-income and immigrant households outward from parts of the core, while surveys through the Community Voices project found that residents placed particular importance on reliable transportation, neighbourhood services, affordable housing, and safety.

% visible minority0%20%40%60%80%100%City of Toronto median: 53.1%Low-turnout Suburbanites median: 80.7%Median pre-tax individual income$20k$40k$60k$80kCity of Toronto median: $38kLow-turnout Suburbanites median: $33k% bachelor's degree+0%20%40%60%80%100%City of Toronto median: 39.1%Low-turnout Suburbanites median: 26.0%

Comparing the five clusters

Election results used to define the clusters

Election results by cluster, compared to the City of Toronto
VariableProgressive
Core
Soft-Left
Belt
Civic
Liberals
Suburban
Conservatives
Low-turnout
Suburbanites
City of
Toronto
2023 mayoral by-election
Olivia Chow53%40%31%25%37%37%
Ana Bailão23%32%37%43%28%33%
Mark Saunders5%6%9%12%11%9%
Anthony Furey3%5%6%7%5%5%
Josh Matlow6%8%6%4%2%5%
Other candidates11%10%11%11%17%12%
2025 provincial election
PC16%24%37%48%42%34%
NDP51%37%8%7%15%23%
Liberal29%36%51%41%39%38%
Green3%3%3%2%2%3%
Other parties1%1%2%3%2%2%
2025 federal election
Liberal62%62%59%51%59%58%
Conservative21%28%37%45%36%34%
NDP15%9%3%3%4%6%
Green1%1%1%0%1%1%
Other parties1%1%1%1%1%1%
Turnout (% of citizens 18+)
2023 mayoral by-election45%44%45%38%30%39%
2025 provincial election54%51%54%46%40%48%
2025 federal election73%76%79%69%63%71%

Census characteristics, used only to describe the clusters

Census characteristics by cluster, compared to the City of Toronto
VariableProgressive
Core
Soft-Left
Belt
Civic
Liberals
Suburban
Conservatives
Low-turnout
Suburbanites
City of
Toronto
Census characteristics (2021)
Population density (per km²)9,0236,1334,0243,0253,6304,196
Age 65+14%16%19%20%16%17%
Average age404143434141
Average household size2.02.02.22.72.82.4
Median pre-tax individual income$44k$53k$47k$42k$33k$42k
Renter households59%53%44%37%47%48%
Spending 30%+ on shelter34%36%33%28%31%32%
Canadian citizens86%83%83%85%79%83%
Visible minority42%39%50%48%79%56%
Migrants in the last 5 years18%23%18%13%16%17%
Bachelor's degree or higher48%55%48%39%29%41%
Commute by car/truck/van42%48%64%71%67%61%
Commute by public transit31%27%26%21%27%26%
Commute 60+ minutes10%9%11%12%16%12%

Data and methods

This project was originally inspired in part by the San Francisco Chronicle's California political groups project, which used clustering to identify recurring political geographies rather than beginning from conventional partisan or demographic categories.

We analyzed all 585 Toronto census tracts using results from three recent elections: the 2023 mayoral by-election, the 2025 Ontario election, and the 2025 federal election. Because voting polls and census tracts do not share the same boundaries, poll-level votes were first interpolated to census tracts using the distribution of Canadian citizens aged 18 and over. Turnout is measured as votes cast relative to that adult-citizen population (Canadian citizens aged 18+), not registered electors, while party and candidate support is calculated from valid votes cast.

We grouped neighbourhoods using their voting and turnout alone – not their demographics. For each election, we considered both how a tract differed from Toronto overall and how it differed from its own ward or riding, giving those citywide and local perspectives equal weight. We also balanced the three elections so that no single electoral contest dominated. We then used principal component analysis (PCA) to condense overlapping electoral measures and k-means clustering to group tracts with similar political profiles.

We tested several alternative approaches and solutions ranging from three to seven clusters. Under the final model, five groups produced the strongest silhouette score, although the difference between three, four, and five was small. Looking at the actual memberships helped distinguish the alternatives. At four clusters, the Progressive Core and Soft-Left Belt remain largely intact, while Civic Liberals largely lose their standalone cluster and are distributed among neighbouring groups. At six and seven, the existing political geographies begin to fragment while statistical separation becomes clearly weaker. Five therefore provides a useful balance between retaining substantively distinct electoral patterns and avoiding unnecessary subdivision. It should still be understood as one useful simplification of Toronto's continuous political geography, rather than evidence that exactly five natural political types exist.

After creating the clusters, we joined the 2021 Census data on age, income, housing, migration, racialization, education, commuting, and other characteristics to understand the places within them. These demographics describe places and should not be read as explaining how any individual voted. The names – Progressive Core, Soft-Left Belt, Civic Liberals, Suburban Conservatives, and Low-turnout Suburbanites – are descriptive shorthand applied after the analysis.

For full details on the data, interpolation, model selection, sensitivity tests, aggregation, and limitations, see our complete Data and Methods and the reproducible analysis notebooks: model exploration, cluster interpretation, and final data processing.

Thanks to Zack Taylor for reviewing the methods, research, and text behind this piece. Jeff Allen helped to review the text and design the graphics.