Carney's Coalition

A segmentation analysis of ridings won by the Liberals. If you're enjoying this Substack, please forward this email to 2 people you know that might enjoy it!

In my previous post, I broke down Pierre Poilievre’s voters, showing that the much-discussed NDP-to-CPC switchers were a red herring. His real gains came from 2021 PPC or non-voters.

This time, we turn to the 169 ridings that sent Mr. Carney to government, and ask: who are his voters, and how will he hold this broad coalition together?

To answer that, I used a k-means clustering algorithm (which is a fancy way of saying: I fed riding-level census data of the ridings the Liberals won into some Python code that grouped these ridings based on how similar their demographic characteristics are). Through this process, you get natural groupings of ridings - or in political speak, voter blocs - that are demographically similar to each other.

Below, you’ll meet five voter blocs: from downtown elites, to 401 stalwarts, to liberal habs fans. While early momentum and a shared appetite for action have masked internal tensions, holding this coalition together will require more than speed; it will demand navigating competing priorities across a sprawling political spectrum.

The chart below compares demographic characteristics of the five clusters against the national average. Each column represents one cluster of ridings, and each row represents one census variable: language, ancestry, diversity, commuting habits, religion, education, housing, household status, etc. Over 75 variables were included in the analysis, and the variables that represent the biggest differences between riding groups are shown below.

Segment 1: Urban working professionals - 28 ridings

This segment contains the highest-income Liberal ridings, with an average personal income of $69,852, far above the national average of $54,450. These ridings are largely in Toronto, Ottawa, Montreal, and Vancouver; ridings like Beaches—East York, Toronto—St. Paul's, Ottawa Centre, Mount Royal, and Vancouver Centre. It has the highest share of renters (55%) and the second-highest education level among all segments (50% hold a bachelor’s degree or higher). Secularism is strong here (42% report no religion), second only to Segment 3. Transit use (24%) is higher than in any other group except Segment 5. Demographically, it sits in the middle on diversity measures, with a 40% visible minority population and 44% first-generation Canadians.

Segment 2: Diverse suburbs - 30 ridings

Segment 2 is the most diverse group by far: 73% visible minority, 53% immigrant, and 59% first-generation, all the highest among the five segments. The vast majority of these ridings are in Brampton, Mississauga, and Scarborough. It also includes the largest South Asian population (29%) and the lowest share of third-generation or more Canadians (14%). English-language usage remains high at 91%. Income is the lowest of all segments at $46,255, and only 31% hold a university degree. These are predominantly newcomer-heavy, suburban ridings.

Segment 3: Rural Canada - 32 ridings

This segment contains very few immigrants; it is largely made up of rural ridings, and contains some indigenous-heavy ridings. These ridings are mostly in Atlantic Canada, and Northern and Eastern Ontario; ridings like Thunder Bay—Rainy River and Churchill—Keewatinook Aski, Acadie—Bathurst and Avalon, and Prescott—Russell—Cumberland. It includes the highest proportion of third-generation or more Canadians (85%) and Indigenous identity (13%), and the lowest levels of visible minorities (6%), immigrants (6%), and first-generation Canadians (8%). It also ranks highest in Christian affiliation (66%) and has the lowest education level (18% with degrees). Public transit use is negligible (2%), and 90% of residents drive to work. Average income is $48,120, the second-lowest among segments.

Segment 4: Commuters - 46 ridings

The largest segment, Segment 4, has the highest share of English first-language speakers (93%) and the second-highest average income at $60,132. This segment represents Toronto, Ottawa, and Vancouver suburbs, southwestern Ontario, and Winnipeg; ridings like Burlington and Oakville East, Kanata and Orléans, New Westminster—Burnaby—Maillardville and South Surrey—White Rock, and Winnipeg South and Winnipeg South Centre. It is moderate across most other indicators: 31% visible minority, 27% immigrant, 34% with degrees, and 32% renters. The third-generation population is 48%, squarely in the middle. This segment also has one of the highest rates of car commuting (83%). It represents balanced, mixed-profile ridings in mid-sized cities and suburban commuter belts. This group looks the most like the Canadian average, and contains lots of traditional Ontario and BC swing ridings.

Segment 5: Suburban Habs fans - 33 ridings

Segment 5 is defined by very low English usage (15%), the highest share of French speakers, and a high third-generation or more population (63%). This segment represents suburban Quebec; ridings like Bourassa, Laval—Les Îles, and Sherbrooke. It is also the second-most Christian (63%) and has the second-highest share of renters (41%). Education levels are relatively low (25% with degrees), and the average income is $51,152, slightly below the national average. Diversity levels are modest: 22% visible minority and 19% immigrant.

This is, without question, a tenuous coalition to build, let alone hold.

Winning 1.5 million more votes than any victorious party in Canadian history means one thing above all: the tent is enormous. And with that size comes complexity. The government is still in its honeymoon phase with sky-high approval ratings and a mandate to move quickly. So far, it has used that runway to its advantage, passing Bill C-5 early, a signal that this government understands one of the few points of near consensus across this diverse coalition: big things need to get built, and fast.

But beneath that consensus lies a scatterplot of competing priorities. There isn’t much shared ground between McGill-educated consultants in downtown Montreal, first-generation families in Scarborough, Indigenous voters in the North, Ford-Carney voters along the 401, and suburban Montreal Canadians fans. This coalition didn’t necessarily come together because they wanted the same things; many came together because they wanted something different than what Mr. Poilievre was offering.

So, as many of you are leaving on summer camping trips, the question becomes: what will hold this tent together?