
Liberals win but just miss out on majority
Strong performance for Conservatives in Ontario keeps Liberals to a minority.
While counting still has to be completed in a handful of ridings, the networks have made their final call: the Liberals have secured a minority government, falling just three seats short of the 172 needed for a majority.
I wanted to hold off before sharing any thoughts until the count was nearly completion as the pace of that count made it easy to draw some faulty conclusions. In at least five ridings last night, I saw that projected winners turned out to be losers as the Liberals closed the gap as more of the advance vote was counted. At one point, the Conservatives were only a few seats behind the Liberals in leading and elected. This afternoon, the gap is 25 seats.
But the Liberals were expected to win a majority government — or at least that was the most likely outcome. So, what happened?
While it is more satisfying to nail an election call, we actually learn more about what happened in an election campaign in the ways that parties surpass or fail to meet expectations. I’ll delve more deeply into how the polls and the projection model did in the coming days, but here are a few initial thoughts I have after my first run through the results:
Outside of Ontario, everything went mostly according to plan
The overall numbers didn’t quite look like the projections for the Liberals and the Conservatives, but that was largely because of Ontario. My projections had the Liberals coming out of Atlantic Canada with 27 seats — they won 25. That’s down two from their targets, but they made up for that by winning 44 seats in Quebec instead of 41. They won seven in the Prairies instead of the projected nine, came up three short of my projection in Alberta, and won 20 seats in B.C. instead of 22. All pretty close.
In Ontario, however, the Liberals were projected to win 82 seats and the Conservatives 38. Instead, the Liberals won 69 and the Conservatives won 53. This was where the surprise came in, primarily in York Region of the Greater Toronto Area and in southwestern Ontario. The Conservatives beat their polls in Ontario by a small amount, but their vote proved more efficient than expected.
And Carleton! It was hard to believe that Pierre Poilievre was really in danger, but the Ottawa area swung hard against the Conservatives. And the Liberals worked hard to win Carleton, too. The combination of an Ottawa-concentrated swing and a good local campaign doomed the Conservative leader. He won that seat by 20 points in 2021!
The Conservatives won more toss-ups than the Liberals
In my last election analysis published on Monday morning, I wrote:
“the Liberals are leading in a lot of projected toss-ups. The party is projected to be involved in 43 tosses and is leading in 28 of them. If those go against them, then the Liberals could find themselves in danger of not cresting the 172-seat mark needed for a majority government.”
In the end, the Liberals won just 14 toss ups. The Conservatives, by comparison, won 25 toss-ups and flipped six “leaning” Liberal ridings. Much of the 2021 Liberal victory could be chalked up to their winning many close races. Missing out on the majority in 2025 was because they didn’t do the same.
NDP incumbents did not withstand the tide
There was much discussion of the resilience of the NDP’s incumbents, but in the end only seven of them survived. None of them were surprises, as all seven NDP wins were in either NDP-leaning or toss-up ridings. Beyond that, though, the party was simply not competitive and, at last count, had only a few hundred more votes than the Bloc Québécois — in all of Canada.
Modern records were set as the race polarized
With 43.7% of the vote, the Liberals had their best performance since Pierre Trudeau got 44% in the 1980 election. The Conservatives’ 41.3% is more than the party ever got under Stephen Harper. Even the combined totals of the PCs, Reform and Alliance in the 1993, 1997 and 2000 elections didn’t crest 40%. The Liberals and Conservatives had the highest combined share of the two-party vote since 1958 and this was the first time that they both earned more than 40% since 1930.
The NDP had its worst ever result with just 6.3% and seven seats. Even the CCF never dipped as low as 6%, though seven seats does match the CCF’s results in 1935 — when there were 98 fewer seats in the House of Commons.
The polls did a good job, for the most part
The final polls of the campaign put the Liberals between 41% and 44%. They had 43.7%. The Conservatives were between 37% and 41%. They had 41.3%. The NDP were between 6% and 11%. They had 6.3%. Some pollsters did better than others (more on that soon) and all parties were at the outer edges of their polling ranges, but there were no surprises in the numbers.
The polls’ errors were frustratingly predictable
And where they were wrong, the errors were predictable. The Conservatives and Liberals have both beat their polling average in the last two elections, and they did so again — the Conservatives a bit more than the Liberals did, again. The NDP under-performed the polls by nearly two points — again. The biggest polling miss at the regional level was an under-estimation of the Conservative vote in Alberta and Saskatchewan — again. Fool me once, shame on you. Fool me twice…
This Parliament will be so unstable that it just might work
How this Parliament will work is hard to figure. The Liberals are three seats short of a majority and will probably need to name a speaker. Both the NDP and Bloc hold the balance of power, but the NDP gives the Liberals only a few seats’ worth of cushion and the Bloc holds the Sword of Damocles (the 2026 Quebec election) over the head of Mark Carney. The Liberals nearly won, but they also nearly lost, and maybe would’ve lost if the election had been a week longer. They shouldn’t risk another election soon. The NDP can’t afford one. The Bloc might not want to risk one, either. The NDP needs a leader. The Conservatives need to figure out what to do with their leader. Could this actually survive longer than we expect? We’ll see!
Over the coming days and weeks, I have a few items planned here at The Writ.
First on the schedule will be my customary grading of the polls and a look at how the Poll Tracker model performed. My first impressions are that the polls did a good job, both nationally and regionally, with a few exceptions. The model also did a good job, even if it overshot the Liberal seat total in the estimated projection. The range bottomed out at 161 so the Liberals finished well above that and all parties finished within their projected range, but it’s always better to be closer to the mark.
The model called the right winner, or had the winner as one of the contending parties in a toss up, in 323 of 343 ridings (94.2%). Nine of the 20 misses were “leaning” calls. It’s the remaining 11 misses that are really interesting — and tell us a little about what happened on the ground. More on that soon.
Then I will do an assessment of how each party performed before taking an even deeper dive into each region of the country. There will be lots of data to parse through.
A new episode of The Numbers will also be up tomorrow, so keep an eye out for that!
And in the coming months we’ll have the fallout from this election to cover, including the race to replace Jagmeet Singh as leader of the NDP and whatever drama will come out of Pierre Poilievre’s failure to win his own seat (and the election). Also, the election caused a number of provincial legislators to vacate their seats, which means we should expect a series of byelections over the next six months. Lots to chew on.
So, stay tuned! And thanks for supporting The Writ during this campaign. I have a feeling it’ll be one we’ll remember for a long, long time.
Really enjoyed your coverage of the unprecedented swing in the arc of this election. Interesting that the polling picked up the narrowing of the gap in the last week which proved directionally correct. I have two questions:
1. What do you think was the biggest factor in the CPC gains in the last week?
2. If the NDP loses party status with just 7 seats, does that provide an opportunity for the Liberals to offer 3 of those "independents" a role in forming a majority government?
Thanks for your analysis! I would like to hear how polling models can be adjusted to account for the errors in the next election. Polls are hard. I appreciate your work!