Projection Update: Gap keeps tightening as Conservatives drop below 200 seats
Odds that the Conservatives fail to win a majority increase to 18%.
It’s been another week of polls showing a significant shift in the federal political landscape, with three surveys putting the gap between the Conservatives and Liberals in single-digits — and that’s without substituting the name of Mark Carney (or anyone else) in place of Justin Trudeau.
Four pollsters released new national polling data since last week’s update: Nanos Research, Abacus Data, EKOS Research and Léger/Canadian Press.
(Polling by Mainstreet Research was also published, but it refers to the Liberals as being led by a ‘new leader’, which enters the realm of hypotheticals and so is not a reflection of voting intentions if the election were held today. It’s excluded from the aggregate for that reason.)
Three of these pollsters showed a small gap. The four-week rolling Nanos poll put the Conservatives ahead by eight points. Nanos’s last independent sample from four weeks ago had the Conservatives ahead by 27 points.
EKOS put the Conservatives five points ahead, though that actually shows a widening gap. This survey was conducted online, rather than via interactive voice response as is usually the case with EKOS. Compared to the last online poll by EKOS, the gap has widened by three points as the Conservatives gained five points and the Liberals gained two since the end of January.
There was also a poll from Léger, which put the Conservatives nine points ahead. When Léger was last in the field at the end of last month, the Conservatives were ahead by 18 points.
The two surveys by Abacus Data, however, told a slightly different tale. They were done back-to-back in order to validate the results, and showed the Conservatives leading by 19 and 20 points. The Liberals were up about four points since the end of January in these polls, but the Conservatives were also up about two or three points. So, while Abacus is still picking up the increase in Liberal support seen by other pollsters, it disagrees that this has come from the Conservatives.
All four of these pollsters agree, however, that the NDP has taken a hit of about two or three points. This has pushed them to a new low in the Poll Tracker, as the Liberals continue to make gains and the Conservatives drop below 200 seats for the first time since Spring 2024.
The Conservatives still lead in the Poll Tracker by a comfortable margin with 42.2% to 26.5% for the Liberals. This is another full-point gain for the Liberals, following gains of three points over the last two weeks. The NDP has dropped again to just 15.9% as things go from bad to worse for them.
With the new polls added to the model ranging from 25% to 34% for the Liberals, one might understandably wonder why the Liberals are still only at 26.5% in the aggregate. Nine of the last 11 polls have the Liberals above that score.
The reason is that the model is still taking into account some older data. The date weighting for each poll decays by only 5% per day, so polls conducted in January still take up about 37% of the total weight of the aggregate. This is deliberate, as it prevents the model from swinging around too wildly outside of a campaign period when public opinion is normally not very volatile. When a campaign begins, the date weighting decays by more and more until it reaches 35% per day by the final weeks.
Changing the date weighting decay rate by even a little bit has a big impact. If it was increased to 10% per day, the model would shrink the margin to 13 points and award the Conservatives 182 seats to 113 for the Liberals. If the decay rate was increased to the 35% that will be used once we approach voting day, the Conservative lead would be only a little smaller but their seat projection would slide to 176, just above the threshold for a majority government.
But here’s the catch: with a 35% daily decay rate, only the Abacus, EKOS and Léger polls would carry any real weight in the model. Instead of a smooth aggregate taking into account dozens of polls, we’d be looking at little more than a three-poll average. The bigger decay rate makes sense when two or three polls are being published every day — not when that many polls are being published every week.
I believe it makes more sense to shifts in public opinion outside of an election campaign with caution (and, more importantly, not to change the way the model works on the fly). The Abacus Data poll that still puts the Conservatives ahead by about 20 points shows why we shouldn’t overreact to new polling trends. The model is taking a calmer approach, which might not please everyone. But in the longer run it usually proves to be the smart way to go.
Nevertheless, the change has still been dramatic. The Conservatives are projected to win 192 seats, down 13 from last week. Their likely range dips down as low as 174 seats, just above the 172 needed for a majority government. This has increased the odds of a Conservative plurality, but not a majority, to 17% — roughly the same as the roll of a die.
The Liberals are now projected to win 97 seats, up 18 from last week. Most of those have come from the Conservatives, particularly in Ontario and Atlantic Canada, but they’ve also come from the NDP and Bloc. The Bloc Québécois is down four to 35 seats, only a handful above their haul from the last election. The NDP is down another seat to just 17, with a likely range of 13 to 25 seats.
There are a few regional shifts that are notable. The Liberals have (narrowly) moved back into second place in British Columbia and are tied with the NDP in Saskatchewan and Manitoba. The gap between the Conservatives and Liberals in Atlantic Canada is little more than a point.
Despite all this movement, it should be emphasised that the Conservatives are still widely favoured. They have a lead of 13 points in Ontario and are dominating in Western Canada. Even in Atlantic Canada the party is poised to make significant seat gains. But we’ve gone from discussing one of the largest victories in all of Canadian history to a “normal” majority — if that. The odds of a Conservative minority had been in the low single digits before Trudeau’s resignation. Now, it is a real possibility. Things have changed quickly.
You can see all the findings from the Poll Tracker here and seat-by-seat projections here.
It’s interesting, for sure. The upcoming election has quickly transformed from a referendum on Trudeau and his key policy legacy to a question on who has the best qualities and plan to lead Canada against an increasingly belligerent US. I would rather be the Liberals on that new question.
I wonder how much of this trend back to the LIberals is a result of the feelings that the country is under threat? For much of the post-war era the threats to Canadian unity were internal, particularly the threat of Quebec succession. In those instances the country trusted the Liberals more than the Conservatives to ‘save the country’.
In 1980 and facing the first Quebec referendum, the country chose Trudeau Sr over Joe Clark to lead the federal opposition to the PQ’s bid for sovereignty association. And it was the Trudeau Liberals that re-created what it means to be a Canadian with a new constitution and the Charter a few years later (despite the fact that elements of that package are now coming back to bite us [cf s.33 and Quebec’s exclusion]). And whatever their flaws, those documents remain incredibly popular.
In 1985 the federal Liberals “saved” medicare with the Canada Health Act which has become the iconic symbol of a social program that constitutes a huge part of the Canadian identity. The reality may not live up to the hype, but, again, it’s the symbolism that is important. Canadians trust the LIberals to defend public health care more than they trust the Conservatives.
By the time of the 1995 referendum the conservative side of the political spectrum was split and ineffective. Charest’s PCs were a rump in the House. Manning’s Reform Party was never seen as a legitimate interlocutor with Quebec and Bouchard, the former right hand to the previous PC prime minister, was appointed the “chief negotiator” for the PQ should they win the referendum. That left the federal Liberals to eke a slim victory for the NO side.
All of this leads me to think that the Liberals may have a built-in electoral advantage when it comes to questions of national unity and the need to stand up against threats to the integrity of the nation. Somewhere in the Canadian psyche is the idea that the Liberals will defend Canada better than will the Conservatives. Combine that with Polievre’s flirtation with Trump-adjacent right-wing populism in the aftermath of the pandemic and the recent upswing for the LIberals makes a lot more sense.
None of this has to “fair” or “rational” or even “true” to be real. Just the perception that the Liberals will do a better job is enough to swing people’s voting intentions. And this makes the question of how the Polievre Conservatives pivot to a new electoral question (i.e Who will save Canada from Trumpism?) that much more important. They are still in the driver’s seat and still poised to win a majority, but their grip on the wheel is loosening.