Election Writ 4/10: Believe the polls
Every poll now has the Conservatives behind.
It would have been nice to get through this campaign without a bout of Poll Trutherism. Alas, here we are.
Some of those who were celebrating the Conservatives’ 25-point lead a few months ago have now turned on the polls as the Liberals move ahead. At Pierre Poilievre’s rally in Brampton last night, a few of them donned sweaters asking the provocative question: “Do you believe the polls?”
They even scored a photo with the Conservative leader. And, according to Robert Fife of the Globe and Mail, this is a view that is shared all the way at the top.
Also in today’s newsletter: The Liberals, NDP and Conservatives all came up one candidate short of a full deck. Will it matter anywhere? Plus, I take a look at Jagmeet Singh’s chances in Burnaby Central.
To the surprise of all, no doubt, I am here to tell you that, yes, you should believe the polls.
We can start with the broad range of polls conducted in different ways that all show the same things. There have been 13 pollsters active since the writ drop:
Six of them (Abacus, Angus Reid Institute, Innovative Research Group, Research Co., Léger, MQO) have polled online using different panels recruited in different ways and all have shown a Liberal lead of between six and 10 points in their latest results — including Abacus Data, which until yesterday’s release was the last pollster still showing the Conservatives tied with the Liberals.
Three of them (Liaison, Pallas, EKOS) have polled using interactive voice response (IVR, or automated phone calls) and they have shown a Liberal lead of three to 15 points.
Two of them (Nanos, Ipsos) are using a combination of traditional live-caller telephone polling and online polling and they have shown a Liberal lead of five to eight points.
One of them (Mainstreet) is using SMS text messaging to invite respondents to complete a survey online, and its latest poll has the Liberals ahead by seven points.
And one of them (Pollara) is combining telephone, IVR and online polling in its sampling, and its last result had the Liberals ahead by eight points.
In addition to all 13 of these pollsters (all of whom have been accused of bias toward one party or another in the past, some of whom have actually worked for one party or another in the past) showing a Liberal lead nationally, every one of them has also shown a Liberal lead in the electorally-decisive province of Ontario of at least six points (and as many as 16). A few weeks ago, we even heard from the Ontario PCs’ pollster who showed a similarly-sized gap in this battleground.
And if you don’t want to believe the voting intentions numbers, you can look at the personal ratings of the leaders. Every poll with those questions included shows that Mark Carney has higher net approval ratings than Pierre Poilievre (Carney is a net positive, Poilievre is a net negative) and leads on preferred prime minister by similar or larger margins than the Liberals lead over the Conservatives. There’s no indication from those personal numbers that there is an underlying weakness that could sap Liberal support or a strength that could boost the Conservatives.
Now, such a broad consensus might not be convincing enough for remaining doubters. After all, the polls have been wrong in the past. Perhaps they are all wrong this time, too.
Just look at Kamala Harris and Donald Trump, amirite?
The last three U.S. presidential elections weren’t terrific showcases for the validity of the polls. But they weren’t as wrong as people seem to believe — especially in 2024.
On average, the polls awarded Harris a lead of about one percentage point over Trump. When all the ballots were counted, Trump won by 1.5 percentage points.
In a tight race, that total 2.5-point error in Trump’s victory margin was hugely impactful and helped him win the electoral college handily. But this is not a tight race.
If the polls are off here by a similar margin, then the Liberal lead over the Conservatives would drop from the currently projected 6.6 points to 4.1 points. According to my model, shifting the margin by that much between the Liberals and Conservatives reduces the chance of a Liberal victory from 98% to 90%. The Liberals would still have a roughly 3-in-4 chance of securing a majority government.
But even a 2.5-point miss in the polls is a lot to ask. In both 2019 and 2021, my Poll Tracker under-estimated Conservative support by 2.7 points. But it also under-estimated Liberal support by 1.1 points — the net error in the projected margin was an identical 1.6 points in the last two campaigns. Simply put, a similarly-sized error won’t change the projected outcome of the election much at all.
The kind of error needed to produce a Conservative plurality of seats would be enormous. Recall that the Conservatives won the popular vote by more than a percentage point in each of the last two elections yet finished far behind the Liberals in seats. The Conservatives probably need to be ahead by three or four points to have a hope of winning more seats — and double that to win a majority. The polling error in the Liberal-Conservative margin needs to be at least nine or 10 points to give the Conservatives a chance to win the most seats. It would need to be off by 15 points or more to make a Conservative majority likely. It would the kind of error we just don’t see in Canadian elections.
Of course, that’s based on where things are today. If the message is not to believe the polls in the sense that the outcome of the election is yet to be decided, I can get on board with that. They aren’t a forecast of where things will be on April 28. The Conservatives still have time to turn things around and it makes sense for them to communicate to their supporters and volunteers that this isn’t over. It’s not.
But if it is true that the poll denialism is not just meant for external consumption — to keep spirits up and to keep volunteers engaged — but is actually how things are viewed internally, then the Conservatives are in real trouble. It’ll be hard enough for the Conservatives to win a campaign that they are currently poised to lose with 2.5 weeks to go. It’ll be nearly impossible if the Conservative brain trust believes they aren’t losing at all.
Candidate slates finalized
The deadline for candidate nominations passed on Monday and we now know the final slates for all parties. The Liberals, Conservatives and New Democrats are all short one candidate, while the Greens and People’s Party are missing many more. Will those absences make a difference in some ridings?