If polls are to be believed, Canadians are about to change government on Monday.
After nine years of a Conservative government, the Liberals appear capable of reclaiming power. A minority is the most likely result, but we can't exclude a surprise victory by Stephen Harper and his party.
The projections below represent our best prediction of what will or can happen on Monday. They are based on the most recent polls published in October. The calculations use past election results as well as the current polls (both national and riding ones) to predict the winner in the 338 ridings. They include regional and incumbency effects.
The confidence intervals and the chance of winning are obtained through the use of 20,000 simulations that account for the uncertainty of the polls as well as for the distribution of the vote and the electoral system. In other words, these simulations try to include every possible scenario, given the information we currently have.
Three key moments in the campaign
This long campaign has had essentially three important moments. The first one, early in August, is when Justin Trudeau beat the (low) expectations during the Maclean's debate. More generally, the Liberal leader managed to re-position his party quickly within a couple of points of the other two. While many will overlook this, we actually believe it was major event. Indeed, had the Liberals failed to do so and continued falling as they had during the summer, they could well have become irrelevant.
Moreover, for the 60 to 65 per cent of the population who have consistently wanted to change government, an early Liberal collapse would most likely have caused the NDP to become the de facto "anti-Harper" party.
Without this early concentration of the anti-government vote around one party, we went a couple of weeks with the NDP ahead, thanks to the party's incredibly good numbers in Quebec. It should be noted that the New Democrats quickly fell to third in Ontario after being in a close three-way race early on. That was possibly a sign of things to come.
The second critical moment occurred in mid-September, thanks mostly to the decision of the Federal Court to allow women to wear the niqab during citizenship ceremonies. Whether we think this debate is irrelevant or racist, there is no denying that it changed the race in Quebec.
Quebeckers are the most opposed to the face-covering veil — although, to be fair, other Canadians aren't much different. Thomas Mulcair's position on this issue — to respect the decision of the court — wasn't shared by a big part of the nationalist voters (former Bloc voters) who had switched parties in 2011.
The New Democrats went from being at about 45-50 per cent and possibly dreaming of sweeping Quebec to slowly but surely falling towards the 30 per cent mark. The Liberals didn't really benefit at first, and it's actually the Bloc and the Conservatives who went back over 20 per cent.
For the Bloc in particular, this issue might have ultimately allowed Gilles Duceppe to save his party, which was heading nowhere and was projected not to have elected a single MP. As it currently stands, Quebec is a four-way race, and it's hard even to gauge which party — the NDP or the Liberals — is ahead.
The niqab debate is not the main issue of this election, but by causing the NDP to fall in Quebec, it changed the narrative of the campaign. After weeks of being favourite, the NDP was now trailing the Conservatives and had the Liberals just behind. By the end of September, it seemed the perfect scenario for Harper was happening, with Liberals and NDP splitting each other's votes. As a bonus, major gains in Quebec were even possible for the Tories.
This is when Ontario decided to pay attention and to rally behind one option in hopes of defeating the Conservatives. During the first week of October and culminating with the Thanksgiving weekend, the Liberals went from being slightly ahead of the Tories to enjoying a 10-point lead and more. Not only were the NDP now far behind in Ontario's polling — sometimes below 20 per cent — it seems Trudeau finally managed to steal voters directly from the Conservatives.
Thanks to the increase in Ontario, the Liberals were now back in the race, and the momentum hasn't stopped since. The Grits can even dream of finishing first in Quebec (in votes) and British Columbia thanks to what appears to be a rallying of some NDP voters to the best option to defeat the Conservatives — that is, voting strategically.
As of the last couple of days, we have observed a remarkable convergence in the polls, with pretty much every one showing the Liberals about 36-38 per cent and the Tories at 30-32 per cent.
Reminder: The polls could be wrong
So, why aren't we calling this a sure win for the Liberals?
Mostly because, while the lead is significant, we need to remember that polls can be wrong. And in order for the Conservatives to win the most seats on Monday, the polls don't even need to be as wrong as they were in Alberta in 2012 or in British Columbia in 2013.
If polls underestimate the Conservatives by three or four points in Ontario, as they did in 2011, that could be enough to give us a 50-50 race. More generally, incumbents have had a strong tendency to be underestimated by the polls in multiple elections in recent years.
Moreover, the polls agree with each other maybe a little too much. Not only do they almost all have the same numbers nationwide, they also show the same gap in Ontario. Forum, Angus Reid, Leger and Mainstreet all have the Tories, for instance, at exactly 33 per cent. Ekos and Nanos are very close as well. Given the sample sizes and the differences in methodologies, the odds of this happening are very low.
We aren't saying it's a bad thing when polls agree, but when they do to this extent, it's a little bit weird. It raises the question whether some pollsters didn't simply decide not to publish some numbers because they were too far off the average. In any case, if polls are off on Monday, please remember this.
The only pollsters showing a tighter race are Ekos and, to a lesser extent, Angus Reid.
Ekos actually has the main two parties tied while Angus Reid has them separated by only four points (one point if looking at likely voters, which accounts for the differences in turnout by age, for instance). Are these firms the outliers or do they see something everybody else is missing? Monday will give us the answer.
Another reason we can't rule the Conservatives out is that their voters are older and more committed. In other words, the Tories don't have to work as hard to get their vote out — although they do have a very efficient machine for that.
Illustrating this advantage are the numbers for the advance vote that took place during the Thanksgiving weekend. Even though the Liberals were surging during that time and turnout was really high, both Ekos and Angus Reid show the Conservatives actually a lot more competitive than we could have expected. Ekos has the Tories at 34.9 per cent and the Liberals at 32.5 per cent. While we completely acknowledge that advance voters are different from the general ones, it remains that Liberal supporters should remain cautious.
Their party hasn't won the election yet.
Let's be very clear, though: a Conservative victory on Monday would represent a major surprise at this point and would seriously hurt the pollsters' credibility (and ours at the same time obviously). A Liberal victory remains the most likely outcome.
What seems less uncertain is the lack of a majority. While the possibility exists — if polls can underestimate the Conservatives, they can also underestimate the Liberals — we are pretty confident that nobody will win outright on Monday. That means that the NDP, despite having a night much worse than supporters were hoping for just a month ago, should at least hold the balance of power.
And this also means we'll most likely have another election in the next 18 months. If the likely outcome happens on Monday, however, that election will probably be with a different leader for the Conservatives.
Bryan Breguet has a B.Sc in economics of politics and a M.sc in economics from the University of Montreal. He founded TooCloseToCall.ca in 2010 where he provides electoral analysis and projections. He has collaborated with the National Post, Journal de Montréal and l’Actualité.
He provided analysis and updates for The Huffington Post Canada throughout the federal election campaign. For riding by riding projections, visit his interactive simulator.
Also on HuffPost