Alright, let's do this again. And by this, I mean making projections for a federal election. I haven't been particularly successful for federal elections (compared to provincial ones) but I decided I had time to cover the upcoming federal election. If you have any questions, just follow and ask me on Twitter.
My model in 2019 performed okay (making 38 mistakes using the actual percentages of votes from the election) but was underestimating the Liberals. This was mostly due to an incredible vote efficiency from the party of the Prime Minister. Not only did they win most of the close races they were involved in (something like 15 out of 20 but don't quote me on this), they also reinforced their urban vote, which allowed them to continue winning not only the entire city of Toronto, but seats like Halifax or many in urban centers in the West.
Since then, I have updated the model with demographics (age, gender, income and education) in the hope of capturing any shift of the electorate (beyond the top race numbers).
Anyway, here are the first projections of this (currently non-official) 2021 election.
LPC: 181 seats
CPC: 103
NDP: 30
Bloc: 22
Green: 2
You can see an interactive version of the map here. Performance is okay (and it varies with your computer and internet connection) but isn't the best, so allow 1-2 seconds after zooming in or out. Sorry but this is the best mapping software I found and that I can use easily.
The Liberals have been polling well over the last year, fully benefitting from a Covid bump like many other incumbents. However, the current lead is significantly less than it was in spring of 2020. It's still enough to currently be projected with a majority but the number of seats (181) might be giving a false sense of how safe this majority is. The seat total for the LPC is only that high because the Bloc is polling, in average, at only 28%. But this party has been polling all over the map, from being as low as 22% to as high as 37%. The Bloc is also a party that would likely benefit from a campaign (more exposure). I'm not saying the polls are wrong, I'm saying Quebec is unpredictable and I'm not sure the Liberals are really ahead there by over 10 points.
This is important because if that's not the case, then the Liberals are barely above the 169 needed for a majority. Yes they'd make gains in Atlantic, the Prairies and especially BC (really good polling numbers!), but they'd also also lose some urban seats to the rising NDP (Davenport in Toronrot, etc). Without Quebec, the Liberals are right on the edge of a majority.
Maybe the most important element is that the Conservatives don't really have a shot at forming the next government, at least not right now. I have their chances at 2% of winning a plurality (which might not be good enough for Erin O'Toole to become PM) and their chances of a majority at a solid 0%. Campaigns matter but right now, even with a massive polling error, there is no path to 170 for the Tories. This is incredibly important as it creates a completely different dynamic from 2015 or 2019. Trudeau cannot, in particular, scare progressive voters into switching from the Green or NDP to the Liberals to prevent a CPC government. Coupled with the popularity of Singh (by far the most liked leader in the country) and the rising NDP numbers and we might have all the ingredients for a mini orange wave. I'm not talking of the NDP reaching 35% and forming the government (not yet?) but the NDP having the clear momentum for the first half of the campaign and eating both the Liberals (in Ontario) and the Tories (BC).
Anyway, I don't intent on doing an overly long analysis right now. So please find the detailed projections here and see you next time.
Projections Canada 2021 5-08-2021 by bryanbreguet on Scribd
