Canada Reset: Ontario, Polls, And Can The Tories Win?

On Maps, Math, And Canada

“What’s my theory about Ontario? I don’t know, everyone’s in fucking Muskoka.”

On Tuesday I took a call with a buddy of mine involved in Liberal politics that I met years ago in university. This friend of mine has always respected my thoughts, so he asked me what I made of the polling, and the upheaval and everything else, and that was my response.

The dynamic in Ontario is weird right now, but getting less weird by the day - Mainstreet and EKOS showed huge anti-Liberal swings on the weekend which were not matched by Abacus, Leger, or Angus Reid, and now those trends are mostly in reverse, with both Mainstreet and EKOS showing decent Liberal improvements in Ontario in their most recent publications. A small Ontario lead isn’t what the Liberals want, but it’s also a far cry from the disaster that a 4% or 5% Conservative lead would represent in Ontario. The former is a Liberal minority with maybe a few less seats than last time, the latter is genuinely an unclear election result. And yet, I never believed the latter would happen.

Abacus, Angus Reid, and Leger are online pollsters, whereas EKOS and Mainstreet call phones. This distinction matters, and in 2019, it was the phone pollsters with more robust Liberal leads in Ontario than the online pollsters. (Ipsos is also an online pollster and showed an CPC lead in Ontario, but they spent the entirety of the 2019 campaign high on the Conservatives, so their polls come with a health warning anyways.) What we saw was the phone pollsters wildly varying their results, with Mainstreet going from a 10% Liberal lead in their Saturday release to a 5% Conservative lead Tuesday, because that really fucking happened. What happened is the polls swung, but let’s be real - did anything actually happen to move the polls that much, and then for half of that swing to reverse itself, all in the span of 5 days? Of course not, don’t be absurd. I have huge respect for Quito and the Mainstreet staff, but it’s clear they got a wonky, wild sample.

Ah, but so did Frank Graves over at EKOS, so it must have been real, right? I mean, what compelling reason would there have been for the Liberals to tank, but not just tank, but tank specifically and singularly in Ontario? It doesn’t make sense for Ontario, and Ontario alone, to become enamoured with Erin O’Toole - unless, of course, they just caught wonky samples on the weekend as the entirety of Ontario takes their first vacation in two summers. Shit happens, and now, the polls have gone from truly unbelievable in the literal sense of the word to “I could see it” levels of doubt. Now that the polls have mostly gotten back to some normal - the range is now roughly Ipsos’ solid lead to Abacus and Angus Reid’s L+6, and a reasonable average would be around a Liberal +2 result in Ontario - the conversation of what it means is easier.

And the answer? Not much.

…

In 2019, the Liberals won Ontario by 8.5%. There are 10 seats the Liberals won at the last election in Ontario with a smaller LPC-CPC margin than 8.5%. A tied province, assuming a uniform swing, would produce 10 LPC to CPC losses - and if the swing followed the pattern from last time, being bigger the further out from the urban centers you get, the losses could be closer to 5, as seats like Whitby, Oakville, and Kanata-Carleton could stay, even on a provincial swing big enough to flip the seats. Now, the Liberals could also lose seats - potentially many, both in Toronto and in many smaller cities - to the NDP, but those wouldn’t change who would be in office, merely the status of Trudeau’s government. But if that’s the sort of provincial result we get, the Tories wouldn’t be getting nearly 20 seats, but probably closer to 8. And that would still leave the Liberals and NDP with enough seats to lock the Conservatives out of power, before even getting to potential NDP gains from the Tories - with Kenora, Oshawa, and Essex at the top of that list.

Let’s say the Tories get that tie in Ontario, and win 10 seats in Ontario. Do I think that's likely? Of course not, but let's go with it. Where else are they gaining seats? Maybe they win a few more out east, but that's unlikely - only Saint John Rothesay, Miramichi, and maybe one of the Cape Breton seats look very appealing, but even then, most of the polling in Atlantic Canada points to the Tories losing West Nova, not them making gains. Call it a wash. There are a couple of Tory targets in Quebec they could get, if the Bloc vote declines, let's be nice and give it to them. They won 121 seats last time, let's call it 133 now. The problem is, they're out of potential gains now.

They're probably going to lose Charleswood to the Liberals, Saskatoon West to the NDP, and even potentially the northern Saskatchewan seat to the NDP, but let's be kind and call it two net losses in SaskyToba. In Alberta, the current expectation is three Liberal gains - Calgary Skyview, Edmonton Millwoods, and Edmonton Centre - all at the Conservatives expense. Now, the Tories are back down to only 128 seats, barely up on last time - and this is with fairly generous assumptions in Central Canada and the East. Oh, and in BC, the Tories only have downside. Fournier’s latest update has the Tories down 2% on their 2019 performance there and the Liberals up 2%, but that’s on the low side of what some pollsters have shown, with EKOS showing 2019’s 8% CPC lead over the LPC down to a tie, and both Abacus and Leger showing LPC leads in BC. Either way, there are 5 seats where the Tory lead over the LPC is under 7% after 2019, and they’re also potentially in real danger in Kootenay-Columbia to the surging NDP, so even if you say they’re not going to lose anything in BC, they’re not gaining anything either.

For the Tories, a fairly rosy assessment of their chances puts them in the range of maybe hitting 130, but beyond that, the road is tough. They’d have to start flipping Liberal seats in Ontario that the Liberals hold by 10%+ margins currently - suburban areas that swung wildly to the Liberals last time, and that are the Canadian equivalents to the areas swinging left everywhere in the fucking world - for them to have a realistic chance to get to 140 seats and a chance at forming government in minority. Even with this rosy assessment of their chances, they don’t get within 20 seats of the Liberals, unless the Bloc either surges hugely (which is neither impossible nor particularly likely, as of right now) or the NDP really eats at Liberal seats in the metros - but even then, the NDP would be on track for 50 seats, meaning that they’d be the kingmakers, and put Trudeau back into office, whatever bullshit attempt at gaining leverage in post-election day negotiations Jagmeet is trying by refusing to explicitly say so now (after saying so explicitly in February).

I wrote in January 2020 that Andrew Scheer’s parting gift to the Conservative Party was an election map so devoid of easy gains that he lost them the next election with his performance in the 2019 one, and this point has been made by these wild polls. Even at the wildest of the polls, the Tuesday Mainstreet forecast - with the Liberals at 19% in BC and down 5.5% in Ontario, they still had the most seats and the Liberals and the NDP still had more than 170 seats between them. That’s how devoid of plausible gains the map is because of 2019, and that’s why I wrote that the next election was over before it was even an idea in Justin Trudeau’s mind. And that’s why I’m still right.

There is no path to the Conservatives winning 140 seats that doesn’t fail a very basic smell test - basically, them winning Ontario by 7 or the Liberals falling to 2011 levels in BC - unless they get there by winning wide swathes of Bloc seats in Quebec, which, putting to one side the question of whether that is plausible or not, wouldn’t make a damn bit of difference to their chances of governing, because so long as the Liberals and NDP have 170 seats, the Conservatives won’t govern. Oh, and while a theoretical Bloc collapse would hand the Tories a ton of northern and regional Quebec seats, the Liberals would sweep all those off-island pseudo-suburban Montreal seats I’ve written about in the past, so that wouldn’t even be a net good for the Tories.

I have tried, deeply, to figure out what the path is that I’m not seeing, I really, truly have. I am looking for this not just out of professional instinct - a more compelling campaign means more people reading this site, which means all the knock on effects of more exposure - but also because I am painfully aware of the fact that I am compromised here. If I say the Tories can’t win - if I boldly stand by the idea that this is already actually in the bag, because this isn’t a contest of narrative, but of math - then I will have every bad 2020 take I ever sent out thrown back at me. I know this, and I know that the smart move is to hedge my bets, wait a bit, wait for more polls, then put my stake in the ground. I don’t care, though, because this site, and my value more broadly, is built on the idea that my thinking at any given moment is valuable, and I’d be lying if I did anything else but gave the honest and true statement of my beliefs. If I’m wrong, there’s little reason anyone should take what I say with much seriousness. But if I am too cowardly to say in public what I say and believe in private, then nobody should give a damn what I have to say anyways.

I cannot find a path to government for the Tories. Maybe one exists that I’m missing, I don’t know. But I can’t fucking find it.