Canada: High Class Bullshit's Limited Value

On The Polls

Is poll analysis this far out from an election still high class bullshit? Yes, of course it is, and anyone who really takes this too seriously should ensure that they understand the English language and most importantly the limitations of this analysis. Plainly, what the polls say now has little value to what they’ll be when the election comes, as the elections of Danielle Smith and Doug Ford remind us, as does the recent rally in Manitoba PC support in recent months as that election closes in. (Manitoba coverage will begin soon, or at least soon-ish!)

So, why am I writing this? Because I can intellectually know this, but Fournier’s site has the Liberals winning the seat count again and the combined left having a majority, and so I broke out my model (on the unofficial but barring changes new lines) and tested it out, because at the end of the day it’s a rather lazy Sunday and I didn’t have anything better to do.

If I just used a straight average of the polls – Nanos, Abacus, Leger, Mainstreet, and Angus Reid – I get the Liberals, NDP, and Greens on 171 seats, one short of the new majority threshold (don’t forget – the new lines increase the Parliament to 343 seats). If I put in a partial byelection effect, that trio goes to 177, meaning that depending on how much we care about the byelections as a predictive measure, the Liberals are probably favourites in an election tomorrow.

Now, that doesn’t mean much – they’re by no means sizeable or big favourites, and I’m sure that if I update in a month the numbers will be different – but I don’t think it’s a bad time to focus on what actually matters for the next election. And, spoiler, it’s not what most people are talking about.

…

The pitch of a more strident Conservative Party is that its stridentness on social issues and their potential deviation from orthodoxy on fiscal matters would give them an opening to win over the kinds of voters who got Doug Ford a seat in Windsor and won the PCs Timmins last time around – working class voters suspicious of the emerging cultural norms. Whether you think that Poilievre will pull off that trick, that’s what the Poilievre leadership is trying to do.

There’s four places where there are seats like that for the taking – Atlantic Canada, where the Liberals lost Scott Simms last time and suffered big swings in their other rural seats, Northern Ontario, Southwest Ontario, and non-Lower Mainland BC. Right now, the Tories are getting the results they need in Northern Ontario and Southwest Ontario they need – seats like Thunder Bay and Sudbury, Windsor Tecumseh and London West all project to fall, even under the byelection adjusted (also known as Liberal friendlier) projections), but that’s where the good news stops.

Atlantic Canada should be fertile ground for the Conservatives, but their polls out there aren’t great right now, and that’s not gonna be good enough for them to win. BC is okay for them, but they need to come close to a sweep of the NDP held seats outside the Lower Mainland, and right now there’s limited evidence the swing will be enough for a sweep.

The problem for the Tories is that yes, they’re not maxed out – they could easily gain a few more seats in Atlantic Canada and BC than I currently have them for even in the less-Liberal friendly projection, but that implies the Liberals are maxed out, and they’re not. Both these projections assume that the CPC will make two gains in Kitchener, one gain in Halton, and the Liberals failing to make gains in winnable seats in Edmonton, Winnipeg, and Quebec.

As likely as the fundamentals tipping over a few more of those rural Newfoundland seats on the basis of the Global Fucking Realignment would be the LPC continuing to outperform uniform swing in the smaller city suburbs like Kitchener or Niagara, let alone the chance of Quebec throwing up some oddity. Now, yes, that could go both ways, but the Liberals have upside as much as it has downside.

If the byelections are even remotely correct as a benchmark then the Liberals are gaining seats and we’re just waiting on a more representative byelection in Quebec to tell us the state of the Bloc-Liberal battleground. But, more than that, where we are right now is a situation where there’s cognitive dissonance between the polls and the byelections, but even then, the polls aren’t anywhere as bad as they appear.

The thing about the electoral map – be it the new one or the old one – is that there are a lot of useless points of movement. The Liberals gaining another 5 points in the Atlantic or the difference between the Liberals at 15 or 25 in SaskyToba or Alberta is actually, fundamentally, 3 or 4 seats, but it’s a swing that makes narratives. Leger – the one that had the Liberals up 2 nationally – wasn’t actually significantly better for the Liberals in seats than Abacus’ C+7 offering, because most of the difference was useless. Abacus has the Liberals tanking where their votes did the least for new seats, and so a big national popular vote difference becomes worth maybe 6 seats.

What matters right now is two things – can this government get their heads out of their asses with the looming reshuffle and the summer break to reset the priorities list, and does Pierre Poilievre look at what happened last week as a success (neutering Bernier) or a failure (Winnipeg and Oxford). The answers to those questions will inform where we end up by the end of the year, and right now only time will tell.

What I feel better about is that the fundamentals of the next election have been set, and what we know is that Pierre Poilievre is not the juggernaut that so many have pretended he is. When faced with the actual prospect of voting for him, voters in 3 LPC-CPC battles have chosen not to. The Liberal vote in all three went up, there is little appetite amongst those voters who actually decide elections to defect from the Liberals, and the Conservative message isn’t inspiring the sorts of swings that Oppositions that are on track to win office get.

The polls right now are high class bullshit, and overanalysing them is a mistake. But the thing is, even the polls don’t actually say Poilievre is on track for anything resembling an impressive result. And if that’s the best the Conservatives can do after this legislative session and the turmoil of the first half of 2023, there might be some value in them after all.