Research Co And The Other Form Of High Class Bullshit
On Polls And Pundits
Last time I wrote about a controversial, potentially outlier poll, I referred to most poll analysis at midterm of a government as High Class Bullshit. I was open and honest about the fact that at many times in Ontario, I was doing a form of it, and I think that whether people know it or not, drawing some form of conclusion about an election based on polls months or years out is mostly an elevated form of elitist bullshit. You can be right - my January 2020 declaration the CPC couldn’t win (what ended up being) the 2021 election was a very well done version of the art, and it ended up being correct in both process and outcome. But to take too much of a victory lap about a take written before I was aware what COVID was, let alone before lockdowns and all that, is the height of bullshit. I was right, but I was also incredibly lucky that none of the entirely unforeseen and unforeseeable things that happened between me writing that and the election ended up biting me in the ass.
I think the concept of High Class Bullshit is important, and I think that’s true when the polls say Trudeau’s doing well as much as when they say Poilievre is inevitable. Normally, I wouldn’t write about Research Co, which had Liberal leads in both Ontario and Quebec and is the only poll so far released with field dates after Global’s Han Dong allegations came out. But, it’s worth thinking about it in the context of the other prominent form of High Class Bullshit - pundits who think they know the answer and refract everything that happens through that lens.
…
One of my favourite things to think about is why the political press and the pundits get things wrong about elections, and there have been two recent British elections where the press and the parties have been wildly wrong. In 2017, everyone assumed that Theresa May would turn a majority of 12 into a majority many multiples bigger, before she would end up losing the majority altogether. Two years later, her successor got the majority she was supposed to, despite two years of chaos in between and a lot of spilled ink about Jeremy Corbyn’s chances of becoming PM.
In both cases, the political press worked backwards from a prior belief – Jeremy Corbyn is incredibly unpopular and will never come anywhere close to power (in the run up to 2017), and the Tories/Boris Johnson can’t win given the age of the government and how chaotic it was (in the lead up to 2019). In both cases, there was some justification for it – in 2017, even as the polls slipped, the fact the Tories beat their polls in 2015 and the historical evidence of how a strongly and unabashedly left wing leader had done in 1983 constituted a decent rationale for the idea May was on pace for a near-100 seat majority.
In 2019, the idea that a government of that age would gain seats was absurd if you cared about history – in 1987, 1992, 1997, 2005, and 2010, all elections when the government was 8+ years old – the party in office lost seats. To go back to a time when a government at or older than 8 years had gained seats, you have to go back to 1959. Obviously, both were wrong, but there were cases for both outcomes at the time that weren’t hackish to make.
What did happen for a lot of that time, however, was reporters would refract everything about politics through the lens of the outcome they knew to be coming, so we’d end up with things like a former No. 10 Downing Street advisor saying at a live event subsequently released as a podcast a version of “I don’t know how, but she’ll win a majority of 100” in a light casualness that is usually reserved for sports. It’s why the New Statesman spent the entirety of the 2017-19 period talking about when Corbyn wins, not if. It’s why even as late as late November 2019 Daniel Finkelstein, Times columnist and Tory peer, was talking about what happens if the Tories win less than 320 seats – a thing that was quite literally never on the cards in that campaign.
What happens quite frequently is people think they know the end state, and they work from the end state backwards. In the Canadian context, it’s that Poilievre will win, according to the many Serious People who remember that it’s rare for Canadian PMs to get 4th terms from their Intro to Poli Sci classes and use that instead of actually thinking about the here and now. “The defences of Trudeau will be what helps elect Poilievre”, Matt Gurney’s classic of the genre, is what sparked all this, because Gurney has decided that Poilievre will win and therefore Liberal partisans defending Trudeau over foreign interference will move meaningful number of swing voters.
The problem with these takes – as someone who did this bullshit for the best part of a year going into 2020, where no matter what happened it was Good News For Joe Biden! – I know what I’m talking about with this shit. It’s toxic to politics, but more importantly it’s analytical garbage. Maybe foreign interference will end up such a problem for the government such that it helps Poilievre, but a lot of pundits have decided it will, but ignore the real possibility that Poilievre and his MPs being unable to stop meeting with Nazis will hurt them. It’s heads I win, tails you lose punditry, because no matter what happens, it’s been decided by the Smart People that the Liberals are losing steam and will lose.
Nothing in this column constitutes an argument either for the Liberals being impervious electorally nor that they should be. I am on the record saying an independent investigation is important here. But just as many times Liberal partisans are blinded into arguing Trudeau and his government is flawless and actual scandals are nothing burgers, a lot of people end up in situations where they refuse to acknowledge that their blind belief that the voters will elect Skippy is lacking in actual evidence.
Anyone who has followed politics in this country for the last 20 years knows that provincial governments are more and more lasting closer to 15 years than 10, that fourth terms are very possible, and that bad oppositions do save governments that maybe don’t on their own record deserve to be re-elected. Acting like a government that would be essentially a coin flip to win again if the election was right now – during a bad economy and the most persistent inflation of my life! – is destined to lose because of vibes or amorphous nothingness is a recipe for sleepless nights or a liver transplant at 40.
No, Research Co doesn’t matter, and pretending that the one poll that says something you like is Very Important is bad faith hackery. But so are the pundits claiming that this government’s demise is inevitable. It’s the other form of High Class Bullshit, and as a lot of people use Chinese interference to project doom on the Liberal grave, it’s worth remembering it for what it is.