History Isn't Decisive

The Past Isn't Predictive

One of the things about Canadians is, as much as we like to make fun of Americans for being terrible, we don't always note just how ass backwards y'all can be. We remember the fact your health care system sucks ass and make fun of you for electing Donald Trump, but we don't really delve deep into how fucked you are. I'm thinking of this, of course, because I'm reminded of how it was only this century that anti-gay sodomy laws were struck down across the US, with 14 states having those laws on the books in 2003. We got rid of those laws in Canada in 1969, and it took y'all to the fucking mid-2000s. Brilliant job, everybody.

And yet, despite this record of bigotry, it only took the US 12 years to get from Lawrence to Obergefell, a stunning turnaround and a truly impressive accomplishment, not that anyone ever would or could have guessed that the march of gay rights forward - so slow and tortured for so long - would speed up so quickly. Things changed, and they changed rather fast. Old rules of how things would work went out the window, because the world of 2015 and the world of even 2003 were radically different. Even more so, 2021 is even radically different from 2015.

As, of course, will be 2022.

…

Yes, this is a politics column, so let's revisit the argument that historical priors matter. They don't. Here's an incomplete list of reasons why.

Democrats overperformed their polls in Texas and Nevada in both 2016 and 2018, in large part because of overly Republican Hispanic samples that didn't end up turning out on the day. In 2020, people who assumed the same thing - also known as me, amongst many others - got burned by a large Texas polling miss that lowballed the GOP, and got the Hispanic vote pretty correct (while botching white voters).

Democrats had always done worse in Georgia runoffs than in voting on election day in November, until this pesky little double Barrel Senate runoff in Georgia in January. Had you assumed that history would repeat itself, you'd have been wrong.

British Conservatives had never won seats like West Bromwich West or Sedgefield before, even in huge, Thatcher landslides, so there was no way Boris could win them. Spoiler, both went Tory in 2019.

On the other side, Putney, a posh London seat, had only been won by Labour in the two Blair landslides, so there was no way they could possibly win it while having such a bad night across the UK, right? Nope, they won it by 10 in their worst result in nearly 40 years.

If you're reading this right now, you probably know who I am, either directly or indirectly, because of those last two calls. I was a cocky bastard because my previous home got the UK, and Canada, very very right, because we allowed the data to guide us, and not the history. My written work was often blowing up that history, and I was right, because people just understood what happened in the past without attempting to understand why that was. Trusting history in those cases was a disaster.

Trusting history would have served you no better in the US as well, as increased partisanship and polarization came home to roost. But, for some reason, after an election cycle of just truly batshit, history breaking events, we're just going back to "well, midterm bad for the President's party?" Vraiment?

If you want to avoid being too wrong - or, more accurately, if your goal is to not be more wrong than the consensus, then fine. Peddle the fiction that the midterms will be better for the GOP in 2022 for sure. But that's not the same thing as it being, you know, true.

Believe me, I know how badly it sucks when you are more wrong than consensus - it sucks! A lot! But that cannot be a consideration in these conversations. Slavish devotion to history as an end in and of itself in political forecasting is about as interesting as Super Bowl betting trends. What happened years ago doesn't mean shit for Chiefs-Bucs, and yet I'm sure we will hear all about it this week.

If you believed history said that Democrats would lose the Georgia races and now you're using history to predict a R-leaning 2022, sit down and shut up. You're not so much looking to history for the answers as looking to find your way to your belief.