Virginia: 5 Days Out, McAuliffe Favoured

2020's Mistake Rearing Its Head Again

Let me describe an election: a candidate and party is acting worried about their chances, despite a small polling lead in the public polls and an opposition feeling like they have the wind at their back. This party has the partisanship of the state at their backs, but the head of their party is languishing in unpopularity and the recriminations are already starting. Pundits are using the supposed closeness of the race to doom about other races, and the conventional wisdom is that if even this race is close, one party’s on the door of a landslide.

What election am I describing?

…

You could think I’m talking about Virginia, but I’m not. I’m thinking about two states south of it, and the worst mistake I made in 2020, namely South Carolina Senate. Lindsay Graham thought he was in trouble, which we know from the way he postulated himself at the foot of the Fox News altar for the final month of the race. I know his campaign thought it was close, but at the end of the day he won by 10% - in other words, he won by the amount South Carolina Republicans win by. Now, obviously, thinking South Carolina was going to be close is much, much more indefensible than thinking Virginia will be, and this isn’t in any way a defence of my worst take. I was dead wrong, and this is in no way an attempt to deflect from that.

But what this is is a reminder that polling is shit. If you take the last 10 polls of the race listed on Wikipedia, the average lead was Graham +2.2%. Morning Consult - not exactly a Dem pollster - had Harrison +2 at one point in late October. Again, none of this is to defend my take, but we have to remember that a lot of polls saying one thing doesn’t mean anything. Why did Graham beat his polls? Because it’s fucking South Carolina, and Democrats don’t lose by 2% or win there. Like, I scoffed at that basic an analysis in 2020, but others were right - it was a red state, and expecting it to sprint that far left that fast was always ludicrous. And so is the idea Terry McAuliffe is in trouble.

Look, I know nobody will give me the benefit of the doubt here, I really, sincerely do. I don’t get that anymore in the US, and I get that, I really do. 2020 killed the notion I will get that for a while, and I haven’t earned a second chance yet. But it’s also clear that McAuliffe will win, and that everyone buying into the Virginia hype is making the exact same mistake I made in 2020.

Let’s look at Wednesday’s two Generic Ballot polls and talk about the national environment, which is either pretty Democratic compared to 2020 (if you believe YouGov) or a point more Republican than 2020 (if you believe Morning Consult). 538’s Generic Ballot tracker is sitting at D+2.5% right now, which would represent a less than 1% swing since the 2020 House result. There was a state special in New Hampshire this week which showed a large overperformance for Democrats compared to Biden’s partisanship, and while there has been a 6% average GOP overperformance in specials since sometime in the summer (compared against 538’s partisanship metric), it’s patchy, and mostly been a story of Democrats doing well in Democratic areas (the seat held Tuesday, the Bedford NH gain, the closer-than-2020 result in the Democratic-trending suburban Iowa seat the night of California’s recall), and getting creamed in GOP areas (and that one uber-wealthy Connecticut seat, which goes against trend). Is this indicative of a national environment where Democrats lose Virginia?

Okay, but what about his terrible approval in New Jersey and New Hampshire and Maryland and Virginia that people were pointing out Wednesday? I mean, US state polls still suck, as I wrote about yesterday, but even beyond that, remember the West Virginia Trump +18 poll that had everyone convinced that the swing really was on? Trump won the state by 39% still. Ah, but approvals are different, right? Well, we don’t have Trump approval for that West Virginia poll, but the NYT/Siena Montana poll had Trump’s favourables there at 52/45, and a 49/43 lead on the ballot test. Trump would win by 16%.

Go to 2018, and 16 of the states that had Governor’s races were reasonably-well polled and had a 3% miss - and 14 of those 16 had a miss in the direction of the 2016 state winner. Like, we know that polling misses are correlated to partisanship, and that Democrats are going to be overstated by polls in Iowa and Ohio and undersold in Democratic states, because that’s what’s happened the last five years. We know what this is, and we therefore know what happens next.

I know a lot of my American readers have their eyes glaze over when I write about or tweet about Canada - like, I get it, it is what it is, I have to manage an audience with three discreet corners - but we saw this in Canada, when arguably Canada’s second best pollster just lost their fucking minds for two weeks in August. I knew the polls were crap and I said as much, and I held firm on the idea the Liberals would be home and dry. Shockingly, they won comfortably, though not a majority - but even then, I always said the majority would come down to Quebec, and I was right.

“The first thing that you want/Will be the last thing you ever need” is playing as I type this (I’m not lying about the amount of Wilco playing these days), and I think it’s fitting - everyone wants the polls to say good things for them, and I include myself in this. My brain knows everything I just wrote, and knows that the polls are almost assuredly wrong, whatever they say. I knew that California Recall SurveyUSA poll was just straight up shit, and yet it still fucked with my head, because I wanted the easy reassurance. I never got it, and I was right there too, as there was a 8% polling miss, undershooting Democratic margins. And yet, here I am, again.

I refuse to say there is zero chance Glenn Youngkin wins, because I have been too wrong about the US in recent years to say that again, but a Youngkin win would be genuinely shocking. And that’s why it won’t happen. The Virginia polls are almost certainly wrong, the national environment still looks good for Democrats, and the state polls that say this is lineball are of worse quality than my attempts to flirt with women back when I was in the closet.

In Canada, I told Democrats to keep breathing, and hold their nerve. I repeat the same advice now. Terry McAuliffe is heavily, heavily favoured to be the next Governor, everyone saying it’ll be close is making the same mistake I made 12 months ago, and it will inevitably end the same way.