Evan Scrimshaw: Georgia, Predictions, And The Finest Tradition Of The Service

(We're once again pleased to have Evan Scrimshaw (@EScrimshaw) write this column today for us as a guest writer. Evan has years of experience with global election forecasting and is an expert on Australian, American, British, Irish, Scottish, and Canadian electoral politics. We're excited to feature his work here.)

In the interminable break between Christmas and New Year's here, and with the Georgia runoffs quickly approaching, there's been a bit of a lull, filled in part by the beginnings of a rewatch of an old classic - the British classic Yes Minister. The story - of the never ending machinations of a Minister and his civil servants - is one of the greatest of all time, and the comedy still holds, almost 40 years later.

One of the common phrases - used both earnestly and sarcastically, by both friend and foe - is that Sir Humphrey is following the "finest tradition of the service" in his dealings with his political master. That line has always struck me as a clever turn of phrase, mostly out of a curiosity for that quintessentially British deference to authority and titles over talent or skill.

We're less than a week out from the Georgia runoff now, and we have some more data that suggests Democrats are the favourites. Lak's precinct model thinks Democrats are about 2 point favourites, some modelling I've done privately suggests the lead is closer to 3 or 4, and the Black share of the electorate is still 31.3%, up 3.6% since November's early voting. If that number is 31% or higher by Monday - when mail ballots are counted, even if in person votes are done until Tuesday - then Democrats have to be seen as clear favourites.

On the polling front, we've had a Trafalgar poll showing Ossoff up by just under 3%, which would scare me as a Republican if I didn't remember that Trafalgar's main sin is not that they're partisan, it's that they're bad. That said, it's a bad sign that even Cahaly can't find it within himself to show a Republican lead. The Election Twitter funded poll came out showing Ossoff up 8%, which has led to a great reimagining of the state of the race, especially amongst those who claimed that this poll would provide real insights into this… sorry, nope, everyone just dismissed this poll that was vital to our understanding of the state just days ago, my bad. Nobody could have seen this coming.

What's the point of a political rating? Is it to properly reflect reality, or is the mere act of being "right" enough? And, depending on how you answer that, what's the right rating for Georgia?

The view of this site, or more precisely, Lak's view, is Lean D. That's consistent with the idea that the GOP have a credible, realistic path forward, but are currently less likely to win than Democrats are. I agree with that view insofar as I agree that you cannot make the case that Democrats are anything less than Lean D. I get the instinct to say Tossup, but the case for it doesn't actually exist. Well, it exists, but if you're going to that well, you're no longer concerned with intellectual honesty, you're just covering your own reputation. The case is that rural turnout is all just waiting to be unleashed on January 5th, which would have some credence if it wasn't for two things - that the rural turnout drops of 2020/21 are the exact same as the rural turnout drops of the 2008 runoff, and that GOP voters already did turn out in person in October, so it's not like there is any reason to suspect that surge is coming. Could it come? Maybe, but that's why Lean isn't Safe. Accepting that something could happen doesn't mean I have to think it is as likely as the more likely outcome, which is that rural turnout falls with Trump off the ballot.

What's the right rating for the race? If you ask me now, I'm going to say Lean D. I'm going to default to it, out of a deference to the idea that arrogance is a mistake here, but I'm not sure if I'm right. I keep hearing that phrase - finest tradition of the service - ringing in my ears, and I'm just paralyzed by indecision. I don't think it will be particularly close, I don't - the GOP are in real, real danger here. The quality polls here don't have a close race at all, and the early vote is a disaster for the GOP. Everything says Democrats win this by a comfortable margin. And yet, hesitation remains.

The GOP are facing an electorate that is substantially Blacker than November, substantially less white than November, and the whites are substantially more educated than November - a trifecta of things that mean they're in trouble. Ah, but the suburbs could trend right without Trump on the ballot, say the Very Smart People, except for the fact that every good poll - and even a couple of bad ones - has Ossoff overperforming Biden with whites. Emerson, InsiderAdvantage, SurveyUSA, and the ET poll all have this to some degree or another, so it's not just some small sample size noise.

If I'm protecting what little of a reputation I still have after a bad 2020 cycle, I'm saying Georgia is Lean D. If I'm being honest, it's Likely D. In the finest traditions of the service, it is time to be honest. Democrats are Likely to win the two Georgia Senate seats next week.


Popular Posts