(Mis-)Understanding The Electorate
A Math Look At 2022, With Some Help
Part of the reason I got the 2020 election wrong was that I assumed that the surge in turnout that caused 2020 to break all previous votes cast records was, fundamentally, turnout to boot the President out of office, which it wasn't. It was a whole lot of people energized by the President, which - as foreign as a concept as that would be to me - clearly was the case, looking at the data afterwards.
That understanding was why I got the Georgia runoffs correct, because I immediately understood that the Trumpian surge would not be repeated on January 5th as it was on November 3rd, because Trump wasn't on the ballot, and that led to me being the first person with any form of anything resembling a platform saying Democrats were favoured in the runoffs. The question of what it means for the 2022 midterms is one that I've nibbled around the edges of in these pages before, but now, we have an actual tool to rely on, courtesy of the lovely Lak.
Take a look at both North Carolina and Georgia, and you see why both states are better for Democrats in 2022 than some believe. Georgia was a narrow victory for Democrats because of the nature of that very strong GOP turnout in the north of the state and the white parts of the south and central bits of the state, and if you get Brian Kemp's turnout levels again, that result blows out by over 1%, because Democrats still turn out and Republicans don't, as much. Bluntly, the GOP need to make inroads back into the well-off white socially liberal enclaves where they've been bleeding votes with a candidate for Governor who is either the guy who talked about rounding up illegals in his truck and will have to take increasingly Trumpian stances to survive a primary, or some nutjob - Doug Collins? MTG? Bueller? - who won't even try to get back the white parts of Gwinnett or Cobb, or the parts of southern Forsyth that are trending against the GOP.
North Carolina is redder, but you see the same pattern emerging, where the GOP does a lot worse with a 2018 electorate than a 2020 one, because that Trumpian turnout doesn't show up when Trump isn't on the ballot. This state that everyone thinks of as this reliably Republican state doesn't look so Republican after all, especially if the relative power of the major urban areas in the state increases in 2022.
This map of Charlotte, from Sabato's Crystal Ball's J Miles Coleman, shows that even in defeat, Joe Biden substantially overperformed Barack Obama in Mecklenburg, meaning that the reason Obama won, and Biden lost, was a much worse position in the non-urban centers in the state, proving once again that it was the rurals which won it for Trump - which we know, but is worth confirming. And, if it was the rurals which won it for the President, then the GOP are in a load of trouble if the rurals don't turn out again, which they didn't in Georgia 2021 or Kentucky 2019 or Louisiana 2019 or Virginia 2017 or 2019, or the midterms generally.
Analyses which begin with the premise that the GOP are going to have a good 2022 start with the basic idea that midterms are bad for the in party, and then hand wave away a lot of the problems with what actually caused 2010 and 2014 to be bad years for Democrats, which is the electorate got more heavily white-degree holding. In 2014, the share of the national electorate that was white, and had a college degree, was 35%. In 2018, it was 33%. 2016 and 2020? 30% and 31%, respectively. That used to benefit Republicans, when some Ancestrally Democratic places were still sort of Democratic and the GOP could rely on winning places like Cobb and Gwinnett, as they did in 2014. Now, that same electorate would be hugely beneficial to Democrats, because all those degree holding, well-off white social liberals are voting for Democrats, like my good friend from Southlake, whose last time voting for a Republican was the 2014 midterms. Analyses which believe that 2022 will be bad for Democrats fundamentally misunderstand the world we live in now, because they don't realize how precarious the GOP coalition of voters is when turnout takes a dive.
The corollary to this is that 2024 is probably going to be a bad year for Democrats, at least in terms of the Senate, where Montana, West Virginia, and Ohio are all up. Fortunately for the interests of honest punditry, I have all three of those as likelier than not to go to the GOP, so this isn't just some pie in the sky Democrats are fine analysis. But midterms, far from being harder for Democrats than general elections as I used to believe, are now, all else being equal, better for Democrats than Presidential years, and Lak's great electorate composition tool shows us that well. Obviously, voters could change their minds, making this exercise irrelevant, but there are a lot of reasons to believe that 2022 will be a very good year for Democrats in both North Carolina and Georgia, if the patterns hold.
Many pundits list redistricting as a reason for democratic pessimism in 2022 and 2024. I'm curious to hear the Scrimshaw take on the relative importance of redistricting.