I scribbled most of this into a note on my phone two months ago while I was on a plane, but then the holidays kicked up and I never got it down, so we’re doing it today. Sometimes, I think that’s the best way to do political thoughts anyway. Let the emotion simmer down. Probably bad for clicks, though.
There are plenty of places where I disagree with the bulk of political commentary, but two things I’d say the prevailing media narrative gets particularly wrong these days are these:
First, not every election is a referendum on Trump. Almost all elections that don’t have Trump in them—whether they be special elections, runoffs, midterms, primaries—are not a referendum on Trump. Trump’s favorability rating has been stable for a year and a half. Results in elections, party by party, have wobbled. That implies two trendlines that are independent of one another. In 2022 midterm exit polling, when asked if their U.S. House vote was to “support Trump” or “oppose Trump,” 54% of voters ended up in the “Trump not a factor” bucket, and of that remaining 46%, there’s little to indicate support or opposition of the former president was a leading cause of the voting decision. Voters do not care about Trump to the degree that Trump discussion draws engagement for media outlets, and they care more about other electoral issues than the prevailing discussion implies, and this shapes political “analysis” in misleading directions.
Second, Trump didn’t perform all that well in the 2016 primary. He won it, which was what mattered, but he won only 44.9% of GOP primary votes, which is part of what made his eventual general election win so stunning. The (admittedly jarring and unusual) phenomenon of the Republican Party broadly, wholeheartedly, unwaveringly supporting Trump developed after those primaries, with over half of Republican voters that year preferring someone other than the eventual president at the time they made their primary vote.
These things bring us to our third thought, which is that as far as it relates to Trump, there are kind of five types of Republican and Republican-adjacent voters in the wake of 2016. This is a broad generalization—there are huge ranges of differences within each type—but hear me out.
The first type is a regular voter who likes Trump. For a stereotype here, think of a financially comfortable white man who believes his vision of America is under threat from liberal ideologies. Again, though, this can look a lot of different ways.
The second type is a voter who tolerates Trump. For a stereotype here, think of that first type’s wife. The Access Hollywood tape where Trump talked about ‘grabbing women by the pussy’ didn’t play very well with women, but a lot of them still voted for Trump. Another stereotype? People who care a lot about tax codes. Yeah, plenty of that type think Trump’s scum, but if he’s going to get the corporate tax rate lowered, they’ll take him.
The third type is a voter who doesn’t like Trump. These are the voters who helped GOP Senate candidates outperform Trump’s margins by a percentage point or so each in Michigan, Minnesota, New Mexico, Colorado, and other 2020 races. The stereotype here is a “Never Trump” Republican. They’re prominent in political media, but they’re rare in the real populace.
The fourth type is also rare, but it’s important: Swing voters who used to lean GOP but have rejected the party’s recent populist shift. From Constitutionalists to moderates, this coalition can swing an election (and did swing 2020, I’d still argue). I belong to this bucket. In 2012, I voted for Mitt Romney. In the 2016 primary, I was ready to vote for Marco Rubio (there was a competence issue in McHenry County, and I was in college, so I never got my ballot). I’d vote for Romney again, but I’d never vote for the Marco Rubio of today, perceiving him to have changed as a political entity. I vote in Democratic primaries now, largely for moderates even though I lean libertarian.
The fifth type is, yet again, rare, but as with the fourth, it’s important: Trump-exclusive voters. Let’s talk about those Georgia Senate races again. Remember this map below, from the Atlanta Journal-Constitution? It’s the turnout dropoff map in 2020 between the Senate elections in November and the Senate runoffs, illustrating the sharp turnout deficit from November to January that plagued David Perdue and Kelly Loeffler in rural areas. There was speculation and anecdotal evidence Trump-provoked doubts about election integrity kept some voters away, but another possibility, one requiring less speculation and relying on fewer anecdotes being treated like data, is that just as Trump’s presence on the ballot brought new voters to the polls in 2016, Trump’s absence from the ballot in other elections left them at home. Why were the polls so correct in 2018 and 2022 despite struggling in 2016 and 2020? Well, what kinds of voters did they struggle to reach in all four of those elections?
This is a lot of speculation on my part and a lot of oversimplification, but compared to the prevailing conventional “wisdom,” is it any more egregious in those departments? For Pete’s sake. Even The Dispatch, of all outlets, was talking about “vibes” in the midterms’ immediate wake. At least this is a coherent collection of thought, and not just a prognosticator misusing a word so they don’t have to admit they’re participating in blind groupthink.
If this five-voter dichotomy is useful, and if elections not involving Trump are not referendums on Trump but elections involving Trump are referendums on Trump, and if at least half of GOP voters were voting for someone else as recently as seven years ago when given the chance, what does this tell us about 2024?
Well, the first of the five types of Republican voters is in on Trump, and the fifth is going to show back up. The fourth isn’t voting in the Republican primary unless they get a moderate making explicit plays for their votes, and that won’t make a difference in primaries that are often winner-take-all, at least at the congressional district level. The third is too small to make a difference, so what matters is what the second—those who tolerate Trump—do, and I guess then the relative combined size of the second and third compared to that of the first and fifth. What determines what the second collection does? How compelling the alternative to Trump is. Which makes the persisting coalescence around Ron DeSantis so persistently interesting.
Currently, in the RCP polling average (which is fairly meaningful) DeSantis trails Trump 44.3% to 30.0%, with Mike Pence at 6.7%, Nikki Haley at 3.3%, Liz Cheney at 2.5%, Ted Cruz at 2.3%, and others consequently making up the remaining 10.9% of voters. Traditionally, from FiveThirtyEight’s research (linked again here), Trump’s number would imply something like an 80% chance of winning the nomination and DeSantis’s number would imply something like a 40% chance of winning the nomination. Those numbers are impossible to exist in conjunction, though—the combined probability of mutually exclusive events can’t be north of 100%—which gets us to the real point here, which is that it’s rare for a presidential primary race to be so binary this early on.
One way to describe the three possibilities of DeSantis as an alternative is that he’s either a placeholder, an idea, or a contender. If it’s the first, a placeholder, voters are telling pollsters he’s their guy because they know he’s a likely candidate and they don’t have another one they’re behind right now. If it’s the second, an idea, voters are telling pollsters he’s their guy because they know something about him—that he’s fighting culture wars but not being the embodiment of political chaos could be one summation—and they like what they hear, but they don’t know him well enough to know for sure yet (I follow politics mediumly closely and I have no idea what Ron DeSantis’s voice sounds like). If it’s the third, he’s in business.
None of these three possibilities is necessarily worse or better than the others for Trump’s 2024 electoral chances. Trump has roughly 44% of the primary electorate behind him right now, and 100% of that electorate knows him well and has an opinion on him. On one hand, that’s pretty good for his chances—back in 2016, he wasn’t consistently hitting that number until more than a month after Super Tuesday. On the other, it leaves a window for the Republican Party to do what the Democratic Party did to Bernie Sanders and Elizabeth Warren and Michael Bloomberg in 2020, rallying around one opposing candidate, if they want that, and boxing out the opposition. It’s a small window, if 44% remains the number, but the window is there, and importantly, that second type of voter is malleable.
