Beware the normie Democrat
The polls missed again, and TV still matters
The polls missed again.
Last night, Francesca Hong narrowly lost to David Crowley by 4,000 votes in the Wisconsin Democratic gubernatorial primary race. Normally, a close primary result for a DSA-backed candidate in a purple state would be seen as an accomplishment. But the Left, and much of the media establishment, expected a blow-out based on Hong’s huge lead in the polls.
This race saw a lot of movement, with two candidates making late-game departures and Crowley dropping out and then returning to the race in mid-July. But the last public poll, by State Navigate on August 7th, had Hong at 44% and Crowley at 22%. Even a poll sponsored by Crowley’s campaign had him trailing by 18 points.
He ultimately won by less than a percentage point. That means the polls overestimated Hong’s support by around five points and underestimated Crowley’s by seventeen.
This polling miss is reminiscent of what happened last Tuesday in the Michigan Senate race, when the polls undercounted the support of Haley Stevens, the moderate candidate, by a number that was far outside the margin of error. Abdul El-Sayed, the progressive favorite, still won. But these two polling misses have been disconcerting to the insurgent Left movement that expected two impressive primary victories.
Should they be worried?
First and foremost, it’s important to understand that primary polling is hard. General election polling benefits from having a stable-ish set of partisanship benchmarks, where you can reliably guess that half(ish) of the electorate will be Democrats and half(ish) will be Republicans (though 90% of the value is correctly estimating that “ish”).
In a primary, you lack these benchmarks entirely. This is one reason why primary polls are so often wrong; it’s just really hard to determine who is going to vote in the primary. Much of the polling in Michigan and Wisconsin very likely undershot the percentage of the primary electorate that is seniors, lower-income, and doesn’t have a college degree—these are demographics that are usually less ideological, and consequently, less interested in factional Left candidates. They are also, famously, less likely to respond to polls. Say what you want about both El-Sayed and Hong’s campaigns, but their supporters were much more energized than their opponents.
You know what fired-up voters like to do? Take polls. But when it comes to election day, their vote counts just as much as their less-fired-up, less-poll-responsive Democratic primary counterparts.
That’s why I think every Democratic primary poll needs to come with this big caveat: Beware the normie Democrat.
Another reason why it’s hard to poll a primary is that candidate choice is simply a lot more fluid. As crazy as it might seem to some online pundits, Democratic candidates generally appear quite similar to the average primary voter. And many of the decisive voters are generally late deciders. According to Anthony Chergosky, a professor of political science at the University of Wisconsin, “It is well known that voters in Wisconsin tend to make up their minds, tend to start paying attention to primary elections pretty late in the game.”
So when a race has a lot of candidate movement, like in Wisconsin, this effect can be magnified. And polls are fundamentally snapshots, which don’t tell you how the electorate will consolidate behind a candidate. Again, the normie Democrat is the voter to watch. They’re going to be less likely to tie themselves to a hyper-ideological candidate like Hong, and will be more persuaded when a candidate is backed by a popular statewide politician (like when Governor Evers endorsed Crowley in July).
You know what else the normie Democrat likes to do? Watch TV.
Hong’s campaign spent only $100,000 on TV ads and another $135,000 on digital.1 Crowley, on the other hand, spent a combined $3.3 million on ads, and most of this was on linear TV.
I am a strong proponent of the idea that winning or losing is only a very rough metric of campaign effectiveness. Hong’s campaign was still generally successful compared to what you would expect from an insurgent candidate in a purple primary. But I think it’s clear (and we saw this in Michigan) that flooding the airwaves with positive ads is still helpful, especially with older voters who both reliably vote and watch TV. This pattern held in the Republican primaries last night too, where two Trump-endorsed gubernatorial candidates lost while being massively outspent on traditional TV ads.
Every election cycle, people forget that TV does, in fact, matter. A few campaigns, usually they’re lefty and young, experiment with an aggressive ground game and ignore the role of traditional TV. Media outlets cover this “new trend” and speculate if traditional campaign strategies had just been upended. Then, election day comes, and it turns out traditional TV still matters. That’s why campaigns still spend so much money on it!
This time, the guilty party was the New York Times, which ran a piece two days ago titled, “Inside the Left’s Playbook for Winning Elections.” The focus was on Hong’s insurgent populist ground game.
Left-wing campaigns also argued that some of their more mainstream primary opponents had erred by spending most of their resources on advertising while failing to actually excite voters with direct outreach or by crafting positions that appeal to them. In Wisconsin, Ms. Hong’s campaign has recruited about 7,000 volunteers, knocked on hundreds of thousands of doors and hosted roughly 500 canvassing events. Her most significant remaining rival, David Crowley, the Milwaukee County executive, held a fraction of that number.
The article cites the Mamdani model as inspiring Left organizers to focus on decentralized organizing-oriented campaign tactics. But guess what? In addition to having a robust ground game, Mamdani invested heavily in traditional TV ads, and he overperformed the polls.
This is reminiscent of Jerusalem Demsas’ messenger class theory—the idea that the people who shape public debate miss the lived experience of the rest of the country. In this case, the media and political elites, a group that I do very much identify with, simply do not watch linear television. But their fellow Americans watch a huge amount of TV every day, especially older Americans, who are disproportionately likely to watch linear TV. Those older people are your peak primary electorate. You have to reach them where they’re at, even if it’s boring.
Here’s the last thing to know about normie Democrats: They like Thanksgiving.
As I noted earlier in the piece, there was scarce recent polling that measured Hong’s support as we got closer to election day and more voters began doing research to make an informed decision. During that time, Hong did more interviews and received more media coverage over some of her previous (Woke 1.0) tweets. Famously, it was revealed that she wanted to cancel Thanksgiving because of its association with the atrocities committed against Native Americans.
Regardless of the merits of that argument (The Messenger Class is probably more inclined to like it), this is just not the sort of take that appeals to voters in suburban and rural Wisconsin. Again, there’s no way to know exactly how Hong’s association with unpopular culturally Left ideas hurt her campaign. But it’s simply true that these issues received quite a bit of coverage over the final few weeks of the campaign, and this was exactly the time when voters were breaking towards Crowley.
It’s also important to note that primary voters can be strategic. They want to nominate a Democrat who can beat Republicans. They want someone with that prized ability to be “electable.” That was the message Crowley and Governor Evers made down the stretch of the campaign, and it was simply not an argument that Hong could make persuasively, especially when she failed to denounce “defund the police” or clearly articulate why she no longer held some of her past culturally woke beliefs.
None of this should be taken to discount Hong’s campaign or the rise of the Left this cycle. But here’s my message to any pundit or political obsessive who consumes a lot of election content on the internet: Beware of the normie Democrat. They vote, and they do not have the same cultural habits as you do.
It’s worth noting that the Republican Governor’s Association backed her with another $3.5 million in “attack ads” that were used to make her more appealing to hard-left Democratic primary voters, but too extreme for general election voters. But these were not the sort of well-designed ads Hong’s campaign would have run on their own.