The state of the parties

Dems as the party of the rich, party ID swings Democratic, the Jersey fundamentals, the science of open-seat Governor's races, redistricting war widens, where Gen Z men and women agree

No. 379 | October 24th, 2025

🐘🫏 There’s more to party affiliation than just issues

One area where I do agree with the “pollingism” critique is that you can’t just tally up all the areas with parties on the issues to get a read on their strength. For example, on the surface, people are more aligned with the Democrats than the Republicans on their prioritization of the top issues.

And you can argue that Democrats have a strong hand on the question driving the shutdown — even with framing talking about how the ACA subsidies are going to higher-income people and were due to expire anyway.

An important clue to why Democrats aren’t completely dominant lies in another question Echelon asked this month: which qualities do you associate with either party? And while Democrats are competitive or narrowly lead on some attributes — coming closer to your views, caring, stability, having the right priorities — where they fall short is on execution, where Republicans have a massive advantage on things like “has a good plan and follows through on it” (R+8), “strong and decisive” (R+20), and “says what they really think, even when it’s not popular” (R+24).

Many of these characteristics are ones that are personally associated with Trump. They may not extend to other Republicans once he’s gone. But as long as he’s here, voters associate the party with strength, decisiveness, and action.

And a better ability to get things done overrides a lot of things, including extremism. Take a look at this question, where voters — especially working class voters — are willing to tolerate some extreme views that they don’t share if they think a party is better at getting things done.

Despite Democrats still being somewhat identified as a party that cares and is aligned with voters on healthcare, the reality of electoral realignment continues to stare us squarely in the face. Brody Mullins brings together an impressive array of data to show that Democrats are now the party of the rich, representing two thirds of the the top quarter of districts by income.

I highlighted many of these same datasets in Party of the People, coming out in paperback and with a new epilogue on Election Day, November 4th.

What’s the bottom line strength of the parties today? Gallup rattled the political world a bit this week with an estimate that leaned party ID had swung all the way to D+7 after Republicans led last year.

This has some directional insight, but I want to use this as a jumping off point to discuss why some polling (like Gallup) moves around more than others.

Party ID isn’t fixed. It moves. But a 13-point swing in net party ID from the 4th quarter of 2024 is unlikely, even if it might be directionally right.

A clue lies in Gallup’s unleaned numbers, where 47% are independent, compared to 27% Democrat and 25% Republican. This high independent number is a wildly high number that you would never see in any regular poll, where independents are generally a third of the sample. And it’s important to note that this is a survey of adults — a population that parties and campaigns rarely care about, since many can’t even vote. Overwhelmingly, nonvoters and the unregistered won’t identify with either party. It’s generally a safe rule that if you care enough to identify with a political party, you’ll vote at least some of the time.

But a big part of the swing to the Democrats, 5 points, comes from people who initially identified as independents. And this independent sample is uniquely politically un-engaged, including more nonvoters and unregistered adults than you’d see in any other poll.

It’s also worth noting that while Gallup still publishes party ID, approval, and issue numbers, they dropped out of the field of election polling after they showed Romney winning in 2012. In that time, election polling methods have been dramatically overhauled, relying much more on voter files that ground pollsters much more in the reality of who actually votes and doesn’t. Registration-based sampling (or RBS) generally prevents huge swings you see in Gallup polling. And Gallup hasn’t needed to adopt these modern election polling methods for the kind of polling it does.

A good rule of thumb is that pollsters that don’t use RBS and have a looser sample of adults will swing around more than polls of likely voters using voter files, both because they don’t have as rich of a data source to weight off and because they’re talking to a much wider audience that’s more likely to swing around.

But that doesn’t mean Gallup is entirely wrong, any more than it means that the same pollsters showing Biden in the low 30s in 2024 were entirely wrong about what was going to happen.