Polls Are Not Reality

Last Wed I did a polling roundup. Polling continued to be remarkably stable. 538 natl average actually ticked up for us that week. We were +4 and stable in Econ/YouGov, +5 in Marist, +4 in Morning Consult, NYT, +3 CBS, ABC. Steady as she goes. 3/https://t.co/wHpYU9qy6j

— Simon Rosenberg (@SimonWDC) October 21, 2024

That’s part of a very long thread on polls which, ironically, accepts the MAGA framing that polls matter more than election results.

Even after "movement towards Trump" here is the @WashingtonPost battleground state polling averages this morning. The Post has a tighter screen on what polls they accept. It shows Harris at 270+ and winning. 5/ pic.twitter.com/Rgre59T1QN

— Simon Rosenberg (@SimonWDC) October 21, 2024

It is reassuring that recent poll reports are just so much dust in the air, but the underlying MAGA point remains: if polls are so important, why are election results so different?

I expect that most of the major media-sponsored/university pollsters are holding their powder for the final two weeks. So prepare to be inundated.

But this narrative of Harris slippage from the last week is (a) very slight, and (b) based mostly on data from low quality pollsters

— Adam Carlson (@admcrlsn) October 21, 2024

Same song, different verse. Although that one draws some interesting responses:

All polling needs to be a dying profession. It's useless nowadays

— TruckerJohn_P (@H1M_16) October 21, 2024

The only polls I will even bother to consider will have more than 2,000 people polled

— Michael Kowsari (@RealMKowsari) October 21, 2024

Frankly, all polling is no more reliable than this:

Nevada: Vice President Harris Predicted to Win
Vice President Kamala Harris is likely to secure victory in Nevada, thereby gaining 6 electoral votes.

Key Insights from the Forecast
• Vice President Harris is expected to garner support from women, LGBTQ+ communities, and… pic.twitter.com/PmizhplAdF

— Vijay Jyotish (@vijayjyotishusa) August 26, 2024

And is touted for pretty much the same purpose.

It is the reliance on polls that underpins MAGA’s claim that elections are stolen. The polls can’t be wrong (although their methods and even data gathering are less than transparent and subject to all manner of arcane gobbledygook:

Polls are using small samples which means smaller subgroups, complex weighting schemes, oversampling, etc to get desired result. It’s unethical to me. Also, I like this guy’s take:

— Lori Thombs (@LoriThombs) October 21, 2024

), but still we all use them to predict the future, and then decide they are the future! And then when they aren’t, we fight the future! Anything but discard the polls as a fiction we make up to make ourselves feel like we’re in control of what we can’t really control.

Which, again, is what MAGA is all about: reacting to changes in a world that it doesn’t like and can’t control. And here we are giving that foolish argument ammunition, all the while worrying about how MAGA is going to affect the future. 

'Driven by falsehoods': Ex-GOP operative shreds latest MAGA election lawsuitshttps://t.co/ji1wrWL3Up

— Raw Story (@RawStory) October 21, 2024

“Many of these suits seem like vehicles to spread conspiracy theories and misinformation,” she said. “In other words, they read more as press releases than serious legal claims.” 

David Becker, who runs the nonpartisan non-profit Center for Election Innovation and Research, meanwhile, similarly told the Guardian that while the lawsuits are "very unlikely to get the relief they’re seeking, this could later fuel claims that the election was stolen." 

Yes, they could. So maybe it’s time to get off this not so merry-go-round. Spurious claims of in-citizens voting (non-white people!) are more easily swatted away as unserious, but we still think polling is serious and reliable. It is neither. Like recognizing racism is real and systemic (and we can do something about that), it’s time to recognize polls are the problem.

And do something about that.