You may have seen the Siena/New York Times poll that was released on Monday, or, because it was so bad for the Democrats, you may have read about it as the media gleefully amplified it. See, it showed that the Democrats were in big trouble and as God as my witness they’ll never write a non-horse race story ever again.
Experts and wags quickly noted a bunch of weird things with the poll’s internals, like the sample was only 792 people. And the sample skewed old and white. The age groups weren’t split along the normal generational lines and they also were weighted in an unusual way. Half of the polling was done via land line, too. Nor did it reflect the gender gap we’ve seen in multiple 2022 post-Dobbs elections.
None of that mattered to the press—bad news for the Democrats after a series of Democratic policy victories was air for their tired DEMS IN DISARRAY narrative, after all. But facts matter, and the problems with the poll itself aside, there are still signs that the Democrats have a chance to hold on.
First, there have been more polls released this week:
New @MorningConsult poll:
– Dems lead 48-45, unchanged from last week
– Biden approval 46%, highest in 2022
– Dems have growing enthusiasm edge, improving scores on the economy 4/https://t.co/eXWnGTs3GH— Simon Rosenberg (@SimonWDC) October 18, 2022
And this:
Know there was a tough NYT poll for Dems but two GOP polls taken on the same days have Dem leads:
Fox 44-41
Public Opinion Strat 45-43A third for Federalist showed it tied, w/Ds gaining.
Majority of polls taken last week had Dem leads. 5/https://t.co/VMR9brskZs
— Simon Rosenberg (@SimonWDC) October 18, 2022
But those signs I mentioned aren’t captured in any of the current likely voter polling. Here’s what’s missing.
One factor is turnout—Georgia has been setting early voting records this week:
Our elections team has reviewed the initial early vote numbers for Day 1 and we know we are north of 125,000. This dwarfs the previous record of 72k from the 2018 midterm 1st day of early voting. We will have exact numbers in the morning. #gapol #PlanYourVote
— Gabriel Sterling (@GabrielSterling) October 18, 2022
And as of Tuesday afternoon:
We have reviewed the turnout yesterday and we did set a midterm 1st day of early voting & we nearly hit the record for a Presidential:
1st Day of Early Voting by cycle
2022: 131,318
2020: 136,739
2018: 70,849
That's an 85% increase from the last midterm. Midterm record! #gapol pic.twitter.com/kHeHlMLoa9— Gabriel Sterling (@GabrielSterling) October 18, 2022
And polling is missing the striking new factor in post-Dobbs elections: the increasing gender gap.
Fwiw, I can find 5 polls that were released publicly for the NY19 special, and none of them had a likely voter model with a 7 point gender gap.
— Tom Bonier (@tbonier) October 17, 2022
And it seems that the registration gains among young people since Dobbs are being ignored by the media and pollsters:
I'm sure many will note that youth reg picks up later in the cycle. All reg picks up later in the cycle. In this super confusing chart I added lines for 2020 – you can see how much bigger the youth spike in reg is this year relative to last cycle. pic.twitter.com/77SsFSl0bL
— Tom Bonier (@tbonier) October 18, 2022
And that the likely voter group needs to be less white:
States w/ substantial early vote reported thus far and where voters of color have increased their vote share over the same point in '18:
CA, FL, GA, ID, IN, MD, MI, MN, NE, NJ, NM, NY, NC, OH, PA, VA.— Tom Bonier (@tbonier) October 19, 2022
Again, I’m not predicting anything. But there’s more to consider than just polling and the media hyping up results that are bad for the Democrats. And in the meantime, we keep up whatever get out the vote activities we’ve been doing.
Barbra can sing us out this time:
I hate to agree with Bill Kristol but he’s describing the worst case scenarists, bed wetter, Nervous Nellie syndrome that I bang on about. It’s the root of poor Democratic messaging. It’s not just the elites.
it promotes nihilism unfortunately.