2022 Generic Ballot / Recruitment / Fundraising / Ratings Megathread

[Fmr. Gov. NickG]

More Dark Brandon Energy!

[Fmr. Gov. NickG]

A tied GCB would still yield a GOP House majority (although not an overwhelming one).

538’s model actually believe the Dems are favored to win the majority with a tied popular vote.  I believe their model tends to overrate incumbency across the board, and Dems would probably lose a tied PV, but it’s pretty close.  I think they’d be favored with a D+1.0 PV.

[Fmr. Gov. NickG]

The way that there are never any articles about the GOP's total collapse with 18-29 year olds/Zoomers/Millennials is stunning. It's almost like they completely get away with not even bothering to try to appeal to that age group.

That's because Zoomers don't read articles.  I imagine there are plenty of TikTok dances on this topic.

[Fmr. Gov. NickG]

https://twitter.com/henryolsenEPPC/status/1564970235157004289

Uh, who do you think all those new Independents are? 

[Fmr. Gov. NickG]

So I agree with prognosticators who say the GOP is a virtual lock for the house.

HOWEVER: With the generic ballot being around tied .. even Dems up a few points. How are others coming to that conclusion with the generic ballot favoring Dems?

“With polling showing Dems tied or even ahead in polls, how can you say Dems are ahead”

What I’m saying is how are people getting to the conclusion the house is a lock for the GOP when the generic ballot is between tied and D+2… that’s 2 points off from a Dem win.

Why are you asking this when -you- are the one saying the GOP is a lock to win the House?

[Fmr. Gov. NickG]

Please check this out...
https://twitter.com/IAPolls2022/status/1571605134911672321

So while the Democrats are GLOATING about the NBC Poll these underlying crosstabs are increasingly bad for them. If this is accurate Democrats are finished certainly in the House.

DeSantis & Abbott looks like were the smart ones here bringing back the Issue of Border/Immigration in their States where Republicans have a decided edge based on those Numbers.

When folks are saying the Cost of Living is more important then Abortion by 22 Points that is a MASSIVE, MASSIVE Red Flag for Democrats particularly on the House Side of things.

Democrats have IMO overplayed their hand with the Abortion Issue.

Do you really think the results of this abortion question support the point you are trying to make?
Do you understand how absolutely huge it is for 37% of voters to be prioritizing abortion over the economy in their vote, compared to historical averages?

If 37% of the electorate is motivated primarily by abortion in November, we're looking at a significant blue wave.

[Fmr. Gov. NickG]

Wow Zogby still exists??

[Fmr. Gov. NickG]

A big question I have about the whole “partisan nonresponse” thing….why isn’t this issue solved by weighting or stratifying on 2020 presidential vote?

I.e. Why wouldn’t a Wisconsin sample be representative if the sample reports being Biden +1 in the 2020 vote?  Aren’t you guaranteeing a representative number of Trump voters in that case?

[Fmr. Gov. NickG]

A big question I have about the whole “partisan nonresponse” thing….why isn’t this issue solved by weighting or stratifying on 2020 presidential vote?

I.e. Why wouldn’t a Wisconsin sample be representative if the sample reports being Biden +1 in the 2020 vote?  Aren’t you guaranteeing a representative number of Trump voters in that case?

https://talkelections.org/FORUM/index.php?topic=408952.0

The sample here was Clinton +1 in MN and Trump +8 in TX and yet the poll was widely off

The fact that you can cherry-pick a couple polls that are wrong from the previous cycle doesn’t explain why they would be wrong in the future, or even why they were wrong in this instance.  (And really Biden +11 in MN wasn’t even that far off.)

If a poll correctly samples or weights on 2020 vote, who exactly arew we worried about over- or under-sampling?

[Fmr. Gov. NickG]

https://twitter.com/allymutnick/status/1580616520870133760?t=QiO9qzjUc30pMpeszzU...

This is one race where the GOP actually has a significant candidate quality advantage over the Dems. The triage probably makes sense despite the district's presidential lean.

[Fmr. Gov. NickG]

A lot of discussion about the new NYT/Siena poll in Nate Cohn's "The Tilt" newsletter this morning.

It begins with a lengthy discussion of how the GOP lead in the LV screen is actually 3 points, not 4; and in fact the difference between the LV and RV screens are only 2.5 points, not 4.

Then a section titled "What’s different about our polls this year?", where I found this interesting note:

We now use additional information about the attitudes of respondents in determining whether they’re likely to vote, including whether respondents are undecided; whether their views about the president align with their party; whether they like the candidate they intend to vote for; whether they back the party out of power in a midterm; and so on, all based on previous Times/Siena polls. At the same time, we now give even more weight to a respondent’s track record of voting than we did in the past. (Emphasis in original)

Doe this mean that new voters (i.e. those with no record of voting in the past) are unlikely to be included as likely voters at all?  

I'd be curious how this impacts the poll in light of how the disproportionate number of newly registered voters who are young women.  Note that this poll has an unusually small gender gap; I wonder if missing these voters might have contributed to that.

[Fmr. Gov. NickG]

This message to Democratic voters is very clear imo, Biden promises Roe codfication WILL happen in January 2023

This is a good start to turning things around.  Dems need to refocus the campaign on abortion.  That's how Ryan won in NY-19, and how they got record turnout in Kansas. 

Quit with the Jan. 6th stuff for now.  Like it or not, swing voters don't perceive a real threat to democracy.  But they do perceive a real threat to abortion rights.

[Fmr. Gov. NickG]

https://twitter.com/SimonWDC/status/1585261022490787840

This is exactly what I've been saying for months. 

Polling errors don't systematically miss Republican voters.  They systematically miss new and infrequent voters. 

Given the new registration trends after Dobbs, this group seems likely to weigh more toward liberal and independent woman this year than conservative WWC.

[Fmr. Gov. NickG]

I don't think there's more variance than normal in generic ballot polls.  It's just we perceive more variance because the race is so close.  Check out the generic ballot polls from 2006...they range from D+4 to D+20 just in the last 3 days of the campaign:

https://www.realclearpolitics.com/epolls/other/2006_generic_congressional_vote-2174.html#polls

[Fmr. Gov. NickG]

Is it weird how many of the polls are finding such a small age gap? I realize it's usually unwise to nitpick crosstabs, but this trend is showing up in so many polls.  In the Q poll, 18-34 is only D+4, an 65+ is only R+1.

[Fmr. Gov. NickG]

This is pretty interesting - it was taken all across the month of October (sample size is insane), but what is the most interesting is that every one of the people who was interviewed didn't get polled on "generic" D or R, they got polled on their specific congress person in their district.

https://twitter.com/stella2020woof/status/1588224064463716354

For a little bit of context, the CCES is -the- gold standard poll in political science academia among professors who study Congress and US public opinion.  It is data that I used in my own PhD dissertation. 

The poll is in two waves: they do a weighted invitational online sample of 30,000-50,000 respondents in the month before every election cycle, and then they re-interview them after the election.  They will publish results that can be broken down by individual state and district, and they include a lot of policy questions, etc.  But these won't be available until well after the 2nd wave is complete.

[Fmr. Gov. NickG]

CCES is one survey where you really -can- dig into the crosstabs even for demographically small groups because the sample size is so large.  A couple interesting things I noticed:

- Republicans lead among "Independents" (Party ID) 45-37, but Dems lead among "Moderates" (Ideology) 52-32.

- The racial breakdown is pretty sensible, unlike a lot of polls we've seen:
-- Whites are 48-42 GOP
-- Blacks are 76-9 Dem
-- Hispanics are 49-36 Dem
-- Asians are 60-25 Dem
The one that caught my eye is the "other" racial category, which is 51-34 GOP, substantially more Republican than White voters! Who are these people?  They can't just be biracial.  I've seen this on a few other polls, but always just dismissed it as a tiny sample.

- The age gap is pretty much what I expected: 65+ are GOP+9 while 18-44 are Dem+21.
- But a smaller gender gap than in 2020 from what I can tell: Men are GOP+5, Women are Dem+11

- Voters who "somewhat disapprove" of Biden are Dem+18!

[Fmr. Gov. NickG]

The CCES is a 50,000+ person national stratified sample survey administered by YouGov. Half of the questionnaire consists of Common Content asked of all 50,000+ people, and half of the questionnaire consists of Team Content designed by each individual participating team and asked of a subset of 1,000 people. In addition, several teams may pool their resources to create Group Content.

The survey consists of two waves in election years. In the pre-election wave, respondents answer two-thirds of the questionnaire. This segment of the survey asks about general political attitudes, various demographic factors, assessment of roll call voting choices, political information, and vote intentions. The pre-election wave is in the field from late September to late October. In the post-election wave, respondents answer the other third of the questionnaire, mostly consisting of items related to the election that just occurred. The post-election wave is administered in November.

In non-election years, the survey consists of a single wave conducted in the fall.

https://cooperativeelectionstudy.shinyapps.io/CumulativeHouseVote/

https://cooperativeelectionstudy.shinyapps.io/HouseVote2022/


Now the toplines, D+2 aren't incredibly interesting, especially since data was covered over a one-month period across October 2022. What is more fascinating is the vast amount of crosstabs and the potential national trends it shows across age, gender, race, religion etc.


Note: All margins below refer to the two-party vote share

Below is probably the most interesting crosstab, as while across most groups (gender, religion, college education) polarization remained the same, here it didn't. This suggests a large swing away from the Democrats by both College-educated and Non-College Educated Hispanic Americans, while College-Educated White voters modestly swing towards the Democratic Party and Non-College White voters have a rather substantial swing towards them.

Image Link

This is a swing of R+22 among Hispanic College-Educated voters
And a swing of R+15 among Hispanic Non College-Educated voters

Almost completely erasing educational polarization within the group

College-Educated White voters are also now more likely to vote for the Democratic Party (by a margin of D+16) than Hispanic voters (at D+15).

Asian voters see a modest swing to the Democratic Party from D+40 to D+41 while Black voters see a moderate swing to the Republican Party from D+82 to D+78.

The Democratic Party also does slightly worse in suburban areas and somewhat better in rural areas, mainly because of a D+7 swing among rural women and an R+3 swing among suburban men.

I'm not sure why this poll doesn't merit it's own thread in the polling forum given that it is a poll of named candidates and not a generic ballot poll.

In any case, racial breakdown graph over time is a good inclusion.  The decline in Latino support for Democrats is really stunning, and needs to force a re-examination of the party's strategies and priorities. E.g. Why are there no prominent Latinos among the national leaders of the Democratic party???

[Fmr. Gov. NickG]

https://twitter.com/NateSilver538/status/1588916015869878272?t=J9Wz9Bfa732UFxcf6...

This critique makes so sense to me. It’s like if you were to complain that the right wing media is symatically dishonest, and I were to reply “Well, the left wing media could always just lie more to make up for it”.

[Fmr. Gov. NickG]

https://twitter.com/NateSilver538/status/1588916015869878272?t=J9Wz9Bfa732UFxcf6...

This critique makes so sense to me. It’s like if you were to complain that the right wing media is symatically dishonest, and I were to reply “Well, the left wing media could always just lie more to make up for it”.

Rather than making stuff up, I think he's saying if the Dems had good internals, they could release them, but the fact that they aren't suggests that they don't have them.

We’re not seeing a lot of real internals from experienced, professional polling firms on either side.  We’re getting fake polls from grifters on the right.  The equivalent scam industry just doesn’t exist to anywhere near the same degree on the left.

Edit: I do wonder why we haven’t seen anything from Center Street recently.

[Fmr. Gov. NickG]

Basically we know nothing.

Don’t we know a lot?  Seems like all the high quality polls are converging (herding?) on a basically tied race.

[Fmr. Gov. NickG]

Sabato/Crystal Ball released its final ratings-

https://centerforpolitics.org/crystalball/articles/final-ratings-for-the-2022-election/

Those Senate ratings are baffling. CCM wins while Warnock loses on the night? Hard to see it happening.

Warnock losing on election night is a...bold call.

I didn't see anything in their prediction specifying that Walker would win on election night.  In fact in their description, they predict his win is "a little likelier than not, on Tuesday or in a runoff".

[Fmr. Gov. NickG]

https://twitter.com/Politics_Polls/status/1589700815899328512

And if it went to R/O? The Senate would be decided in January for a second time.

The run-off is in early December this year.  I'm not sure why it was in January last year (maybe because of the simultaneous special election). Tbh it seems like it should not be legal to hold a run-off after the Senate is sworn in on Jan. 3.

[Fmr. Gov. NickG]

https://twitter.com/RonBrownstein/status/1589767818895319040

We've seen this pretty consistently - all of the Latino-only polls this cycle have generally been around D+25-30. This also lines up with Civiqs too FWIW (D+27)

Not sure who is right, but it's interesting that both Latinos and young voters are considerably more Dem when we get these big sample polls vs. the small subsamples.

Interesting then that Hispanics are only D+13 in the CCES, which had the national congressional vote overall at D+2.  I wonder if the difference in terms between Hispanic and Latino yields different results.

Whites were only R+6 in the CCES, while “Other” race was much more Republican.  (Remember that the CCES is an enormous sample.)