Thread by @dogemperor

@exelephant2017 @ForecasterEnten This is where I remind folks:
a) CNN was taken over in a *hostile takeover* (like this hellsite) by a Trumper CEO who actively promoted incestuous Quiverfull families as entertainment
b) This does *not* match polling especially for under-45s (and polling oversamples over-55s)

@exelephant2017 @ForecasterEnten And just so I can bring ALL my receipts:

Even New Republic has noted the takeover of media by right-wing companies newrepublic.com/post/178256/ba…
theguardian.com/commentisfree/…
thehill.com/homenews/media…
vox.com/2022/8/26/2332…

(That's for CNN and its hostile takeover)

1/x

@exelephant2017 @ForecasterEnten On polls being a HOT mess in methodology (phone-based polls first):
threadreaderapp.com/thread/1845207… (NYT/Sienna)
threadreaderapp.com/thread/1845509… (NBC)
threadreaderapp.com/thread/1848830… (Suffolk)
threadreaderapp.com/thread/1848419… (Emerson)

NBC, Suffolk, Emerson oversample evangelicals and/or South

2/x

@exelephant2017 @ForecasterEnten On polls being a HOT mess in methodology (online "paid portal" polling--YouGov, Ipsos, Cint):
threadreaderapp.com/thread/1845536… (CBS/YouGov)
threadreaderapp.com/thread/1845568… (ABS/Ipsos)

Also: all "paid survey" portals are opt-in, subject to "pool poisoning" efforts

3/x

@exelephant2017 @ForecasterEnten YouGov in particular has had particular issues with both apparent "pool poisoning" (there are things showing up in demos in surveys for 18-30 that do *not* match up with other surveys indicating flooding of a polling pool) AND some very conservative-skewed surveys

4/x

@exelephant2017 @ForecasterEnten A lot of people consider TIPP a good poll but I do not (on a fundamental level) trust *any* poll that keeps methodology, crosstabs, etc. behind a paywall

Marist is a good example of a Well Designed Poll (and even IT likely skews a bit, as I'll note)

5/xthreadreaderapp.com/thread/1846583…

@exelephant2017 @ForecasterEnten All polling on a fundamental level is likely to skew a bit more R than the usual populace, based on largely unavoidable issues that present across *ALL* polling (political or otherwise) on who is likely to answer a poll to begin with

6/xthreadreaderapp.com/thread/1842984…

@exelephant2017 @ForecasterEnten In general *all* surveys for people under 50 are impacted by what I have termed the "Nassau Straw Market Problem"--aggressive advertising/telemarketing/scams/spams are SO ubiquitous people block ANY attempts at unsolicited contact, whether polling or phishing

7/x

@exelephant2017 @ForecasterEnten (I've named this issue after an infamous straw market at the entrance of the Port of Nassau that existed until about a year ago, and which one *had* to go through if exiting a cruise ship to the city; it was a veritable gauntlet of *extremely* aggressive street vendors.)

8/x

@exelephant2017 @ForecasterEnten (This gauntlet also had occasional pickpockets targeting it, and it got so bad cruise forums were recommending staying on the ship rather than visiting bars and museums in Nassau; Nassau moved the "straw market" about a year or two ago, but it had a Rep when it existed)

9/x

@exelephant2017 @ForecasterEnten (And basically because extremely aggressive advertising, spams, scams, and telemarketing made a Digital Nassau Straw Market Problem, most people under 50 aggressively use callblocker apps, don't join "paid survey" forums, and are just harder to reach than over-55s.)

10/x

@exelephant2017 @ForecasterEnten The people who *do* answer polls are also the ones that are more likely to answer telemarketer calls (including phishing calls), are more likely to be targets of affinity fraud, do trend to be more conservative, are older, and even are more likely to *spread* disinfo

11/x

@exelephant2017 @ForecasterEnten Overweighting towards 55+ can skew a poll R. Overweighting towards the South can significantly skew a poll R (due to the Christian Nationalism Belt; the single highest predictor of a Trump voter is belief in "Seven Mountains" theology).

12/x

@exelephant2017 @ForecasterEnten Even what seems like an innocuous rewording of a question pre-Dobbs vs post-Dobbs can have an impact of 10 points or so rightward (as shows up in the ABC/Ipsos poll with pre-Dobbs and post-Dobbs wording of an abortion related question)

13/x

@exelephant2017 @ForecasterEnten So yes, it IS actually hard to reach certain contingents (which are also likely to be a particularly high pool of "unaffiliated" voters, including swing states--in part because of automatic voter registration in some states where "unaffiliated" is the default).

14/x

@exelephant2017 @ForecasterEnten In fact, I'd even caution strongly about even taking early returns with a grain of salt, in part because there is likely to be at least a segment of R-registered voting D *and* large numbers of "unaffiliateds" in states largely doing vote-by-mail and absentee

15/x

@exelephant2017 @ForecasterEnten There is also the matter of the Dobbs Effect--women voting re reproductive health issues--which has caused a CONSISTENT 5-9% underestimate of D votes since the Dobbs decision, and is likely driven by women who are unwilling to share this information w/ pollsters

16/x

@exelephant2017 @ForecasterEnten One hint of how significant of a Dobbs Effect we may be looking at is Marist's poll of white evangelicals (around 15% of total, which is in line with national averages from Pew Forum and PIRR of number of self-identified white evangelicals broadly)

17/x

@exelephant2017 @ForecasterEnten Marist is the sole poll (with public access to data) I've found that asked self-identified white evangelicals their sentiments on positive vs negative views of candidates, AND on likely voter contingent (instead of merely asking if they identified as "white evangelical")

18/x

@exelephant2017 @ForecasterEnten *Religious* studies by PIRR, Pew Forum, and others indicate (again) something like 66% of self-identified Evangelicals identify as some form of Christian Nationalist, and something like 69% believe in SOME form of NAR distinctives

19/xreligionnews.com/2024/05/06/how…

@exelephant2017 @ForecasterEnten Some surveys, including PIRR's study on Christian Nationalism () even put people who espouse distinctives of NAR theology as around 69% total, with closer to *80%* of white Evangelical Protestants that are "hard" Christian Nationalists

20/x prri.org/wp-content/upl…

@exelephant2017 @ForecasterEnten That same PIRR study has those populations skewing VERY hard Republican, 64% as registered R and only *6%* registered Democrat

The single largest predictive factor for Trump voters is "Does a voter espouse NAR theology distinctives like 7 Mountains Mandate?"

21/x

@exelephant2017 @ForecasterEnten According to at least some surveys, something like 40.9% of American churches espouse some form of blatant NAR "Seven Mountains" Christian Nationalism, which is HEAVILY concentrated with white evangelical Protestants religioninpublic.blog/2024/05/13/bel…
religionnews.com/2024/05/06/how…

22/x

@exelephant2017 @ForecasterEnten So, we're dealing with a VERY Trumpy subpop

And this is important in light of that Marist survey because *its* poll of white Evangelicals as a subset showed a *20% approval rating for Harris*, and *18% of self-professed white Evangelicals who said they'd vote for Harris*.

23/x

@exelephant2017 @ForecasterEnten There is also a lot of *anecdotal* info from door-to-door canvasses (and especially "deep canvassers", who will visit for a period of 20 mins or more) of women noting they will vote for Harris but *do not* want their spouses to know whom they voted for

24/x

@exelephant2017 @ForecasterEnten And there's even a post-Dobbs effect that almost certainly impacts polling in the South and parts of the Midwest--in areas that have post-Dobbs abortion bans, there's also a push by state AGs to criminalize almost anything RELATED to abortion or abortion initiatives

25/x

@exelephant2017 @ForecasterEnten For an example related to this year's election, there have been efforts by the state of Florida to make even *promotion* of voting "yes" for Amendment 4 (which would enshrine reproductive freedom in the state Constitution) illegal and attacks on the "Yes on 4" campaign

26/x

@exelephant2017 @ForecasterEnten Dobbs "ban states" also have increasingly moved towards use of things like search engine data, and even menstrual cycle data from health apps, as a means towards potential prosecution of people seeking information on reproductive health services

27/x

@exelephant2017 @ForecasterEnten And yes, having a MAGAt husband and/or living in a state with a post-Dobbs abortion ban where the AG is actively pushing to make *assisting* a woman in getting an abortion a felony DOES have a dampening effect in regards to survey responses

Possibly...as high as 9%

28/x

@exelephant2017 @ForecasterEnten There's one more factor dampening polling. A lot of junk "narrative driving" polls that are explicitly skewed ARE bombarding swing states--not unusual to get 15+ calls a day in PA on these, over 40 groups involved. That's probably driving a Straw Market Problem itself

29/end

@exelephant2017 @ForecasterEnten @threadreaderapp unroll