Thread by @dogemperor

@ArseGrammatica @BeschlossDC a) A lot of Narrative Driving Polls (which are meant as Noise Polls to spread exactly this kind of FUD) are being pushed
b) There are entire demographics (18-29 set, women voting Harris with Trumpy husbands) who are LITERALLY unpollable by any means
c) Dobbs Effect

1/x

@ArseGrammatica @BeschlossDC d) Polling, in and of itself, has so many methodology issues that *even high quality polls* really shouldn't be considered trustworthy and should be considered to have significant margins of error

e) The only poll that matters at the end of the day is the ballot box

2/x

@ArseGrammatica @BeschlossDC f) Regarding that last one--apparently there are FAR higher numbers of both Dems that have requested early ballots and *Returned them* than at this same point in 2020 in Pennsylvania (we should absolutely NOT rest on our laurels, but THAT's something I trust more than polls)

3/x

@ArseGrammatica @BeschlossDC g) As always, PEOPLE vote, not polls. We don't win if we don't come out in NUMBERS--focus on GOTV!

Check your reg! vote.gov
Find early voting info vote411.org
MAKE A VOTE PLAN: threadreaderapp.com/thread/1763057…

And then VOTE and SHARE THIS INFO

4/x

@ArseGrammatica @BeschlossDC h) If you are a woman you can actually help out SO MUCH. Point your friends who may have MAGAt husbands towards Vote Without Fear

Post-it notes in bathrooms reminding women nobody can see their vote w/ Vote Without Fear num (888-291-3027) help!

5/xfreshstartsregistry.com/vote-without-f…

@ArseGrammatica @BeschlossDC MANY MANY canvassers and even pollsters have reported that women are sometimes blocked from answering pollsters or canvassers, a significant population CAN'T answer polls for their own safety. Encourage GOTV for them in other ways, and focus on THAT

6/x

@ArseGrammatica @BeschlossDC i) To bring my receipts on how polling is an absolute hot fucking mess right now and how "narrative polling" is a Thing and how even the best polls are basically the equivalent of gazing at a goat's liver to see the future, I'll go ahead and post stuff in the postscripts

7/end

@ArseGrammatica @BeschlossDC In regards to "narrative polls": There is quite a business now of right-wing polling agencies explicitly trying to inject noise into polling to create an appearance of a false "red wave"

threadreaderapp.com/thread/1840026…
x.com/SimonWDC/statu…

pp/1

@ArseGrammatica @BeschlossDC NYT/Siena polling has some definite issues of methodology (of a nature very hard to correct for in polling) right down to the design of questions asked that may inadvertently or deliberately skew answers more conservative (from a discussion on a saner social media network)

pp/2

@ArseGrammatica @BeschlossDC There are even deeper issues that are effectively impossible to correct for, including the fact that one of the largest cohorts of likely Harris/Walz voters basically has been rendered unpollable due to advertisers inflicting unsolicted calls/email/banner ads/metadata abuse

pp/3

@ArseGrammatica @BeschlossDC Younger populations (including GenZ and Gen Alpha, and even millenials) are generally more privacy focused, are significantly more likely to be cellular-only (which impacts polling), are far more likely to use callblocker apps or even whitelisting to block survey calls

pp/4

@ArseGrammatica @BeschlossDC And at the same time, these groups are also much more likely to actively use adblockers to block ads online, are significantly less likely to participate in paid online surveys (like Ipsos surveys or Yougov), and in general are more distrusting of how metadata is used

pp/5

@ArseGrammatica @BeschlossDC Yes, you're reading this right: Telemarketers, obnoxious banner ads, autoplay videos, surveys where customer data has been insufficiently anonymized, hacks involving leaks of user databases, and so on pretty much have permafucked polling for at least 2 generations

pp/6

@ArseGrammatica @BeschlossDC Which leads to ANOTHER aspect of polling that pollsters have actively tried to correct for (since around 2012) without much success:

People who answer polls from random numbers are, in general, easier marks for a LOT of things and trend more conservative.

7/x

@ArseGrammatica @BeschlossDC So, I've actually done a much longer thread on this (at the link below), but *in essence* people who are more likely to pick up a random phone call at any age (and who don't do call screening) tend *in general* to be worse at "reality testing"

8/xthreadreaderapp.com/thread/1842984…

@ArseGrammatica @BeschlossDC The same person who'll pick up a random call from a pollster (or answer a random SMS message or join a paid survey group) is ALSO significantly more likely to answer phishing calls, general telemarketing scams (including timeshares), and so on.

9/x

@ArseGrammatica @BeschlossDC Said people are more likely to be targets of forms of affinity fraud, are more likely to click on banner ads/phishing emails/etc. and to not use adblockers, and are also (notably) more likely to be targeted by charity scams, coercive groups, and so on.

10/x

@ArseGrammatica @BeschlossDC And there is a *known* association, backed up by multiple sociological studies, that people who are worse at "reality testing" (and thus tend to be "easy marks") also are not only more likely to vote conservative but ALSO more likely to actively spread disinformation

11/x

@ArseGrammatica @BeschlossDC And unfortunately for a poller, these are LITERALLY the people most likely to actually answer the phone and complete a poll, versus blocking the number outright using a call-blocking app or whitelisting (where they'll get the "this number is not in service" or similar).

12/x

@ArseGrammatica @BeschlossDC And I'm actually going to lay out, with a pair of actual crosstabs from a NYT/Sienna poll for AZ and PA (that I am sure people will panic about) just HOW this kind of narrative, and even "where calls happen", can seriously skew results

pp13/x

@ArseGrammatica @BeschlossDC So AZ crosstabs.

First off: All of the NYT/Sienna polling covers less than 900 people in a sample for an entire state. In AZ's case only 808 people, out of a state population of 7.369 million

That's 0.01% of the population

But it gets worse

pp14/xnytimes.com/interactive/20…

@ArseGrammatica @BeschlossDC Out of that sample: 61% of their sample was over 45 years of age, and over 30% were over 65

Only 14% of respondents--only 147 respondents--were 18-29, and 1% of respondents REFUSED to be surveyed in that age range

Only 166 (20%) were 30-44 years of age.

pp15/x

@ArseGrammatica @BeschlossDC 62% of respondents were white non-Hispanic--519 people who answered that phone. 148 (19%) were Hispanic, and 116 (16%) were non-white non-Hispanic...and 2% of non-white non-Hispanics refused to complete the survey, in a state with a sizable First Nations community

pp16/x

@ArseGrammatica @BeschlossDC It. Gets. Worse.

Fully 61% of calls were done to *MARICOPA COUNTY*, and the Phoenix metro area in Maricopa County includes MULTIPLE 55+ only retirement communities...which trend conservative at the best of times. (Imagine a whole bunch of "The Villages" as suburbs)

pp17/x

@ArseGrammatica @BeschlossDC The "party ID" and "Party Registry" thing gets *even worse*. 30% are registered to "Other" (vs 38% for GOP and 32% for Dems), and for "Party ID" "Independent" got 34% (more than GOP at 32% or Dems at 27%), and for votes, equal 41% Dem/GOP and 13% "non voters"

pp18/x

@ArseGrammatica @BeschlossDC And if you are aware at all of how the primary systems work in Arizona and in Pennsylvania (which that cross-tab also shows similar issues) your Spidey-sense should already be tingling.

See: Arizona has semi-closed primaries, and Pennsylvania has closed primaries.

pp19/x

@ArseGrammatica @BeschlossDC In Arizona, you have to be a member of a party to vote in a primary, but the one exception is that if you're listed as "Unaffiliated" you can *at the time of a primary* join a party at the polling station and get a ballot for that party's primary

pp20/x

@ArseGrammatica @BeschlossDC Per AZ voter reg statistics, 35.41% of AZ voters were registered Republicans, 29.10% registered Democrats, 0.75% registered Libertarians, 0.71% registered No Labels, 0.08% registered Green Party, and 33.95% unaffiliated or other

pp21/xazsos.gov/elections/elec…

@ArseGrammatica @BeschlossDC And to make things worse yet, one of the other major metro areas in Arizona (Pima County, aka Tuscon et al) had something like a 3% refusal rate (with a sample of only 119 people) and west Arizona (sample of only 105 people) ALSO had a refusal rate of 2%

pp22/x

@ArseGrammatica @BeschlossDC This same poll also claims almost dead heats or slight edges for Harris in literally every polling average of voters who will actually come out (almost certain, very likely, somewhat likely, and not very likely) with the ONLY diff being voters not at all likely to vote

pp23/x

@ArseGrammatica @BeschlossDC Even here, it was all of 1% of Harris voters, vs less than 0.5% of TFG voters (2% of who refused to complete the survey) and *7%* of unaffiliateds (including Libertarians, Green Party, etc.)

It's also notable the "not at all likely" seems to be entirely *men*

24/x

@ArseGrammatica @BeschlossDC So. You have a phone poll (which is by its *very nature* going to skew R), with wording issues (which will make it skew R), where the primary place being polled is known for 55+ retirement communities that skew R AND are also more likely to answer your random call

25/x

@ArseGrammatica @BeschlossDC AND you are literally basically calling and getting responses from a whole shitload of white retirement communities (a significant part of that 45-64 contingent quite possibly IS younger folks in those retirement communities) who liked what Joe Arpaio was doing

26/x

@ArseGrammatica @BeschlossDC The ONLY thing you should be learning from this is to focus your efforts on younger voters (who won't be answering polls), women in general who'll be driving the Dobbs Effect (including women who may need a reminder that nobody can see who they voted for)

pp27/x

@ArseGrammatica @BeschlossDC and also outreaches to Hispanic *and* First Nations communities

Especially since this polling also shows that women are MUCH more likely to vote for Harris, esp. under 45 years of age (again a sign they must have hit retirement communities in polling).

pp28/x

@ArseGrammatica @BeschlossDC And even this crosstab poll shows signs of a STRONG potential "Hidden Dobbs Effect"--over 6% of people 30-44 and 3% of people 18029 outright REFUSED to answer who they'd vote for. Over 10% of non-white, non-Hispanic voters refused to answer

All these poll for Harris

pp29

@ArseGrammatica @BeschlossDC There's also some definite weirdness going on with listings of 2020 voting of who they'd vote for now vs then, I have my doubts that 9% of people who voted for *Biden* in 2020 are going towards TFG (in a sample of 299 Biden voters)

pp30/x

@ArseGrammatica @BeschlossDC And the polling also indicates, again, that we should be concentrating effort on outreach to people 18-44 (who were more likely to answer "Probably", and more likely to answer in Pima and west parts of Arizona--which also had higher general refusal rates for specifics).

pp31/x

@ArseGrammatica @BeschlossDC (As an aside: This poll also proves that Jill Stein does not have quite the spoiler effect she thinks she does--only 4% of undecided voters would go for her, and only 1% of both Trump and Biden voters in 2020. There are actually more respondents who refused to answer)

pp32/x

@ArseGrammatica @BeschlossDC And interestingly, when the pollers tried to push someone to say who they'd vote for, they had *a higher rate of overt refusals in almost every cohort than affirmatives for either candidate* (and some indications that younger cohorts basically started trolling them)

pp33/x

@ArseGrammatica @BeschlossDC And there were EXTREMELY small pools--as few as seven or eight and in some cases as few as four, and HEAVILY skewing towards independent-identifying, unaffiliated respondents (which is a pretty good indication of "completely unpollable demographic").

pp34/x

@ArseGrammatica @BeschlossDC The poll also points out an op for enhancing the Dobbs Effect--Arizona ALSO has a prop on its ballot (Proposition 139) which would restore Roe v Wade on a statewide basis, Hispanic women are STRONGLY in favor of Proposition 139.

pp35/x

@ArseGrammatica @BeschlossDC And as to exemplify what I mean by a likely Hidden Dobbs Effect--the Proposition 139 question actually had the *second highest* rate of refusal to answer, next to the handful of "independents" being forced to pick a candidate who were likely Libertarian voters

pp36/x

@ArseGrammatica @BeschlossDC In almost every cohort except people 30-44 (which even then was at 8%), Dem-identifying respondents/registered respondents, and Hispanic & non-white non-Hispanic cohorts, you had refusal rates for the Prop 139 question above 10% and in some cases above 20%

pp37/x

@ArseGrammatica @BeschlossDC And in Arizona, this means that there is a significant population there that are probably in the "Women with Trumpy husbands who can't actually safely answer the question". (Might point those to Vote Without Fear resources, just sayin'.)

pp38/x

@ArseGrammatica @BeschlossDC Also an interesting eccentricity of the Arizona poll:
a) Almost no white respondents from 18-29, and very few 30-44 (most respondents were Hispanic or non-white non-Hispanic).

The white skew is almost ENTIRELY from 45 years old on up, also a demo that votes Trumpy

pp39/x

@ArseGrammatica @BeschlossDC And out of younger, more diverse cohorts...most of the voter registration has been Dem or independent (and even HERE, the ONLY cohorts where there's even more than 40% listing *at all* is for those 45+ up, so it's kind of obvious how things skewed)

pp40/x

@ArseGrammatica @BeschlossDC And that same poll notes that among people 18-44 we REALLY need to drive the Dobbs Effect and the things that Kamala Harris will be doing for the working class (that same poll also indicates that the older set that is voting Trumpy is *largely* being driven by racism)

pp41/x

@ArseGrammatica @BeschlossDC Now, let's compare the crosstabs of the Times/Philadelphia Inquirer/Siena poll in Pennsylvania:

A bit saner but, as we'll note, still a Hot Mess in multiple respects, which is why I advise caution re polls in this day and age

pp42/xnytimes.com/interactive/20…

@ArseGrammatica @BeschlossDC Still a TINY sampling size, and if anything even a smaller percentage and number of 18-29 yr olds at 141 (13%) and 30-44 yr olds at 186 (21%), 33% (271) are 45-54, 29% (236) at 65+

Still, over 60% are 45+ years old, which in and of itself skews to R

pp43/x

@ArseGrammatica @BeschlossDC THIS poll shows more reasonable splits between "Almost certain" and "very likely", very few refusals on that end...and the *same* skew towards 45+ trending more conservative, but with a MASSIVE 81% of African-American likely voters for Harris, and 77% within Philadelphia

pp44/x

@ArseGrammatica @BeschlossDC There's also a BIG difference in this survey--there is more of a general split between rural areas, but--notably--central Philly and the *SUBURBS* are explicitly split out, as are the Alleghenies (Pittsburgh) vs the rest of the West of Pennsylvania

pp45/x

@ArseGrammatica @BeschlossDC Again, there's some definite weirdness (still don't see 7% of Biden voters claiming they'd vote for TFG) but what's ALSO notable is (AGAIN) the definite rate of refusals when people are asked to pick a candidate at all

On the LOW end it's 2-3%, often as high as 5%

pp46/x

@ArseGrammatica @BeschlossDC A note: Pennsylvania has 12.97 million people, of which 8.86 million are registered voters.

Less than 0.01% of voters are being polled for this, and fully 1 out of 20 out of that 0.01% are noping out on the question of "who would I pick for POTUS".

pp47/x

@ArseGrammatica @BeschlossDC For the nervous: Those numbers are almost certainly *not* Jill Stein voters. Only 1% of Dem voters even hinted at Stein; 4% from "other", again likely Green Party registered. Again, there's LITERALLY as high a margin of people who refused to answer.

pp48/x

@ArseGrammatica @BeschlossDC And there's a fairly strong trend in voters 30-44 to have upwards of 5% or more refusal rates of almost ALL questions, and (again, reassurance for the Concerned) Stein has almost *no* support among 18-29, the 3% of support she has is in 30-44 cohort

pp49/x

@ArseGrammatica @BeschlossDC And again, persons below 45 years of age were a bit more likely to say "Probably", which makes sense because there tend to also be higher refusal rates to answer questions and answering "probably" is often a bit of a hedge of "please get on with the question"

pp50/x

@ArseGrammatica @BeschlossDC And, again, you had handfuls of people who--when pressed--resulted in trolling and the highest rates of refusal in the survey was when people who weren't already going for Harris or TFG (in fact, higher than for literally either candidate).

pp51/x

@ArseGrammatica @BeschlossDC (What is PARTICULARLY impressive is that Northeast/Lehigh Valley, one of the Trumpier parts of Pennsylvania, had a *94%* refusal rate and Philadelphia suburbs and west PA had 73% and 77% refusal rates respectively, with a general 73% refusal rate re Dem-identifying folks)

pp52/x

@ArseGrammatica @BeschlossDC Again: This points to a likely "Hidden Dobbs Effect". The kind folks at @BucksCoBeacon can give a whole-ass essay on this, but there is likely a non-negligible silent majority of likely voters for Harris in MAGAt households who CAN'T answer a poll to this effect.

pp53/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And there's some of the same odd skewing of poll demographics as showed up in Arizona--even though 77% of the respondents to the poll were white, only 11% of respondents 18-29 were; while only 7% were African-American poll-wide, 19% 18-29, and 28% other ethnicities/races

pp54/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And likewise, for 30-44, 21% of respondents were white, 25% were African-American, and 21% of other ethnicities/races; even for 45-64, 31% white, 42% African-American, 36% of other ethnicities

There's also another point in both the AZ and PA cross-tabs to point out

pp55/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon I'm going to point out a thing in both these cross-tabs that shows a point I have CONSISTENTLY noted--Younger Demographics Are Increasingly Unpollable.

First, let's look at Arizona by age and by region, because that in and of itself is going to show what I speak of

pp56/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon Millennials, Generation Z, and Generation Alpha are the largest generational cohorts since the Boomers

Yet 18-29 polling...15% of respondents in Maricopa County, 9% in North/East polling area, 12% in Pima County, 13% in West Arizona

Maricopa County was 61% of this poll!

pp57/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon For 30-44: 22% of Maricopa County respondents, 13% for North/East, 18% in Pima County/Tucson, and 17% in West

Again: Maricopa County is *over half of this total*

And it keeps getting worse and worse from here, as we'll show

pp58/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon Maricopa County is the only country that has a category *other* than 65+ with a majority.

45-64 polling: Maricopa Co. 30%, North/East 31%, Pima Co./Tucson 28%, and West AZ 31%

And again: Maricopa County is *61% of all of the respondents*. Tucson metro is just *16%*.

pp59/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And finally, 65+ polling: Maricopa Co. 27%, North/East 44%, Pima Co. 36%, and West AZ 33%.

For those of you doing the math: Parts of AZ are having *75%* of their respondents being 45+, most likely being 55+, and even Maricopa Co. has 57% of respondents 45+.

pp60/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon They might as well be calling up The Villages in Florida (a 55+ community in Florida that has had a history, until quite recently, of being "so red it's infrared"; it was considered Progress when residents actively noted their intent to vote for Dems openly!)

pp61/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon "Well, ok, but Arizona is the new Florida where people go to die", you might say. "Surely it's better in Pennsylvania, right?"

"RIGHT?"

It is, I am afraid, *worse* in Pennsylvania, *not* better, and even more ludicrously so considering the existence of State College.

pp62/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon So first off, I should note how much of a percentage of polled people are in each area, and numbers

10% Alleghenies/Pittsburgh (88 total polled)
14% Central (116 total polled)
16% Northeast/Lehigh Valley (146 total polled)
10% Philly (98 total polled)

pp63/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon 22% (the majority) Philly suburbs (183 total polled)
15% South Central (117 total polled)
13% West (109 total polled)

So keep this in the back of your head while I describe the actual percentages by AGE. Maybe get a calculator. And a stiff drink of your choice

pp64/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon So. 18-29 polling percentages across regions:

Alleghenies 16%, Central 7% (wtf isn't this where State College is?), Northeast/Lehigh Valley 13%, Philly central 16%, Philly suburbs 16%, South Central 9%, West 15%

Yes, possibly as few as *EIGHT* respondents from Central

pp65/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon As few as *FOURTEEN* from Alleghenies. MAYBE 19 from Northeast/Lehigh Valley (the furthest-right part of PA in this poll). MAYBE 16 from Philly central. 29 from the Philly suburbs. As few as *ELEVEN* for South Central. 16 for West

Perhaps 113 people 18-29 *TOTAL*.

pp66/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon For people 30-44:
17% from Alleghenies, 16% from Central, 17% from NE/Lehigh Valley, 35% from Philly central, 24% from Philly suburbs, 19% from South Central, 22% from West PA

Yes, outside of Philadelphia & burbs and West, over 60-70% of respondents are over 45

pp67/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon Philly Central is LITERALLY the only area where the percentage of people 65+ answering are less than 18-29; the Philly suburbs actually have over-65s as the largest category (at 30% of respondents)

THIS is what I mean by "Young Folks Are Unpollable".

pp68/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon Congratulations, Advertisers and Demographics Metadata Harvesting Industries. You played yourselves; people don't actually WANT to participate in this garbage, so now you get pools of maybe 800-900 people to represent a whole state, most of whom are actually RETIRED.

pp69/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon Literally the ONLY useful info you're going to get out of polling nowadays is "What would be the best issue to GOTV about", and even THIS is restricted because you have entire demographics of silent majorities who either refuse for privacy or refuse for safety

pp70/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And that's why I say "don't sweat the polls; polls don't vote, people do." Concentrate on voting in NUMBERS

The BEST thing you can do is making a voting plan, make a voting buddy, and make sure you get that ballot in

(also plz to share this!)

pp71/endthreadreaderapp.com/thread/1763057…

@ArseGrammatica @BeschlossDC @BucksCoBeacon @threadreaderapp unroll

@ArseGrammatica @BeschlossDC @BucksCoBeacon And it has deeply fucked polling in SUCH a way that effectively *ANY* polling will only reach a population of people who are more willing "marks", and polling agencies think it is possible to correct for this when it is probably *not* in fact possible to

2/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And because of that, the polling pools we have are SO small (especially in regards to the precise demographics of voters that both candidates would love to get as lifetime voters) that at most what we can get are "what issues do people find important"

3/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And even *this* gets really hard to poll for with certain specific issues, where in certain cases women actually can't safely answer a poll (there's a lot of ancedotal info from canvassers of a silent cohort of female Harris voters with MAGAt husbands, for instance)

4/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon Effectively there is only so far you can safely extrapolate from, especially with small pools of voters

And when you're talking *candidate preferences*, the "pollable respondents" skew *so* hard towards demos that tend to vote R that it skews the whole poll rightwards

5/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon I think *at best* you can look at questions that are at interest to young people *and* questions that have high refusal rates (outside of candidate preference) but for the issue of "Who are you gonna vote for" effectively all states have become KY (a VERY hard state to poll)

6/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And I do think that there are mirages of "red waves" that show up in polls for this reason (especially those with methodology issues, or wording--I'm not even getting into "narrative polling"), especially when you bring the Dobbs Effect into play

7/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon That is actually one thing you CAN glean from polls--abortion and birth control is both VERY important to demos under 45, AND also has a high refusal rate for questioning (meaning people may think it's too personal OR they cannot safely answer the poll question)

8/x

@ArseGrammatica @BeschlossDC @BucksCoBeacon And THAT (especially if you have a high refusal rate from people registered as *Republicans*) is an indication of a potential "Silent Majority" you should trigger messaging to. (Including reminding women "Nobody has to know who you voted for".)

9/end

@ArseGrammatica @BeschlossDC People are panicking over polls that are contacting maybe 800 people total in a state, are often weighted towards areas with a lot of conservative voters who are older, where in some cases over 70% of the people answering are GenX or Boomers

@ArseGrammatica @BeschlossDC And yes, you can find out what older conservative likely voters (who are the most likely to pick up the phone on a random phone poll) prefer, but the biggest potential pool of voters (voters under 44, and esp. under 30), not so much (just because polling pools are SO small)

@ArseGrammatica @BeschlossDC My point is more that "Pretty much the main demographics you actually WANT to get more info about are actively hostile to polling, so that does skew polls and there are limits to what meaningful info we can get from polling as a result"

@ArseGrammatica @BeschlossDC Like, 18-29 are next to unpollable, in a lot of parts of the US not in rural areas 30-44 are difficult to poll, as people age you are going to get more people in the 45-64 demo that are going to be just as difficult to poll

And your pool of pollables is going to shrink

@ArseGrammatica @BeschlossDC How does one conduct a statistically valid poll when a larger body of people a year (including the exact demographics you want to get a pulse on) *actively* block any attempt at being polled due to privacy reasons?

And meanwhile the sampling pool shrinks and self-selects

@ArseGrammatica @BeschlossDC And part of sampling and modeling is knowing when *you do not have enough of a data set to do accurate modeling*.

Know what you can get accurate modeling info and trends from, know what you can't, and work from there towards the goal

@ArseGrammatica @BeschlossDC Get some heavy anti-telemarketing laws that include hard enforcement on SHAKEN/STIR as a protocol, blocks from the PSTN on CLEC and VoIP providers that forge CID or violate SHAKEN/STIR, yes that includes blocking int. transit and not letting AT&T get that sweet carriage fee

@ArseGrammatica @BeschlossDC Pass laws putting in HARD restrictions on what can be used with harvested user info including in surveys, including requirements for deanonymization and destruction of data.

Hard penalties (including "corporate death penalty") if compromise occurs leading to ID theft

@ArseGrammatica @BeschlossDC Full disclosures on what is done and how long data is kept for professional paid-survey firms like Ipsos and Yougov, restrictions on metadata.

Requirements for third party audits of any non-government agency (akin to HIPAA) for any data collection including telephone numbers

@ArseGrammatica @BeschlossDC Do you think you can do that? Because honestly *that* is what it's going to take for most people under the age of 55 years to actually remotely *consider* allowing calls other than a potential employer through a call-blocking app or adding to a whitelist.

@ArseGrammatica @BeschlossDC That's what it will require for a lot of people 18-29 to be willing to do an online survey or be in a focus group.

And again, that's assuming they're willing even THEN; there is an *actual* cultural shift on some things just not being shared with third parties

@ArseGrammatica @BeschlossDC So, unless you can somehow summon the Privacy Fairy to have Congress actually pass the privacy regulation that computer security experts & privacy advocates have wanted since the 90s...you'll *have* to deal with the fact the pool of the pollable is shrinking & will continue to

@ArseGrammatica @BeschlossDC Now. I can say my OWN personal read on things is that there IS probably a silent majority of younger folks AND "Dobbs Effect" Harris voters that is *explicitly not being picked up in polling*, but again, you NEVER want to assume

All gas, NO FUCKING BRAKES.

@ArseGrammatica @BeschlossDC Note: This is LITERALLY how Christian Nationalists BUILT a political engine and how we ended up in the shit nationally.

They didn't give a FUCK about polls

They gave, and give, a FUCK about getting every fucking warm body to VOTE for Project 2025 to happen

@ArseGrammatica @BeschlossDC Polling can call for things but if people don't get their asses out to VOTE it is meaningless.

Christian Nationalists realized that shit in the late 70s, made sure to build enthusiasm AND make sure people got to the polls EVERY election, and now we're staring down Gilead

@ArseGrammatica @BeschlossDC And basically if we want to beat them we have to ALSO make sure we go all gas no brakes EVERY year, EVERY election, and come out in NUMBERS.

Don't care what the polls say. If she's ahead, act like she's 20 points behind.

THAT is what you do.

@ArseGrammatica @BeschlossDC Functionally, polls at the end of the day are a distraction from the Assignment, and the Assignment is GET YOUR ASS OUT TO THE POLLS AND VOTE.

And in a post-polling world *this doesn't change*, imagine that!

So I say, Ignore the distraction, and focus on the Assignment. :D

@ArseGrammatica @BeschlossDC @BucksCoBeacon NB. A few more notes on what we CAN glean from younger folks in polls--many of them will state that abortion access is a major issue and THIS kind of outreach is the Way in reaching those voters in a post-polling world

pp1/

https://x.com/Amy_Siskind/status/1845097388089754052

@ArseGrammatica @BeschlossDC @BucksCoBeacon A note on "narrative polls" aka "noise polls"/"FUD polls". At least a few of these ARE specifically trying to fluff up TFG, and in some cases even high-quality polls are leading to states being targeted by FUD polling which can reduce polling effectiveness altogether

pp2/

@ArseGrammatica @BeschlossDC @BucksCoBeacon Here's a thread specifically talking about FUD polls trying to flatter TFG, and at least SOME of what we're seeing with polling is literally a psych-out game at this point (involving a malignant narcissist, yet)

pp3/threadreaderapp.com/thread/1845119…

@ArseGrammatica @BeschlossDC @BucksCoBeacon A form of "polling" that the NYT et al doesn't capture is that a lot of "accelerationist" rhetoric has ramped up since TFG's poor showing at debates, and (at least to their own) Christian Nationalists ARE starting to mildly panic:

pp4/peoplefor.org/rightwingwatch…

@ArseGrammatica @BeschlossDC @BucksCoBeacon In general, these accelerationist calls are both a mix of "Rosh Hashanah Panics" (related to appropriation of Jewish High Holidays for apocalyptic imagery) AND legitimate concern they Will Not Win the election legitimately

pp5/threadreaderapp.com/thread/1546663…

@ArseGrammatica @BeschlossDC @BucksCoBeacon In general, you don't see things like literally threatening to bring armed men into polling places for voter intimidation if the GOP think they've got the Electoral College wrapped up

(That said: Even more reason for all gas, no brakes.)

pp6/rollingstone.com/politics/polit…

@BeschlossDC @BucksCoBeacon And an additional observation from OP:

pp7/

https://x.com/BeschlossDC/status/1845160864959045989

@BeschlossDC @BucksCoBeacon Another gentle reminder on FUD polling:



pp8/

https://x.com/SimonWDC/status/1845103624420016528

@BeschlossDC @BucksCoBeacon A gentle reminder that FUD polling is, in and of itself, used as a disinformation narrative to try to discourage turnout:

pp9/

https://x.com/SimonWDC/status/1844887796235825372

@BeschlossDC @BucksCoBeacon More on the use of skewed polling and junp polling as an actual disinfo narrative (including in "agents of hostile foreign power" narratives):



pp10/

https://x.com/cbouzy/status/1845185159026430345