Democrats see polls as overestimating Trump's strength
Source: The Hill
03/09/24 6:00 AM ET
Former President Trump led President Biden in a recent national New York Times/Siena College poll by 5 percentage points, prompting a campaign spokesperson for the Democrat to argue polling consistently overestimates Donald Trump while underestimating Joe Biden. The remark is emblematic of a deeper skepticism of Trumps strength in certain Democratic circles.
Biden allies point to primary results this cycle that show Trump underperforming his polling numbers and failing to capture moderate voters, as well as nonpresidential cycles under Biden where Democrats outperformed projections. Weve got a long way to go. Weve got a lot of work to do. But I think everyone in this town
is overestimating their strengths and underestimating ours, said Democratic strategist Simon Rosenberg. Some Democrats were quick to dismiss The New York Times poll in particular.
Strategists were puzzled at the polls finding that women were equally split between Trump and Biden at 46 percent each, given exit polls in 2020 found Biden won women by 15 percentage points. Democrats also questioned the polls findings that Trump was leading Biden by 6 points among Latino voters because Biden won that group by 33 percentage points in 2020, according to exit polls. Multiple Democratic strategists and pollsters noted 97 percent of interviews with Latinos surveyed for the poll were conducted in English.
Even Rep. Dean Phillips (D-Minn.), who ended his long-shot primary bid against Biden on Wednesday, was skeptical of the polls findings. When the NYT/Sienna poll shows me at 12%, you better believe it is flawed, he wrote on X, formerly known as Twitter. Only 5% even know who I am. The Biden campaigns belief that Trump is overestimated by polls is largely driven by the fact that voter behavior has not matched projections, including during this years primary elections.
Read more: https://thehill.com/homenews/campaign/4518985-democrats-see-polls-as-overestimating-trumps-strength/
onecaliberal
(32,888 posts)The maths and methods are always wrong and suspect.
PSPS
(13,614 posts)DLCWIdem
(1,580 posts)I am not a pollster. But short Stack was looking at the cross tabs. I heard this on David Pakman show and he had a link to the article. Pakman was uncertain about the weight but I read the article and it seemed to me they were taking that into consideration.
Think. Again.
(8,367 posts)...it's a fairly simply concept:
Dem's see reality,
while rightwinger's see weird shit they make up in their bitter minds and share among themslves.
Edit to add:
and most pollster's start each work day thinking about how they're gonna sell polls to advertising-money-dependent media.
IbogaProject
(2,840 posts)The younger generations answering machines through mobile phones have developed a different relationship w calls than those before. So phone polls under report those less likely to be rah rah GOP. But we can't get complacent, we have to run like we are behind especially with needing two win the trifecta of the Electoral College, the Senate and Congress to try and fix things.
groundloop
(11,521 posts)DON'T GET COMPLACENT.
Response to BumRushDaShow (Original post)
GP6971 This message was self-deleted by its author.
PortTack
(32,791 posts)Ppl just blindly accept them! Why?
Republican candidate Mitt Romney continues to lead the Gallup national tracking poll of the presidential race. Mondays edition shows Romney with 51 percent to Obamas 46 percent, an uptick for Romney who led Sundays Gallup poll 50 percent to Obamas 46 percent.
This was dated just a few days b4 the election 10/29/2012 Any questions?
https://talkingpointsmemo.com/livewire/gallup-tracking-romney-by-5-points-nationally
BumRushDaShow
(129,409 posts)Romney 49%, Obama 48% in Gallup's Final Election Survey
Early voting so far breaks 49% for Obama and 48% for Romney
Gallup Editors
PRINCETON, NJ -- President Barack Obama and Republican challenger Mitt Romney are within one percentage point of each other in Gallup's final pre-election survey of likely voters, with Romney holding 49% of the vote, and Obama 48%. After removing the 3% of undecided voters from the results and allocating their support proportionally to the two major candidates, Gallup's final allocated estimate of the race is 50% for Romney and 49% for Obama.
The survey was conducted as part of Gallup Daily tracking Nov. 1-4.
(snip)
https://news.gallup.com/poll/158519/romney-obama-gallup-final-election-survey.aspx
And eventually the mea culpa had to happen and they released a final analysis of what went wrong (PDF), their partner USA Today officially dropped them as a partner after that, and Gallup no longer does this poll -
Martha T. Moore
USA TODAY
Published 2:24 p.m. E.T. June 4, 2013 | Updated 5:02 p.m. ET June 4, 2013
WASHINGTON Pollsters at Gallup said Tuesday they have identified flawed methods that contributed to their incorrect prediction that Mitt Romney would win the 2012 presidential election, but they are still working to determine how to better identify who is likely to vote.
The survey firm undertook a far-reaching review of its operations after its surveys came up short in the election: Gallup's final pre-election estimate showed Romney with 49% support to Obama's 48%, with a margin of error of +/-2%. Most polls estimated Obama would win the popular vote by 1 percentage point. Obama won the popular vote by 3.85 points.
In pre-election polling, Gallup consistently showed Romney with a 3-percentage point lead over Obama. When Gallup switched to surveying only "likely voters," Romney's edge increased to 4 percentage points.
Gallup, with researchers from the University of Michigan, will experiment with ways to better identify likely voters in surveys during the 2013 governor's races in New Jersey and Virginia. Gallup asks seven questions in its phone surveys to determine whether people are likely to vote a questionnaire that may rely too much on past voting and on how much "thought" voters have given to the election, Gallup Poll editor in chief Frank Newport said. Though all polling outfits showed an increase of support for Romney among likely voters vs. registered voters, Gallup's bump for Romney was the most extreme. "We really are re-evaluating that from square one," Newport said.
(snip)
https://www.usatoday.com/story/news/politics/2013/06/04/gallup-poll-election-obama-romney/2388921/
And more recently, the 2022 "Red Fizzle" -
https://fortune.com/2022/11/16/pollsters-got-it-wrong-2018-2020-elections-statistical-sophistry-accuracy-sonnenfeld-tian/
https://www.vox.com/policy-and-politics/2020/11/10/21551766/election-polls-results-wrong-david-shor
https://www.scientificamerican.com/article/why-polls-were-mostly-wrong/
https://www.pewresearch.org/methods/2021/04/08/confronting-2016-and-2020-polling-limitations/ - skewed analysis for 2016/2022
https://www.nytimes.com/2022/12/31/us/politics/polling-election-2022-red-wave.html
https://nul.org/news/why-pundits-and-junk-polls-got-midterm-elections-wrong
https://www.thedailybeast.com/pollsters-have-no-fcking-idea-whats-going-to-happen-during-2022-midterm-elections
https://people.com/politics/why-midterm-polls-falsely-predicted-red-wave/
And a more recent skew - https://www.theguardian.com/us-news/2023/sep/24/washington-post-abc-poll-trump-10-points-over-biden
Farmer-Rick
(10,206 posts)Geez, so they know all their problems but do nothing, or very little, to fix them. It sort of like the GOP think if they keep telling everyone they are going to win, and making up polls to say that, then it's a done deal.
This headline from Fortune at your link says it all: "Pollsters got it wrong in 2018, 2020, and 2022. Heres why political polling is no more than statistical sophistry."
Though this one from Vox is the most interesting: "One pollsters explanation for why the polls got it wrong.....The kind of people who answer polls are really weird, and its ruining polling."
Thanks for the good info BumRush
Novara
(5,851 posts)That's pretty much all you need to know.
Pollsters can't robocall cell phones (it's illegal), so they don't get much representation from people whose only phone is their cell phone. Yeah, some people get polls on their cell phones but that's when they've given out their number to other entities who sell their number to pollsters. The great majority of cell phone-only people never get a polling phone call.
A lot of people, when they do agree to answer a poll, don't even finish the poll.
Pollsters need people with time on their hands to respond. That eliminates busy people who are working. Who has a lot of time? Older people (who tend to be conservative), people who don't need to work.
And then there are biased questions that pigeonhole answers the pollsters want. When faced with a loaded question, people often pick the less onerous answer although it does not represent how they feel, just because they feel they have to pick one.
Some polls use a tiny polling population, and those results have a ton of error. A small group isn't representative.
There are a LOT of reasons why polls have not been accurately predictive for a long time.
The much better predictor: look at the actual recent election results. According to polls, Democrats consistently "overperform." Republicans consistently do much worse than predicted.
So take polls with a grain of salt. Look at the polling entity. Some are very biased, like Rasmussen. They lean so far to the right it tilts the earth. Look at the sample size. Look at the way the questions are phrased. Or if you don't feel like doing all that work, just ignore polls. I do.
NanaCat
(1,237 posts)Then why are 70% or more of their responses, industry-wide, from cell phone contact?
It's not the 90s anymore, you know.
Demsrule86
(68,657 posts)Novara
(5,851 posts)Marketers, pollsters, companies can't robo-call cell phones. It is illegal. That isn't to say that no cell phones are ever called for polls. People give their numbers out to various entities, and lists of cell phone numbers get sold to pollsters. So yes, cell phones are called for polls. But not a large, diverse, representative sample of all cell phone holders are among the cell phones that do get called. This is obviously a smaller subset of cell phone holders.
Saying 70% of responses are from cell phone holders doesn't tell you anything about who this subset of cell phone holders is, and how diverse a population it is - or isn't. It's still a small subset, and it's people who have voluntarily given their numbers out to various entities that then sold their numbers to pollsters.
A lot of people - including me - don't give out their cell phone numbers to very many people. Me? Friends, work, medical providers. I don't give my number to stores or anyone selling anything. As a result, I am never contacted by pollsters. Most people I know have not been contacted either.
That 70% is not 70% of all cell phone holders. It's 70 % of TOTAL RESPONSES OF THE POLL, with 30% of land line responses. 70% of poll responses is not 70% of all cell phone holders.
Magoo48
(4,720 posts)ificandream
(9,385 posts)I've never understood how so much emphasis is put on polls. I'll concede they're information, but only in a light sense. We take them way way way too seriously now. The wide difference in polls should be enough to lessen their influence. But it doesn't.
I do believe there is a lot more disturbance in the Trump force than the polls would allow us to believe. I think moderate Republicans are very dissatisfied with him and will not support him. And I think the SOTU helped consolidate that. Granted, there are magas who won't be changed. But I think there are less. He's not getting the crowds he got for this campaigns in 2020 and 2016. But he'll do every trick in the book to make it appear the country loves him. But we as a country don't.
Fiendish Thingy
(15,652 posts)It should read:
Sentient humans who can read and understand polling methodology see polls as overestimating Trumps strength .
BumRushDaShow
(129,409 posts)SWBTATTReg
(22,156 posts)Forecasting doom and gloom for Democrats? The article is titled "Why Did Polls Prepare Us for a Red Wave? Experts Weigh In on the Surprising Midterm Election Results", by Amy Eskind Published on November 10, 2022 02:09PM EST, PEOPLE Magazine
Two days after the election, the majority in Congress has yet to be determined. If midterm polling had been accurate, Republicans would already be celebrating a dual-chamber sweep. The "red wave" that pollsters were predicting before the midterms turned out to be more of a red trickle. Expectations that Democrats could lose as many as 35 seats in the House of Representatives have been disproven, and if Republicans do clinch majorities in the House or Senate, it will be by a razor-thin margin. (WE ALL KNOW NOW MARCH, 2024 THAT MARGINS IN THE HOUSE ARE RAZOR THIN).
Democrats did better than history would have predicted the best a leading party has done in the midterms in 20 years.
...
My question here: ARE THESE so-called POLLSTERS 'GAMING' THEIR DESIRED RESULTS? Being that that the MSM seems to highlight djt's accomplishments etc. (that is, a majority of MSM is owned by republicans obviously).
Some polls were way off. Michigan's Democratic Gov. Gretchen Whitmer had a very comfortable lead, ... Then a poll was released on Oct. 31, well into the state's early voting period, finding the race was a virtual tie, and that half of independent voters were supporting Dixon, while less than 30 percent planned to vote for Whitmer. It was conducted by Insider Advantage, a Republican firm ...
NEED I SAY MORE?
BumRushDaShow
(129,409 posts)2022
https://fortune.com/2022/11/16/pollsters-got-it-wrong-2018-2020-elections-statistical-sophistry-accuracy-sonnenfeld-tian/
https://www.vox.com/policy-and-politics/2020/11/10/21551766/election-polls-results-wrong-david-shor
https://www.scientificamerican.com/article/why-polls-were-mostly-wrong/
https://www.pewresearch.org/methods/2021/04/08/confronting-2016-and-2020-polling-limitations/ - skewed analysis for 2016/2022
https://www.nytimes.com/2022/12/31/us/politics/polling-election-2022-red-wave.html
https://nul.org/news/why-pundits-and-junk-polls-got-midterm-elections-wrong
https://www.thedailybeast.com/pollsters-have-no-fcking-idea-whats-going-to-happen-during-2022-midterm-elections
https://people.com/politics/why-midterm-polls-falsely-predicted-red-wave/
And a more recent skew - https://www.theguardian.com/us-news/2023/sep/24/washington-post-abc-poll-trump-10-points-over-biden
Rather than correct their errors, methods, and question formatting, they have doubled, tripled, and quadrupled down on including push-poll type questions in order to continue the anti-Democratic Party narrative.
SWBTATTReg
(22,156 posts)are still wrong, e.g., look at tRUMP, and his ongoing string of losses. And, this is a big reason why I don't take polls or the like (it's not like they are beating down my front door to do so, ha ha heh).
I wouldn't be surprised that tRUMP pays off these pollsters (or if they're republican owners, they do it (alter results of polls) to keep tRUMP happy. there is some hand in the pocket somewhere.
I've always wondered how they were so far off, they've (the polls) have been fine-tuned, to excess, to avoid such pitfalls and to see how much some of them (the skewed polls) fail so badly? It's like a fog in front of their faces, or they can't see the one tree in the forest right in front of them.
I love your description of the 'push poll' type of questions, that is, to solicit the wanted answer to skew the poll results. When I was an instructor, we had to be very careful when we were designing our tests for students, in order to not skew the results inadvertently. Not an easy thing to do (that is, to skew test results).
BumRushDaShow
(129,409 posts)and have been trying to over-correct ever since. Literally dismissing every other demographic except the so-called "rural/white working class" and oversampling them. And when called on it, they claim they "correct to the numbers in the population", which is a big fail as well.
There were a couple huge threads around the 2022 election and perfect examples in 2023 -
https://www.democraticunderground.com/100218254030
https://www.democraticunderground.com/100218253358
That CNN had the push-poll questions (I had grabbed a snapshot of some) -
And they even tracked how well the propaganda was going with that -
I think they even asked "Hunter Biden" questions as well - keeping the narrative going.
Other than results from "exit polls" that might better reflect "reality", the main takeaway is this -
Seeking Serenity
(2,840 posts)Just take that to the bank. No one that the establishment doesn't approve of will ever be allowed to take office again.
That said, screw the polls. They're wrong, and will be shown to be wrong. A 2d term is Biden's for the acceptance.
Demsrule86
(68,657 posts)"Red Flag #1: Trump Keeps 97% Of His Supporters
According to the poll:
Mr. Trumps ability to consolidate the Republican base better than Mr. Biden has unified the base of his own party shows up starkly in the current thinking of 2020 voters. Mr. Trump is winning 97 percent of those who say they voted for him four years ago, and virtually none of his past supporters said they are casting a ballot for Mr. Biden. In contrast, Mr. Biden is winning only 83 percent of his 2020 voters, with 10 percent saying they now back Mr. Trump.
The poll result does not match up with the results of the Republican primary.
Red Flag #2: Trump And Biden Are Tied With Women
Here is how The New York Times describes the results of the poll with women, The gender gap, for instance, is no longer benefiting Democrats. Women, who strongly favored Mr. Biden four years ago, are now equally split, while men gave Mr. Trump a nine-point edge.
Biden beat Trump by 15 points in 2020 among women, and now they are tied.
This poll has arrived on the same week when Trump announced his support for a national abortion ban. Trump brags about taking away reproductive healthcare rights from women, and his party is struggling as Alabama banned IVF, and that is just one issue.
Red Flag #3: Trump More Than Doubles His Support With African-American Voters
The New York Times poll also found that there is going to be a historic African American voter realignment. The poll found that 23% of African Americans support Trump.
If 26% of black voters voted for Trump, it would be the most radical realignment in African American voting since the Civil Rights Act was passed.
The Fox News theory is that black voters are supporting Trump because he got indicted, a point which manages to be both stupid and racist at the same time.
The black vote for Trump has been non-existent in the Republican primary, while Biden has shown his usual strength with black voters in Democratic primaries in South Carolina and Michigan."
https://tcinla757.substack.com/p/why-the-new-nytsiena-poll-is-bullshit