They think they are voting in a primary: someone else.
This is true.
I’ve used it for predicting a Trump win, based mostly on belief that voters have made up their minds unusually early this time, but it is fair to call that misuse of the polls.
If this was the case, I would expect Cornel West and/or Stein/Greens to be surging since October 7. I do not see it.
Also worth considering – and this is not misuse of polls – voters say they trust Trump more than BIden on the Israeli-Palestinian conflict. I think what is going on here is that some voters blame any bad news on the incumbent president, the exception being a big attack on American soil (link to document post-9/11 Bush poll bounce).
To be clear, I do not personally think Biden was to blame for the inflation bounce, or for war in the Middle East. I’m just reporting what non-broken polling tells me.
Serious question, and I do know the answer for a certainty. Are there any mainstream pollsters – the kind that get an A, B, or even C pollster rating at 538 – who do political polling face to face at malls, or anywhere else? I think not but could be wrong.
Putting that aside, Trump beat the polling averages when he won in 2016. And he beat them again when he lost in 2020. It wasn’t by a big enough margin to say polling is broken, but we did see the shy Trumper effect twice. If something has happened to cause the F-150 scampering effect to switch directions this cycle, I would love to know it. On my supposition that Trump will again beat polling averages, I really want to be wrong.
WTAF? I can’t believe anyone would trust Trump to run a casino in Atlantic City, let alone a delicate, tense international conflict involving Arabs and Jews. And that polling doesn’t even show a plurality of adults trusting Trump over Biden- it’s 50% of registered voters. I understand in Michigan, we may be seeing Muslim voters expressing a backlash against Biden for his current stance, but I still find these results mind-blowing. Again, WTAF?
I think that there is a global and dismal failure or disappearance of even the most basic critical thinking, because of which I think that we are screwed.
This weekend I informally polled the nine Rs in my family. They were turning now against Trump because of his legal problems so were considering DeSantis, Christie or Haley instead. I don’t think there were any other candidates they thought could run in the spring primaries.
Polls are conducted by cold-calling people. People my age and younger don’t answer the phone if we don’t recognize the number, because we know it’s likely to be a scammer, bill collector, or just plain someone we don’t want to talk to.
So true Smapti! Whenever I’ve completed a poll the caller has always been looking for a different sample , a different population.
Nate Cohn said about the poll I linked to that 99.6% of 18-29 year olds in the poll used cellphones. They were not statistically unusual, since their stated preferences matched voting results.
While it’s much harder to get a sufficient sample for polling, top firms still get it done. There have always been “throw the rascals out” elections in which the issues are less important than pure replacement. 2024 may be that kind of election.
Or not. Tons of news can be packed into a year.
Yeah, I was going to ask how they are attempting to obtain their representative samples. They have to do a lot of finagling with the raw data because of how unlikely people are to respond. This requires guessing the actual make up of the population (either voting or overall) so they can use the numbers they have for each set of people who actually respond and extrapolate to the whole population including those who don’t respond. This in turn requires defining groups that you expect to act similarly in terms of voting but act differently in terms of whether they decide to answer the pollster. If they just post their raw data, there’s presumably going to be something like 90% “didn’t respond” which makes it hard to make conclusions. I assume that they can make some sort of conclusions statistically, but it depends a lot on whether their assumptions and stratification are reasonable.
Not just people your age and younger. I’ve long employed this practice. And I think I’m older than dirt you.
Is that how polls are made in the USA?
That makes more sense to me. May I ask: Do pollsters call mobile phones or land lines? If only one or the other: what kind of people are more likely to have land lines, what kind of people only have mobiles nowadays? Do pollsters have access to data indicating what age or race or gender or income level or whatever they use to make the sample representative the numbers holders are?
It may be that polling is virtually impossible today, and not just in the USA.
But I am also very worried that the more obnoxious right wing (and it does not get any more obnoxious than Trump to me) has historically fared better at the elections than at the polls, this I have observed personally since Berlusconi’s early days, and in the Brexit referendum, and many more times besides, time and again. I don’t understand how Trump even has a chance at all. Every sensible person I know shares this feeling with me, many on this board do too, but here we are. It looks like he has a chance. Frightening and unbelievable.
Maybe the US is really broken. But if this is so you are not the only ones. A weak consolation, I know. Seems like the liberals (in the good sense of the word) are in for their share of "End of Times Feeling"™, like the religious zealots have been enjoying for some time. Only I can’t enjoy it, it is sad beyond words.
Sure, but I think part of the reason people are so incredulous is that the “replacement” is someone who is a known quantity who was a rascal already voted out. That makes it unusual among such elections.
There has been a tremendous decline in the percent of Americans who will talk to a pollster. You’d think that would break polling, but it hasn’t. Going back several decades, I see well-documented findings such as this:
The Polls Were Historically Accurate In 2022
Some cycles polling is more accurate, and some less, but there’s no accuracy trend.
Of course, the quality control of comparing final polls to election results only directly applies to final polls.
Methods vary some, but IMHO it is unfair to call it finagling or guessing.
There are ways to objectively know how many many Americans are in various cohorts, such as cell phone and land line users. And the pollsters ask for demographic information, such as age, race, and gender. Then results are weighted accordingly.
Maybe the poll respondents are just screwing with ya:
Sure does.
You have to dial back to very large considerations to find parallels, but they do exist.
This group is precisely those that will maximize the “answering polls to send a message”.
“Biden is supporting the Isreali war crimes I can’t vote for him”. “Biden isn’t doing enough for climate change I can’t vote for him.”, 'Biden isn’t doing anything about guns, I can’t vote for him."
Hopefully once the campaign really gets underway they will realize that as far as Biden is from their dream candidate, the alternative is much much worse in all categories.
Yes, it is unusual.
One way Trump is fighting history: Election losers usually lose the rematch, too
And in the most recent presidential example, Eisenhower beat Stevenson more strongly the second time.
Donald Trump lost the election.
He lost his sexual assault jury trial.
The way his New York civil fraud trial is going, he’s almost surely going to lose the right to do business in New York.
And there’s a fair chance he’ll lose a criminal trial before the next election day.
Will all that losing bury Teflon Don?
I fear not. I just can’t get over that the jury decided he sexually penetrated a woman against her will in
a Fifth Avenue store, and it did nothing to his numbers. That’s awful, awful close to proving correct his claim that he can shoot someone on Fifth Avenue and get away with it. He’ll just say it was self-defense. And the polling is not broken.
The one time I ever got polled over the phone (which was 20 years ago), it was a robocall that identified itself as from ABC News and asked me to push buttons to record my opinion - there was no option to justify my choice or send a message, just press 1 for this, press 2 for that, and so on.
I disagree with the news media is at fault. Polling a year out is not particularly accruate, but it more accurate then anything else. News Media conducting and sharing polling is good. Certainly better then having pundits make senseless predictions.