In a previous article (https://www.glibertarians.com/2024/10/a-look-at-current-presidential-polls/), I summarized poll predictions for the 2024 presidential race and compared them to their accuracy in 2020 (as measured by Atlas Intel, https://atlasintel.org/media/atlasintel-was-the-most-accurate-pollster-of-the-2020-presidential-election). How did they do in 2024?
Pollster | 2020 Accuracy | 2024 Prediction for Battlegrounds |
Atlas Intel | 1.94% | +2.4% |
Trafalgar | 2.31% | +1.7% |
Insider Advantage | 2.78% | +1.4% |
Rasmussen | 3.28% | +2.3% |
Emerson | 4.15% | +0.9% |
The Hill/Emerson | 4.15% | +0.4% |
Marist | 4.79% | -0.7% |
Morning Consult | 4.99% | +0.7% |
New York Times/Siena | 5.47% | -0.6% |
CNN | 7.16% | -1.6% |
Quinnipiac | 8.62% | +0.1% |
Note: Redfield and Wilton was included in the original article, but not here as they did not release a poll in the last few weeks of the election. Removing them did not change the results much. The battleground averages give equal weight to each state, and favored Trump by 2.8% in the actual vote. Quinnipiac would have been negative except for an anomalous +6% for Trump in Georgia.
Taking the top four versus the bottom five on a state-by-state basis:
State | Actual vote | Top 4 | Bottom 5 |
AZ | +6.0% | +3.3% | +1.5% |
GA | +2.2% | +2.8% | +1.6% |
MI | +1.4% | +1.0% | -2.6% |
NC | +3.3% | +2.5% | -0.4% |
NV | +3.3% | +1.3% | -0.7% |
PA | +2.1% | +1.8% | 0.0% |
WI | +0.8% | +1.3% | -2.0% |
Arizona still had less than 80% of the vote counted as of this writing, while the others are at 98% or more. I didn’t want to wait around for more complete numbers.
In every case but one, the top four did better, and often substantially better, than the bottom five. Only for Georgia were they off by about the same amount, albeit in opposite directions. Note that the top four as a group predicted a sweep of the battleground states, and therefore a convincing Trump win, while the bottom five predicted a Harris win in the Electoral College, even if Trump had pulled out the predicted tossup in Pennsylvania.
A New York Times article published a few days before the election cast aspersions on mostly unnamed “partisan” polls (as determined by fivethirtyeight.com) that favored Republicans compared to the “nonpartisan” polls such as those commissioned by various traditional media. The article was skeptical of polls that were indicating a Trump sweep of the battleground states. But those polls seem to have been more accurate this year.
Also interesting is how Trump’s percentage differences in each state changed from 2020. But that is a topic for a later post.
Thanks, whiz, this really puts polls into perspective. What a racket. Or should I say, propaganda tool.
I wonder if there’s any correlation between percentage changes and inflow/outflow.
Sorry, population inflow/outflow of each state.
Some more commentary. Particularly interesting why Ann Selzer in Iowa failed so badly.
Most Polls Underestimated Trump. Here’s Why Ours Didn’t
https://www.wsj.com/opinion/most-polls-underestimated-trump-heres-why-ours-didnt-election-voters-575e1aa2?st=bBkhpg&reflink=desktopwebshare_permalink
Mind you they are hawking their own business after the fact, but still interesting.
Random number dialing was passe back when I was taking a public opinion polling class in college in 2010. Ann Selzer knows full well that she ran a shit poll.
I’m still skeptical that any blend of online, text messages, phone, and the like are going to be representative. There is always going to be this 2 or 3% fudge factor on top of the statistical error due to not having a good mechanism for getting a randomized sample of broad populations now that landlines are dead.
My personal feeling is there are many more people like me who now refuse to answer any poll.
On the vast majority of them the questions are also ridiculously worded to shape the desired response.
Also – I thoroughly enjoyed this bit of cluelessness by our journalistic betters. They questioned her!
https://www.wsj.com/opinion/most-polls-underestimated-trump-heres-why-ours-didnt-election-voters-575e1aa2?st=bBkhpg&reflink=desktopwebshare_permalink
As I noted when I shared this before – I most certainly feel comfortable with journalists ability to understand math and statistics.
There is always going to be a fudge factor, there is no way to do this without, for reasons alluded to by Sensei. And this is where each polling company comes in with their ideas of how to get past this and be the most accurate.
That is, unless the company is a propaganda firm at some level, like Seltzers is.
My personal feeling is there are many more people like me who now refuse to answer any poll.
This is a factor that I think is underappreciated.
I could have been tapped 100 times to be in various polls, but I don’t answer the phone if I don’t know the number, I adblock when on the internet, and I delete robo-texts. That’s all before the part where I refuse to do surveys or polling due to the way such data is weaponized to harm people.
I don’t think the Iowa Poll is necessarily biased — it’s been fairly accurate in the past, at least as much as you could expect.
trsh –
Yeah, I wonder if there is a bias that needs to be or even can be measured there.
Do people who refuse to answer polls or who are never asked in the first place (like me – I have never been asked in my life) vote differently?
@whiz
Which is why they tried so mightily to use it as a propaganda tool right before the election. Even right thinking journalists like Megyn Kelley mentioned that the Des Moines poll is considered a “gold standard” in polling.
Which could mean a few things:
1. She just got a bad draw on respondents.
2. People are actively misleading polls.
3. It was a shit day and shit happens to even the most studious.
4. It was a purposefully misleading poll for political purposes.
1 can definitely happen. There’s definitely a chance that a pollster could get bad leaning results because they didn’t run into any people willing to be honest about voting Trump.
2 Seems highly likely. Although it’s disappearing, particularly after a landslide election, but there are people who simply don’t want the harassment that comes with admitting you’re a Trump voter.
3 can happen to any one of us. It’s just that our shittiest products aren’t generally out there for the world to see.
4 seems possible. But I’m not sure why you’d go out of your way to ruin what appears to be a stellar reputation in your field of work. Maybe one could argue that risking the credibility of your business is worth not having OMB in office, and those who rely on polls (journos mostly) will give her a pass for a valiant effort, but the point still stands.
I don’t think the Iowa Poll is necessarily biased
Not biased, clumsy.
She used a sledgehammer to do an appendectomy and then is wondering why the spleen is all crushed up.
The Iowa Poll used to be the only one I trusted.
It was so obvious that it was wrong this year. It killed what little trust I had left.
Clearly, she missed – the appendix and spleen are on opposite corners of the abdominal cavity.
A New York Times article published a few days before the election cast aspersions on mostly unnamed “partisan” polls (as determined by fivethirtyeight.com) that favored Republicans compared to the “nonpartisan” polls such as those commissioned by various traditional media. The article was skeptical of polls that were indicating a Trump sweep of the battleground states. But those polls seem to have been more accurate this year.
The New York Times relies on experts. They decline to lend credibility to wishful thinking.
The NYT was right. The voters were wrong.
Great post!
Thank you! I just got back from a couple of my wife’s appointments, so am just now able to comment.
From the morningnthread, Kamala Harris raised a billion bucks and spent a couple hundred million more…
Someone needs to look into her campaign and into act blue.
Presidential campaigns are hard dollars, if I recall. So all of those billionaire donors don’t help, unless they are laundering contributions through straw donors.
A quick check says it is now $3,300 per individual.
That would mean that Harris had to have at least 300k donors to her campaign.
Clearly impossible.
So they used accounting tricks to get around limits.
Should be an easy prosecution.
I fully believe the rumors that act blue has been structuring donations.
I’m also not sure why it’s such a big deal to get so much money into the campaign itself post-Citizens United. It seems like super-PACs would be the way to go.
Some of their payments look like fraud too. $10 day 1, $1 day 2, $500 day 3 for A LOT of donors.
I looked up my parents’ names in to see if they had “donated” to Act Blue. They had not, but others with the same names had. It looks like some of the donations might be transaction fees or donations to Act Blue itself. Pretty consistently there was one donation accompanied by a second donation that was 10% of the first on the same date. The timing on some of the recurring donations was strange, but that might be explained by donations to multiple candidates. For example I donate monthly to candidate A on the first of the month and money to candidate B on the 7th of the month. So while I can believe there are shady goings on at Act Blue, I also think there are explanations for some of it.
Didn’t Obama receive a record number of individual contributions that were below the reporting threshold?
Don’t forget about the Clinton donations, back when it was Bill and China.
I wonder how much of that was looting before the Titanic went down. If all your internal polling is telling you you’re down I can see some people feathering their bed.
So the “polls are for strippers” crowd is correct?
Thanks, Whiz. Interesting stuff!
Husband won a Raising Cane’s contest with 28 free combo boxes. XX FLUVS Raising Cane’s. Us, not so much. We can get a bag of chicken fingers from Sam’s Club for cheaper. Wondering if he’s going to give it to her.
Congrats to the husband on winning, but I’m with you. I can’t quite figure out how Raising Cane’s seems to be everywhere when their chicken strips seem to be the most flavorless in the market.
Flavorless and no gravy to dip them in. They looked at me cross eyed when I asked “we have Cane’s sauce” they said. Ugh, chicken strips need gravy.
Cheaper than free?
Well, I mean, as in, going there and spending money for chicken strips.
Must be a libertarian thing.
The “premium chicken strip” joint is weird to me too. We have PDQ, Raising Canes and i think another one or two.
The wives and kids always want to go there. People have about the sauce.
Raising Canes sauce seems like mayonnaise with black pepper in it. I don’t get it.
They are nice places with good service, but as you say…. it is chicken strips.
Not only are the frozen options fine, around here the Publix chicken strips are every bit as good.
Clearly we are out of step with the normal folks on this one.
OBE cooks > Kroger/Costco/Etc > Fast Food chicken joints.
That is how Mrs OBE handles the decision of fried chicken.
I thought Raising Caine’s was the traditional ketchup + mayo blend.
I will stand up for one of our local places which started with just chicken wings and sides, added tenders (and sandwich options based on that), and when the price of chicken got higher, came up with “Vegan wings” (mixed pieces of cauliflower and tofu). They won a Cruelty Free Award from PETA they used to keep in the bathroom, and one of the owners commented, ‘we did it for price reasons, and are glad to took off. Pretty sure the chicken isn’t cruelty based though.”
I still miss them having chicharrones (pork rind variant, not fried cubed pork variant).
We have a Zaxby’s too.
I think the dipping sauce has ketchup in it as well as some unidentified spice, but it wasn’t enough to blow my socks off, and is sure as hell is not the ambrosia that is fry sauce.
Yeah I just cook Bell & Evans frozen. But I am solo so I don’t eat out much.
https://www.npr.org/2024/11/12/g-s1-33817/shell-dutch-appeals-court-overturns-climate-ruling-carbon-emissions
THE HAGUE, Netherlands — A Dutch appeals court on Tuesday overturned a landmark ruling that ordered energy company Shell to cut its carbon emissions by net 45% by 2030 compared to 2019 levels, while saying that “protection against dangerous climate change is a human right.”
The decision was a defeat for the Dutch arm of Friends of the Earth and other environmental groups, which had hailed the original 2021 ruling as a victory for the climate. Tuesday’s civil ruling can be appealed to the Dutch Supreme Court.
MAGA is everywhere.
Reality eventually wins. Europe – in particular the UK and Germany – have really screwed the pooch wrt energy.
The same week that the Dutch restart the Jew-hunts, they stab Gaia?
In fairness it isn’t the Dutch.
Technically, it wasn’t the Dutch in WWII, either.
Although it was almost certainly a Dutcher who ratted out Anne Frank.
People are shockingly ignorant when it comes to things like polls. Many years back, several companies ago, I was working on a project with a PM who had attended all the right schools, had an MBA and had never heard of the difference between stated and revealed preferences.
It immediately became something he used in his meetings. As in the next day.
It was a paradigm shift for him. He needed to leverage best of breed terminology to optimize synergies.
Thank you, Jaime, for reminding why I am glad to not spend my days in the corporate world any more.
Well, he bubbled it up, so once upper management got ahold of it, it could rain down on him from the 5000 foot level.
Posted before, but too good to miss another opportunity. The first part doesn’t mention polls, but it’s got a great line in: https://www.youtube.com/watch?v=ahgjEjJkZks
In a marketingt class I took long ago, the professor handed out a survey concocted by one of her other classes. i can’t remember what thy were selling. I filled it out as a wealthy freshman girl whose favorite color was pink, et c. At the bottom I put, “Never forget. People lie on surveys.”
The only thing I need to know about polling is the “spam call” on my caller ID. Immediate decline.
The people who answer polls are lonely and don’t have anyone to talk to.
I did a couple of the automated ones this election season, but declined a live person one.
I did a live one the afternoon of the presidential debate with one of the companies…famous name, but I don’t recall it and I don’t think its on the list in the article.
For most of them, I send them to the LLM call screening, which very few people even attempt to interact with.
I told you – I do not talk to computers.
A.I. Neph should become a reality to answer calls and go on long diatribes without supplying straightforward answers.
So, an AI trained on Kamala?
^This. That is also why surveys lean negative for businesses. Only angry people will fill them out.
I recall from chats with QM friends, that the generally accepted ratio was 1/10 for good reviews, and 1/3 for bad reviews. As in 10% of people who have a good experience will spend the time to comment on it, while 33% of people with bad experiences would spend the time.
Let the hate flow through you, young customer…
Businesses are getting extremely needy these days. “Do you have a moment to answer 29 questions about your experience with us?” Can I just buy a chicken sandwich or get my car repaired and get on with my life?
Yes, polling lonely boomers.
I was actually called for the Iowa Poll the previous month. I did participate.
Tendentious nonsense
The tradition of Fed independence aims to give the central bank the ability to shape monetary policy decisions — such as raising or lowering interest rates — based solely on the future health of the U.S. economy.
Yes, like propping up the bond market to allow virtually unlimited borrowing at zero interest in order to fund out of control fiscal policy.
If a poll is favorable to the Republicans, it’s a partisan poll. If it’s favorable to the Democrats, it’s non-partisan. This is similar to the fact checking method of reaching out to the nearest Democrat to get the “facts”.
Polls: Taken by people who sell ads and need to keep eyes glued to screens
Fact checking outfits: Owned by leftists who have never argued in good faith in their lives.
Paragons of truth all of them. If you dont believe it just ask ’em.
Hopes and dreams!
https://www.cnbc.com/2024/11/08/toyota-california-ev-mandates-impossible.html
Battery-electric, fuel cell and, to an extent, plug-in hybrid electric vehicles qualify as zero emission under the regulations.
Fuel cells generate water vapor. Isn’t H20 the single most effective “greenhouse gas”? Why does California hate Gaia?
The green movement is all a scam and none more so than the EVs. With regards to energy all they are doing is playing a shell game.
No matter how much sleight of hand they use the same amount of energy comes from the same place and goes to the same place. The problem of course is that to play the shell game requires many times more energy, being much less efficient and thus less ‘green’.
We are drowning in an ocean of grifters. I have been dozing on and off all morning in front of the TV and hearing ads. Right now they are hawking ‘super beets’. Earlier it was smart pills. Supplemental insurance. Some kind of rub on your face magic potion that makes you younger. It never ends.
Maybe that is why I feel like I have had an extra black pill this morning.
Oh fuck. Now it is a magic clam that only grows in a secret place over the rainbow. Apparently the elixir of it’s oil cures inflammation.
Goddammit people, stop giving these fucking snake oil salesman money!
Do not respond to anyone using an emotional appeal, fear or appeals to authority – and I mean in politics, charities and products.
It will simplify your life and your wallet will get thicker.
Wait a minute! It was my last chance to donate to holocaust survivors. Do you want holocaust survivors not to survive? I didn’t think so. Pony up, better still send it to me to include in with my generous donation.
Wait, you have magic clams?
Damn right Suthen, save that money for those limited edition gold (plated) Trump commemorative coins!
Do I have magic clams? No, but over the years I have run into a few young ladies that did.
‘..gold plated…’
You made me spit my tea.
*I do have some collectables…the coffee mug with Trumps mugshot on it…we have a pair of those. Also, wife could not resist the little ‘Trumpinator’ doll. We have one on the mantle.
lol I used to think we were past the days snake oils and elixirs. Nope.
I forgot about the holocaust survivors. Those poor old women in Russia wearing sack cloth with nothing to eat but bugs and wild roots. I love those ads…the pitch girl with the overly affected pitiful tone standing next to them wearing a $5K Overland coat is priceless.
She’s working the Ukraine beat now.
Good article.
There is only one poll that counts. The many polls that precede it tend to keep people anxious and confused. After the one true poll results are revealed it is easy to see how obvious it was all along.
I haven’t paid the slightest attention to polls since 2016.
“It was my understanding that there would be no math.”
No, that’s my other series of articles 🙂
Of course, a confounding variable in polling v results is that polls sample voters, while results count ballots. We’ll never know, of course, how many ballots were, umm, introduced into the counting in the battleground states. While there could be an innocent explanation for the jump in Harris ballots in WI, it sure looks like a dump. Just not a big enough one to turn the election.
I meant to put a link to the graph here, but I can’t seem to find it now.
https://x.com/EndWokeness/status/1856120816544473433
This one?
That’s the one. Thanks, Sean.
That does look suspiciuous. OTOH, Milwaukee was late, and would certainly go for Harris. Looking at the graph, there was a much smaller jump for Trump at the same time, so if those precincts went 80/20 or 90/10 for Harris, it might look like that.
Big jumps in the Dem without any change in the R (like apparently happened in 2020) is what really makes you wonder.
I can’t comment without knowing the process that a ward or city has their votes added to the state’s totals. If for example a ward close to Marquette university counts votes locally, then send the counts to the greater Milwaukee count in one batch that would make a big bump for a Democrat. If Madison WI holds all of their votes and only sends them to the state for counting in a big pile once the voting polls are closed and all wards have reported, then a big dump could look suspicious.
If they are continuously counting votes and reporting automatically as they are counted, then this would make less sense.
If…maybe…it could be…
NO. They cheat. They have been caught too many times.
Cop: “How do you happen to be in possession of this knife, this ski mask and the wallet that was just mugged off of that guy over there?”
Guy with neck tattoos: “I…uh…I found them! They were just laying there next to the dumpster I was hiding behind!”
Yeah, sure buddy.
STEVE SMITH HAVE LONG POLE. HIM ALWAYS ACCURATE!
Gotch ya pole right here!
Scrolling through the “business” headlines, I see the NYT is assiduously poisoning the well with regard to Yrumponomics.
We’ll all be in rags nd starving by the Fourth of July. Well, except for right wing MAGA billionaires, anyway.
Trumponomics, too.
Y change it?
At least you didn’t call it tRumponomics or Drumpfonomics. Those would be the Block Insane Yomamas of economics.
I had blessedly forgotten about xem. Did xey make the crossing?
Ughhhh IIRC I think that person got the hammer
I thought you were talking about that Greek guy who’s been banned from speaking at college campuses.
Socrates?
*sigh*
On a vendor webinar that includes people from other clients.
I forget how dense the general public can get at times.
Reading the manual for this Mental Health First Aid course I’ll be taking next week. It’s pretty horrifying in the signalling and the shibboleths. It could have been written on tumblr ca. 2005.
“Pandering to this stuff triggers my PTSD from the First and Second Great Meme Wars.”
“Certified peer specialist.” This is a thing? I have a certificate that says I’m a peer? Is this like a royal grant of title?
“Hello fellow peers!”
The ‘r’ is a typo. It’s a German handbook.
https://nypscb.org/
Holy fuck. That photo of ‘generic people I wish I was friends with’ with the phony smiles. *shivers*
That’s creepy.
Our expectations are high. https://hwfo.substack.com/p/autistic-rationalist-goes-to-dei
I think that’s the post that got me following HWFO. He’s a treasure.
I scratched myself off the list. I don’t want to get stuck with that crowd.
Answer: It doesnt matter who goes or stays. If those are your standards for choosing people we are all doomed.
What was I saying earlier about drowning in grifters?
Thanks whiz. This has been interesting.