Plus: Voters trust Democrats more than Republicans to handle the economy; a majority of voters say Trump is acting above the law; and 62% say the country is off on the wrong track
Thanks as always for the comprehensive data and analysis. I'm really enjoying your newsletter, it's a breath of fresh air from the pundit narratives divorced from reality.
One question if you're able to answer it: Since the percentage of respondents who said they attended No Kings rallies is clearly far higher than for the population as a whole, does that suggest that the anti-Trump sentiment in the poll overall is overstated? Or have you adjusted for that? Contrarians to your conclusions may argue that polls have long underestimated Trump/MAGA support, and if far more are saying they attended No Kings rallies than in the population as a whole, the same may be true for Trump's overall disapproval?
I found it interesting how tight the margins trusting Democrats are with voters on the economy and jobs at +2% for democrats and on prices and inflation at +5% for democrats. These 2 items are the most important issues for 48% of voters, so once again how the economy is doing seems to dominate the concerns of most voters. Those 2 items plus health care and government spending and social programs which most concern another 24 % of voters, highly favor democrats, so does that mean the importance of the economy, jobs, prices, and inflation are being undervalued when polls ask which party voters favor or trust most or are parties simply becoming less important to the voters?
If 11% of respondents said they participated in a No Kings protest, which is way above the reported numbers, doesn't that suggest a very skewed sample?
There is still debate about the 6-7 million conclusion.
No offense, Elliott, but that number is almost certainly low, based on my personal experience on that day.
We counted and stickered 118% of the "official count" in my town alone. The news reports and analyses didn't believe it was possible to get that many people in a town our size, so they underreported, even when faced with photo crowd counts that estimated even more than we counted!
I've also heard anecdotal accounts like mine all over the country. If even half the small protests were reported at 5 of every 6 attendees ... Well, you do the math.
"A plurality of Americans say they get their news from social media, podcasts, or online news aggregators/newsletters"
Let's find out WHICH podcasts/aggregators/newsletters
AND
Let's find out which organizations/donors fund and control those "aggregators". I've always wondered about that, since finding a specious "24,000 Home Depots are going out of business!" news post on my phone from the U.S. Sun, which is a subsidiary of the UK Sun (then spent an unrecoverable amount of time figuring out how to block U.S. Sun notices from my phone).
Encouraging polling but I suspect things will even out again once Congress sorts SNAP/healthcare funding and Congress puts its foot down on Trump's idiotic tariff schemes.
True! But a bipartisan group of Senators blocked Trump's Canadian tariff scheme so that's a tiny glimmer (really tiny haha) of hope. Would need the House to go along tho so...
I'm still hung up on the strategists' fallacy and the attempt to relate election prediction to polling results based on aggregation of voters responses to detailed policy questions. I realize that as the poll is dissected, results become less meaningful and the risk of "torturing the data" becomes great.
Are there polls that intentionally identify & report on low information/low interest voters? Is this the group that generates the largest swing in an election? Is the low information group swung by the high information group (my friend told me to vote for ....) or are they swung by some gestalt? What is the best query(s) to assess voter engagement? So many questions ......
Hi Elliott, Excellent article, thank you. I wonder if I might ask a question please? I do my best to follow polling/elections and I wondered if you had a view about the respective predictive qualities of the 'Nowcast' models used by Dave Trotter of Voting Trend and conventional opinion polls? Many polls are showing a very tight race in New Jersey, whilst Mr trotter's analysis using his Nowcast model suggests it could be the Democrat by double digits. Just wondered if you had any thoughts?
Hi Elliott, Please see: https://www.youtube.com/@votingtrends I'd be interested in your thoughts about whether there is value in this with regard to SIN etc. Thank you.
Ok Russell, I haven't been able to find a concrete explanation of his methodology. I did see this, in which he criticizes polls in VA and NJ based on the performance of polls in the NYC mayoral primary. I do not think that is a good critique of the polls; polling primaries is very hard and not the same as polling a GE.
Quick note. In the section "Trump issue approval and direction of the country: both underwater", you talk about the Democrats' "2016" election strategy. I assume you mean "2026"....
Thanks as always for the comprehensive data and analysis. I'm really enjoying your newsletter, it's a breath of fresh air from the pundit narratives divorced from reality.
One question if you're able to answer it: Since the percentage of respondents who said they attended No Kings rallies is clearly far higher than for the population as a whole, does that suggest that the anti-Trump sentiment in the poll overall is overstated? Or have you adjusted for that? Contrarians to your conclusions may argue that polls have long underestimated Trump/MAGA support, and if far more are saying they attended No Kings rallies than in the population as a whole, the same may be true for Trump's overall disapproval?
I found it interesting how tight the margins trusting Democrats are with voters on the economy and jobs at +2% for democrats and on prices and inflation at +5% for democrats. These 2 items are the most important issues for 48% of voters, so once again how the economy is doing seems to dominate the concerns of most voters. Those 2 items plus health care and government spending and social programs which most concern another 24 % of voters, highly favor democrats, so does that mean the importance of the economy, jobs, prices, and inflation are being undervalued when polls ask which party voters favor or trust most or are parties simply becoming less important to the voters?
Elliott, please read: https://drive.google.com/file/d/1uQkcwWcv80D4PyEDiEDXBLzIA38nu3xG/view
If 11% of respondents said they participated in a No Kings protest, which is way above the reported numbers, doesn't that suggest a very skewed sample?
Polls have a problem with over-representation of politically engaged folks.
There is still debate about the 6-7 million conclusion.
No offense, Elliott, but that number is almost certainly low, based on my personal experience on that day.
We counted and stickered 118% of the "official count" in my town alone. The news reports and analyses didn't believe it was possible to get that many people in a town our size, so they underreported, even when faced with photo crowd counts that estimated even more than we counted!
I've also heard anecdotal accounts like mine all over the country. If even half the small protests were reported at 5 of every 6 attendees ... Well, you do the math.
"A plurality of Americans say they get their news from social media, podcasts, or online news aggregators/newsletters"
Let's find out WHICH podcasts/aggregators/newsletters
AND
Let's find out which organizations/donors fund and control those "aggregators". I've always wondered about that, since finding a specious "24,000 Home Depots are going out of business!" news post on my phone from the U.S. Sun, which is a subsidiary of the UK Sun (then spent an unrecoverable amount of time figuring out how to block U.S. Sun notices from my phone).
Encouraging polling but I suspect things will even out again once Congress sorts SNAP/healthcare funding and Congress puts its foot down on Trump's idiotic tariff schemes.
It's a big ask from this prostrate-to-the-king congress to get off their bellies and exert their constitutional authority.
True! But a bipartisan group of Senators blocked Trump's Canadian tariff scheme so that's a tiny glimmer (really tiny haha) of hope. Would need the House to go along tho so...
And Unholy Mike has hitched his political power to his Carrot Colored Charlatan. No way he will move an inch.
I'm still hung up on the strategists' fallacy and the attempt to relate election prediction to polling results based on aggregation of voters responses to detailed policy questions. I realize that as the poll is dissected, results become less meaningful and the risk of "torturing the data" becomes great.
Are there polls that intentionally identify & report on low information/low interest voters? Is this the group that generates the largest swing in an election? Is the low information group swung by the high information group (my friend told me to vote for ....) or are they swung by some gestalt? What is the best query(s) to assess voter engagement? So many questions ......
The methodology section talks about a September poll but the narrative refers to an October poll
I am sure you are juggling a lot of balls and I so appreciate your work. Thank you!
Sorry, I updated that in the topline document but not the methods section of the post. Fixed!
How can you describe this as a “new” poll when it was conducted more than a month ago (September 15-19)?
Says at the top that the poll was October 24-29.
Hi Elliott, Excellent article, thank you. I wonder if I might ask a question please? I do my best to follow polling/elections and I wondered if you had a view about the respective predictive qualities of the 'Nowcast' models used by Dave Trotter of Voting Trend and conventional opinion polls? Many polls are showing a very tight race in New Jersey, whilst Mr trotter's analysis using his Nowcast model suggests it could be the Democrat by double digits. Just wondered if you had any thoughts?
Hi Russell. I’ve never heard of this model or Mr Trotter. Could you send a link please?
Hi Elliott, Please see: https://www.youtube.com/@votingtrends I'd be interested in your thoughts about whether there is value in this with regard to SIN etc. Thank you.
Ok Russell, I haven't been able to find a concrete explanation of his methodology. I did see this, in which he criticizes polls in VA and NJ based on the performance of polls in the NYC mayoral primary. I do not think that is a good critique of the polls; polling primaries is very hard and not the same as polling a GE.
https://www.youtube.com/watch?v=Y0QItTgqgQM
Many thanks for taking a look Elliott. It will be interesting to see what happens on Tuesday. Thanks again, Russ.
I’ll check it out and report back. Great question!
Thank you for including the most important question, IMO, which is a ranking of which issues are most important to voters. That was very helpful.
Quick note. In the section "Trump issue approval and direction of the country: both underwater", you talk about the Democrats' "2016" election strategy. I assume you mean "2026"....