The 2024 CES pre-election survey, based on a huge sample of nearly 80,000 adults, puts Kamala Harris ahead of Donald Trump 51-47% among likely voters. She's ahead 52-46% among very likely voters. Here's their breakdown by age and gender:
Harris is tied or ahead with every single group except one: old men, where she's behind by 12 points. But what can you do about that? This is the core Fox News demographic, and they've been told for years that Kamala Harris is a radical leftist fanatic who's determined to let in a ravaging horde of Mexicans who will destroy America. There's not much hope of ever getting through to them.
I am unclear if, CES is a valid sample of likely voters nationally, or their track record (2016 or 2020) of historic polling.
https://tischcollege.tufts.edu/research-faculty/research-centers/cooperative-election-study
I can't imagine becoming so old that I turn into a Trump supporter/Fox Entertainment junkie.
Hang in there. I’m 76 and happy to say I’m more liberal than when I was younger.
I don't see myself as becoming conservative to be honest. I think those that become more conservative were probably conservative leaning to begin with.
Meanwhile: https://www.cnn.com/2024/10/28/business/trump-social-media-stock-election/index.html
They quote a venture capitalist saying if TFG wins, the price will appreciate further. That wouldn’t be on fundamentals, but it would be a fine way to launder money and bribe a sitting President. Remember, though, today’s buyers have no more actual information than any poll aggregator/election modeler. They’re placing bets. Most bettors lose money.
I hope to dog this is right. I really do!
Look at all those women voting for Trump. Hilarious. I was thinking that Dems were the party for women and white urban men, but that's giving women too much credit. Go figure.
Time for a policy shift... (I know, J Chait) But these groups were wrong! They called Obama the "deporter in chief" for gods sake.
This episode ought to prompt deep self-reflection. Immigration-rights groups spent years convincing Democrats that the key to cementing the loyalty of Latino voters was to follow the immigration-rights agenda on border enforcement. I personally believed this as well. But those groups turn out to have imposed their own view, luring Democrats into adopting politically toxic policies that repelled them in the mistaken belief it would win them over.
But rather than concede error, the groups are continuing to attack the administration for its harsh enforcement approach, and — even now! — urging Harris to abandon her hard-line stance that is supposedly alienating Latinos.
https://nymag.com/intelligencer/article/kamala-harris-donald-trump-television-ad-trans-prisoners-activists.html
Do you think the number of trans immigrants seeking medical treatment in jail is that high? If so you're part of the problem. It's virtually an ant hill made out to be Mt Everest by the right wing echo chamber.
I don't have any information on that. I was pointing out that immigration activists were wrong about it being good politics. It is, in fact, bad politics. Probably bad policy too, but that's not my concern.
I'm amazed too, that Justin has actually stumbled across something realistic in his usual echo chamber of hate and nihilism.
Here's a fun fact, women can be misogynists too. And racists, and both ignorant and profoundly stupid.
Your brain worm is still pretty healthy, I see.
I am shocked by the women 40-59 numbers. More than shocked!
me too. if they're right, it really exposes how we're all in our own social and political bubbles.
I keep reading how women are going to save us from Trump. If that poll is correct it will be young men and women. I just hope they vote. It would also seem that abortion is not the issue we thought it was.
I’m a parishioner at a very large Catholic Church. The women there are much more active and vocal about being pro-life than men. Some group within the church periodically organizes pro life demonstrations (this is an extremely small segment of the of the church members that participate). The majority of the organizers and participants are women.
I just want to check in as a member of "Old White Dudes for Kamala". We do exist!
71-year old male for Harris here. Though as a Brooklyn-born Jew (in my golden years, I've relocated all the way to northern NJ, where I can see the Empire State and Chrysler Buildings from my window), I'm not sure that I'm counted--or want to be--as "white" in certain orange-tinged circles. Stephen ("America for Americans") Miller really needs to check the company that he keeps.
68-year-old Christian in South Dakota, strongly for Harris. I think I "count", even if you don't. Even my 100-year-old dad said he couldn't vote for Trump again (he did twice before).
Give your dad my thanks - he & Jimmy Carter are appreciated & admired!
Only 54 but I would consider myself a member of the "Old White Dudes for Kamala" group! I think once a younger person calls you sir, you're old! And that's already happened to me!
Unless one has the State level information, and the state level regarding the swing states, it is quite useless. The nasty reminder of 2016, you have Electoral College 50 seperate elections not an actual national aggregate.
Madame Harris may very win a popular vote on national and lose the electoral (or not, it is not a prediction, it is a scenario potentiality). It is engaging in self-deception to analyse by national.
As I’ve mentioned before, I have a hard time believing any poll that shows Trump and Harris tied or Trump in the lead nationally. I base this on the fact that Trump lost by 7 million votes in 2020. I very seriously doubt he has turned half of those Biden votes his way. Particularly since 5 million of those voters reside in California which I don’t see going Trump’s way in any significant numbers.
But based on our electoral college system there is a very real chance of a Trump victory. Maybe by a lot.
i believe i read nate silver a few weeks back saying that apparently there has been a shift in the electoral college math, that trump has picked up support in states that are solidly blue and means little in terms of the outcome.
i believe i read nate silver a few weeks back saying that apparently there has been a shift in the electoral college math...
For sure there's significant sentiment to the effect that Trump's Electoral College advantage has diminished.
If Harris wins—despite what the poll aggregators are now saying (Harris up by 0.9)—it'll probably be a combination of one or more of:
(1) polls oversampling Trump vote.
(2) a diminished EC advantage for Trump.
(3) late momentum shift toward Harris that isn't or isn't fully picked up by polls (we saw something similar in 2016). That hate rally in NYC doesn't seem to be helping his cause with persuadables...
At this point if I were given a free bet, my money would probably be on Trump. But I don't have much confidence in that. I think it's perfectly plausible the polls could be off significantly. We'll soon find out!
Fifty separate but hardly independent (in the statistical sense) elections. This year the national polling tells us only that the election is very likely to be close, as the polling difference has nearly always been within the margin of error. But in years that the late polling difference is large, it has nearly always predicted the EC winner, and 2000 was the only year in which late polls mispredicted the winner of the popular vote.
Unless one has the State level information, and the state level regarding the swing states, it is quite useless. The nasty reminder of 2016, you have Electoral College 50 seperate elections not an actual national aggregate.
This is false. As KenSchulz points out, this argument rests upon the idea that the vote in each state is an independent event. This is, in fact, the error that led a lot of analysts (and specially Sam Wang) forecasting that it was basically impossible for Trump to win in 2016. Of course, each state's vote is not an independent event; there is a strong correlation between them.
The national numbers are not meaningless; you just have to be careful in assessing what they do mean. Winning the national vote does not necessarily mean that you'll win the electoral college. But any increase in the national vote makes it more likely that you will.
Good points. I might have noted that Sec'y Clinton's 2.1% margin in the 2016 popular vote was by far the largest achieved by an EC loser; the next largest was VP Gore's 0.5% margin in 2000. Putting it another way, only one popular-vote winner with a greater than 0.5% victory has ever lost the vote in the Electoral College. Of course, even late polls are imperfect predictors of the vote margin -- late polls predicted Clinton by 3 - 4%. This, plus the high correlations of voting among the states, is what leads predictors to assume that, say, a 4 - 5% margin in late polls points to an almost certain win.
I want to see the 18-39 demographic broken down into 18-29 and 30-39. I'm having a hard time believing that men in the 18-39 group are going 3:2 for Harris because everything I've ever experienced in my life between three red, blue, and purple states is that even younger men are more 50/50, and the youngest cohort of men is disturbingly more right wing than I would like.
Remember, loud does not equate to large numbers. They're loud, not numerous.
+1
🙏🏻🙏🏻🙏🏻
The numbers back it up, though, which is what I was saying - and I'm not an idiot, I know loud != numerous. This CES result showing 60-40 for Harris is the aberration and I don't think it should be trusted.
https://nymag.com/intelligencer/article/trump-harris-gender-gap-election.html
As Time magazine recently reported, John Della Volpe of Harvard University’s Kennedy School found surveys of “voters under 30 show that young men are shifting towards Republicans by a 14-point swing compared to 2020.”
"In four years, he cut what was once a 19-percentage-point Democratic margin among registered young male voters (50 percent Mr. Biden, 31 percent Mr. Trump) roughly in half (48 percent Ms. Harris, 38 percent Mr. Trump) in our poll."
And that's ^ just Gen Z, which isn't the whole 19-39 bracket.
This old white dude is proud to buck the trend of old white guys.
51 - 47 eh? that's what happened when Obama beat Romney, after the final polls showed only a single point advantage for Obama.
I don't see Trump reaching even 46% this time. According to James Carville, voters tend to break one way or another in the final week. I just can't see them breaking for Trump this time. I expect that maybe about 1% of Republicans are turned off by Trump's evident insanity, and will stay home.
So, I've made my final contributions and my yard signs are out. I will leave GOTV efforts to younger people this year. I will read a minimum of political news from now until the election.
Two caveats. As Heraclitus rightly observed, we can never step in the same river twice. And the electoral college is not our friend.
You quit yougov when it started tightening.
The CES has a much larger sample, nearly two orders of magnitude. Makes an easy case for a post.
I just followed Kevin’s link. The data for the CES was collected by YouGov.
You quit yougov when it started tightening.
More basically, Kevin has consistently cited that poll as opposed to looking at at poll aggregators (or indeed multiple aggregators). That seems a mistake, and I really don't get that choice. Per my periodic averaging of the five aggregators I track, Harris's lead has declined from 3 points on October 1st to 0.9 points as of now—a decline of two full points. On the other hand, it's possible that the bulk of the polling organizations have been overcorrecting for the past two cycles' under-sampling of Trump voters. In other words, an erroneous, pro-Trump herding effect can't be ruled out.
If, say, Harris's lead was actually 5.6 points on October 1st, her real lead at this point might still be 3.5 points—more than enough to comfortably win the election, even assuming the full two point swing toward Trump is legitimate, which it might not be (presumably the same over-correction would still apply). And one fundamental weakness of all polls is they're of limited effectiveness in picking up late swings in momentum, for obvious reasons.
Not going to lie, I'm terrified of what might happen if Trump wins. And yet, despite what my own analysis of the raw numbers tells me, I'm a bit more hopeful than I was heading into the weekend, because I feel I now have a better handle on what might precipitate polling error in Harris's favor. I also think we can't rule out a very late momentum shift back toward her. Finally, a lot of MAGA seems overconfident and complacent, so, if there's any lazy "I'm not going to bother to get out and vote on this cold night because my candidate has this thing in the bag" sentiment, I strongly feel it's much more likely to hurt Trump than Harris. Nobody likes to be scared—it's a highly unpleasant feeling—but a terrified Democratic electorate is an effective one from my perspective because it banishes complacency. This is in stark contrast to the MAGA tribe, which really does seem to have settled into a smug "Trump has already won" dynamic. And Trump GOTV machinery seems to be utterly shambolic.
There’s another potential source of error when looking at trends in aggregators, and that is composition effects. That is, weighting on recency means that at different times, the aggregate may comprise a different mix of polls. If Simon Rosenberg is correct, that Republican-leaning pollsters are ‘flooding the zone’ lately to create an illusion of momentum for TFG, that’s how it would work. One of KD’s reason for choosing to focus on YouGov was just that, that it allows clean comparisons over time, to see trends.
Two counterpoints:
1. Electoral College
2. State vote counting and elector shenanigans
Future quote: "The Harris campaign today announced it will stop contesting the electoral process in the interest of national unity. 'It is important for the nation to heal by looking forward. We must end the current chaos and instead focus on passing legislation to ensure the will of the voter is accurately reflected in all our election results going forward. We congratulate President Trump and urge the nation to unite behind our new president.'"
I really hope to be embarrassed by this prediction in January, 2025.
No offense, but so did I!
Princeton Election Consortium just dropped like a rock for Harris this afternoon. 256 EC votes. And the house is tied. And the Senate is down two.
Hoping that Sam Wang has to eat another bug …
Aha! Kevin is a LIAR! He says Fox News accuses Harris of letting in a "ravaging horde of Mexicans." FALSE! They say it's a "ravaging horde of Mexicans AND Venezuelans AND Hondurans AND Nicaraguans AND Somalians AND Iranians AND Congolese AND Martians"! And "Hannibal Lecter."
So there.
For men of my 60+ cohort the Republicans represented not much that you had to think about and a lot of room for ambition. That was about it. Anything else felt like it grossly limited their options.
I'm not one of 'em, but I noes does guys. Well put!
Interestingly, Harris' lead expands to 6 when the pool is filtered to those who will definitely vote or have already voted.
In which states?
I am somewhat confused by the polls suggesting that "late" voters (read NOT EARLY) will "break for Harris.
Is that even close to the truth?
Would THAT explain the extent at which policy statements by Trumps crowd warmers at his campaign stops are going so far over board with the rhetoric?
Is it they have nothing to lose and everything to gain?
Is the reference to ‘late voters’ or ‘late deciders’?
But what about the unlikely voters?
And the very unlikely voters?
You know, the ones who registered so they could sit on a jury.
https://youtu.be/q5uztpW5xjU?si=0zkmrgUQtrDxaFUy
This is an odd post from Drum. Until now Drum has always given these posts with some context as to why he has chosen this poll rather than something else and how to understand it. That is useful. Here he simply gives a poll, says it is large and then presents it as if it should be taken at face value. Does this poll have a history of greater reliability?
I hope that this poll is accurate. Actually I hope it has a Trump slant and Harris is winning by more. But I am not sure why we should think that this is more revealing than the final NYTimes poll or a different one. I wish Drum had given some explanation as to why this poll was worth reporting.
If the sampling is properly done, the margin of error for a poll this large is far smaller than the typical 1200-respondent polls run throughout the campaign.
Kevin didn't mention this, but the reason he used it is because this is the polling outfit he's used consistently all year. You have to dig a bit, but this is based on data collected by YouGov.
It's not just men over 60 going for Trump. The majority of white women voted for Trump in 2016 and 2020. In 2024, they’ll likely do the same.
2016 --- 52% for Trump, 43% for Clinton (per CNN exit polls)
2020 --- 55% for Trump, 44% for Biden (per NBC exit polls)
2024* --- 51% for Trump, 47% for Harris
* See link Kevin sent, sort by race/ethnicity and gender.
If that's accurate then it would sugggest Harris will win. After all, Trump lost to Biden when white women went 55-44 for Trump. If it's only 51-47 this time I dojn't see how Trump can win.
Every single poll has women much more contra Trump than men. Here it is 50 : 50. I doubt these figures are correct, no matter the source.
I'll be curious about the exit polls and what they say about how the two genders voted.