The more "you people" think and act on beliefs like this, the more normal America will turn their backs on you. There are lots of normal Americans, many highly educated and successful Americans, who truly believe the people pandering this fascist nonsense have truly either lost their minds or been brainwashed by leftist media. How many of you are still wearing covid masks in public?
Well, I'm 72, and Covid doesn't like old people. So, I wear a mask when I'm in the supermarket, or in the audience at a performance. The one time I forgot to wear a mask at a performance, I caught it. Fortunately it was a very mild case. But I have no interest in catching it again, especially as repeated exposures increase one's risk of getting long Covid. I have better things to do with what's left of my life than lying around, trying to breathe.
Well, I am 73 and we had largely ditched masks in Texas (ok maybe not Austin) six months into the pandemic. Our grandchildren did not go to school wearing them - you do know they won't stop a virus don't you? My physician had great advice. If one is worried about catching Covid - don't get tested. Then all one has to worry about is a cold.
What I do know is that N 95s are effective at stopping viruses, due to their electro static charge. Nothing is 100%, but I’ll take whatever measures improve the odds. YMMV.
Masks were never touted as a panacea. Medical grade masks do and did, in fact, reduce the risk of the wearer infecting others. That is why surgeons and ICU nurses wear them. The coronavirus SARS-Co-2 and its mutations are borne on the droplets broadcast as much as 6 feet away in a conical pattern by normal human respiration--coughing, sneezing, exhaling. Droplets are caught in the mask or muted at the source before it dissipates. There was a dramatic reduction in other airborne viral maladies --colds, flu, and possibly pityriasis rosea--while people were wearing masks and isolating during COVID. They have now returned to pre-COVID levels. Ideology is no substitute for public health.
Belief in ideologies is due to a lack of critical thinking.
The masking thing is always going to repel people for any number of reasons that outweigh exposing one's self and others to pathogens. It's interesting because in Asia, masking is taken as a matter of course when needed.
I kind knew that we were in for something bad when many of the Asian members of our crew and on other shows starting wearing masks to work in late 2019.
It is also a matter of airflow dynamics and how particles move in the airflow of a blown N95 mask a bit like the way a catch boom works to pick up trash in a river.
I cannot think of anything less interesting to argue about than whether someone else is wearing a mask. I thought Republicans were the ones who say people should be free to do as they please.
Your physician doesn't happen to make money from treating sick people does he. I knew three people, two elderly parents and their son who all caught that cold and the paramedics had to wear masks with camphor to cover the smell of decomposition when they came to their house. I would like to know where you got the would not stop a virus fact from. Corn Flakes are good for breakfast but you should not read the box.
I’m 72 and sympathize. In 2022 I attended a small applied math conference at Oxford and modeling disease and analyzing data was a now favored topic. The data/model presented on masks is that it would reduce the spread by 50%. I’m interested in further benchmarking. Data/model showed, not talking will reduce the spread by 10x (that is, instead of down to 0.50x it is down to 0.10x). So wearing a mask where everyone is talking is much worse than minimizing time in a chat zone. And yes I although I was vaccinated I got COVID traveling via Heathrow on the way back like your performance.
You’ll catch it again. Your better bet is catching it early via later flow testing and then popping antivirals quick (highly effective in initial infection but quickly wane by symptomatic presentation).
But I am curious to read the long Covid data. What most people don’t realize is that the phenomena isn’t new to Covid. What did everyone think post viral tussis (cough) was?
In 2020/2021 Donald Trump lost a fairly run election, subsequently promoted conspiracy theories while continuing (to this day) to state that he actually won this election, encouraged protestors to storm the capitol, and pressured Pence (and everybody else he could think of) to help overturn the results.
Given this behavior, do you *really* think it's "nonsense" to question whether such a person would intend to relinquish power in the future?
Maybe "don't vote for a person who does this stuff" is really a losing message, and America seemed to turn its back on it this time as you suggest. But the concern is very much not unwarranted.
Really the thing to say here if you really want to help people believe Trump will not extend his presidency beyond this term is "I'm a Republican, and I voted for Trump and I think he's going to be a great president for four years. However, if he seeks the GOP nomination in 2028 I will vote against him and if he is somehow on the general election ballot again as the Republican nominee in 2028 I will vote for whoever the Democrats nominate."
I will say that it's hard for me to game out how he could possibly extend his term beyond these 4 years, but it doesn't seem all that unlikely that he will attempt to do so (just as it didn't seem unlikely that he would run for president in 2024 -- once a person has done a thing, it's typically plausible that they'll do that same thing again). Nobody needs to really agree on whether he will try to remain in power, though. We just need to agree that we will all do our part to stop him if he does try.
Everyone I've seen on the right is very quick to disparage the idea that Trump will try to stay president beyond 2028, but I've yet to see anyone who enthusiastically voted for him already say that there's no possible way they or anyone else would vote for him if he ran again. If he does give it another try, his voters are going to get behind it.
Yea this year I’ve lived in Pennsylvania and California, both city and rural. I haven’t seen almost anyone wearing a mask outside hospitals. I feel like this is just propaganda you’re spreading.
I was at a grocery store in Austin 18 months ago (granted not a Whole Foods, because I don’t like to burn money), and I don’t think I saw anyone wearing a mask. Or at least not a notable percentage.
I also live in blue NJ in a relatively liberal area and only see the occasional person who may be recovering or immune suppressed.
So I very much doubt your story is representative of anything but conservative media exaggerations.
I live Oakland, CA. A reliable 5% still mask everywhere, especially outdoors. Two children in my kid’s first grade class are still masked daily by their parents. And, no neither these kids nor their families have any comorbidities which would recommend around-the-clock infection prophylaxis.
That said, I'm aware of one attempt by a governor to unilaterally cancel an election. It was in 2020 and the courts quickly put an end to that obviously bad idea.
Saying all progressives is admittedly too harsh. But this a recurring refrain elements of the left put forward every gop admin, without any evidence.
The repetition of such baseless fears against all gop politicians also just diminishes the charge. Kinda like how Reagan, Bush I, Bush II, McCain, Romney, and Trump were all compared to nazis at some point.
This is just completely made up. There was absolutely no widespread fears that Reagan, Bush, Bush, McCain, or Romney were threats to democracy or that they would attempt to subvert the results of their elections. And then, in the case of Bush I, McCain, and Romney, none of them did challenge the election results and Romney is among the people who condemn Trump. You cannot be serious here.
Re: Romney & McCain specifically, yes, that's the point. You don't have to look hard to find examples of democratic campaigns calling comparing them to nazis and Hitler.
And, yes, in 2004 and 2008 elements of the left circulated conspiracy theories Bush II was going to cancel the elections. The nation article in my original post was an example from 2008, 2004 leaned more into Bush would use a Madrid bus style terror attack to permanently cancel all elections.
And I honestly cannot tell if you are serious because there are literally books written about why every republican is a threat to democracy. That's the point that this line of argument isn't persuasive against Trump because it's used against every republican.
I can’t tell if you’re serious because nobody in the mainstream left was saying these things about previous Republicans and now Mitt Romney is saying it about Trump. This is such a preposterous false equivalency
This person said on camera "I committed election fraud?" Link? The only person I'm aware of who was convicted of election fraud in recent memory was McCrae Dowless in North Carolina (a Republican).
No evidence? Trump literally refused to accept the 2020 election results and incited a deadly insurrection to try and halt certification. He then peddled his demonstrably false election conspiracy theories for months afterwards, trying every avenue he could to steal the election himself. He's repeatedly talked about wanting to be President beyond his 2 terms, he praised Xi Jinping for becoming "President for Life" and suggested they should do the same in the US.
But we're all supposed to ignore that because he hasn't succeeded "yet". We have to wait until it's too late before we can point out the obvious.
Exactly. The problem for them is there is ever a wolf (for which I’m confident most would claim Trump is) then no one will take them seriously.
If everyone is a nazi or commie or whatever the nom de jour is, then it has absolutely no rhetorical value and is arguably counterproductive. Telling a family paying 25% more for their groceries to look at Trump as a rapist/indicted felon/fascist/etc only makes them tune them out. Had they stolen a page from the blame China handbook, they maybe could have allayed inflation sentiments. Then again, calling it transitory and saying don’t worry about it were woefully naïve in electoral presentation.
If I were a betting person, I'd be willing to place a bet that he won't be. Aside from reasons why any other person wouldn't be on the ballot after 2 terms as president, he's also 78. A lot of men die before they get to 82.
Zero chance he’s on the ballot in 2028. In order to do so, he’d have to pass an amendment, which would have the side effect of allowing Obama to run against him.
He didn't try to amend the constitution in 2020 to stay in power, he just came up with some legal hocus pocus to justify it and tried to force his way back. If he wants to run in 2028, he won't amend anything, he'll just have his lawyers cook up some reason the existing amendment doesn't apply, and then date anyone to stop him.
We seem to be forgetting that Trump already was disqualified under another amendment. The 2-term limit is one of the few guardrails I'm confident will hold, but I'm not 100% confident.
Yeah, I hear you. But Hugo Chavez died and Venezuela isn’t exactly a thriving democracy under Maduro. Dictatorships start with strong personalities but end with systems.
I think Trump's age will be the biggest impediment to seeking a third term. Comparing 2020 Biden to 2024 Biden there was a pretty big deteriorating effect from 4 years, and Trump is already looking pretty old. Even if he's not dead he might just not have the energy for it in 2028.
Trump running for a third time takes one of two things:
- A constitutional amendment
- The assent of the military
No one can get constitutional amendments passed even for much more broadly acceptable things. There is at most a 1-2% chance he could get it done even with major successes in the states and Congress.
And while Trump's fairly popular among the rank and file in the Army, the officer class does not like him. It does not like him at all. Military officers are classic Romney-Clinton voters, and eight more years hasn't changed that. And though many of them are still Republicans, they take their oaths seriously, and "I will support and defend the Constitution of the United States against all enemies, foreign and domestic; that I will bear true faith and allegiance to the same" comes before "obey the orders of the President" for a reason, a reason they all know.
Things might get worse. That's a real possibility. Trump running for a third term is much, much less likely than that.
Trump's lack of military officer support scuppered his plans in 2020, but he knows that as well as anyone. It's little wonder though that he's placing much higher value on loyalty with appointments this time and trying to purge the unbelievers. Maybe he doesn't succeed completely but it's comforting to get to a scenario of having to rely on unelected military leaders to force out the civilian president and his cabinet
That is an excellent example of a way things could plausibly get much worse while still falling well short of enabling Trump to run for a third term.
The independent apolitical military will not fall apart in four years. A 'warrior board' would do enormous damage and might make it possible for three terms of *Vance*, but not three of Trump; it couldn't work fast enough.
It’s kinda blowing my mind that Nate, Pfeiffer, and pretty much every other election analyst out there is acting like the party who was openly running on “Facism for everyone! Wheee!” Is going to be super cool with elections as usual.
It's the Democrats who cried wolf effect. Democrats had an absolute cow between 2016 and 2020, the sky didn't fall in, and it seems like the voters preferred 2016-2020 to what the Democrats offered, unfortunately.
A million dead people may dispute your analysis and there are a lot of republican members who are calling hands off on Biden's policies because they know how their districts benefited.
A million dead people as compared to what number? Trump handled
COVID like a moron, but I totally disagree with the implication that number would have been lower under say a President Clinton. What would have been substantially different in people’s reaction to COVID if she were President? Would people who were flouting COVID restrictions in Spring 2020 have been double masking and Tweeting #flattenthecurve under Clinton? I don’t think so.
The sky did fall in-- the judicial branch is dead. It is a delayed effect, but there have already been plenty of nakedly partisan rulings even outside SCOTUS which is completely compromised.
I agree; Trump is definitely a threat to our institutions. But those institutions have been around for a very long time, and it's not going to be so easy for him to subvert them. This isn't like Venezuela or Germany in the thirties; we've had our Constitution for 200 years, and even the justices on the Supreme Court, some of whose decisions I've strongly disagreed with, are committed to the rule of law (with maybe 1 or 2 exceptions). The next 4 years are certainly going to be interesting, but I still give us about a 80-90% chance of surviving with our democracy more or less intact. The likelihood of it not surviving is much higher than it's been with any other President, and I don't like it one bit; nevertheless, it's too soon to assume the worst.
That’s a totally valid worldview, I guess I just think it’s odd that Nate, who’s long been an advocate for probabilistic thinking (and that proselytizing has had a pretty big impact on me personally) is kinda writing off a pretty obvious set of worst case scenarios that one would normally price into their risk assessment.
If he didn’t want to be too alarmist, for example, a simple caveat of “assuming elections proceed normally” would probably be a sufficient disclaimer that addresses the legitimate 10-20% risk you’re describing here.
To me this probability estimate seems too optimistic for two reasons:
1. For most of Trump's first term, he didn't know how to work around the limited guardrails; now he does after his 1/6 trial run and armed with the Agenda 47/Project 2025 playbooks. Plus many of the authors are now announced as part of the incoming administration.
2. The Supreme Court has immunized him from prosecution if there's a pretense that his actions were official acts.
It’s one of the ways I’m morbidly grateful for having holocaust survivors in my family tree (grandmother). It really gets you to internalize the capacity for countries to become the worst possible version of themselves very, very quickly.
It's due to comments like this that Trump won and in an electoral landslide increasing his margins across all demographics. If progressives keep this up 2026 may actually result in further losses
I don't know how serious you are, but to the extent you are serious it's part of a broader problem of liberals continuing to treat Trump as an existential danger to democracy when the majority of voters, looking back over his publicly available record of governance, simply don't see it.
Also, not only is Trump constitutionally barred from seeking another term, he wouldn't want to even if he could. His ego has been sated, and he's lazy.
No, I don't think that will be an issue. I think we need to come to terms with the fact that over the past two years, partisan liberals really have been engaging in lawfare against a political opponent. Given that the primary motivation behind Trump's trials was to stop him from becoming president again, there will not be nearly enough political will to prosecute a two-term former president - who will likely still be popular with at least half the country - when he leaves office in January 2029.
He’s already been convicted of 34 felonies so there doesn’t need to be a political will to prosecute. They delayed the sentencing because he was running for president, and will delay again because he is going to be president. It wasn’t lawfare, it’s trying to hold everyone to the same standard. He’s a convicted felon and should be sentenced .
I agree that Trump would probably like to be on the ballot in 4 years, I just don't think he realistically could achieve that. They can get up to a lot with a majority in congress and a friendly SCOTUS, but they probably won't be able to amend the constitution.
I personally think the fear of Trump's corruption and envy of dictators is warranted, but I think the democratic backsliding will be a bit more subtle than Trump literally running for a third term.
Yeah, it's hard to see how he would manage to get into a position where he can actually run for another term. That said, given that SCOTUS has said anything a president does in their role as president is immune from criminal prosecution, he has little in terms of disincentives from trying to suspend the electoral process or otherwise remain in power beyond the end of his term out of some concocted "national security" or "election integrity" concerns. I realize a lot of what he says is just an old man who doesn't understand how our government operates rambling, but he's used phrases such as "suspend the constitution," so it feels like there is some degree of possibility of stuff like that going on (as to what percent chance? Maybe like 5-10% chance he tries and then some lesser percent that he's successful? 🤷🏻♂️)
To me the least likely outcome is that he serves a full four year term and then there is an uneventful 2028 contest and he goes into normal presidential retirement in the way Obama, Bush, and Clinton did (and that presumably Biden will).
You're assuming that Trump will still be President in 2028. I think that you can flip a coin on that. A lot of the GOP would love to see him gone, and will be watching for an opportunity to push him out the airlock. Then you can freak out about JD Vance.
Would they like him gone? Yes. But they didn't even have the guts to vote to impeach him in 2021, at which point they could have barred him from office. Mitt Romney said many GOP didn't vote guilty for fear of their own personal safety.
Would the House impeach him? 9 of 10 Republicans who voted to do so were driven from office by MAGA. Would the Senate this time vote for removal? That would require (in the first two years) that at least 19 GOP members voted for removal.
More of a concern than Trump running again is whether we can have free and fair elections in 2028 under an embolden Trump administration. I used to discount those who had this concern in prior election cycles. But after 2020 and later actions, it's a real concern now.
I think the motivation for him to retire would be (1) a clean pardon for all criminal acts that he's facing prosecution for. Sure, he can try to pardon himself, and that might ultimately be effective. But that wouldn't keep him out of court in the way a pardon from a President Vance would and (2) he's really pretty lazy, so if it gets difficult or if he feels like he's done all the things he wants to do maybe he's good with passing the baton to Vance and still being able to basically dictate the things he cares about under the threat of primarying Vance in 2028. I'm not sure if it's a coin flip, but the odds of him serving his full 4-year term have to be lower than the average president's would be
I suppose that enough Trump chaos might cause these gutless and corrupt Republicans to dump Trump to save themselves. It certainly won't have anything to do with virtue. That's a quality lacking in all political spectra and in a large portion of the population.
Their big fear is being primaried and MAGA has an iron grip on the GOP voters. Elected officials aren't turning against Trump; it's political suicide for them if they do.
There's probably some who privately would like him gone, but their numbers have been waning in the 8 years that he has completely dominated the party and more and more of them came in either riding his coattails or else just living in MAGA world for so long
Down ballot Republicans lost most of the swing state statewide races this year in an election where Trump was on the ballot against an incredibly unpopular incumbent administration amidst record inflation. The hardcore MAGA base doesn't care enough to show up for down ballot races, and the reluctant Trump voters who chose him over Harris because of inflation and the border aren't necessarily conservative, which is why Gallego, Slotkin, Whitmer, Ossoff, and Warnock are in the senate.
I think a big problem is that while the data was good, the analysis was terrible. And that occasionaly includes here. The google searches article was a classic example. Other forecasters (such as Split Ticket) talked about the horrible campaign Trump had run, when, lets be real, he ran a pretty effective one. The MSG rally and other things were just Harris supporters hoping against hope the narrative would change. Their big example of a terrible campaign? Focusing on people who just weren't going to vote... right up until they did.
Others are choices from a marketing perspective. Focusing more on the 50/50 nature of the forecast rather than understanding more the likeliest outcomes probably did good for subscriber numbers (and the liberal audience who wanted to believe it really was closer than it was).
I also think the problem wasn't so much with polls as it was the "gold standard" polls. Polling had a good year because of a lot of those polls which were routinely viewed as garbage or partisan hackery ended up being a lot closer than the gold standard polls. Emerson got lots of shit (including from Nate), but they performed a lot better than the gold standards. It really feels like the pollster ratings need a more dramatic revision than is probably called for, but most of the gold standard guys blew it even worse than 2020.
I actually think Silver Bulletin did better than most, but retros are good and healthy.
"The MSG rally and other things were just Harris supporters hoping against hope the narrative would change. "
The funny thing about the MSG rally thing is that I have since heard that Puerto Ricans are one of the groups that broke big for Trump, particularly in Florida. I guess they can take a joke. Don Rickles made a career on comedy that I guess would be beyond the pale now.
Nobody so far has been able to explain how that joke actually works as it doesn't have any thing humorous or witty about it. Don Rickles would say that there was no joke there. I don't think that the Island of Garbage would pass the three year old test and they will laugh at anything even close to poop. A bit like Vance and the Grilled Dogs in Springfield, saying that people couldn't take a joke. But if you find that funny then laugh away for everyone has a different standard when it come to humor.
You asked for someone to explain how the joke works. Someone explained it. This is all well-known if you follow Puerto Rican politics at all. Alas, the side that loudly proclaims how outward-looking and empathetic they are only ever looks inward and has no theory of mind for people who are different to themselves.
Nobody explained how the joke worked they just mentioned a category. Now that you have informed me that it was directed at a select sub group I am still trying to get someone to explain how it worked. I am familiar with the double entendre which is saying something while meaning another generally with a bawdy overtone but there was no double entendre the delivery was done as straight fact hence my questioning how the joke worked. The three year old who says poop understands that it is going to work with other three year olds but if you want to expand the positive reception you have to do a little more work. I have a great sense of humor but I rarely waste my time laughing at things which are just not funny. You must extrapolate on this theory of mind.
Thanks! Claim on wiki is that it is over 33% Puerto Rican, and other sites say the high concentration there goes back several decades.
It would be fascinating to know if this was a case of them not caring, or actively being irritated, for example at Democratic performative outrage while ignoring more important issues.
I don’t think saying Trump’s campaign was good is “being real.” It’s not clear it was good.
My guess, though I ofc don’t know for sure, is that he ran a pretty bad campaign and won because inflation mattered much more than campaign effects. A good campaign is worth maybe 2% of vote share, and I think 2024 tells us your opponent’s administration presiding over 20% higher prices is worth a fair bit more than that.
While I agree that there were pollsters that were massaging the results a bit for fear of blowback from a liberal media subscriber base, I don’t think it is true for the analysts. Somewhat different subscriber base looking for a different product.
Nate himself has said that liberal subscribers would get angry when he said something they didn't want to hear. I think he relishes that, but many analysts are more inclined to pander.
The Seltzer poll was actually very good the only problem is that those who were polled failed to hold their mud in the booth. The reasons for the numbers were quite clear for Blind Freddy, Soy Beans and Corn. The stockpiles are at high numbers and the price is at low numbers and when you start to impose tariffs there will not be many incentives for foreign customers to buy our beans n corn because other countries have stockpiles too. This will deliver a lot of good farm acreage to Corporate Ag. Have a look back and see how the farm sector weathered the last set of tariffs, they had to go on major farm welfare to the tune of billions of dollars. Seltzer got the numbers right but underestimated the cowardice of the respondents.
Rosenberg had the worst combination of being a transparent partisan hack, and actually having a fair bit of reach with the completely unreliable crap he was slinging each day. I would read his blog just to see how thick the walls were in the echo chamber.
Even now, he will make every excuse on earth for the results except that he was wrong - which is the truth. He should have no followers and no influence, and if he really cares about helping Democrats win in the future, he should shut up and sit back and let some rational humans influence the next race.
This is why, even though I didn't like it, I kept reading. It is why, being a Trump voter, I also read FDB, Klippenstein, and many others. I did not do that in 2020. I was in a bubble. Never again. You can only be fooled once. After that, you are choosing to be ignorant.
Well, that’s the kindest thing I’ve ever heard from a non-Trump voter. Have faith, my fellow citizen. You might be pleasantly surprised. If I’m wrong, I’ll do an apology tour and community service.
1) Matt Gaetz: I'm sorry, but isn't a "criminal" someone who's been convicted? The DOJ didn't even prosecute Gatz.
2) RFK Jr.: Don't lie. He and his family are vaccinated for smallpox, polio, and other diseases. Okay he's not too big on the mRNA vax. And for your sake, I hope you didn't get 3 more jabs of that particular vax.
3) Tulsi: You Fred should be shot. Tulsi is an American Hero, and if she is too peace-loving for you, then you should volunteer in Ukraine or Gaza.
RFKJr runs one of the leading antivax misinformation factories. His spreads all sorts of debunked antivaxx bull like the lie about autism. If he does accept a few vaccines, that doesn’t forgive all the lies he spreads about the others… far more than just Covid mRNA.
Tulsi actions as a veteran do not make up for her frequently parroting anti-American Russian propaganda. Whether she’s doing it knowingly or is just a fool, she assists the Kremlin in damaging our nation.
3. I'm not 100% sure on Gabbard, but I do think putting someone in that position who has never stepped foot in a War Zone is a horrible idea and I know she cares very much about the actual bodies used to fight
I see what you are getting at here, but I am fully aware of what I voted for and against. I am prepared to accept those consequences. Trump's first term was a complete shite-show. Obviously, I can't control what happens, but I am comfortable and informed about my choice.
Why would you vote for someone whose first term was a complete shite-show? Do you think being 4 years older (and 8 years older by the end of his term) would help?
Right, but how did they do in 2022? If they were great in one and mediocre in the other, that wouldn't justify considering them some sort of oracle. Nate's pollster ratings did have them as one of the top pollsters, but not far and away the best, which I think was reasonable.
People were really saying that AtlasIntel was wrong because they weren't capturing the "silent republican woman" vote that would secretly be cast for Harris.
I downloaded the last Atlas Intel swing state polls, just before the election, and they were fascinating: women made up between 51% and 55% of the polling sample in each state and Trump won at least 47% of the women vote in each.
If anything, Atlas Intel may have captured that many women were secretly voting for Trump, not Harris. The CNN exit poll showed Trump won married women by 2% (but lost single women by 21%).
This is what the actual analysts and number crunchers who don't waste a lot of time are more worried about.
Most of use view Trump as a unique factor, but he's either not going to be on the next election or he will be picking up the unique baggage of running on a questionably legal constitutional interpretation, so attempting to adjust polling for this election is mostly a waste given that the polls were 'okay'.
But believe it or not, those of us looking at Atlas Intel basically are looking at the numbers and seeing herding at the end towards a expected average, which means that they *guessed right* in which polling to publish, but it was still a guess.
What if next time, they guess wrong?
What if they guess wrong with my Clients polling in other areas?
For those who don't see the danger, let me sum it up like this.
The last person who we had this concern with was Selzer.
Her 'Intuition' had been really good for a really long time, until it wasn't.
And I can't have a client get burned that badly and expect not to get hell for suggesting a pollster in that case.
I've lived long enough that I've seen a number of populous figures try a number of questionable or outright illegal tactics to stay in the ring. The weaker the party is without a individual, the more creative they get.
So I'll hedge my bets, even though I don't think Trump has enough strength in his years or in his administration to pull anything beyond this term as president (and I have my doubts that his heart will hold up under the stress he signed up for.)
The theory that I heard from a pollster is that about 3% of voters are low-trust and love Trump. They will refuse to answer when the pollster calls, but will reliably show up for Trump. That explains the bias in all 3 cases.
My brother's version is that they aren't necessarily low-trust, they are just rural and not always available by phone.
Regardless, a small and difficult to poll group with strong opinions would perfectly explain the result. No matter what the explanation. Based on that, both my brother and I arrived at the actual result by modifying the output of your model in the same basic way. And if that is the cause, then you should be able to find the size of that group by comparing polls with past elections (including primaries), and therefore fix your model to be robust to this source of error.
Though, honestly, Donald Trump is the only politician that I'm aware of who seems to show such an effect in practice. So it might not come up again for a long time. It's not that other politicians don't try to reach that group. It's that they don't realize, for example, the importance of simplifying your language down to emotional terms that that group seems to respond to.
You are peddling hopium. Nearly every county in the US swung towards Trump. Including Republican strongholds like NYC (irony intended). The “only rural hicks” narrative is BS but hopefully liberals will continue to believe it.
okay, someone’s grumpy! how about you drink a juice box and calm down, and talk to us when you’re ready to have a reasonable response to why you think he’s wrong.
I am sympathetic to Ben's view, and I don't even disagree with most of what he said (I liked his post). But you asked for a reasonable counter-point, so I thought I'd defend Gary a bit:
(1) This is the same story most have told about the polling error since 2016. It's certainly a plausible story. But if it were truly the reason polls were missing, wouldn't they have figured out some solution by now (8 years later)?
(2) It seems to confirm the priors of democrats too squarely to be credible. "Why did Trump underperform the polls?" "Well, his voters can't figure out phones and only understand simple talk." (This part is, obviously, an extremely unfair version of Ben's point, and I don't argue that that was his contention or his view. But it is how this sort of argument comes off to many people.)
(3) The theory assumes that this group of hard-to-reach voters with extremely similar worldviews were, in prior elections, either (i) not voting at all or (ii) voting in divergent ways from one another. That could be true, but this would require some further explanation.
None of this entirely disproves the theory Ben proposes, and I don't really have a strong view on it. But I do love a good juice box, and I hate to hear juice-lovers insulted.
Eh, that would require reading, and that's hard these days and doesn't give the same endorphin levels as just stating what they believe and exiting stage left and declaring it a win.
So, as a analyst, allow me to say that while I think you're on the right track, as the final numbers come in that 3% shifts to about 1.7%, which is big but not nearly as ground breaking.
I'd also argue that Rural is less the defining factor, as smartphone and app interaction is the course de'jour these days (and yes, those of us who live out in the 'wild lands, beyond the reach of city lights', do have and use cell phones) and a lot of money since Trumps first election has been in that direction to reach them, whereas in this election it appears that a general lack of enthusiasm amongst traditional voters led to a measurable loss from those quarters as trump pulled in a larger group non traditional voters this election, aka 'People who don't normally vote came out and voted the Trump / Anyone but BidenHarris.'
The 'Voting for Trump' factor is really, REALLY hard to catch without oversampling, and because they are non traditional and don't show up for midterms or any other non-Trump election *so far*, the industry is stuck in a unwinnable situation:
Over adjust to Trump ruins all other polling efforts ('The Red Wave' talk was a result of such a case), OR Do nothing and get hammered whenever Trump ends up being a force in the vote.
For what it's worth, the actual results of the polls were fine.
They missed in the same direction as last time, but by less. I'm actually concerned that without Trump on the docket for the next few elections we're going to see the rubber band snap back in the opposite direction, ESPECIALLY if there is a popular figure who comes out of nowhere on the Democratic side and there is continued herding, and that people who don't understand polling will take it as a sign that 'polling is broke' when in fact there has just been such a heavy attempt to capture the Trump voters that it is making the mistake of assuming that Trump voters are the same as reliable voters.
I wish I cared more about polls, but I don't, except as barometers of the electorate, which, hate me now, I consider largely politically hazy, uninformed, and having an instinct for shooting itself in the head. This is no doubt informed by 50+ years of voting and watching people make impulse decisions over careful research.
We spend more time on our wardrobes than we do vetting political candidates.
So I'm not surprised at the result.
Kitchen table issues, like the cost of food, mean more to people than rhetoric, and certainly more than polls. Inflation was, and still is, a big issue. But people had no knowledge regarding the causes of this latest round. They saw it as a Biden failing, when it was a global problem, caused by a confluence of supply chain chaos meeting surging demand as things opened up in 2021. Add to that a dash of greedflation, as noted by the Kansas City branch of the Fed, and way too liberal credit policies, which make a lot of money for banks, while keeping people in a hole of their own digging.
Democrats did a terrible job of messaging on inflation. It was almost as if they thought that by ignoring it they would do better. Oops! Then again, Democrats do a terrible job of messaging anyway.
And letting Biden run again was a guarantee of a loss. After the pullout from Afghanistan, which didn't resemble the fairy tale version that magical thinking Americans expected, Biden's approval ratings plummeted, and never recovered.
And though a person's character matters to some of us, it doesn't matter to most of us, at least not as much as the cost of a carton of eggs. Trump is our first Presidential felon, maybe our first Presidential rapist. He ran a particularly disgusting campaign, which might have cost him with a stronger opponent.
Thrusting Harris into the role of carrying forth the banner, partly necessitated by Biden running out the clock, proved to be a poor choice. She didn't differentiate herself from Biden in any detailed way, relied on sloganeering to replace a cohesive set of policy statements, and proved to be too cautious. Yes, she wiped the floor with Trump at their debate, and he was smart enough not to give her another chance, just as he was smart enough to avoid the primary debates. So, Harris lost whatever momentum she had because she lacked to tools to capitalize on it.
Given the freak parade that Trump is proposing to head government departments, 2026 could be an interesting election, provided that Democrats stop feeling sorry for themselves and focus on those kitchen table issues. Trump is going to give them lots of opportunities. Get the sound bites ready.
My problem with the "polls are fine" is not that errors of 2-4% are either unexpected or bad, but that for the third presidential in a row it is going the same way, indicating a systemic Dem bias.
I myself have recognized that and gave a prediction of my own in my FB group Political Lane of a GOP trifecta, with 52 Senators and +8 in the HOR, and close to nailed it. My methodology was to assume some Dem bias of somewhat less than 2020 as some pollsters likely adjusted but since the real problem has not been solved, it would still be there.
I continue to use the RCP aggregates as my benchmark because in the three elections they have had less of an R underestimation, even as they lack the methodological sophistication of Nate. The excuse of looking at 2022 or 2012 does not seem valid to me as mid terms are a different critter, and 2012 was a different universe where low propensity voters and reluctant poll responders were just as likely to be Democratic, but no longer.
In Nate's analysis on the morning of election day, he said the race was essentially a tie. In each of the last 3 elections Trump has outperformed the polls; granted, not by a huge amount in any of them, but when it happens 3 times in a row one begins to wonder. I don't think the pollsters are deliberately screwing it up, but they may be underestimating the number of Trump voters who refuse to talk to them.
Yes, SB final forecast was the probability of Harris winning at 50.0 and Trump winning was 49.6. I guess the other 0.4 was a tie.
But that probability was based on his poll aggregations which for the national was off 2.5 and for the 7 swing states averaged off 2.13. That is pretty good for accuracy but every poll was off towards the Dems which hints at a systemic bias.
Now RealClearPolitics a site I otherwise dont like for its right politics also underestimated Trump in all three elections but to a lesser degree and where now in all three elections Nate has gotten 9 states wrong and all have been states where his poll aggregations had the Ds ahead and went R, RCP has had 7 wrong and 6 where the Ds ahead went R and 1 the other way (called GA red in 2020). This years national polling at RCP had Harris at +0.1 where as SB had it at +1.0 and the different swing states were about that much tilted towards the Dems.
Now as far as I know RCP just averages the recent polls, where SB does various juitso on it, which analytically makes sense, and Nate is smarter than me, but, regardless, has had the effect of skewing it more to the Dems than a straight average.
Janice , the Dem bias is that in the last three presidentials the polling aggregations of Nate, prior 538 now here, were consistently towards the Dems. They have led to 7 miscalled states in the three and every one was a state that the polls predicted D and the vote was R. An error of 2% I agree is far from bad but if it only goes consistently one way, that is problematic
I agree that it seems to make sense to use the RCP average as a benchmark. The only question I have is how do they define the line for inclusion in their average, so that polls in the average have weight 1.0 and polls outside the average have weights 0.0. Otherwise it is very transparent and looks like the SP500 average which stock pickers have a hard time beating. (Okay Nate would point out SP500 itself is weighted…)
My question: perhaps this is something that can’t be answered from the data, but indications were that Trump was gaining over the final month or so. Was there a meaningful movement in his favor (rather than random fluctuations)? And if so, were there any events or circumstances that could explain it?
I think the VP debate had a negligible effect on the race, but it coincided with the beginning of Kamala’s decline in Nate’s forecast, and I remember 2 key moments from the debate - Tim Walz bragging about Dick Cheney’s endorsement and the online left reacting exceptionally negatively to it - and JD Vance effectively being “normal” and giving a reasonable looking image to his far-right policy agenda.
Ron Bauer this is an excellent question and maybe something Nate could discuss in a paragraph or section in one of the posts, or in the introduction to the 2026 or 2028 models. We know the final result and 2024 and the polling results of all these pollsters during the campaign, both before and after Biden drops out. In addition to the Trump vs. Harris polls which is the focus of Nate’s work (and RCP poll average, to take a more transparent baseline), there are the betting markets and also the opinion polls on topics like what matters to you more inflation or abortion (etc. ). And they did exit polls on Election Day (don’t know if they do exit polls on early voters or mail-in voters?). I was going to assert that Trump was always in the lead because opinion polls showed people thought the US was on the wrong track, inflation, border. It might be that careful exit polls might answer if that was true. One angle Nate might know is what % of people have made up their mind in the “early” polling, and in the final Nov election what percent was “likely voters” and what percent were “unlikely” voters and what was the Harris/Trump split in each bucket.
The odds of three heads in a row is 1/2^3 = 12.5%. Likewise the odds of three tails in a row is 1/2^3, or 12.5%. The odds that either one of these happens is the sum of the two, since there is no overlap between these outcomes.
Yup if your referring to it to either candidate before it has happened once. But what if you are talking about it happening to Trump twice after it has already happened once. Or happening to Trump a third time after it has already happened twice? That final analogy is reality.
Something has been bothering me about election polling commentary like this. Maybe someone can correct my misunderstanding.
"The polling averages were right in 48 of 50 states." Fair enough. But there was no reasonable doubt about many of those states. Nearly all election coverage, campaign attention, etc. was focused on about 7 (not 50) "swing states." Would a more accurate gauge of accuracy be 5/7? (Still pretty good.) Or, being generous, there are about 15 states that have "swung" in the past 25 years (see 270's 'same since' maps). Even being extremely generous, and using states that have "swung" in my lifetime (somewhere in the range of 30-40 years), the denominator would be less than 25.
In any event, that's not my gripe. A consistently-applied 50-state denominator as the gauge of polling prediction accuracy would be fine--even if it inflates accuracy, the year-by-year comparisons of accuracy would at least be consistent.
My gripe is mostly that every time the issue is raised of directional misses--i.e., the polling bias favoring D in every presidential election in the trump era--I am told that the REAL denominator for the fraction is ONE, not 50. E.g., "It's just three elections, and heads three times in a row does not tell us much." Fine, again, in a vacuum. But what happened to 50? If we use 50 states as the sample size, the data would tell a pretty significant (and convincing) story of systemic polling/aggregator bias. In 2016, 2020, and 2024, the D lean was present in almost every single state. (I haven't checked in a couple days, but I'm not sure the polls missed TOWARD trump in a single state this year.)
Perhaps there is an explanation why 50 is the correct denominator for determining whether polls got the answer right, but not for determining whether polls systemically miss one direction or the other. But I haven't heard it, and it makes things like this difficult to read.
Polling errors are usually systemic, so if they were off by 2 or 3 points in all 50 states, it wouldn't be that surprising. Does that help address your question?
Also, the state-by-state directional results certainly have a systemic connection, but I'm not sure it's as direct as the last few elections would suggest. In 2012, for example (the last year polls generally leaned toward R(omney)), Obama still underperformed the polling averages in more than 1/3 of the states.
Not really. But I'm not a stats guy, so maybe I'm missing it. The logic seems to be that: if polls miss left/right in AK, then they will probably miss left/right in WV as well. And that seems probably true enough. But wouldn't the necessary flip-side be: if polls get the answer correct in AK, then they'll probably do the same in WV? In other words, a systemic relationship between the state-level polls doesn't address why state-level polls should be the metric for some accuracy measurements (e.g., binary accuracy) but not others (e.g., directional miss).
Where are you seeing that state polls are the metric for only some accuracy measurements? I think a person could look at either state or national polls (or both), but that they should be consistent in which polls they look at.
My main takeaway from this article was that Nate cares A LOT more about progressive viewpoints than I do. Uncertain whether that's because he cares too much, or I care too little.
I believe it’s because you care too little. Nate, for being the relatively centrist liberal he is, spends a great deal of time consuming the ideology and logic of both extremes in order to further develop and strengthen his own personal worldview to be able to cut through the emotionally-driven narrative that exists on the insurgent right and “wokeist” left
As a matter of situational awareness, it's worth being /aware/ of what's up on the progressive left; but being aware, I currently struggle to care. They've been playing defense for nearly a decade, and since 2016 most of their intellectual achievements have been some variation of denying America's political realignment. It's not at all clear they'll have much claim to being the left's thought leaders after this year.
The quotes in Nate's article are a case in point - judging purely on the merits, nothing they had to say warranted acknowledgement from the likes of Nate Silver. Yet he seems to care more than I do. It's possible I'm too dismissive of this once-influential group.
Tbh, I think a lot of Nate's substack-era punditry is driven by his personal experiences. He is an upper class gay man who lives in NYC and works in media. His social circle is likely very exposed to liberal/progressive viewpoints, and he has a sort of inherent contrarian personality, so he has a lot more opportunity to get rubbed the wrong way by it. I'm sure if he lived in a small rural town in West Texas and was primarily exposed to the viewpoints of his community Facebook page, the set of viewpoints that were constantly bugging him would be quite different.
I also don't think people understand polls and how they work and how easy it is for them to be off. People are people. Sometimes they lie, sometimes they don't understand the question as asked, sometimes there are no good answers or not enough expansion is done on a question. I know this because I'm a regular poll taker. I also send in feedback on questions. Example: Many polls this year asked about "protecting democracy" but they didn't ask what people meant by that. Many people were thinking about free speech, lawfare, etc. Others were thinking about institutions. See?
I am not sure how you can lead with "She's good at this" in that post and then write an entire post that Harris was a replacement level candidate. To Nate's credit, only a handful of posts the entire cycle came across as fawning (any positive reference to coconuts comes to mind), but it is something I hope the Silver Bulletin team (which hopefully grows) does an honest internal look at.
FInally, Nate should take a victory lap on calling out some of the late campaign narratives on polling - margin of error, typical accuracy issues, and the 'flooding the zone' narrative that inexplicably caught on. I would love to see a response/follow-up to Taibbi's write-up last week:
I like the optimism that Trump won’t be on the ballot in 4 years.
The more "you people" think and act on beliefs like this, the more normal America will turn their backs on you. There are lots of normal Americans, many highly educated and successful Americans, who truly believe the people pandering this fascist nonsense have truly either lost their minds or been brainwashed by leftist media. How many of you are still wearing covid masks in public?
Well, I'm 72, and Covid doesn't like old people. So, I wear a mask when I'm in the supermarket, or in the audience at a performance. The one time I forgot to wear a mask at a performance, I caught it. Fortunately it was a very mild case. But I have no interest in catching it again, especially as repeated exposures increase one's risk of getting long Covid. I have better things to do with what's left of my life than lying around, trying to breathe.
Well, I am 73 and we had largely ditched masks in Texas (ok maybe not Austin) six months into the pandemic. Our grandchildren did not go to school wearing them - you do know they won't stop a virus don't you? My physician had great advice. If one is worried about catching Covid - don't get tested. Then all one has to worry about is a cold.
What I do know is that N 95s are effective at stopping viruses, due to their electro static charge. Nothing is 100%, but I’ll take whatever measures improve the odds. YMMV.
Masks were never touted as a panacea. Medical grade masks do and did, in fact, reduce the risk of the wearer infecting others. That is why surgeons and ICU nurses wear them. The coronavirus SARS-Co-2 and its mutations are borne on the droplets broadcast as much as 6 feet away in a conical pattern by normal human respiration--coughing, sneezing, exhaling. Droplets are caught in the mask or muted at the source before it dissipates. There was a dramatic reduction in other airborne viral maladies --colds, flu, and possibly pityriasis rosea--while people were wearing masks and isolating during COVID. They have now returned to pre-COVID levels. Ideology is no substitute for public health.
Belief in ideologies is due to a lack of critical thinking.
The masking thing is always going to repel people for any number of reasons that outweigh exposing one's self and others to pathogens. It's interesting because in Asia, masking is taken as a matter of course when needed.
I kind knew that we were in for something bad when many of the Asian members of our crew and on other shows starting wearing masks to work in late 2019.
It is also a matter of airflow dynamics and how particles move in the airflow of a blown N95 mask a bit like the way a catch boom works to pick up trash in a river.
I cannot think of anything less interesting to argue about than whether someone else is wearing a mask. I thought Republicans were the ones who say people should be free to do as they please.
Your physician doesn't happen to make money from treating sick people does he. I knew three people, two elderly parents and their son who all caught that cold and the paramedics had to wear masks with camphor to cover the smell of decomposition when they came to their house. I would like to know where you got the would not stop a virus fact from. Corn Flakes are good for breakfast but you should not read the box.
Paging the DHOTY Awards!
I’m 72 and sympathize. In 2022 I attended a small applied math conference at Oxford and modeling disease and analyzing data was a now favored topic. The data/model presented on masks is that it would reduce the spread by 50%. I’m interested in further benchmarking. Data/model showed, not talking will reduce the spread by 10x (that is, instead of down to 0.50x it is down to 0.10x). So wearing a mask where everyone is talking is much worse than minimizing time in a chat zone. And yes I although I was vaccinated I got COVID traveling via Heathrow on the way back like your performance.
Why? We literally know it doesn't work. Like, you seriously get no benefits from it.
You’ll catch it again. Your better bet is catching it early via later flow testing and then popping antivirals quick (highly effective in initial infection but quickly wane by symptomatic presentation).
But I am curious to read the long Covid data. What most people don’t realize is that the phenomena isn’t new to Covid. What did everyone think post viral tussis (cough) was?
In 2020/2021 Donald Trump lost a fairly run election, subsequently promoted conspiracy theories while continuing (to this day) to state that he actually won this election, encouraged protestors to storm the capitol, and pressured Pence (and everybody else he could think of) to help overturn the results.
Given this behavior, do you *really* think it's "nonsense" to question whether such a person would intend to relinquish power in the future?
Maybe "don't vote for a person who does this stuff" is really a losing message, and America seemed to turn its back on it this time as you suggest. But the concern is very much not unwarranted.
Really the thing to say here if you really want to help people believe Trump will not extend his presidency beyond this term is "I'm a Republican, and I voted for Trump and I think he's going to be a great president for four years. However, if he seeks the GOP nomination in 2028 I will vote against him and if he is somehow on the general election ballot again as the Republican nominee in 2028 I will vote for whoever the Democrats nominate."
I will say that it's hard for me to game out how he could possibly extend his term beyond these 4 years, but it doesn't seem all that unlikely that he will attempt to do so (just as it didn't seem unlikely that he would run for president in 2024 -- once a person has done a thing, it's typically plausible that they'll do that same thing again). Nobody needs to really agree on whether he will try to remain in power, though. We just need to agree that we will all do our part to stop him if he does try.
Everyone I've seen on the right is very quick to disparage the idea that Trump will try to stay president beyond 2028, but I've yet to see anyone who enthusiastically voted for him already say that there's no possible way they or anyone else would vote for him if he ran again. If he does give it another try, his voters are going to get behind it.
How many have you seen? I see almost no one wearing masks outside of hospitals atp
They’re still everywhere in Westchester. I still see them driving cars with masks on. Alone and with the windows up.
Which is weird because I thought those people loved the smell of their own farts.
Yea this year I’ve lived in Pennsylvania and California, both city and rural. I haven’t seen almost anyone wearing a mask outside hospitals. I feel like this is just propaganda you’re spreading.
You should come enjoy the spectacle at a Whole Foods in Austin.
I was at a grocery store in Austin 18 months ago (granted not a Whole Foods, because I don’t like to burn money), and I don’t think I saw anyone wearing a mask. Or at least not a notable percentage.
I also live in blue NJ in a relatively liberal area and only see the occasional person who may be recovering or immune suppressed.
So I very much doubt your story is representative of anything but conservative media exaggerations.
I was at an urban running store the other day in a gentrified area and about half the people had them.
I have lived in two heavily liberal areas and I find this so hard to believe. 50% is absurd man. I don’t even see 1% masked around me.
I live Oakland, CA. A reliable 5% still mask everywhere, especially outdoors. Two children in my kid’s first grade class are still masked daily by their parents. And, no neither these kids nor their families have any comorbidities which would recommend around-the-clock infection prophylaxis.
wow you even know about their medical histories and the medical histories of their families!
Why do you repubs have to make it so obvious you're lying?
How is it even possible for you to know that? Do you just assume no one has comorbities
This is the same nonsense progressives peddle seemingly every republican administration. It's almost like they wish they were living under Franco.
https://www.thenation.com/article/archive/bush-postpones-2008-election/
That said, I'm aware of one attempt by a governor to unilaterally cancel an election. It was in 2020 and the courts quickly put an end to that obviously bad idea.
https://www.washingtonpost.com/politics/wisconsin-governor-suspends-in-person-voting-in-tuesdays-elections-amid-escalating-coronavirus-fears/2020/04/06/9d658e2a-781c-11ea-b6ff-597f170df8f8_story.html
Saying all progressives is admittedly too harsh. But this a recurring refrain elements of the left put forward every gop admin, without any evidence.
The repetition of such baseless fears against all gop politicians also just diminishes the charge. Kinda like how Reagan, Bush I, Bush II, McCain, Romney, and Trump were all compared to nazis at some point.
This is just completely made up. There was absolutely no widespread fears that Reagan, Bush, Bush, McCain, or Romney were threats to democracy or that they would attempt to subvert the results of their elections. And then, in the case of Bush I, McCain, and Romney, none of them did challenge the election results and Romney is among the people who condemn Trump. You cannot be serious here.
Re: Romney & McCain specifically, yes, that's the point. You don't have to look hard to find examples of democratic campaigns calling comparing them to nazis and Hitler.
https://www.google.com/amp/s/thehill.com/blogs/blog-briefing-room/news/124572-romney-campaign-tells-obama-to-rein-in-his-supporters-on-nazi-comments/amp/
And, yes, in 2004 and 2008 elements of the left circulated conspiracy theories Bush II was going to cancel the elections. The nation article in my original post was an example from 2008, 2004 leaned more into Bush would use a Madrid bus style terror attack to permanently cancel all elections.
And I honestly cannot tell if you are serious because there are literally books written about why every republican is a threat to democracy. That's the point that this line of argument isn't persuasive against Trump because it's used against every republican.
https://press.georgetown.edu/Book/Reaganism-and-the-Death-of-Representative-Democracy
I can’t tell if you’re serious because nobody in the mainstream left was saying these things about previous Republicans and now Mitt Romney is saying it about Trump. This is such a preposterous false equivalency
I mean, we literally had a Democrat in PA *admit* on camera to election fraud in Bucks County.
You can't talk your way out of that one. She should rot in jail.
This person said on camera "I committed election fraud?" Link? The only person I'm aware of who was convicted of election fraud in recent memory was McCrae Dowless in North Carolina (a Republican).
No evidence? Trump literally refused to accept the 2020 election results and incited a deadly insurrection to try and halt certification. He then peddled his demonstrably false election conspiracy theories for months afterwards, trying every avenue he could to steal the election himself. He's repeatedly talked about wanting to be President beyond his 2 terms, he praised Xi Jinping for becoming "President for Life" and suggested they should do the same in the US.
But we're all supposed to ignore that because he hasn't succeeded "yet". We have to wait until it's too late before we can point out the obvious.
An ex-sabermetrician like Nate would probably call this the Designated Hitler Rule.
Exactly. The problem for them is there is ever a wolf (for which I’m confident most would claim Trump is) then no one will take them seriously.
If everyone is a nazi or commie or whatever the nom de jour is, then it has absolutely no rhetorical value and is arguably counterproductive. Telling a family paying 25% more for their groceries to look at Trump as a rapist/indicted felon/fascist/etc only makes them tune them out. Had they stolen a page from the blame China handbook, they maybe could have allayed inflation sentiments. Then again, calling it transitory and saying don’t worry about it were woefully naïve in electoral presentation.
If I were a betting person, I'd be willing to place a bet that he won't be. Aside from reasons why any other person wouldn't be on the ballot after 2 terms as president, he's also 78. A lot of men die before they get to 82.
Zero chance he’s on the ballot in 2028. In order to do so, he’d have to pass an amendment, which would have the side effect of allowing Obama to run against him.
He didn't try to amend the constitution in 2020 to stay in power, he just came up with some legal hocus pocus to justify it and tried to force his way back. If he wants to run in 2028, he won't amend anything, he'll just have his lawyers cook up some reason the existing amendment doesn't apply, and then date anyone to stop him.
It’s not the same.
I wouldn't put pass the current SCOTUS some creative reading where consecutive terms disqualify you but non-consecutive terms don't :)
Read the amendment, it’s very short and very clear. We would need another amendment to allow Trump to stay on.
We seem to be forgetting that Trump already was disqualified under another amendment. The 2-term limit is one of the few guardrails I'm confident will hold, but I'm not 100% confident.
Current scotus has no particular love or fear of Trump
Obama would lose.
Not a chance
Obama couldn't even persuade the brothers to vote for Kamala.
He is a relic of an earlier time.
You’re looking at this backwards. Kamala was so unappealing to Black men that even Obama couldn’t persuade them
Yeah, I hear you. But Hugo Chavez died and Venezuela isn’t exactly a thriving democracy under Maduro. Dictatorships start with strong personalities but end with systems.
Odds are for him making it past 82 since he's already 78 and that's less than half average life expectancy for a 78 year old.
For a 78-year-old man? I've known plenty of women who made it that long but 0 men.
You must know some unhealthy 78 year olds.
https://www.ssa.gov/oact/STATS/table4c6.html
9.5 years expected from here.
I think Trump's age will be the biggest impediment to seeking a third term. Comparing 2020 Biden to 2024 Biden there was a pretty big deteriorating effect from 4 years, and Trump is already looking pretty old. Even if he's not dead he might just not have the energy for it in 2028.
About 23% of 78-year-old men fail to make it to 82, according to the social security administration's actuarial life tables.
Trump running for a third time takes one of two things:
- A constitutional amendment
- The assent of the military
No one can get constitutional amendments passed even for much more broadly acceptable things. There is at most a 1-2% chance he could get it done even with major successes in the states and Congress.
And while Trump's fairly popular among the rank and file in the Army, the officer class does not like him. It does not like him at all. Military officers are classic Romney-Clinton voters, and eight more years hasn't changed that. And though many of them are still Republicans, they take their oaths seriously, and "I will support and defend the Constitution of the United States against all enemies, foreign and domestic; that I will bear true faith and allegiance to the same" comes before "obey the orders of the President" for a reason, a reason they all know.
Things might get worse. That's a real possibility. Trump running for a third term is much, much less likely than that.
Trump's lack of military officer support scuppered his plans in 2020, but he knows that as well as anyone. It's little wonder though that he's placing much higher value on loyalty with appointments this time and trying to purge the unbelievers. Maybe he doesn't succeed completely but it's comforting to get to a scenario of having to rely on unelected military leaders to force out the civilian president and his cabinet
The military will follow whoever signs their paychecks, they're not gonna take a principled stand on anything.
People don't join the officer corps and stay there for a paycheck.
Great take on this. On your second point, the recent Hagel Op-Ed in NYT was quite sobering. https://www.nytimes.com/2024/11/15/opinion/trump-military-politics.html
That is an excellent example of a way things could plausibly get much worse while still falling well short of enabling Trump to run for a third term.
The independent apolitical military will not fall apart in four years. A 'warrior board' would do enormous damage and might make it possible for three terms of *Vance*, but not three of Trump; it couldn't work fast enough.
It’s kinda blowing my mind that Nate, Pfeiffer, and pretty much every other election analyst out there is acting like the party who was openly running on “Facism for everyone! Wheee!” Is going to be super cool with elections as usual.
It's the Democrats who cried wolf effect. Democrats had an absolute cow between 2016 and 2020, the sky didn't fall in, and it seems like the voters preferred 2016-2020 to what the Democrats offered, unfortunately.
A million dead people may dispute your analysis and there are a lot of republican members who are calling hands off on Biden's policies because they know how their districts benefited.
A million dead people as compared to what number? Trump handled
COVID like a moron, but I totally disagree with the implication that number would have been lower under say a President Clinton. What would have been substantially different in people’s reaction to COVID if she were President? Would people who were flouting COVID restrictions in Spring 2020 have been double masking and Tweeting #flattenthecurve under Clinton? I don’t think so.
Trump knew 3 weeks before the first official (then) death that Covid was airborne and much more deadly than the flu.
He did nothing.
Reacting faster would have at least slowed the initial exponential slope.
Look at the fatality rates in other countries that were more aggressive.
The sky did fall in-- the judicial branch is dead. It is a delayed effect, but there have already been plenty of nakedly partisan rulings even outside SCOTUS which is completely compromised.
>the judicial branch is dead.
Yeah, if fat Elena Kagan can't make all the rules in society, the judicial branch is dead. I totally get it.
Do you people hear yourselves?
I think they feel that a fascist dictator has never taken over the government before so we will be fine.
Of course before 2009 we never had a black president before. Or every other thing that has happened for the first time.
I agree; Trump is definitely a threat to our institutions. But those institutions have been around for a very long time, and it's not going to be so easy for him to subvert them. This isn't like Venezuela or Germany in the thirties; we've had our Constitution for 200 years, and even the justices on the Supreme Court, some of whose decisions I've strongly disagreed with, are committed to the rule of law (with maybe 1 or 2 exceptions). The next 4 years are certainly going to be interesting, but I still give us about a 80-90% chance of surviving with our democracy more or less intact. The likelihood of it not surviving is much higher than it's been with any other President, and I don't like it one bit; nevertheless, it's too soon to assume the worst.
That’s a totally valid worldview, I guess I just think it’s odd that Nate, who’s long been an advocate for probabilistic thinking (and that proselytizing has had a pretty big impact on me personally) is kinda writing off a pretty obvious set of worst case scenarios that one would normally price into their risk assessment.
If he didn’t want to be too alarmist, for example, a simple caveat of “assuming elections proceed normally” would probably be a sufficient disclaimer that addresses the legitimate 10-20% risk you’re describing here.
To me this probability estimate seems too optimistic for two reasons:
1. For most of Trump's first term, he didn't know how to work around the limited guardrails; now he does after his 1/6 trial run and armed with the Agenda 47/Project 2025 playbooks. Plus many of the authors are now announced as part of the incoming administration.
2. The Supreme Court has immunized him from prosecution if there's a pretense that his actions were official acts.
It’s one of the ways I’m morbidly grateful for having holocaust survivors in my family tree (grandmother). It really gets you to internalize the capacity for countries to become the worst possible version of themselves very, very quickly.
And that because Trump left power the previous time, it will just happen again. To the hell with the actual way on which he left office.
I mean, Democrats were running on s3xualizing children. What do you want from people?
Just because Daniel Liberson likes it doesn't mean the world does.
It's due to comments like this that Trump won and in an electoral landslide increasing his margins across all demographics. If progressives keep this up 2026 may actually result in further losses
This was not a landslide, regardless of how many times you (and pretty much nobody else) say it was.
Agreed. It wasn't nearly as big as Obama's 2008 win, and I'd say that's borderline for being a landslide.
I don't know how serious you are, but to the extent you are serious it's part of a broader problem of liberals continuing to treat Trump as an existential danger to democracy when the majority of voters, looking back over his publicly available record of governance, simply don't see it.
Also, not only is Trump constitutionally barred from seeking another term, he wouldn't want to even if he could. His ego has been sated, and he's lazy.
He still wants to avoid jail though, and he’ll be sentenced when he leaves the presidency.
No, I don't think that will be an issue. I think we need to come to terms with the fact that over the past two years, partisan liberals really have been engaging in lawfare against a political opponent. Given that the primary motivation behind Trump's trials was to stop him from becoming president again, there will not be nearly enough political will to prosecute a two-term former president - who will likely still be popular with at least half the country - when he leaves office in January 2029.
He’s already been convicted of 34 felonies so there doesn’t need to be a political will to prosecute. They delayed the sentencing because he was running for president, and will delay again because he is going to be president. It wasn’t lawfare, it’s trying to hold everyone to the same standard. He’s a convicted felon and should be sentenced .
I agree that Trump would probably like to be on the ballot in 4 years, I just don't think he realistically could achieve that. They can get up to a lot with a majority in congress and a friendly SCOTUS, but they probably won't be able to amend the constitution.
I personally think the fear of Trump's corruption and envy of dictators is warranted, but I think the democratic backsliding will be a bit more subtle than Trump literally running for a third term.
Yeah, it's hard to see how he would manage to get into a position where he can actually run for another term. That said, given that SCOTUS has said anything a president does in their role as president is immune from criminal prosecution, he has little in terms of disincentives from trying to suspend the electoral process or otherwise remain in power beyond the end of his term out of some concocted "national security" or "election integrity" concerns. I realize a lot of what he says is just an old man who doesn't understand how our government operates rambling, but he's used phrases such as "suspend the constitution," so it feels like there is some degree of possibility of stuff like that going on (as to what percent chance? Maybe like 5-10% chance he tries and then some lesser percent that he's successful? 🤷🏻♂️)
To me the least likely outcome is that he serves a full four year term and then there is an uneventful 2028 contest and he goes into normal presidential retirement in the way Obama, Bush, and Clinton did (and that presumably Biden will).
You're assuming that Trump will still be President in 2028. I think that you can flip a coin on that. A lot of the GOP would love to see him gone, and will be watching for an opportunity to push him out the airlock. Then you can freak out about JD Vance.
Would they like him gone? Yes. But they didn't even have the guts to vote to impeach him in 2021, at which point they could have barred him from office. Mitt Romney said many GOP didn't vote guilty for fear of their own personal safety.
Would the House impeach him? 9 of 10 Republicans who voted to do so were driven from office by MAGA. Would the Senate this time vote for removal? That would require (in the first two years) that at least 19 GOP members voted for removal.
More of a concern than Trump running again is whether we can have free and fair elections in 2028 under an embolden Trump administration. I used to discount those who had this concern in prior election cycles. But after 2020 and later actions, it's a real concern now.
I think the motivation for him to retire would be (1) a clean pardon for all criminal acts that he's facing prosecution for. Sure, he can try to pardon himself, and that might ultimately be effective. But that wouldn't keep him out of court in the way a pardon from a President Vance would and (2) he's really pretty lazy, so if it gets difficult or if he feels like he's done all the things he wants to do maybe he's good with passing the baton to Vance and still being able to basically dictate the things he cares about under the threat of primarying Vance in 2028. I'm not sure if it's a coin flip, but the odds of him serving his full 4-year term have to be lower than the average president's would be
I suppose that enough Trump chaos might cause these gutless and corrupt Republicans to dump Trump to save themselves. It certainly won't have anything to do with virtue. That's a quality lacking in all political spectra and in a large portion of the population.
Their big fear is being primaried and MAGA has an iron grip on the GOP voters. Elected officials aren't turning against Trump; it's political suicide for them if they do.
There's probably some who privately would like him gone, but their numbers have been waning in the 8 years that he has completely dominated the party and more and more of them came in either riding his coattails or else just living in MAGA world for so long
I have a feeling 2026 will be landslide for Republicans and then it will be fun to see the reactions
Down ballot Republicans lost most of the swing state statewide races this year in an election where Trump was on the ballot against an incredibly unpopular incumbent administration amidst record inflation. The hardcore MAGA base doesn't care enough to show up for down ballot races, and the reluctant Trump voters who chose him over Harris because of inflation and the border aren't necessarily conservative, which is why Gallego, Slotkin, Whitmer, Ossoff, and Warnock are in the senate.
>lost most of the swing state statewide races this year
Only in states with no ID and endless mail-in vote counting.
lol good one
I think a big problem is that while the data was good, the analysis was terrible. And that occasionaly includes here. The google searches article was a classic example. Other forecasters (such as Split Ticket) talked about the horrible campaign Trump had run, when, lets be real, he ran a pretty effective one. The MSG rally and other things were just Harris supporters hoping against hope the narrative would change. Their big example of a terrible campaign? Focusing on people who just weren't going to vote... right up until they did.
Others are choices from a marketing perspective. Focusing more on the 50/50 nature of the forecast rather than understanding more the likeliest outcomes probably did good for subscriber numbers (and the liberal audience who wanted to believe it really was closer than it was).
I also think the problem wasn't so much with polls as it was the "gold standard" polls. Polling had a good year because of a lot of those polls which were routinely viewed as garbage or partisan hackery ended up being a lot closer than the gold standard polls. Emerson got lots of shit (including from Nate), but they performed a lot better than the gold standards. It really feels like the pollster ratings need a more dramatic revision than is probably called for, but most of the gold standard guys blew it even worse than 2020.
I actually think Silver Bulletin did better than most, but retros are good and healthy.
"The MSG rally and other things were just Harris supporters hoping against hope the narrative would change. "
The funny thing about the MSG rally thing is that I have since heard that Puerto Ricans are one of the groups that broke big for Trump, particularly in Florida. I guess they can take a joke. Don Rickles made a career on comedy that I guess would be beyond the pale now.
Nobody so far has been able to explain how that joke actually works as it doesn't have any thing humorous or witty about it. Don Rickles would say that there was no joke there. I don't think that the Island of Garbage would pass the three year old test and they will laugh at anything even close to poop. A bit like Vance and the Grilled Dogs in Springfield, saying that people couldn't take a joke. But if you find that funny then laugh away for everyone has a different standard when it come to humor.
I think the joke was a double entendre about Puerto Rico being unable to handle its garbage.
https://globalpressjournal.com/americas/puerto-rico/trash-crisis-leaves-puerto-rico-brink/
If you have to think it it wasn't a joke it was a critical commentary. You don't have to think jokes that's why they are fun hence funny.
You asked for someone to explain how the joke works. Someone explained it. This is all well-known if you follow Puerto Rican politics at all. Alas, the side that loudly proclaims how outward-looking and empathetic they are only ever looks inward and has no theory of mind for people who are different to themselves.
Nobody explained how the joke worked they just mentioned a category. Now that you have informed me that it was directed at a select sub group I am still trying to get someone to explain how it worked. I am familiar with the double entendre which is saying something while meaning another generally with a bawdy overtone but there was no double entendre the delivery was done as straight fact hence my questioning how the joke worked. The three year old who says poop understands that it is going to work with other three year olds but if you want to expand the positive reception you have to do a little more work. I have a great sense of humor but I rarely waste my time laughing at things which are just not funny. You must extrapolate on this theory of mind.
Perhaps they isolated that the comment was by some dunderhead on the undercard that spoke hours before Trump did.
Puerto Ricans loved the joke because it was a clever topical reference to their long-standing trash problem: https://www.theenvironmentalblog.org/2024/10/puerto-rico-trash-problem/
Source for the Florida PR data?
Osceola County went outright for Trump after Biden won it by 14 in 2020. Heavily Puerto Rican.
https://floridapolitics.com/archives/706853-after-disappointing-night-for-florida-dems-has-osceola-county-gone-red-forever/
Thanks! Claim on wiki is that it is over 33% Puerto Rican, and other sites say the high concentration there goes back several decades.
It would be fascinating to know if this was a case of them not caring, or actively being irritated, for example at Democratic performative outrage while ignoring more important issues.
I don’t think saying Trump’s campaign was good is “being real.” It’s not clear it was good.
My guess, though I ofc don’t know for sure, is that he ran a pretty bad campaign and won because inflation mattered much more than campaign effects. A good campaign is worth maybe 2% of vote share, and I think 2024 tells us your opponent’s administration presiding over 20% higher prices is worth a fair bit more than that.
Matt Taibbi discussion, references and disagrees with Nate
https://www.racket.news/p/how-americas-accurate-election-polls
And Rasmussen. https://x.com/Mark_R_Mitchell/status/1858903406972166160
While I agree that there were pollsters that were massaging the results a bit for fear of blowback from a liberal media subscriber base, I don’t think it is true for the analysts. Somewhat different subscriber base looking for a different product.
Nate himself has said that liberal subscribers would get angry when he said something they didn't want to hear. I think he relishes that, but many analysts are more inclined to pander.
I know Nate has said his audience is pretty liberal. I don't think it was blowback so much as implicit bias
The Seltzer poll was actually very good the only problem is that those who were polled failed to hold their mud in the booth. The reasons for the numbers were quite clear for Blind Freddy, Soy Beans and Corn. The stockpiles are at high numbers and the price is at low numbers and when you start to impose tariffs there will not be many incentives for foreign customers to buy our beans n corn because other countries have stockpiles too. This will deliver a lot of good farm acreage to Corporate Ag. Have a look back and see how the farm sector weathered the last set of tariffs, they had to go on major farm welfare to the tune of billions of dollars. Seltzer got the numbers right but underestimated the cowardice of the respondents.
Lol. Okay
Didn't just just ring up people at random out of the phone book? Maybe that doesn't work anymore.
Rosenberg had the worst combination of being a transparent partisan hack, and actually having a fair bit of reach with the completely unreliable crap he was slinging each day. I would read his blog just to see how thick the walls were in the echo chamber.
Even now, he will make every excuse on earth for the results except that he was wrong - which is the truth. He should have no followers and no influence, and if he really cares about helping Democrats win in the future, he should shut up and sit back and let some rational humans influence the next race.
This is why, even though I didn't like it, I kept reading. It is why, being a Trump voter, I also read FDB, Klippenstein, and many others. I did not do that in 2020. I was in a bubble. Never again. You can only be fooled once. After that, you are choosing to be ignorant.
Well, shame on you for being a Trump voter, but at least you are an intelligent and analytical Trump voter. Hope you like 1933…..
Well, that’s the kindest thing I’ve ever heard from a non-Trump voter. Have faith, my fellow citizen. You might be pleasantly surprised. If I’m wrong, I’ll do an apology tour and community service.
What part of sex criminal Matt Gaetz, anti-vaxxer RFKJr and Russian stooge Tulsi Gabbard makes you think we’ll be pleasantly surprised??
I’m not one who thinks he’ll likely end democracy, but can’t imagine being pleasantly surprised unless the bar is in the Mariana Trench
1) Matt Gaetz: I'm sorry, but isn't a "criminal" someone who's been convicted? The DOJ didn't even prosecute Gatz.
2) RFK Jr.: Don't lie. He and his family are vaccinated for smallpox, polio, and other diseases. Okay he's not too big on the mRNA vax. And for your sake, I hope you didn't get 3 more jabs of that particular vax.
3) Tulsi: You Fred should be shot. Tulsi is an American Hero, and if she is too peace-loving for you, then you should volunteer in Ukraine or Gaza.
RFKJr runs one of the leading antivax misinformation factories. His spreads all sorts of debunked antivaxx bull like the lie about autism. If he does accept a few vaccines, that doesn’t forgive all the lies he spreads about the others… far more than just Covid mRNA.
Tulsi actions as a veteran do not make up for her frequently parroting anti-American Russian propaganda. Whether she’s doing it knowingly or is just a fool, she assists the Kremlin in damaging our nation.
1) is a valid point.
2) So RFK jr is a hypocrite, and you are part of the antivax disinformation cult regarding mRNA.
3) you finish off with a vague death threat.
I guess you couldn't come up with a fourth point to escalate your rhetoric even further.
1. Gaetz - if the DOJ could have charged him with a crime, they would have done it
2. Rfkjr - here's an article
https://open.substack.com/pub/bariweiss/p/rfk-jr-health-human-services-flouride-vaccines-covid-trump-europe?utm_source=share&utm_medium=android&r=b7l5e
3. I'm not 100% sure on Gabbard, but I do think putting someone in that position who has never stepped foot in a War Zone is a horrible idea and I know she cares very much about the actual bodies used to fight
Lol get with the program and join the party of KINDNESS!
"You can only be fooled once". The irony.
I see what you are getting at here, but I am fully aware of what I voted for and against. I am prepared to accept those consequences. Trump's first term was a complete shite-show. Obviously, I can't control what happens, but I am comfortable and informed about my choice.
Why would you vote for someone whose first term was a complete shite-show? Do you think being 4 years older (and 8 years older by the end of his term) would help?
In short, one would have been better off just reading the Atlas Intel national and swing state polls and dispensing with everything else.
It's easy to see which polls were most accurate after the fact; not so easy to predict before. Hindsight is 20/20.
Atlas Intel was also the most accurate in 2020 so it's mostly common sense and foresight rather than hindsight
https://atlasintel.org/media/atlasintel-was-the-most-accurate-pollster-of-the-2020-presidential-election
Right, but how did they do in 2022? If they were great in one and mediocre in the other, that wouldn't justify considering them some sort of oracle. Nate's pollster ratings did have them as one of the top pollsters, but not far and away the best, which I think was reasonable.
It's not clear from the website they did midterm elections.
But they did national+7 swing states in 2020 and 2024 and seem to have been the best in both.
Much more accurate to predict Trump came into Nov 5 very strong than that they were about equal and could go either way, as the model did.
Atlas intel didn't have him "very strong," he was within the margin of error if their polls too.
People were really saying that AtlasIntel was wrong because they weren't capturing the "silent republican woman" vote that would secretly be cast for Harris.
I downloaded the last Atlas Intel swing state polls, just before the election, and they were fascinating: women made up between 51% and 55% of the polling sample in each state and Trump won at least 47% of the women vote in each.
If anything, Atlas Intel may have captured that many women were secretly voting for Trump, not Harris. The CNN exit poll showed Trump won married women by 2% (but lost single women by 21%).
The interesting thing to me is the lack of within margin of error results that overestimated Trump's win.
Even RCP's curated list couldn't find a Trump +4 poll, but had plenty of ties or Harris+1, not to mention beyond that.
Nate's math about herding applies here.
The lack of outliers leaning towards Trump shows a broad polling problem, which makes the methodology of the "accurate" polls worth exploring.
THIS.
This is what the actual analysts and number crunchers who don't waste a lot of time are more worried about.
Most of use view Trump as a unique factor, but he's either not going to be on the next election or he will be picking up the unique baggage of running on a questionably legal constitutional interpretation, so attempting to adjust polling for this election is mostly a waste given that the polls were 'okay'.
But believe it or not, those of us looking at Atlas Intel basically are looking at the numbers and seeing herding at the end towards a expected average, which means that they *guessed right* in which polling to publish, but it was still a guess.
What if next time, they guess wrong?
What if they guess wrong with my Clients polling in other areas?
For those who don't see the danger, let me sum it up like this.
The last person who we had this concern with was Selzer.
Her 'Intuition' had been really good for a really long time, until it wasn't.
And I can't have a client get burned that badly and expect not to get hell for suggesting a pollster in that case.
Are you thinking he'll run as VP?
I've lived long enough that I've seen a number of populous figures try a number of questionable or outright illegal tactics to stay in the ring. The weaker the party is without a individual, the more creative they get.
So I'll hedge my bets, even though I don't think Trump has enough strength in his years or in his administration to pull anything beyond this term as president (and I have my doubts that his heart will hold up under the stress he signed up for.)
I really hope they run Vance / Trump 2028, with the intent for Vance to resign. That'd be amazing!
The theory that I heard from a pollster is that about 3% of voters are low-trust and love Trump. They will refuse to answer when the pollster calls, but will reliably show up for Trump. That explains the bias in all 3 cases.
My brother's version is that they aren't necessarily low-trust, they are just rural and not always available by phone.
Regardless, a small and difficult to poll group with strong opinions would perfectly explain the result. No matter what the explanation. Based on that, both my brother and I arrived at the actual result by modifying the output of your model in the same basic way. And if that is the cause, then you should be able to find the size of that group by comparing polls with past elections (including primaries), and therefore fix your model to be robust to this source of error.
Though, honestly, Donald Trump is the only politician that I'm aware of who seems to show such an effect in practice. So it might not come up again for a long time. It's not that other politicians don't try to reach that group. It's that they don't realize, for example, the importance of simplifying your language down to emotional terms that that group seems to respond to.
You are peddling hopium. Nearly every county in the US swung towards Trump. Including Republican strongholds like NYC (irony intended). The “only rural hicks” narrative is BS but hopefully liberals will continue to believe it.
okay, someone’s grumpy! how about you drink a juice box and calm down, and talk to us when you’re ready to have a reasonable response to why you think he’s wrong.
I am sympathetic to Ben's view, and I don't even disagree with most of what he said (I liked his post). But you asked for a reasonable counter-point, so I thought I'd defend Gary a bit:
(1) This is the same story most have told about the polling error since 2016. It's certainly a plausible story. But if it were truly the reason polls were missing, wouldn't they have figured out some solution by now (8 years later)?
(2) It seems to confirm the priors of democrats too squarely to be credible. "Why did Trump underperform the polls?" "Well, his voters can't figure out phones and only understand simple talk." (This part is, obviously, an extremely unfair version of Ben's point, and I don't argue that that was his contention or his view. But it is how this sort of argument comes off to many people.)
(3) The theory assumes that this group of hard-to-reach voters with extremely similar worldviews were, in prior elections, either (i) not voting at all or (ii) voting in divergent ways from one another. That could be true, but this would require some further explanation.
None of this entirely disproves the theory Ben proposes, and I don't really have a strong view on it. But I do love a good juice box, and I hate to hear juice-lovers insulted.
Mmm juice boxes.
I think his response was very reasonable. We can rule out the "rural voters" effect because the effect is not limited to rural areas.
All you are arguing for is that my position is more likely right than my brother's. Thanks, I guess.
Now why don't you reread what I wrote and look for something that disagrees with my actual position.
Eh, that would require reading, and that's hard these days and doesn't give the same endorphin levels as just stating what they believe and exiting stage left and declaring it a win.
So, as a analyst, allow me to say that while I think you're on the right track, as the final numbers come in that 3% shifts to about 1.7%, which is big but not nearly as ground breaking.
I'd also argue that Rural is less the defining factor, as smartphone and app interaction is the course de'jour these days (and yes, those of us who live out in the 'wild lands, beyond the reach of city lights', do have and use cell phones) and a lot of money since Trumps first election has been in that direction to reach them, whereas in this election it appears that a general lack of enthusiasm amongst traditional voters led to a measurable loss from those quarters as trump pulled in a larger group non traditional voters this election, aka 'People who don't normally vote came out and voted the Trump / Anyone but BidenHarris.'
The 'Voting for Trump' factor is really, REALLY hard to catch without oversampling, and because they are non traditional and don't show up for midterms or any other non-Trump election *so far*, the industry is stuck in a unwinnable situation:
Over adjust to Trump ruins all other polling efforts ('The Red Wave' talk was a result of such a case), OR Do nothing and get hammered whenever Trump ends up being a force in the vote.
For what it's worth, the actual results of the polls were fine.
Part 2 because of hidden character limit:
They missed in the same direction as last time, but by less. I'm actually concerned that without Trump on the docket for the next few elections we're going to see the rubber band snap back in the opposite direction, ESPECIALLY if there is a popular figure who comes out of nowhere on the Democratic side and there is continued herding, and that people who don't understand polling will take it as a sign that 'polling is broke' when in fact there has just been such a heavy attempt to capture the Trump voters that it is making the mistake of assuming that Trump voters are the same as reliable voters.
Isn’t he talking about polls?
I was the one presenting "low trust voters". My brother thought "rural" voters instead.
But in both cases, Trump beats his polls by a hard to predict, but important, margin.
Rosenberg is such a hack. You would think he could afford a better wig with all of the hopium Substacks he sold.
I wish I cared more about polls, but I don't, except as barometers of the electorate, which, hate me now, I consider largely politically hazy, uninformed, and having an instinct for shooting itself in the head. This is no doubt informed by 50+ years of voting and watching people make impulse decisions over careful research.
We spend more time on our wardrobes than we do vetting political candidates.
So I'm not surprised at the result.
Kitchen table issues, like the cost of food, mean more to people than rhetoric, and certainly more than polls. Inflation was, and still is, a big issue. But people had no knowledge regarding the causes of this latest round. They saw it as a Biden failing, when it was a global problem, caused by a confluence of supply chain chaos meeting surging demand as things opened up in 2021. Add to that a dash of greedflation, as noted by the Kansas City branch of the Fed, and way too liberal credit policies, which make a lot of money for banks, while keeping people in a hole of their own digging.
Democrats did a terrible job of messaging on inflation. It was almost as if they thought that by ignoring it they would do better. Oops! Then again, Democrats do a terrible job of messaging anyway.
And letting Biden run again was a guarantee of a loss. After the pullout from Afghanistan, which didn't resemble the fairy tale version that magical thinking Americans expected, Biden's approval ratings plummeted, and never recovered.
And though a person's character matters to some of us, it doesn't matter to most of us, at least not as much as the cost of a carton of eggs. Trump is our first Presidential felon, maybe our first Presidential rapist. He ran a particularly disgusting campaign, which might have cost him with a stronger opponent.
Thrusting Harris into the role of carrying forth the banner, partly necessitated by Biden running out the clock, proved to be a poor choice. She didn't differentiate herself from Biden in any detailed way, relied on sloganeering to replace a cohesive set of policy statements, and proved to be too cautious. Yes, she wiped the floor with Trump at their debate, and he was smart enough not to give her another chance, just as he was smart enough to avoid the primary debates. So, Harris lost whatever momentum she had because she lacked to tools to capitalize on it.
Given the freak parade that Trump is proposing to head government departments, 2026 could be an interesting election, provided that Democrats stop feeling sorry for themselves and focus on those kitchen table issues. Trump is going to give them lots of opportunities. Get the sound bites ready.
Trump coming back to win after all that was thrown at him . . . will go down in history next to Alexander the Great, Caesar, and Napoleon.
You mean because the others were tyrants who enslaved their populations and in the end, failed or were assassinated?
Or is it because they all fellated their microphones?
My problem with the "polls are fine" is not that errors of 2-4% are either unexpected or bad, but that for the third presidential in a row it is going the same way, indicating a systemic Dem bias.
I myself have recognized that and gave a prediction of my own in my FB group Political Lane of a GOP trifecta, with 52 Senators and +8 in the HOR, and close to nailed it. My methodology was to assume some Dem bias of somewhat less than 2020 as some pollsters likely adjusted but since the real problem has not been solved, it would still be there.
I continue to use the RCP aggregates as my benchmark because in the three elections they have had less of an R underestimation, even as they lack the methodological sophistication of Nate. The excuse of looking at 2022 or 2012 does not seem valid to me as mid terms are a different critter, and 2012 was a different universe where low propensity voters and reluctant poll responders were just as likely to be Democratic, but no longer.
I don't see the glaring Democratic bias. Nate's analysis of Kamala's prospects was uniformly grim. In vain did I search for the slimmest hope.
In Nate's analysis on the morning of election day, he said the race was essentially a tie. In each of the last 3 elections Trump has outperformed the polls; granted, not by a huge amount in any of them, but when it happens 3 times in a row one begins to wonder. I don't think the pollsters are deliberately screwing it up, but they may be underestimating the number of Trump voters who refuse to talk to them.
Yes, SB final forecast was the probability of Harris winning at 50.0 and Trump winning was 49.6. I guess the other 0.4 was a tie.
But that probability was based on his poll aggregations which for the national was off 2.5 and for the 7 swing states averaged off 2.13. That is pretty good for accuracy but every poll was off towards the Dems which hints at a systemic bias.
Now RealClearPolitics a site I otherwise dont like for its right politics also underestimated Trump in all three elections but to a lesser degree and where now in all three elections Nate has gotten 9 states wrong and all have been states where his poll aggregations had the Ds ahead and went R, RCP has had 7 wrong and 6 where the Ds ahead went R and 1 the other way (called GA red in 2020). This years national polling at RCP had Harris at +0.1 where as SB had it at +1.0 and the different swing states were about that much tilted towards the Dems.
Now as far as I know RCP just averages the recent polls, where SB does various juitso on it, which analytically makes sense, and Nate is smarter than me, but, regardless, has had the effect of skewing it more to the Dems than a straight average.
If anything, Nate's Nov 5 analysis was the brightest he had sounded on Harris since the DNC.
Janice , the Dem bias is that in the last three presidentials the polling aggregations of Nate, prior 538 now here, were consistently towards the Dems. They have led to 7 miscalled states in the three and every one was a state that the polls predicted D and the vote was R. An error of 2% I agree is far from bad but if it only goes consistently one way, that is problematic
I agree that it seems to make sense to use the RCP average as a benchmark. The only question I have is how do they define the line for inclusion in their average, so that polls in the average have weight 1.0 and polls outside the average have weights 0.0. Otherwise it is very transparent and looks like the SP500 average which stock pickers have a hard time beating. (Okay Nate would point out SP500 itself is weighted…)
Did you correctly predict Trump winning the popular vote?
I didn't predict a pop vote but I did assume some Dem bias of 1-2%.
My question: perhaps this is something that can’t be answered from the data, but indications were that Trump was gaining over the final month or so. Was there a meaningful movement in his favor (rather than random fluctuations)? And if so, were there any events or circumstances that could explain it?
I think the VP debate had a negligible effect on the race, but it coincided with the beginning of Kamala’s decline in Nate’s forecast, and I remember 2 key moments from the debate - Tim Walz bragging about Dick Cheney’s endorsement and the online left reacting exceptionally negatively to it - and JD Vance effectively being “normal” and giving a reasonable looking image to his far-right policy agenda.
Ron Bauer this is an excellent question and maybe something Nate could discuss in a paragraph or section in one of the posts, or in the introduction to the 2026 or 2028 models. We know the final result and 2024 and the polling results of all these pollsters during the campaign, both before and after Biden drops out. In addition to the Trump vs. Harris polls which is the focus of Nate’s work (and RCP poll average, to take a more transparent baseline), there are the betting markets and also the opinion polls on topics like what matters to you more inflation or abortion (etc. ). And they did exit polls on Election Day (don’t know if they do exit polls on early voters or mail-in voters?). I was going to assert that Trump was always in the lead because opinion polls showed people thought the US was on the wrong track, inflation, border. It might be that careful exit polls might answer if that was true. One angle Nate might know is what % of people have made up their mind in the “early” polling, and in the final Nov election what percent was “likely voters” and what percent were “unlikely” voters and what was the Harris/Trump split in each bucket.
I thought chances of heads or tails three times in a row was 12.5 out of 100….
The odds of three heads in a row is 1/2^3 = 12.5%. Likewise the odds of three tails in a row is 1/2^3, or 12.5%. The odds that either one of these happens is the sum of the two, since there is no overlap between these outcomes.
Got it. So 12.5+12.5=25 out of 100 for either heads or tails to come up three times in a row. Thank you.
Indeed. But Trump is heads so it's 12.5%
But the point is that if it is just a random error it could have gone either way.
Yup if your referring to it to either candidate before it has happened once. But what if you are talking about it happening to Trump twice after it has already happened once. Or happening to Trump a third time after it has already happened twice? That final analogy is reality.
I mean, in the coinflip analogy, the chance of it "happening to Trump a third time after it has already happened twice" is 50% lol
Something has been bothering me about election polling commentary like this. Maybe someone can correct my misunderstanding.
"The polling averages were right in 48 of 50 states." Fair enough. But there was no reasonable doubt about many of those states. Nearly all election coverage, campaign attention, etc. was focused on about 7 (not 50) "swing states." Would a more accurate gauge of accuracy be 5/7? (Still pretty good.) Or, being generous, there are about 15 states that have "swung" in the past 25 years (see 270's 'same since' maps). Even being extremely generous, and using states that have "swung" in my lifetime (somewhere in the range of 30-40 years), the denominator would be less than 25.
In any event, that's not my gripe. A consistently-applied 50-state denominator as the gauge of polling prediction accuracy would be fine--even if it inflates accuracy, the year-by-year comparisons of accuracy would at least be consistent.
My gripe is mostly that every time the issue is raised of directional misses--i.e., the polling bias favoring D in every presidential election in the trump era--I am told that the REAL denominator for the fraction is ONE, not 50. E.g., "It's just three elections, and heads three times in a row does not tell us much." Fine, again, in a vacuum. But what happened to 50? If we use 50 states as the sample size, the data would tell a pretty significant (and convincing) story of systemic polling/aggregator bias. In 2016, 2020, and 2024, the D lean was present in almost every single state. (I haven't checked in a couple days, but I'm not sure the polls missed TOWARD trump in a single state this year.)
Perhaps there is an explanation why 50 is the correct denominator for determining whether polls got the answer right, but not for determining whether polls systemically miss one direction or the other. But I haven't heard it, and it makes things like this difficult to read.
Polling errors are usually systemic, so if they were off by 2 or 3 points in all 50 states, it wouldn't be that surprising. Does that help address your question?
Also, the state-by-state directional results certainly have a systemic connection, but I'm not sure it's as direct as the last few elections would suggest. In 2012, for example (the last year polls generally leaned toward R(omney)), Obama still underperformed the polling averages in more than 1/3 of the states.
Not really. But I'm not a stats guy, so maybe I'm missing it. The logic seems to be that: if polls miss left/right in AK, then they will probably miss left/right in WV as well. And that seems probably true enough. But wouldn't the necessary flip-side be: if polls get the answer correct in AK, then they'll probably do the same in WV? In other words, a systemic relationship between the state-level polls doesn't address why state-level polls should be the metric for some accuracy measurements (e.g., binary accuracy) but not others (e.g., directional miss).
Where are you seeing that state polls are the metric for only some accuracy measurements? I think a person could look at either state or national polls (or both), but that they should be consistent in which polls they look at.
My main takeaway from this article was that Nate cares A LOT more about progressive viewpoints than I do. Uncertain whether that's because he cares too much, or I care too little.
I believe it’s because you care too little. Nate, for being the relatively centrist liberal he is, spends a great deal of time consuming the ideology and logic of both extremes in order to further develop and strengthen his own personal worldview to be able to cut through the emotionally-driven narrative that exists on the insurgent right and “wokeist” left
As a matter of situational awareness, it's worth being /aware/ of what's up on the progressive left; but being aware, I currently struggle to care. They've been playing defense for nearly a decade, and since 2016 most of their intellectual achievements have been some variation of denying America's political realignment. It's not at all clear they'll have much claim to being the left's thought leaders after this year.
The quotes in Nate's article are a case in point - judging purely on the merits, nothing they had to say warranted acknowledgement from the likes of Nate Silver. Yet he seems to care more than I do. It's possible I'm too dismissive of this once-influential group.
Wow, I think they're very much on offense, and they're very dangerous.
Tbh, I think a lot of Nate's substack-era punditry is driven by his personal experiences. He is an upper class gay man who lives in NYC and works in media. His social circle is likely very exposed to liberal/progressive viewpoints, and he has a sort of inherent contrarian personality, so he has a lot more opportunity to get rubbed the wrong way by it. I'm sure if he lived in a small rural town in West Texas and was primarily exposed to the viewpoints of his community Facebook page, the set of viewpoints that were constantly bugging him would be quite different.
All Hail West Texas!
I also don't think people understand polls and how they work and how easy it is for them to be off. People are people. Sometimes they lie, sometimes they don't understand the question as asked, sometimes there are no good answers or not enough expansion is done on a question. I know this because I'm a regular poll taker. I also send in feedback on questions. Example: Many polls this year asked about "protecting democracy" but they didn't ask what people meant by that. Many people were thinking about free speech, lawfare, etc. Others were thinking about institutions. See?
Yes.
The exit polls showed a large number of Trump voters who felt protecting democracy was important.
The left wing doesn't understand the perspective that they might be considered to be on the wrong side of this issue.
I'm prefacing my quick criticism of Nate with a simple note that I subscribe to him, so pretty obviously find some value in his analysis and thoughts.
Some of the post-election analysis veers into ignoring what was written in August.
For example:
https://www.natesilver.net/p/kamala-harris-is-not-going-back-to
I am not sure how you can lead with "She's good at this" in that post and then write an entire post that Harris was a replacement level candidate. To Nate's credit, only a handful of posts the entire cycle came across as fawning (any positive reference to coconuts comes to mind), but it is something I hope the Silver Bulletin team (which hopefully grows) does an honest internal look at.
FInally, Nate should take a victory lap on calling out some of the late campaign narratives on polling - margin of error, typical accuracy issues, and the 'flooding the zone' narrative that inexplicably caught on. I would love to see a response/follow-up to Taibbi's write-up last week:
https://www.racket.news/p/how-americas-accurate-election-polls
Replacement level is not really an insult.
It is definitely sounds like it until you look into their actual performance level.
Given Biden's approval numbers and the global anti-incumbent push, "replacement level" is more about needing a star to beat Trump.