American Election 2024

Is Polling Broken?

November 1, 2024 at 3:03 pm

Nate Cohn: “It’s hard to overstate how traumatic the 2016 and 2020 elections were for many pollsters. For some, another underestimate of Mr. Trump could be a major threat to their business and their livelihood. For the rest, their status and reputations are on the line.”

“If they underestimate Mr. Trump a third straight time, how can their polls be trusted again? It is much safer, whether in terms of literal self-interest or purely psychologically, to find a close race than to gamble on a clear Harris victory.”

“At the same time, the 2016 and 2020 polling misfires shattered many pollsters’ confidence in their own methods and data. When their results come in very blue, they don’t believe it. And frankly, I share that same feeling: If our final Pennsylvania poll comes in at Harris +7, why would I believe it? As a result, pollsters are more willing to take steps to produce more Republican-leaning results.”
 
Not a bad summary:

Let’s look at what it would mean for each candidate to lose.

If Kamala Harris loses:

  • It will be very hard to argue that she blew the election. She’s led a remarkably disciplined campaign, decisively outperforming Trump in their only debate and making herself accessible through a broad range of interviews. She had virtually unlimited resources. She’s run as strong a campaign as anyone could. Hers might be the best presidential campaign of the modern era.
  • The underlying fundamentals of the current political climate—an unpopular incumbent, a migrant problem and years of inflation—were likely too challenging for any Democrat to overcome.
  • The media’s persistent struggle to cover Trump fairly and objectively is a big factor. A recent example is three days of attention to a misplaced apostrophe that made it seem like Biden called Trump supporters “garbage,” while statements from Trump’s former top aides calling him as “fascist” received far less coverage.
  • If younger Black and Latino men ultimately move toward Trump, it will underscore the erosion of the Democratic party’s traditional coalition.
If Donald Trump loses:

  • First and foremost, he’s unlikely to concede, instead insisting on widespread fraud without evidence. The big question is whether his supporters will turn to violence in response.
  • Though overturning the result is highly unlikely, Trump’s refusal to accept defeat will once again exact a real cost on his supporters and the Republican party.
  • This would mark the fourth consecutive election where Trump has undeniably hurt the GOP’s prospects.
  • Trump will instantly become the frontrunner for the 2028 Republican nomination. If we’ve learned anything over the last decade, his base will remain loyal — even if he’s in prison. Republicans will likely face yet another cycle defined by Trump
 
And an interesting take on polls by Nate Silver.….7 states are not all within 1 pt.

Nate Silver Has Lost Trust in Pollsters

November 1, 2024 at 9:11 am EDT By Taegan Goddard 306 Comments

Nate Silver accused pollsters of putting their “fucking finger on the scale” and “lying” in an episode of his Risky Business podcast.

Said Silver: “I don’t think we’re going to learn very much in this last week of the polling. In fact, I kind of trust pollsters less, they all, every time a pollster says ‘Oh, every state is is just +1, every single state’s a tie,’ No! You’re fucking herding! You’re cheating! You’re cheating!”

He continued: “Your numbers aren’t all going to come out at exactly 1-point leads when you’re sampling 800 people over dozens of surveys. You are lying! You’re putting your fucking finger on the scale! I will not name names, but some pollsters are really bad about this… Emerson College. Whoops! Was that recorded? Oops, sorry.”
 
Trump will instantly become the frontrunner for the 2028 Republican nomination. If we’ve learned anything over the last decade, his base will remain loyal — even if he’s in prison. Republicans will likely face yet another cycle defined by Trump
The party won’t close the page on the chapter? Losing two elections isn’t enough for them to drop the experiment? And he’d be four years older. He’s already losing it.
Surely they can find someone moderate that can give Americans a genuine conservative alternative.
 
The party won’t close the page on the chapter? Losing two elections isn’t enough for them to drop the experiment?
All of these elections have been close. It takes a blowout loss or perhaps a longer period of sustained losses to get a fervent fan base to move on. And even before the election happens, Trump primes his fans in advance with the "we wuz robbed!" narrative.

Surely they can find someone moderate that can give Americans a genuine conservative alternative.
Trump does not fit the "conservative" tradition of Reagan. Free trade and resisting the "Evil Empire" were pillars of the Republican Party for decades. Not any more!
 
It takes a blowout loss or perhaps a longer period of sustained losses to get a fervent fan base to move on.
Although not the base, but most Reps were willing to cut him lose after Jan 6, but almost all of them feel back in line.
Trump does not fit the "conservative" tradition of Reagan. Free trade and resisting the "Evil Empire" were pillars of the Republican Party for decades. Not any more!
Wonder if the mob would have killed Pelosi and Pence during the insurrection; these “cowards” would probably still support him, Hats off to Cheney, Kelly, Pense, Bolton and others that came out against Trump.
Trump is the Dream candidate for the “Evil Empire” , praising Putin, XI and Kim. Reagan is spinning so fast in his grave.
 
but most Reps were willing to cut him lose after Jan 6
Obviously not enough of them. The Senate could have banned Trump from elective office permanently via an impeachment conviction (which is far less problematic than via criminal prosecutions), but the Vichy Republicans led by Marshal Petain Mitch McConnell flinched.
Wonder if the mob would have killed Pelosi and Pence during the insurrection;
My gut feeling is that it would have taken an elected official being injured or killed on Jan. 6 to get an impeachment conviction.
 
Last edited:
Obviously not enough of them. The Senate could have banned Trump from elective office permanently via an impeachment conviction (which is far less problematic than via criminal prosecutions), but the Vichy Republicans led by Marshal Petain Mitch McConnell flinched.
I don’t know what this means exactly but I’m guessing you people have factional battles within your two major parties that dilutes the effectiveness of their governing just like we do.
 
What it means is that Mitch McConnell and many Republican senators think Trump is a joke but they are too cowardly to stand up to him, alienate the base and risk not being reelected. Sort of like how Marshal Petain, hero general of WWI, agreed to govern Vichy France during WWII on behalf of Hitler. The term "Vichy Republicans" was coined by historian Ken Burns giving the Stanford commencement address in June 2016. I'll bet Burns had no idea how much worse it would get.
 
Last edited:
Honestly I'd be shocked if this was real.
If you read ChrisC' link to Nate Silver's comments, you realize that with single state polls typically of about 800 voters we should expect some variability and a few outliers. Silver said that the consistency of polls among swing states all and repeatedly showing 0-2% margins is highly suspicious and likely indicative of tweaking results toward herd mentality.

So I'll say that Iowa poll was at least conducted and reported honestly. That of course does not mean it's not an outlier that's way far off. However the article showed that Trump's lead in Iowa was only 4 points a month ago. So it's seems likely that Iowa will be considerably closer than Trump's 9.41% margin in 2016 and his 8.20% margin in 2020. There is also the issue of pollsters trying to correct for their misses in 2016 and 2020, and the possibility that some of them adjusting raw results not enough or too much.
 
I can’t believe this is still a ‘race’. The bloke is pretending to suck off a microphone for heaven’s sake. I can’t see how he could get ten million votes let alone 70 million.

Despite all these close polls I’m going to stick my neck out and predict a clear convincing win (both popular and EC) to the candidate that isn’t a lunatic. The women are going to make it happen.
 
I can’t believe this is still a ‘race’. The bloke is pretending to suck off a microphone for heaven’s sake. I can’t see how he could get ten million votes let alone 70 million.
While many articles try to explain his appeal to the non-MAGA world (and why the clubfooted Dems have been unsuccessful in defeating him decisively after ten years), this one is especially impressive. I was able to click on the link once without it being paywalled but the second time it blocked me.
 
If you read ChrisC' link to Nate Silver's comments, you realize that with single state polls typically of about 800 voters we should expect some variability and a few outliers. Silver said that the consistency of polls among swing states all and repeatedly showing 0-2% margins is highly suspicious and likely indicative of tweaking results toward herd mentality.

So I'll say that Iowa poll was at least conducted and reported honestly. That of course does not mean it's not an outlier that's way far off. However the article showed that Trump's lead in Iowa was only 4 points a month ago. So it's seems likely that Iowa will be considerably closer than Trump's 9.41% margin in 2016 and his 8.20% margin in 2020. There is also the issue of pollsters trying to correct for their misses in 2016 and 2020, and the possibility that some of them adjusting raw results not enough or too much.

I remember asking—somewhat continuously—my MBA Market Research Professors, 'Why not just model your expected outcome and fit survey data to it?'

They were always highly suspicious of destroying the random sample set.

For example, if you only have a 5% Hispanic response rate for an Election Poll - just bump it/extrapolate it to 10%. I have no idea what they are doing to massage hidden Trump voters. It also feeds into the narrative of "generals fight the last war" - trying to use what the electorate looked like last time. (i.e., Last war improbabilities - Just look at how Ukrainian Drones can keep the Entire Russian Balck Sea fleet hidden away and out of combat after losses. Or using fake internet 'gateways' to mimic command-and-control centers. Or the ability to intercept pagers going to Lebanon and imbed explosives).

After hitting the correct demographics in polling outreach (I could not understand to what extent), Ann Selzer uses a straightforward likely voter screen—1. voted or 2. highly likely to vote—and does count the probable voters. This gives a new electorate that is not modeled on the previous electorate. She said her numbers moved not due to the loss of support for Trump but the addition of new voters for Kamala - entering the likely voter screen.
 
This is mind blowing. Just one poll, but a well respected pollster.


Honestly I'd be shocked if this was real.
When I liked this post I hadn't noticed the last sentence.

Seems like Trump is already preparing for loss: https://www.nytimes.com/2024/11/02/us/politics/trump-kamala-harris-campaign.html and

Professor who has predicted winner of 9 of the last 10 elections (the one he got wrong was Gore where he disputes the outcome) says Harris will win: https://www.cnn.com/2024/11/04/politics/video/lichtman-harris-prediction-trump-election-digvid
 
While many articles try to explain his appeal to the non-MAGA world (and why the clubfooted Dems have been unsuccessful in defeating him decisively after ten years), this one is especially impressive. I was able to click on the link once without it being paywalled but the second time it blocked me.
I got part way through (up to Reagan and Nixon being good bullshitters too) before it asked me to subscribe.
Sadly I missed the punch line as to how this skill of lying can win such huge appeal with voters.
 
Back
Top