US Election Day 2020

[Innula Zenovka]

 

[Free]

From a recently heard discussion (on NPR or BBC World, I think) of the problems of polling in this age of extreme right-wing politics, a significant percentage of Trump supporters are unwilling to participate in polls.

And these are the supposed "shy Trump voters." J. Ann Selzer is arguing her data doesn't show this to be the case, or that if it did, they should see more "randomness" in it than they find. I don't know what the facts are here, but I do know Selzer is no slouch when it comes to polling.

 

[Innula Zenovka]

Seems to me there's an alternative explanation for the phenomenon of the "shy Trump voter."

That is, that the sampling method is somehow at fault and that existing methods simply fail to adequately to sample a particular Trump-leaning demographic.

Have they tried going back shortly after the election, and asking people if they actually voted and, if so, for whom? It might be interesting to compare the outcome with the pre-election polls (with the huge caveat that people's recollection of how they voted, should they be prepared to share it, may not be accurate).

 

[Govi]

And these are the supposed "shy Trump voters." J. Ann Selzer is arguing her data doesn't show this to be the case, or that if it did, they should see more "randomness" in it than they find. I don't know what the facts are here, but I do know Selzer is no slouch when it comes to polling.

The common description of a "shy Trump voter" that I've heard over the last four years has been a Trump voter who is ashamed to admit that s/he voted for Trump; s/he lies to the pollster. Frank Luntz (a Republican pollster) on NPR, arguing differently:
The Polls For This Election Were Off Again. Is This The End Of The Industry?
He argues that many Trump voters believe that pollsters won't tell the truth, that the pollster will mock them for their answers to political questions or present their views with a false slant. Thus, they do not lie to the pollster; they don't answer the pollster at all. This means that the pollster will get a lower response rate from the population of Trump voters than from the population of Biden voters. It's Luntz's argument, presented without proof, and needs to be tested, somehow.

 

[Veritable Quandry]

This election had an extraordinary turnout on both sides. I suspect a large amount of error that we will see (it's already closer to predictions than we saw on election night because of a difference in voting methods by Trump and Biden voters) will be in the models used to turn raw responses into "likely" voters. The Trump campaign reported that a large number of people who they collected data from at the rallies in the last weeks of the campaign were new voters, and I would seriously look into that. If they did in fact expand their electorate then that countered the unprecedented Democratic turnout, just not enough. But those voters may be one-time voters with Trump losing the election, especially if he is off the ticket in 2024.

 

[Free]

Seems to me there's an alternative explanation for the phenomenon of the "shy Trump voter."

Maybe it doesn't exist?

Have they tried going back shortly after the election, and asking people if they actually voted and, if so, for whom?

You're basically asking for exit polls, which have issues with the idea of random sampling. As for polls taken soon after an election...who is managing the costs for all these polls?

 

[Free]

Frank Luntz (a Republican pollster)

Luntz is a bit of a boob.

GOP Pollster Appears To Blame Hotel's Broken Elevator On Socialism

“Is it socialism when Frank Luntz is publicly owned?” joked one wag on Twitter.

www.huffpost.com

 

[Dakota Tebaldi]

lol. Okay, fine: nobody is answering polls untruthfully (or refusing to answer them at all). But, the polls are still wrong. Obviously bad information is entering the system from somewhere. So where, then?

If it's "a whole demographic of Trump voters that isn't being sampled", what demographic is that, and where is the evidence that they're not being sampled (as opposed to excluding themselves from the sampling)?

 

[Dakota Tebaldi]

I'm just saying - we already know that masses of people coordinate to deliberately skew online poll responses for funsies or rhetorical purposes. We know this. It's like...this aspect of online culture that everyone is perfectly aware of. So I don't get why everyone is so incredulous that this can happen to telephone-based polling as well.

 

[Free]

Stop arguing with yourself, Dakota.

 

[Free]

 

[Innula Zenovka]

Maybe it doesn't exist?



You're basically asking for exit polls, which have issues with the idea of random sampling. As for polls taken soon after an election...who is managing the costs for all these polls?

I'm not asking for anything. I'm simply suggesting that it might be instructive to see how the way people say they voted reflects the way the polls in that area turned out. Whether or not it would tell anyone anything useful I have no idea, but it seems the sort of data it might be worth having.

That's not an exit poll -- you don't need one, since you've got the actual results. You're comparing what the sample tell you they remember doing with what actually happened.

As to who should pay for it, the people who want it, I guess -- the polling companies themselves, I guess, so they can improve the accuracy of their product, which is what they're selling. Or the political campaigns, consultants and news organisations who commission them, since they're presumably interested in assessing the quality of the surveys for which they pay, and in whatever else they can glean from the follow-up data.

 

[Free]

As to who should pay for it

Rhetorical.

 

[Kara Spengler]

I wouldn't pay him 20 dollars to spit on him.

can i take your place in line then?

 

[Innula Zenovka]

Very interesting blog piece (for people who are interested in legal commentary, that is) on Giuliani's floundering yesterday, since he's an experienced, though out of practice, lawyer and he's not stupid

A bad day in court for Rudolph Giuliani – the possible significance of his inability to answer one important question from the judge

18th November 2020 Yesterday Rudolph Giuliani appeared in a Pennsylvania court, on behalf of the Trump campaign, seeking to somehow challenge the presidential election result for that state. Accord…

davidallengreen.com

It links to this related analysis

 

[Free]

I like Stephen Colbert's description: A before and after photo overlay.

 

[Free]

Might want to go and read the past day of tweets from her about her homelessness et al.

 

[Free]

Sign the Petition

Add Four Seasons Total Landscaping to the national register of historic places.

www.change.org

We as a nation need to remember where the travesty of the Trump administration died with a whimper.

 

[Jolene Benoir]

The Trump Campaign Is Paying For A Recount In Wisconsin — But Only In Areas With Large Black Populations

Trump’s lawyers are asking for tens of thousands of ballots to be tossed out in a long-shot bid to flip Wisconsin. But even if it succeeds, Joe Biden would still win the election.

www.buzzfeednews.com

Yep, when Wisconsin told him that they would do a recount for $8 million, I knew he was going to take them up on it. That they're letting him apply it only to heavily black areas is a fucking travesty. It should have been all or none, instead they let him negotiate both the areas and the cost?

Rather than request a statewide recount, Trump’s team filed a petition Wednesday to the Wisconsin Elections Commission to recount votes in Milwaukee and Dane counties — the state’s two most populous and racially diverse counties, which both voted heavily for President-elect Joe Biden.

 

[Free]

The gloves are coming off and the racism is going back on.

Wayne County Republicans ask to ‘rescind’ their votes certifying the election results - WP NewsPaper

In affidavits signed on Wednesday evening, the two GOP members of the four-member Wayne County Board of Canvassers allege that they were improperly - WP NewsPaper

wpnewspaper.com

In affidavits signed on Wednesday evening, the two GOP members of the four-member Wayne County Board of Canvassers allege that they were improperly pressured into certifying the election and accused Democrats of reneging on a promise to audit votes in Detroit.

“I rescind my prior vote,” Monica Palmer, the board’s chairwoman, wrote in an affidavit reviewed by The Washington Post. “I fully believe the Wayne County vote should not be certified.”

William Hartmann, the other Republican on the board, has signed a similar affidavit, according a person familiar with the document. Hartmann did not respond to a message from The Post.

Jonathan Kinloch, a Democrat and the board’s vice chairman, told The Post that it’s too late for the pair to reverse course, as the certified results have already been sent to the secretary of state in accordance with state rules. He lashed out at the Republicans over their requests.

“Do they understand how they are making us look as a body?” he said. “We have such an amazing and important role in the democratic process, and they’re turning it on its head.”