17 Comments

Were these normal times, Trump would lose as badly as did Goldwater or Mondale. I am hoping he does no damage to himself in the debate tonight. We know from McLuhan and others that television can sway attention away from the facts to how a candidate presents himself, and Trump came off as a bully in the first debate. Trump can say that Biden used his son to make money in both Russia / Ukraine and China, but he has to say it in the right way.

In Center City Philadelphia there are only signs for Biden. I told my neighbor across the street that I was not putting up a Trump sign on my lawn for fear that people would steal it. At least she did not give me a lecture about being for Trump. Really I am not as much for Donald Trump as for the policies he has maintained during the past four years as President; I don't know that the Trump family has broken with their habit of building an empire for themselves.

Expand full comment

Due to distrust of everything “public”, Trump voters tend to ignore poll surveys. They consider voting to be a private matter. Consequently, poll numbers are as inaccurate as MSM narratives. I believe the President will do very well on 11/3.

Expand full comment

I heard a LA County supervisor on the radio yesterday say they have 5million+ eligible voters in the county and have had 1million vote already. If that’s true, and if it’s indicative (two big ifs), that tells me something special is happening.

Expand full comment

In all the talk about enthusiasm for Trump and lack of enthusiasm for Biden, no one seems to mention the anti-Trump enthusiasm. That's the big unknown.

Expand full comment

Rick, My view is that there are all kinds of unknowns and people making bold election predictions are premature. I agree there is a good deal of anti-Trump enthusiasm. IMO, not much of it is new since 2016, as Trump did virtually what he said he was going to do and his character traits are not a surprise to anybody. In other words, the people who hated Trump in 2016, hate him now, but I suspect they voted for Hillary in 2016 and are a virtual NOP. But I also think some amount of milder anti-Trump enthusiasm sat out the 2016 election and that this group will vote Biden in 2020.

On the other hand, I think there is a good deal of 2016 anti-Trump enthusiasm in 2016 that is now pro-Trump because he basically did what he promised to do (as one 2016 Never Trump argument was that he would govern as a Democrat, which was clearly not the case). IMO, there is also evidence that a set of never-before voters are aligning with Trump and that some amount of Hispanics/Blacks that are swinging in Trump's direction because of his policies. The question is whether Trump has lost more votes than he has gained. IMO, that is something the polls are not going to be able to predict.

Have a good day!

Expand full comment

Good points. Have a good day as well.

Expand full comment

Rick, Somehow my comment was logged to your post and not my reply to Erick. My mistake.

Expand full comment

No problem.

Expand full comment

2018 was not a national election year making the polling very different. The Democrats won in the House by dumping big money into suburban districts upset with tax changes reducing SALT deductions, and with ballot harvesting in CA flipping elections long after election day, and with court redistricting that moved urban voters into suburban districts (all unique to 2018). But in 2018, the Republicans won Senate races and increased their majority despite the polls predicting a huge blue wave (remember the Beto-craze in TX and how Democrats were going to take GA house seats in special elections?). In about two weeks, we will know whether 2016 or 2018 is the anomaly in national elections, and if the the huge Biden leads predicted by this years polls are as wrong as the huge Clinton leads were in 2016.

Expand full comment

Don't forget all the early retirements brought on us by never trumpers and people believing the Russian collusion hoax. That was not a straight election.

Expand full comment

As I have stated about you before: you continually challenge what I would often prefer to think. This is your profession, and even when you say things I do not like, I have deep respect for your position and your mind. Yet as you and I both know, God will be on His throne no matter what happens on November 3rd.....

Expand full comment

Ok, so Biden is going to win. What does that mean for America? How long will he be an active president before Harris - and the far Left - take over? If the Republicans lose the senate also, will we be on a runaway train toward socialism? Should I start studying the culture of Venezuela?

Victor Davis believes America, as we know it, is toast if Biden wins.

Expand full comment

Once again, Rasmussen was the most accurate nationwide poll in 2016 election with a final poll result of Clinton +2 (versus 2.1 for the actual vote count). One can see that Rasmussen was one of the most accurate polls in 2016 over a 6 month interval by looking at the poll spreads on Real Clear Politics from May to October, with its range exactly centered on the final result (i.e. Clinton +2 with +/-4 points of sampling error).

Rasmussen: Clinton +6 to Trump +2

Bloomberg: Clinton +9 to Trump +2

Economist You/Gov: Clinton +6 to Trump +5

ABC Washington Post: Clinton +12 to Trump +1

Fox News: Clinton +6 to Clinton +2

Monmouth: Clinton +13 to Clinton +6

Gravis: Clinton +4 to Tie

NBC News Wall Street Journal: Clinton +11 to Clinton +1

CBS News: Clinton +9 to Clinton +4

Similarly, Trafalgar was the most accurate state poll for MI, WI, PA as the other battleground state polls were not even close. The 2020 election results may not match 2016. But there are good reasons to believe the polls may once again be missing Trump supporters. In about 2 weeks, we should know if that is the case.

Expand full comment
author

Zogby had a great year in 2004 and was downhill. In 2018, Rasmussen was the least accurate pollster in America. Rasmussen has done a remarkably good job of going where the trendlines are at the end. Regardless, the polling average is a way better resource than the individual polling out there. And surprise, the polling average in 2016 was way better overall than any individual pollster.

Expand full comment

Erick, I also mistakenly posted a reply regarding the difference between a 2016 national election and the 2018 midterm election to Rick Williamson's comment, but it was meant as a reply to you, as that is one of the items you leave out of your analysis.

I also believe your comment about poll average accuracy is not correct. In 2016, the RCP poll average was Clinton +3.2 with the election result being Clinton +2.1. Rasmussen's final poll was Clinton +2.0 (about a point closer than 3.2). If you average Rasmussen's range of Trump +2 to Clinton +6, it also has a +2 Clinton average.

I think Rasmussen does have a slight GOP bias in contrast to the bulk of other polls that have a Democratic bias. But if you look at Nate Silver's Poll ratings ( https://projects.fivethirtyeight.com/pollster-ratings/ ) he gives ABC/WP an A+ rating (for correctly calling 72% of the races), NBC an A- rating (for correctly calling 73% of the races), and Rasmussen a C+ rating (for correctly calling 78% of the races). So Rasmussen calls more races accurately than either ABC/WP or NBC/WSJ but gets a lower rating. Nate Silver can do some fancy averaging to claim ABC/WP and NBC/WSJ are more accurate than Rasmussen because he is weighting 2018 results more heavily than 2016 results (or prior Presidential election results), but that is comparing apples and oranges.

Expand full comment

Erick, take it from a statistics junkie. Garbage in, garbage out. While an average would/should <i>theoretically</i> eliminate errors, that's only true of a large <i>random</i> sample. RCP is anything but random. Nearly all the polls contained therein are skewed left, greatly oversampling dems and carefully choosing non-dems for desired headlines. Did anyone believe Biden was ever up 9, 12 or 16?! No, not even Biden did. Better to look at party registrations, enthusiasm and polls that have historically teased out results, such as; better off than four years ago, who do you think will win, and who are your neighbors voting for.

Expand full comment

Oops, didn't know html specialized text is a no-go. : )

Expand full comment