Why are the Trump defectors having such little impact on the polls and Trump's chances?

I’ve been reading about various Republicans who are unhappy with Trump and plan to either vote for Biden or go third party this election.

There are prominent names like John Kasich and Rick Snyder, former GOP governors of the rust belt states of Ohio and Michigan, who have endorsed Biden. There are the more than 70 former GOP national security officials who have endorsed Biden. There’s Cindy and Meghan McCain, John McCain’s widow and daughter, who have de facto had John McCain endorse Biden from the grave. There is the Lincoln project as well, consisting of several current former and current Republicans opposed to Trump.

Then there are ordinary voters who are also coming out saying they plan to change their vote from Trump to Biden. I’ve seen stories about evangelicals saying the plan to vote for Biden after having voted Trump in 2016. Here is an article that details some voters who plan to change their vote.

Despite all this, Trump still has around a 1 in 3 chance of winning according to 538’s model. This is roughly the same chance 538 gave him back in 2016, and we all know how that turned out.

So the question for this thread is why are all these efforts seemingly having such little effect? Is it an electoral college thing, where the voters that are changing to D are either in very blue states, making the deep blue states even bluer, or maybe in deep red states making deep red states only slightly less red? Is there a sizable group of people who voted for Clinton last time and plan to vote for Trump this time that are being missed by the media discussion? What do you all think is the explanation for why Biden has basically the same overall chances that Clinton had in 2016?

Because ~63M are perfectly fine with the bigotry, corruption and incompetence. Would not surprise me if it’s more than that, now that they’ve been emboldened.

Yes, it’s very much an electoral college thing, which I just finished addressing in a post on a different thread. I, too, was guilty of overstating Biden’s strength until I started looking at the poll numbers going back to last year (using 538) and the electoral college map (270towin.com).

I think Biden should win the national vote and he should also be able to win enough votes to earn a projected victory in the electoral college (which Clinton was not able to do). However, once you look at the individual state contests, the odds aren’t nearly as long as the national polls would indicate. As Silver and others have noted before, polarization for and against Trump account for some “wasted” votes in certain states. Biden has more wasted votes because he is polling better in places like California and New York, which are two of the top five most populous states and are decidedly in favor of Biden.

If Biden can win Ohio, Florida, and Georgia, then Biden is not just a winner, but a decisive winner. But if Trump wins those states, Biden might win the EC, but there’s a good chance that the race would be impacted by lawsuits and other dirty tricks. It’s even possible that Trump could lose the election and still find a way to win it through some constitutional hardball, which I’ve also discussed on other threads.

Yeah I wonder how accurate the state-level polling is this time; the lack of good state-level data was something that contributed to the shock factor in 2016.

I wonder how many rural or suburban Trump supporters hear “Hello, my name is Tom, and I’m calling on behalf of the Washington Post and we’re doing a survey on --”

“ZOMG! The Fake News, Bezos-owned, Soros-controlled Washington Post! ZOMG! Eff off and leave me along you pedophile librul media hack!”

Point being, I wonder how many Trump voters are actually predisposed to sit through an interview long enough to give out good data.

I think you are misinterpreting what 538 is saying. They are saying that based on what we know today, on Nov 3 there is a roughly one in three chance that Trump wins. That cannot be directly compared to the forecast on election day in 2016. For example, when the model for 2020 was released it was pointed out that the “if the election were today” chances for Trump were much lower (I can’t seem to find it off-hand, but I think it was single digits).

They leave a lot of room for “drift” between now and election day.

Also, about 25% of the forecast is still based on “fundamentals” rather than polling.

Put all of that together and it explains a lot of why the forecast is ~30% chance of a Trump victory.

If you want more of a “now-cast” snapshot, the national polling average (currently 50% to 43%) along with the state-by-state averages (currently showing a 4.4% lead in PA, for example) are probably the best thing to look at. It’s also helpful to think about what the maximum national popular vote loss for Trump could still result in a EV victory. I tend to think that number is maybe 5% tops.

Very few I would imagine. I suspect it’s a matter of pride and no small amount of F-U resentment for non right-wing friendly media.

And the reason that the model is less confident for a date two months in the future than it is for right now is that there’s the possibility of some unanticipated big event between now and then that help Trump greatly. Which is in fact possible.

Of course, it’s also possible that something will happen that helps Biden to the same degree. But that doesn’t matter as much, because it’d probably just take him from a narrow victory to a larger victory, and our system doesn’t make any distinctions by size of victory.

@asahi, the effect you describe should, if anything, cause the polls to be unduly optimistic about Biden’s chances, because they’d be losing some Trump supporters in their sample, but keeping all of their Biden supporters. Which is the opposite of explaining what we’re seeing.

And yet, some of the worst polling for Trump has come just a few days ago from FOXnews pollsters.

A friend of mine once asked his friends to defend and explain the thousands and thousands of documents lies from the POTUS. There was silence.

I then asked my friend what reason he had to believe that 45’s supporters had ever even heard about these lies.

I consume media from 80% of the spectrum, studiously avoiding the far left and the far right. I don’t think either side does a lot of outright lying, per se. It’s far more about what they do and don’t cover, and – clearly – the way they present each story.

Conservative media has this guy incapable of doing any wrong.

Liberal media tends to portray this guy as incapable of doing any right.

As I’m fond of saying: if you were a juror in a criminal trial … if you only ever listened to the prosecution … or … if you only ever listened to the defense …

Lots of guilty people would go free.
Lots of innocent people would go to jail.

I think the endless cries of “FAKE NEWS” are, critically, the second line of defense for His supporters. I think abject ignorance – not being exposed to critical or derogatory stories – is the first line.

I understand that it’s the opposite of what we think we’re seeing. My point is, Biden’s lead nationally, which stands at about 6-8% depending on the poll, may actually be inflated for that reason.

Now I’m sure that the survey data crunchers do their best to account for this phenomenon, but having done phone-based survey research myself, I know enough to know that this type of respondent behavior can have at least some impact on the data at the back end. It throws another layer of uncertainty into the polling.

FOX and Trump have been on the outs lately. I wonder if that has played a role in their sampling?

TBH, I don’t know that I have a good explanation for FOX polling results. Something I watched this morning showed that Trump is within the margin of 3 electoral seats when it comes to the electoral college results. That is what gives me pause and makes me think Trump voters are far from feeling defeated just yet.

I think it’s the news division that Trump has had problems with, not so much the network. If Trump had some brains for this kind of thing and wasn’t so impulsive, he’d recognize the role that Fox’s news division plays in legitimizing him. Yes, Fox has a news division that is staffed with capable and professional journalists; their proximity to Hannity, Ingraham, and Carlson is what makes the network so effective in peddling lies. Fox News viewers can dismiss criticisms of the network by saying “What do you mean they’re in the bag for Trump? Shep Smith just practically cussed on TV talkin’ about Trump.

To answer the OP, Trump voters are a hard-baked in crowd. They are almost un-budgable in their views and their loyalty. Almost no number of Republican defectors could sway them.

I do wonder, tangentially, why the D’s can’t ever muster that loyalty. Not even Obama had such a hard-baked-in crowd for him.

I don’t know what this means. Did you mean within the margin of error in states that total enough to lose by 3 electoral votes? That could be true, depending on what polls you use in your averaging. But like I said up-thread, the tipping point state (which gives Biden 279 EV) is currently PA, where in the polling average he is leading by over 4%. For comparison, on election night 2016 the RCP average in PA was Clinton +1.6%.

A 4-5% miss in the polling aggregate is certainly possible at the state level. Wisconsin missed by almost 7% in 2016. That was one of the larger misses in a very long time, I believe, and a big part of the miss was third-party candidates and undecided voters breaking for Trump. Clinton’s vote share was almost exactly predicted (46.8 in polling, 46.5 actual).

Getting back to the OP -

I believe the Trump defector narrative is representative of a very small number. The narrative is “sexy” for the media because it is part of the bigger “horse race” narrative, particularly for more left leaning publishers. But I don’t see the numbers moving the needle at this point. It is still early for a lot of voters, they aren’t paying close attention yet (unlike us news junkies). And, as others have pointed out, there are probably a lot of FOX viewer types who are completely unaware that any Trump defectors exist at all - because they haven’t been told about them.

I agree. I think people have the impression that there are more Trump defectors than there really are, because 1) “Republican rejects Trump” is a man-bites-dog story and thus tends to make headlines; and 2) the Republicans who have rejected Trump are, disproportionately, the sort of Republicans who tend to be well-represented in the media: either people who actually work in the media, like George Will, David Brooks, or Jennifer Rubin, or longtime, well-known public figures like John McCain or Michael Steele. That doesn’t mean they’ve taken the average Trump voter with them.

I have said it before, but I think that trump people lie to pollsters out of either shame, paranoia, or perversity.

I’m not sure that they necessarily lie, although some probably do. They could be just harder to reach and less responsive to traditional polling. A lot of people simply don’t answer their phones if they don’t recognize the number, fearing that it’s a scam.

Going back to the thread topic, something I wanted to add earlier is that those 70 republicans who criticize Trump now mean nothing in terms of the party. This is why I reject the idea that Republicans like McConnell or McCarthy will stop Trump if he tries to go authoritarian on election day. The party no longer belongs to Bush or anyone from his administration; it’s Trump’s party.

Not sure where you get that. I’m seeing individual states listing more line 40 to 49% as fundamentals still.

It’s pretty simple: It’s the same reason progressives are going to hold their noses and vote for an old white man who has a record of sketchy behaviour with women and hawkish foreign policy impulses coupled with a mind that may be starting to slip, backed by a running mate who threw 1500 pot smokers in jail and who said that she believes Biden’s sexual assault accuser.

Because no matter how bad your side is, you firmly believe that it is miles better than that of those other bastards.

Such is the state of partisanship in the U.S. Both sides have sucky candidates, and neither side will admit it - or care.