Polling is broken, right? Or is it the news? Or all of the US?

The attitude of a lot of Biden supporters reminds me of sports fans who aren’t worried about their team falling behind by a touchdown or two in the first quarter “because, aw shucks, it’s only the first quarter.”

Sure, technically they have an entire three quarters remaining to catch up. But this is a dangerous way to think. You never want to play from behind.

This is not universally true. There are a lot of people in my social media sphere who are excoriating Biden over the Israeli war in Gaza. I think for some progressives and “progressive adjacents” in this moment, there is something to be shy about, because voicing support for Biden could be a trigger for being labeled a “war-mongerer” or worse.

No, the attitude is that this is essentially the preseason, which doesn’t matter, and that (in general) there’s not much more Biden and the Democrats could be doing than what they already are (like, say, give a good quality SOTU). In 2012 Obama took out a big ad purchase in July and that was considered unprecedentedly early. At this point, all the naysaying and media stuff about Biden being too old, out of touch, or otherwise not doing what he should be doing is punditry noise. It means absolutely nothing.

You and other skeptics (presuming this describes you) haven’t actually presented anything realistic that they/we should be doing that would help.

Yeah, there and have been homeless encampments in my general neighborhood in L.A., but it’s mostly US citizens from other parts of the country. White people.

Thr sentence quoted above is true regarding me.

The Biden campaign should run focus groups, with voters who went for Joe four years ago but now lean toward Trump, and try messaging alternatives to see if anything works. But they must already be doing this.

Unfortunately, telling swing/uncertain voters the truth about Republicans generally fails to sway them:

The entire section of that article should be quoted to make clear the part you posted is from over 20 years ago:

… in the aftermath of the 9/11 attacks, Congress eyed measures intended to give the economy a boost, and House Republicans were only too pleased to pitch their ideas. By any fair measure, the GOP bill included little more than tax cuts for corporations that the party wanted anyway. Even the editorial page of The Wall Street Journal conceded the plan “mainly padded corporate bottom lines.”

Democrats, eager to go on the offensive, convened focus groups to help sharpen their message. The party, however, quickly discovered a problem: Voters literally couldn’t believe that Republicans would respond to deadly terrorism by pushing corporate tax breaks. As The New York Times’ Paul Krugman explained at the time, the Republican proposal “was so extreme that when political consultants tried to get reactions from voter focus groups, the voters refused to believe that they were describing the bill accurately.”

EDIT: My mistake – the part you quoted dates from Obama’s 2012 campaign – so 11-plus years ago :

More than a decade later, it happened again: A super PAC supporting Barack Obama’s re-election informed focus group participants about Paul Ryan’s budget plan and Mitt Romney’s support for it. As the New York Times reported soon after, respondents “simply refused to believe” what they were hearing, despite the fact that what they were hearing was true.

As New York magazine’s Jon Chait summarized at the time, focus group participants were receiving accurate descriptions of real GOP proposals, but the truth “struck those voters as so cartoonishly evil that they found the charge implausible.”

Is there even a faint idea about how large this cohort can possibly be?

I do know that CNN bends over backwards to find people who will go on camera and say this. “I voted for Biden last time … but [something something something] I’m going with Trump this time. I Just Like His Policies[TM]. Oh … you know … His Policies[TM]. Policies … yeah, that’s the ticket.

(Skillfully edited in with the minority voter who says “Biden is taking our vote for granted! What has he ever done for us?”)

I bet CNN has to ask 100s of people to find that one quoteworthy person.

I should have credited Jonathan Chait for the words I quoted from his July 6, 2012 article on the Obama - Romney race.

Whether twenty years ago or twelve years ago, figuring out what changes minds is damn dificult.

Biden was up by 6.3 percent in the RealClearPolitics rolling polling average on March 12, 2020.

Today, March 12, 2024, Biden is down by 1.7 percent.

So – 8 percent shift as of today. Over the past year, the 2020 to 2024 swing against Biden has swung between about 6 percent and 10 percent, so going with 8 percent makes sense.

Biden got 81 million votes in 2020. Trump got 74 million.

8 percent of 81 million is 6.5 million. But coming back in to edit this post, I think the real starting number I should have used is the 155 million 2020 two party total. So – 8 percent of 155 million, which is 12 million.

Is the number you asked for exactly 12 million? No, deaths, first time voters, those who sat out last time, and those switching in the opposite direction all could nudge this number up or down. But 12 million is a lot more than a faint idea.

The question was not about the shift you describe, though – the question is narrower. Specifically: how many 2020 Biden voters today lean Trump?

The 8% shift you describe is more complex because six of the nine polls Real Clear Politics uses for their average allow a choice other than Trump or Biden. So all of the following cohorts (as of 3/10/2024) come into play:

  • Biden 2020 to Trump 2024
  • Trump 2020 to Biden 2024
  • Biden 2020 to Undecided/3rd Party
  • Trump 2020 to Undecided/3rd Party

So, combine all that together, and you get Real Clear Politics’ 8% shift (as of 3/10/2024). I don’t believe we have the data to tease out the “Biden 2020 to Trump 2024” cohort apart from the others.

I agree it is complex.

I intended to use the poll questions where the only choices were Biden, Trump, or possibly undecided, even if there were more options presented in other poll questions. If I find out otherwise, we could try some sort of sensitivity analysis to see if it really matters.

Re my 12 million figure, consider that the oldest voters trend strongly GOP, and the youngest trend Democratic. And in the last four years, many of the oldest have died, while the youngest have become eligible voters. Taking that into account, doesn’t it suggest that the 12 million switching from Trump 2020 to Biden 2024, calculated as if 2020 and 2024 voters were the same people, was an underestimate?

I acknowledge that some of the 12 million or so will not vote at all, or will vote third party. This is only half as bad for Biden as a Trump vote, But it still suggests the median voter theorem AKA election decided by centrists.

There are still people who are “voting” in the polls. They think they are of some influence. They think they can still nominate someone other than Biden or Trump by… answering polls.

Polls 2024 are not polls 2016. It was simpler times. “But her emails!..crooked Hillary.”

Yes. We still see the occasional “Another Democrat will yet push Biden aside and slay Trump” on this very message board.

Voters in Michigan. Biden handling border and economy badly. But…

The survey found that 65 percent of registered voters in Michigan described the economy as either “not so good” or “poor,” with 28 percent and 37 percent of responses, respectively. At 61 percent, most respondents still said their personal financial situation is either “excellent” or “good,” with 9 percent and 52 percent of responses, respectively.

https://thehill.com/elections/4532668-trump-tops-biden-michigan-poll/

They must claim a personal win for their excellent state under poor Biden economy.

People who respond to polls - are they telling the truth or lying because of fear that someone will find out and target them? I know it sounds farfetched, and it probably is, but fear can be a strong motivator for a lot of reasons. All this talk of revenge, is it hyperbole or a real threat? Trump sounds a lot like Putin lately, “vote for me or suffer the consequences.” I’m not feeling good about the future.

Given all the evidence that polling isn’t broken, my first answer has to be – telling the truth.

However, your post led me to wonder a little about why there seemed to be a small shy Trumper effect in 2016 and 2020 polling vs. actual results. Could it be that Trump does extremely well among the small portion of the population who are both seriously mentally ill and suffering from paranoid delusions?

However, I do not want to push this too far. There’s research showing that the severely mentally ill are less likely to vote, and I suppose they are also less likely to be reached by pollsters.

It’s interesting to me that even after both nominations are clinched, some voters are still finding it worth their time to essentially cast protest votes against both candidates – but especially Trump.

It’s no secret that significant numbers of Democrats and Republicans don’t want to nominate Biden and Trump (again).

Even now, 21% of potential GOP primary voters prefer “someone else” over Trump, according to a new Yahoo News/YouGov poll — while a full 35% of their Democratic-leaning counterparts prefer someone else over Biden.

But will that resistance keep materializing at the ballot box? Or will it subside now that the contested phase of the 2024 primary season is petering out?

Tuesday offered some clues. In Florida, Gov. Ron DeSantis and former South Carolina Gov. Nikki Haley together drew just under 20% of the vote, despite having dropped out of the race.

In 2020, Trump won nearly 94% of the Florida primary vote as an incumbent president facing two long-shot opponents. This year, he captured about 81% of the vote.

The results were similar in other states. Haley won about 20% of the vote in Arizona, 16% in Kansas and 14% in Ohio and Illinois. In Kansas, an option for “None of the Names Shown” took another 5% of the vote.

The only way to interpret these continued votes for DeSantis, Haley and other options is as a protest against Trump.

This opposition could extend to the general election. In Ohio on Tuesday, half of Haley’s supporters said they would back Biden over Trump in November, according to preliminary exit polls.

Trump is weak, and the numbers don’t lie

Nikki Haley dropped out of the presidential race after Super Tuesday. But that hasn’t stopped many voters in the GOP primaries from voting for her anyway as a kind of protest against Trump. As of the time of this writing, for example, in the crucial swing state of Arizona, Haley has received nearly 19 percent of the vote—over 108K votes.

And in populous Maricopa County, Arizona, which was key to Biden’s win in 2020, her performance was even stronger. She has captured around 21 percent of that county’s vote so far at the time of this writing. Remember, it was swing voters in that county that helped deliver the governorship to Democrat Katie Hobbs over Kari Lake and return Mark Kelly to the Senate instead of sending Blake Masters (Both Lake and Masters were MAGA extremists …).

… of particular interest to me was Florida, which held a Republican primary contest while the Democrats canceled theirs. While other states such as Arizona, Ohio and Kansas held “partially closed” primaries, meaning independents could vote in them, Florida held a completely closed primary, meaning only Republicans could participate. You’d think Trump would hit the high 90s in such a state, but he again fell short, with nearly one in five voters choosing [someone else.]

As Ron Filipkowski observed, Trump continues to struggle with suburban white voters, the kind that live in Maricopa County or the counties around big cities like Philadelphia, Milwaukee and Detroit. Trump won those voters in 2016, but he lost them in 2020, and it looks like most are not coming back in 2024. Indeed, Trump has given them no solid reason to come back and many reasons not to—from the loss of abortion rights to his attempted coup, from scores of criminal charges to a civil jury verdict against him on sexual assault.

If one wants to be optimistic, one can always point to subgroups that are trending your way.

If one wants to be pessimistic, one can always point to subgroups that are trending the other way.

I live in Chester County, a Philadelphia suburb which has the highest percent of college graduates, and the highest median household income, in the state. I will be surprised if Biden doesn’t do better here than last time, because the county has been trending Democratic for decades. But I also will be surprised if Joe doesn’t do worse than last time, statewide, with those who tell pollsters that they never attended college. It is part of the fabled Big Sort where similarly situated Americans, more and more, tend to vote the same way. Add it tells me nothing about who will carry our state this November.

I’d say polling is broken, since as I understand it, professional pollsters are still working almost entirely via cold calling. Who answers unknown/unexpected calls anymore, except in specific situations where you have a good reason to?

I’d say until everyone had mobile phones, people generally answered any call, at least if they had no reason to expect a call from someone they didn’t want to talk to. Obviously, the absence of spammers, robocalls, and Caller ID were significant. Back then, getting a phone call was a bit akin to a social media “like”; it meant you were connected to things and people wanted to talk to you. If your caller turned out to be a pollster, then it was your chance to be an influential voice.

What’s more, I can’t think of any alternative method of soliciting responses that would be as effective as cold calling once was.