Why are people assuming Trump will outperform his polling by less than he did in 2020?

[cherry mandarin]

Why do people all of a sudden seem to think the polls are picking up shy Trump voters who refused to answer the polls in 2016 or 2020?

It's clear that "shy Trump voters" were a problem for the polls in 2016 and especially 2020. His supporters simply had a lower response rate than Clinton or Biden voters, and no amount of demographic weighting was going to obscure this issue.

Why are we suddenly assuming this isn't the case anymore, or at least is less of a problem than it used to be? What is your reasoning for thinking these shy Trump voters' response rates are now higher than they used to be?

I don't think the 2022 comparisons make for a great argument, simply because turnout dynamics are just so different in midterm years compared to presidential ones, and because we know for a fact there are many voters who will only turn out when Trump is on the ballot. When you look at the demographics, the turnout dynamics for midterm years now favour the Democratic Party more than they do for the Republican Party, relative to presidential turnout rates. This is a total reversal from where we used to be even just a short decade ago.

[heatcharger]

There has to be some way for Biden to win.

[Amenhotep Bakari-Sellers]

There has to be some way for Biden to win.

Lol Trump is an insurrection, we don't care what silly polls say we still have to vote

[iceman]

Why do people all of a sudden seem to think the polls are picking up shy Trump voters who refused to answer the polls in 2016 or 2020?

It's clear that "shy Trump voters" were a problem for the polls in 2016 and especially 2020. His supporters simply had a lower response rate than Clinton or Biden voters, and no amount of demographic weighting was going to obscure this issue.

Why are we suddenly assuming this isn't the case anymore, or at least is less of a problem than it used to be? What is your reasoning for thinking these shy Trump voters' response rates are now higher than they used to be?

I don't think the 2022 comparisons make for a great argument, simply because turnout dynamics are just so different in midterm years compared to presidential ones, and because we know for a fact there are many voters who will only turn out when Trump is on the ballot. When you look at the demographics, the turnout dynamics for midterm years now favour the Democratic Party more than they do for the Republican Party, relative to presidential turnout rates. This is a total reversal from where we used to be even just a short decade ago.

most people assumed already that pollsters have learned their lessons from disastrous performances in recent cycles and are using new methodologies to factor in those so called “shy trump voters”. Still there’s reason to believe that Trump may overperform again his polling, especially in the rustbelt.

[cherry mandarin]

most people assumed already that pollsters have learned their lessons from disastrous performances in recent cycles and are using new methodologies to factor in those so called “shy trump voters”. Still there’s reason to believe that Trump may overperform again his polling, especially in the rustbelt.

And what would said "new methodologies" happen to be? I'd like to hear about them.

[LAKISYLVANIA]

Why do people all of a sudden seem to think the polls are picking up shy Trump voters who refused to answer the polls in 2016 or 2020?

Short answer:

Wishful thinking

[ProgressiveModerate]

Because basically no other metric besides polling suggests any sort of Trump landslide. Midterms and special elections. Donations and fundraising. Campaign offices and ground game. Demographics and demographic changes in key states.

[LAKISYLVANIA]

Because basically no other metric besides polling suggests any sort of Trump landslide. Midterms and special elections. Donations and fundraising. Campaign offices and ground game. Demographics and demographic changes in key states.

Because none of these metrics matter. And demographic change is pointless when within that demographic you see a mass drop in turnout.

If demographic change in states was all that mattered, really you could write in the winning party from 2024 to 2052 basically with 100% confidence.

[The Economy is Getting Worse]

Because basically no other metric besides polling suggests any sort of Trump landslide. Midterms and special elections. Donations and fundraising. Campaign offices and ground game. Demographics and demographic changes in key states.

Biden approval is sh**t. Most people think the economy is bad and getting worse. A bunch of foreign policy failures. Migrant crisis. Midterms and special elections are irrelevant when the same downballot Democrats are outperforming Biden by double digits.

[The Economy is Getting Worse]

Some people do not want to believe Trump is way ahead, or that 9% inflation, 8% mortgage rates, and foreign policy failures do hurt your support by more than 1%.

[cherry mandarin]

Basically no other metric besides polling suggests any sort of Trump landslide.

It doesn't have to be a landslide even if Trump outperforms his polling. The polls just have to narrow between now and Election Day (or, more likely, flip to a Biden lead instead) in order for that to happen.

Demographics and demographic changes in key states.

I think if you look at the trends within each group individually, on the whole there are more promising signs for the GOP than the Democratic Party long-term.

[Neo-Malthusian Misanthrope]

It's not unreasonable to believe that pollsters are attempting to correct for shy/unreachable Trump voters and avoid the same misses as the last two times. Have any explicitly said this is their methodology?

[The Economy is Getting Worse]

It's not unreasonable to believe that pollsters are attempting to correct for shy/unreachable Trump voters and avoid the same misses as the last two times. Have any explicitly said this is their methodology?

Downballot Democrats are way outperforming Biden in the same polls. It looks bluer than 2020/2022 downballot.

[cherry mandarin]

Have any explicitly said this is their methodology?

This was my motivation for creating this thread—I'm not convinced in the least that any of the changes they've purported to make can come anywhere close to accounting for the lower response rate of hidden Trump voters. Sure, Biden could end up outperforming his polls or they could be bang-on, but at least from what I've seen, that looks like it'd be more a product of good fortune on the pollsters' part rather than an actual fix to the systemic issues that plague their industry (at least when it comes to Trump as a candidate).

[Progressive Pessimist]

Because his polling is a little too good for being the polarizing, obnoxious nightmare he is.

[cherry mandarin]

Downballot Democrats are way outperforming Biden in the same polls. It looks bluer than 2020/2022 downballot.

Exactly; this is another reason why I'm not too skeptical of the figures showing Trump ahead right now (although this doesn't mean I necessarily expect him to maintain his current lead).

[Aliens]

It's good for Biden that there are strong candidates for Senate in some of the swing states that he needs to win, since there's not much ticket splitting anymore.

[Progressive Pessimist]

Downballot Democrats are way outperforming Biden in the same polls. It looks bluer than 2020/2022 downballot.

Exactly; this is another reason why I'm not too skeptical of the figures showing Trump ahead right now (although this doesn't mean I necessarily expect him to maintain his current lead).

Do you expect Biden to catch up to Senate Democrats?

I mean, even if Democrats miraculously maibtain the Senate while Trump wins the election-it wouldn't be the worst outcome-but still a bottom five outcome that would infuriate me for four years. Kari Lake is a step too far, but Trump is fine!? What the s*** is wrong with this country!?

[cherry mandarin]

Kari Lake is a step too far, but Trump is fine?

Evidently there are people who feel this way.

If Democrats miraculously maintain the Senate while Trump wins the election

Oh I'd be shocked if that happens, especially in a world where Trump wins. All I'm saying is that polls show a substantial number of voters choosing Trump or third-party presidential candidates who also intend to vote for down-ballot Democrats. I'm not sure we have any good reasons to believe that these voters are either lying or will change their minds en masse between now and November.

Particularly in states like MT, OH, PA with popular Dem incumbents running, Biden's margin deficit remains in the high single digits or even as many as 15 points relative to those down-ballot Democrats. The same general trend holds in virtually every other competitive state, just not to the same degree.

[ProgressiveModerate]

Because basically no other metric besides polling suggests any sort of Trump landslide. Midterms and special elections. Donations and fundraising. Campaign offices and ground game. Demographics and demographic changes in key states.

Because none of these metrics matter. And demographic change is pointless when within that demographic you see a mass drop in turnout.

If demographic change in states was all that mattered, really you could write in the winning party from 2024 to 2052 basically with 100% confidence.

You can make the same argument polls don’t matter because they aren’t actually real like election results and donation numbers.

Also Demographic change does matter because in a state like GA, it gives Biden a bit of a cushion to perhaps like with black voters for instance. It doesn’t guarantee he’ll win the state but certainly helps him.

[ProgressiveModerate]

Another big issue with polling is it suggests a pretty massive re-alignment despite not all that much changing since 2020; the candidates are the same and general messaging is simillar from both sides. There hasn’t been some massive recession or world war.

Elections that prompt large re-alignments have unique circumstances from previous elections like the recession + Obama in 2008 or going from Romney —> Trump in 2016.

[cherry mandarin]

You can make the same argument polls don’t matter because they aren’t actually real like election results and donation numbers.

Polls have way greater predictive value than past results and especially fundraising figures, especially in areas with a large number of high-quality polls (virtually all the battleground states down the stretch + the national polls).

As for the re-alignment narrative, I’m not sure why you don’t buy into it. I’ve made posts detailing why I think it’s real, but I guess only time will tell.

[kwabbit]

Because basically no other metric besides polling suggests any sort of Trump landslide. Midterms and special elections. Donations and fundraising. Campaign offices and ground game. Demographics and demographic changes in key states.

Demographics can't be helpful everywhere for Democrats though. If generational turnover and demographic change is helpful everywhere, then it's helpful nowhere. The whole country is getting more diverse and the whole country has old people dying and new people entering the LV pool, but Trump has still gained 5 pts in margin since 2020. In PA, WI, MI, there's little change, NV, NC, and AZ are diversifying but not in a strictly helpful or powerful way. GA I would concede that demographic changes are genuinely helpful for Democrats.

[LAKISYLVANIA]

Because his polling is a little too good for being the polarizing, obnoxious nightmare he is.

which is again wishful thinking and purely based on your personal convictions.

Despite, still half of the country basically still would hate him, even if he wins, even if he wins by 55-45, 45% still hates him - including all the people who vote third party or don't care enough/refuse to turn out.

[LAKISYLVANIA]

Another big issue with polling is it suggests a pretty massive re-alignment despite not all that much changing since 2020; the candidates are the same and general messaging is simillar from both sides. There hasn’t been some massive recession or world war.

Elections that prompt large re-alignments have unique circumstances from previous elections like the recession + Obama in 2008 or going from Romney —> Trump in 2016.

Pretty sure some people would disagree with that or would ask if you were living under a rock.

Sure... it might not be a technical world war. We haven't had a world war since the 1940s. But we've had the first territorial war on European soil since the 1940s and the bloodiest conflict to ever take place in Israel.

And lots of people would disagree with that recession. Things haven't been really great. It's not Great Depression-tier of recession. But we've had economic hardship since basically covid pandemic broke out. And at the end purchasing power of people matters when evaluating presidents in the USA traditionally.

The world is dramatically changing over a short period of time, we might not all realize it. But AI, social media, fake news, all weren't even a thing 20 years ago. The whole mood, terms like woke and level of political campaigning has changed dramatically. The world is changing faster than people realize. Sure, lots of these things were already around in 2020 but in a slightly longer timescale, things have changed quite a lot. This isn't 2004 anymore.