Political polling and the subconscious

[amadeus]

Popular vote indicates Trump is was more popular than progressive identity politics+ warmed up leftovers the Democrats served up.

Until the left figures out the root cause of that..... Trump's a symptom not the cause.

Number 1# issue was the economy. Stop trying to redirect to culture war nonsense.

I didn’t want to take the Trump administration thread into too much of a tangent but I had a question about this, and it is one irrespective of any person’s individual politics:

Does or can political polling reflect the true feelings of its respondents? Is there an element beyond consciousness that we are failing to account for?

I’m not sure what to believe, whether we can say with certainty that a poll is accurately reflective of the true feelings of people regardless of how ostensibly well the polling is done.

 

[Lexicus]

I don't think anyone can argue that anything in the culture war fundamentally changed between 2020 and 2024 that explains the differing outcome of those two elections. In fact, the only real culture war difference between these two years is that the Democrats got a lot more right-wing on various culture war issues in 2024 compared to 2020. I don't think this supports the thesis that Trump won the 2024 election on the strength of culture war issues.

 

[Zardnaar]

I don't think anyone can argue that anything in the culture war fundamentally changed between 2020 and 2024 that explains the differing outcome of those two elections. In fact, the only real culture war difference between these two years is that the Democrats got a lot more right-wing on various culture war issues in 2024 compared to 2020. I don't think this supports the thesis that Trump won the 2024 election on the strength of culture war issues.

What we are saying if left wing cultural stuff was that important they would win more. It drowns out stuff online.

NZ once upon a time went economy first thos people were not social liberals in the modern term. What they did do was create the environment which made NZ more liberal.

Right here right now economic populism (left wing) narried to a milder form of nationalism woukd probably clean up.

 

[Zardnaar]

I didn’t want to take the Trump administration thread into too much of a tangent but I had a question about this, and it is one irrespective of any person’s individual politics:

Does or can political polling reflect the true feelings of its respondents? Is there an element beyond consciousness that we are failing to account for?

I’m not sure what to believe, whether we can say with certainty that a poll is accurately reflective of the true feelings of people regardless of how ostensibly well the polling is done.

Polling just a snapshot. It can influence people votes though eg vote for the winner.

 

[Arwon]

I don't think anyone can argue that anything in the culture war fundamentally changed between 2020 and 2024 that explains the differing outcome of those two elections. In fact, the only real culture war difference between these two years is that the Democrats got a lot more right-wing on various culture war issues in 2024 compared to 2020. I don't think this supports the thesis that Trump won the 2024 election on the strength of culture war issues.

I would actually dispute that the 2016, 2020 and 2024 elections were even particularly different. On any sort of global/historical comparison they were nearly the same election three times in a row. About half the vote went each of the two permitted directions. Most voters voted the same way all three times, most US states delivered pluralities in the same direction. The swings were mostly just a relatively small number of votes at the margins, and probably explainable as much by incumbency factors and specific people not bothering to vote as anything else.

Any discussion of political "mandates" gets a bit silly because it presupposes that the handful of people who swung at the margins for heterogeneous, poorly understood and often incoherent reasons get to decide the mandate. Rather than the two broad pretty stable masses of voters who make up most of the two parties' consistent voting bases and who form a clear majority of the people who voted for the party that won.

To get specific, most of the people who voted for Trump in 2024 also did so in 2016 and 2020, and they were supporting the same general ideology all three times. It's quite possible or even probable that some combo of economic conditions, creeping vote suppression and the Democrats being generally bad caused a couple percentage points of swing and put Trump in office again rather than barely losing again. But that's not why most ppl who voted that way did so, and it foesn't explain the underlying widespread support for the party and its current ideology. It also doesn't tell us anything about why the party has radicalised in the way it has.

Also while we're at it, in the context of contemporary western electoral politics, most people who say "the economy" is why they vote are just working backwards from who they vote for anyway, and have an extremely feelings and identity based understanding about what "the economy" even is. It's a truism bordering on meaninglessness. They're all mostly operating in the same economic system, it's not like anyone electorally relevant is trying to change the whole structure.

 

[Marla_Singer]

I don't think anyone can argue that anything in the culture war fundamentally changed between 2020 and 2024 that explains the differing outcome of those two elections. In fact, the only real culture war difference between these two years is that the Democrats got a lot more right-wing on various culture war issues in 2024 compared to 2020. I don't think this supports the thesis that Trump won the 2024 election on the strength of culture war issues.

It's all a question of what topics were attracting the attention of public debate at any given time. In 2020, people still had in mind the large riots which followed the murder of George Floyd. That was all forgotten in 2024. It's not about people switching side in the culture war, but rather how motivated they are on each side to go vote.

 

[Archon8]

James Carville was a political strategist for Bill Clinton's successful 1992 campaign for president and he had a saying, "It's the economy, stupid." Ultimately, the state of the economy and the relative appeal of the economic plans of the candidates is a larger motivator for voters than anything else. There are always single issue voters for other things, but the economy always wins in priority among voters.

 

[Klaus Hergersheimer]

I too don't see the 2024 election being a mandate on much of anything besides dissatisfaction with the Joe Biden interregnum (for lack of a better word).
I still maintain that the decisive voters were not given sufficient enough time understanding his replacement--Kamala Harris--as being that much different than him in terms of substance, and they instead went with the devil they knew in Trump.
The choices were bad ones but also very easy.

 

[Zardnaar]

I too don't see the 2024 election being a mandate on much of anything besides dissatisfaction with the Joe Biden interregnum (for lack of a better word).
I still maintain that the decisive voters were not given sufficient enough time understanding his replacement--Kamala Harris--as being that much different than him in terms of substance, and they instead went with the devil they knew in Trump.
The choices were bad ones but also very easy.

It's understandable why Trump won. Don't listen to what hesays but what he's talking about.

If he actually tries doing what he talks about it seems to blow up.

 

[Marla_Singer]

I still maintain that the decisive voters were not given sufficient enough time understanding his replacement--Kamala Harris--as being that much different than him in terms of substance, and they instead went with the devil they knew in Trump.

Switching the candidate so late was clearly a bad move. It's like the Democrats admitting themselves that their candidate is bad and they need to put a secondary choice instead. That was a panic move. It definitely more contributed to Trump's victory than it prevented it.

If the Democrats had doubts about the ability for Biden to run for a second mandate, that should have been settled a year before.

 

[amadeus]

Polling just a snapshot. It can influence people votes though eg vote for the winner.

If in case you are unfamiliar, exit polling is done after voting, so the respondents’ votes cannot be influenced by the poll even if it were possible to provide the response data in real time.

I don't think anyone can argue that anything in the culture war fundamentally changed between 2020 and 2024 that explains the differing outcome of those two elections.

I don’t think it was my intent to do so. My question is entirely as to whether polling can accurately gauge the importance people truly place on some issues.

Some of it could be leading questions, even if they are not intended to be—“what is the most important issue?” can be a different question from “what feelings are influencing your decision?” Depending on terms, that could change the outcome.

But beyond that, my initial question was one of consciousness, and whether when we answer polls we are unknowingly being less than fully truthful about ourselves.

 

[Voidwalkin]

Trump won voters under 50k outright, and Gen Z men outright, to proclaim there was no meaningful shift is simply inaccurate.

In fact, the only real culture war difference between these two years is that the Democrats got a lot more right-wing on various culture war issues in 2024 compared to 2020. I don't think this supports the thesis that Trump won the 2024 election on the strength of culture war issues

Explanation for losing Gen Z men, voters under 50k, a large swing in Hispanic men, black men, while holding steadier with women?

Pattern appears to undermine an argument the culture war was not more relevant than 2020.

Switching the candidate so late was clearly a bad move. It's like the Democrats admitting themselves that their candidate is bad and they need to put a secondary choice instead. That was a panic move. It definitely more contributed to Trump's victory than it prevented it.

Did you happen to watch the infamous debate? Biden confirmed the right-wing attack narrative against him, on stage in front of seventy million people.

 

[Klaus Hergersheimer]

It's understandable why Trump won. Don't listen to what hesays but what he's talking about.

If he actually tries doing what he talks about it seems to blow up.

What he was talking about is also why he lost in 2020 so I don't follow. His tune hasn't changed much: "the only problem with the world is I don't have my finger in every pie."

Switching the candidate so late was clearly a bad move. It's like the Democrats admitting themselves that their candidate is bad and they need to put a secondary choice instead. That was a panic move. It definitely more contributed to Trump's victory than it prevented it.

If the Democrats had doubts about the ability for Biden to run for a second mandate, that should have been settled a year before.

One of the better (though still unconvincing) explanations that I ever read on these pages was that Joe Biden picked Kamala Harris as a VP so she could run as his replacement in an election.
The fatal flaw in this reasoning is that that is not the job of the VP; the VP is supposed to serve as a replacement during the president's current term; not within a new term.
The Dems shot themselves in the foot because of a lack of a competitive primary. And I hope they and everyone learns a lesson: as much as Americans complain over "the imperial presidency," they can quite easily put a stop to that when they detect that a candidate has been "anointed" for them rather than chosen by them...

 

[Ajidica]

I don't think anyone can argue that anything in the culture war fundamentally changed between 2020 and 2024 that explains the differing outcome of those two elections. In fact, the only real culture war difference between these two years is that the Democrats got a lot more right-wing on various culture war issues in 2024 compared to 2020. I don't think this supports the thesis that Trump won the 2024 election on the strength of culture war issues.

Eh, I think the difference was that by and large people were no longer scared like in 2020, so they would get obsessed by stupid stuff. The economy was just bad enough for some people that large swathes of the country who were doing just fine could point to those areas and say they were voting over the economy.

 

[Akka]

Does or can political polling reflect the true feelings of its respondents? Is there an element beyond consciousness that we are failing to account for?

I’m not sure what to believe, whether we can say with certainty that a poll is accurately reflective of the true feelings of people regardless of how ostensibly well the polling is done.

I think that people will consciously and even unconsciously suppress what they consider "bad thoughts", even in anonymous survey (and as such, much more in face-to-face ones), so there is always an undercounting of the proportion of "bad" opinions (depending on what is generally considered "bad" in the location where the polling is happening).
So if there is a significant delta between anonymous and face-to-face survey on some aspects, I'd say you can increase a bit this delta to get a better approximate of the actual true feeling of respondents.

 

[Archon8]

I think that people will consciously and even unconsciously suppress what they consider "bad thoughts", even in anonymous survey (and as such, much more in face-to-face ones), so there is always an undercounting of the proportion of "bad" opinions (depending on what is generally considered "bad" in the location where the polling is happening).
So if there is a significant delta between anonymous and face-to-face survey on some aspects, I'd say you can increase a bit this delta to get a better approximate of the actual true feeling of respondents.

I'm sure you're already aware, but for others who want to read up about this:

Bradley effect - Wikipedia

en.wikipedia.org

 

[amadeus]

I think that people will consciously and even unconsciously suppress what they consider "bad thoughts", even in anonymous survey (and as such, much more in face-to-face ones)

I wonder then what the gap is between those kinds of polls. I don’t have the energy right now to do such a thorough look-through though!

I'm sure you're already aware, but for others who want to read up about this:

Bradley effect - Wikipedia

en.wikipedia.org

That kind of, to me, raises another question: if polling of a constituency who have less filtering of unpopular views more accurately reflect their choices, and vice-versa for those who have high sensitivity towards being ostracized.

 

[Voidwalkin]

I wonder then what the gap is between those kinds of polls. I don’t have the energy right now to do such a thorough look-through though!


That kind of, to me, raises another question: if polling of a constituency who have less filtering of unpopular views more accurately reflect their choices, and vice-versa for those who have high sensitivity towards being ostracized.

When studies were done regarding the Shy Tory effect, they found the polls were not properly representative IIRC. Meaning, they didn't call enough right wingers. Similar poor representation is thought influential in why Trump overperforms the polling.

Besides "people lie", there does exist another explanation: the pollsters find what they want to see, deliberately. I suppose if you wanna take it to exit polling, the way that would manifest would be the pollster choosing not to interview the gnarly, tattooed up guy in a leather jacket, choosing instead to interview the meeker dude in biege with a light brow.

 

[Zardnaar]

Polling here basically predicts the winner.

I've picked the winner every election since 1999. Kinda wrong in 1996 because it was the first proportional/messy election.

Basically media picks up on xyz party doing well and it becomes a self fulfilling prophecy.

This often leads to big blowouts once it's clear what major party has won. Often it's to give the incoming party more power vs the more extreme left/right parties.

Eg if you're a national voter (center right) you'll voter for Labour (centre left) or a centrist party (potential coalition partner for labour) to minimize the Greens (progressives) influence.

Last seen 2020 Labour got 51% of the vote in proportional.

When this happens the main opposition party collapse to 21-25% support.

You also know whose won before the votes are counted (except 1996).

 

[Arwon]

When studies were done regarding the Shy Tory effect, they found the polls were not properly representative IIRC. Meaning, they didn't call enough right wingers. Similar poor representation is thought influential in why Trump overperforms the polling.

Besides "people lie", there does exist another explanation: the pollsters find what they want to see, deliberately. I suppose if you wanna take it to exit polling, the way that would manifest would be the pollster choosing not to interview the gnarly, tattooed up guy in a leather jacket, choosing instead to interview the meeker dude in biege with a light brow.

Herding is a thing but also, polling can be hard. Stratifying a sample to be representative breaks down when a previously meaningless distinction within the population becomes politically salient.

We had an example of this with sustained, consistent polling misses for the 2019 election, where basically all the pollsters were about 2 or 3 percentage points away from the actual national two party preferred result for a good year or two.

Late swing wasn't a factor, there's potentially other issues, but a lot of the analysis concluded that they had failed to notice two new variables becoming meaningful. Those were level of political engagement, and education level. Evidently people started voting differently along both those lines, with less engaged and less educated voters (for a given age/gender/income/location) voting more for the Coalition than their exact peers, where previously those variables didn't matter to the result.