Were there any red flags that indicated the race would be closer than polling suggested?

[Tekken_Guy]

Where there any red flags you saw that indicated the race would be close and not the landslide polling said it would?

[Hope For A New Era]

Well, there's the classic "YouTube comment method," but I dismissed the high volume of MAGA comments as a cultural difference between the supporters. It turned out to be more than that.

[Beet]

The massive boat rallies of thousands of boaters that Trump got. Boaters statistically are a middle class cross-section of society.

[支持核绿派 (Greens4Nuclear)]

Aside from pre-COVID personal anecdotes from friends, coworkers, and Reddit- for me it was

1) WA state referendum 88's narrow defeat in 2019, even though Yes got the lion's share of endorsements. Not related to the GE but it got me thinking about non-response bias
2) Biden's surprise comeback on Super Tuesday as the least "progressive" and the most "populist" candidate
3) Trump doing absurdly well in Univision's post-debate polls
4) this 2020 graphic of Asian American voter identification patterns from 2012 to 2020 (source).


There were also these Feb 2020 Atlas threads:

• Is the Iowa turnout a bad sign for democrats chances?
• Why is Sanders under performing?
• Why did gop turnout balloon in nh?

[sting in the rafters]

Trump consistently getting 80%+ support among conservatives. The problem is that most pollsters do not weight by ideology, which is the single most important determinant in voting preference. For example, the final YouGov poll had a top-line of 49-40 with registered voters and 53-43 with likely voters.. However, the sampling on this poll was hilariously skewed with liberals outnumbering conservatives. hThis is a drastic contrast to Gallup's polling data which has long-held that conservatives make up 36-38% of adults, liberals make up 24-26%, and the remaining moderates are 36-40% of the population. Say we split the difference and have a split of 37-25-38. That gives a margin of 47.5-42 and 51-45, respectively. Both obviously are much more in line with final of 51-47.



I'm inclined to buy this line of reasoning because it fits the narrative around both candidates. Biden had the highest appeal of any candidate in the last 8 presidential election cycles among moderates, even more than Obama's in 2008. The reason Obama's win was bigger was because he had higher liberal turnout+actually won a high percentage of conservatives for a Democrat. The popular vote remained closer than expected because Trump won conservatives by a historically large margin, fitting the explanation that polls were missing conservatives due to COVID, not passing likely voter screens, etc.


This also falls in line with the partisan splits. Poll screens tended to have more Democrats than independents, whereas most partisan surveys of the nation at-large tend to have independents slightly outnumber Dems (Gallup+Pew for example).

Much like 2012 when most pollsters failed because they underestimated Obama's GOTV operation, I think the most likely explanation is that Trump's GOTV operation got more conservatives to the polls than anticipated.





Tl;dr:poor turnout models didn't factor for high GOP enthusiasm, or as I like to call it, "variable-T"

[ctrepublican512]

Covid lockdown backlash coupled with school re-openings

[Annatar]

Voter registration trends suggested a closer race than the polls.

For me personally, the main reason I believed Trump would do better than the polls is that in most states, in particular the Midwest, Republicans had outperformed state polls 3 elections in a row, 2014, 2016 and 2018 and there was no sign pollsters had changed their methodology to remove this bias. Hence I believed Trump would do better in states with a lot of non-college whites than the polls were suggesting just as Republicans had done in 3 previous cycles in a row.

I also currently believe Republicans will outperform polling in states with a lot of non-college whites in 2022 just as they have for 4 cycles in a row.

 

[Motorcity]

This might be a personal thing, but the unusual high number of undecided

Biden was never leading by 8 points

The polls were usually

Biden: 51-53%
Trump: 43-45%
Undecided/Not sure: 4-7%

Everything from my personal experience was "undecideds" were usually Trump supporters who didn't want to look racist.

In my opinion, any poll that has a high number of undecided voters is not a reliable snapshot of the ground.

[Oregon Eagle Politics]

Yes, The Biden Campaign said that they were not up double digits and that the race was a nail-biter, and that the polls were inflated.

There is also the fact that Biden didn't try in AK/MO/TX/IA where the polls showed a close race/dems leading.

And the fact that polls underestimated 2018 Rs in the lower midwest and florida.

https://www.ntd.com/we-are-not-ahead-by-double-digits-biden-campaign-manager-reportedly-admits-polls-inflated_518308.html

[Amenhotep Bakari-Sellers]

Lol, there wasn't an Insurrectionists, we haven't had a natl Election since the Insurrectionists, we just have to wait until after Midterms to find out, but Trump hurt R chances, not helped it

[Obama-Biden Democrat]

Trump consistently getting 80%+ support among conservatives. The problem is that most pollsters do not weight by ideology, which is the single most important determinant in voting preference. For example, the final YouGov poll had a top-line of 49-40 with registered voters and 53-43 with likely voters.. However, the sampling on this poll was hilariously skewed with liberals outnumbering conservatives. hThis is a drastic contrast to Gallup's polling data which has long-held that conservatives make up 36-38% of adults, liberals make up 24-26%, and the remaining moderates are 36-40% of the population. Say we split the difference and have a split of 37-25-38. That gives a margin of 47.5-42 and 51-45, respectively. Both obviously are much more in line with final of 51-47.



I'm inclined to buy this line of reasoning because it fits the narrative around both candidates. Biden had the highest appeal of any candidate in the last 8 presidential election cycles among moderates, even more than Obama's in 2008. The reason Obama's win was bigger was because he had higher liberal turnout+actually won a high percentage of conservatives for a Democrat. The popular vote remained closer than expected because Trump won conservatives by a historically large margin, fitting the explanation that polls were missing conservatives due to COVID, not passing likely voter screens, etc.


This also falls in line with the partisan splits. Poll screens tended to have more Democrats than independents, whereas most partisan surveys of the nation at-large tend to have independents slightly outnumber Dems (Gallup+Pew for example).

Much like 2012 when most pollsters failed because they underestimated Obama's GOTV operation, I think the most likely explanation is that Trump's GOTV operation got more conservatives to the polls than anticipated.





Tl;dr:poor turnout models didn't factor for high GOP enthusiasm, or as I like to call it, "variable-T"

Liberal turnout went down and conservative turnout went through the roof from 2016. Trump's hidden deplorables were real. What saved the country was moderates voting for Biden by 64%-34%. Moderates are disgusted by the authoritarian MAGA cult. The Democratic party is a coalition of moderates and liberals, while the Republicans are a extremist party.

[Bootes Void]

Trump consistently getting 80%+ support among conservatives. The problem is that most pollsters do not weight by ideology, which is the single most important determinant in voting preference. For example, the final YouGov poll had a top-line of 49-40 with registered voters and 53-43 with likely voters.. However, the sampling on this poll was hilariously skewed with liberals outnumbering conservatives. hThis is a drastic contrast to Gallup's polling data which has long-held that conservatives make up 36-38% of adults, liberals make up 24-26%, and the remaining moderates are 36-40% of the population. Say we split the difference and have a split of 37-25-38. That gives a margin of 47.5-42 and 51-45, respectively. Both obviously are much more in line with final of 51-47.



I'm inclined to buy this line of reasoning because it fits the narrative around both candidates. Biden had the highest appeal of any candidate in the last 8 presidential election cycles among moderates, even more than Obama's in 2008. The reason Obama's win was bigger was because he had higher liberal turnout+actually won a high percentage of conservatives for a Democrat. The popular vote remained closer than expected because Trump won conservatives by a historically large margin, fitting the explanation that polls were missing conservatives due to COVID, not passing likely voter screens, etc.


This also falls in line with the partisan splits. Poll screens tended to have more Democrats than independents, whereas most partisan surveys of the nation at-large tend to have independents slightly outnumber Dems (Gallup+Pew for example).

Much like 2012 when most pollsters failed because they underestimated Obama's GOTV operation, I think the most likely explanation is that Trump's GOTV operation got more conservatives to the polls than anticipated.





Tl;dr:poor turnout models didn't factor for high GOP enthusiasm, or as I like to call it, "variable-T"

Liberal turnout went down and conservative turnout went through the roof from 2016. Trump's hidden deplorables were real. What saved the country was moderates voting for Biden by 64%-34%. Moderates are disgusted by the authoritarian MAGA cult. The Democratic party is a coalition of moderates and liberals, while the Republicans are a extremist party.

liberal turnout didnt go down whatsoever, Biden got 17m more votes than Clinton.Anyways I think alot of people look at fascists and sjw wokies as the same rather than differntiate

[If my soul was made of stone]

A decent amount of poll crosstabs showed Trump doing better with nonwhite voters, which was concerning despite the fact that they generally showed a very inflated Democratic white vote that pointed towards a landslide.

[Interlocutor is just not there yet]

A decent amount of poll crosstabs showed Trump doing better with nonwhite voters, which was concerning despite the fact that they generally showed a very inflated Democratic white vote that pointed towards a landslide.

Yeah, there were all those "Is Trump really winning Florida Hispanics?" articles from places like CNN during the last month or two.

I admit that I chalked it up to "typical" pollsters undersampling Hispanics, even though it seemed to be a bit on the money in places like Florida & Texas.

[DaleCooper]

The generic congressional ballot always made me uneasy.

I don't know if this is a red flag about the election specifically, but in hindsight there was also a lot of generally unprofessional behavior among many of the big polling agencies that should've been red flags about their credibility. The fact that they'd advertise their new polls directly to "election Twitter" like they were trying to sell New Coke or something was odd to me, and the way many of them directly smeared Ann Selzer because her poll was different from their consensus also seemed classless. People like Rachel Bitecofer cynically trying to make names for themselves as election gurus were also a dime a dozen, which makes it tougher to trust people's credibility, though to be fair that was happening on both the left and right.

[Adam Griffin]

Every set of crosstabs for 6+ months prior to ED measuring their opinions...

(from September)

[Motorcity]

According to the book “Lucky”, the Biden campaign internal polls showed a much closer race. That’s why they sent both Biden and Obama to Michigan those last few days. They also knew they were never competitive in Texas, Ohio, and Iowa.

The week before the election, they gave up on Florida. They did not expect to carry it by Halloween. To their credit, they did predict winning Georgia.

At the end of the day, the only state they were wrong about were North Carolina but they expected it to be a toss up anyway

[sguberman]

According to the book “Lucky”, the Biden campaign internal polls showed a much closer race. That’s why they sent both Biden and Obama to Michigan those last few days. They also knew they were never competitive in Texas, Ohio, and Iowa.

The week before the election, they gave up on Florida. They did not expect to carry it by Halloween. To their credit, they did predict winning Georgia.

At the end of the day, the only state they were wrong about were North Carolina but they expected it to be a toss up anyway

Although even they were surprised by Miami Dade

[Drew]

The Selzer poll in IA a few days before Election Day.

[Orser67]

Some of these have been mentioned by others in this thread, but these are the things that I would consider to be red flags:

• Many national polls showed a tightening of the race between mid-October and late October/early November
• Selzer poll
• Relatively weak Democratic performance in Washington's top-two primaries (hat tip to Sean Trende)
• Relatively weak Democratic performance in 2020 special elections (iirc both Trende and 538 had noticed this)
• The Biden and Trump campaigns generally treating the race like it was still competitive (at the time, and arguably even now, it wasn't clear if they really believed it was a closer race than polls showed or if they were simply being cautious/optimistic)
• Biden's fairly weak performance with Hispanic voters in the primaries

One other thing that I think is worth mentioning, and that I still don't completely understand, is the huge number of people (not all of them conservative Republicans) who thought that a Trump victory was inevitable. Obviously they were wrong, and I think in many cases they were just politically unsophisticated people reading too much into the last election, but I wonder if some of these people were legitimately picking up on something that media/polling wasn't.

[Pericles]

One important thing was that the special elections didn't show great swings to the Democrats like they did before the 2018 midterms, but did actually indicate a more neutral environment. Mike Garcia's 10 point win was particularly alarming. Still, it was not too hard to just handwave it by saying that even if that was meaningful, public opinion by election time had swung against Trump. Unfortunately, it just wasn't true.

With the House, I did always have a niggling doubt that people were going a bit wild with their predictions. If the Democrats were supposed to win the House popular vote by about the same or less than in 2018, how were they supposed to improve in pretty much every swing seat and make like 20 net gains? Democrats did distribute their vote better in the House than in 2018. However, on a uniform swing, Democrats would have only gotten a few more seats than in 2018, not even 240, if the generic ballot was exactly right. And even if they had gotten a D+8.6% margin, they would have only gotten 241 seats (disturbingly, Republicans got that exact seat total in 2016 with a 1.1% popular vote win).

[Non Swing Voter]

The Selzer poll in IA a few days before Election Day.

This.  When I saw that it felt like things were reverting back to normal/2016 numbers.

[Motorcity]

I know this isn't too relavent to the discussion, but is it weird that I felt better as election day approached?

Throughout the campaign season, I was a nervous wreck. I was scared it was going to be a repeat of 2016. But I am just a overly cautious person.

But on November 2nd, I did a poll. I teach 6th grade Social Studies. The results were 39 votes for Trump and 35 for Biden. Mind you, I live in a small southern town, almost all white, and poor. The principal said he would have expected 75-90% for Trump when I told him. Such a close margin in the heart of Trump country made me feel better.

Another thing. All the people I knew who did not vote because of #bothsides or voted third in 2016 were voting for Biden. This was the key to a Biden victory, consilating the ampathy vote of 2016

[Motorcity]

And even if they had gotten a D+8.6% margin, they would have only gotten 241 seats (disturbingly, Republicans got that exact seat total in 2016 with a 1.1% popular vote win).

This is disturbing. A huge part of this is gerrymandering, but the system favors republicans in widespread districts not Democrats in jam packed cities

[Figueira]

The fact that Biden was spending a lot of time and energy on Michigan despite the public polls showing it as a guaranteed win.