last updated 6 Nov 2024

Contents: Polling: | Outcome forecast | Beting on the outcome | Undecided Voters Could Make the Difference | Ethnic Community inflence | My speculation | Polling methods and accuracy | Ethnic Community inflence: | Links:

I'll post the demographics at 2024-Results when all the counting is done.
A month ago I posted in My speculation below. "In my humble opinion I'm afraid that biases, misogyny and xenophobia may have more to do with the outcome than democracy, the economy, immigration, abortion, international relations and other real issues."
I still think that had a big part to palay, but it was just one piece of the puzzle which I don't understand.


In most cases you can click on a poll chart below to go to the current version of it or find a web link to it above or below the chart.
The polls I list here all got it wrong, by underestimating Trump support.
The Washington Post does mention that if you apply compensate for the Trump underestimate in the 2020 polls to the 2024 polls, Harris would loose all the battleground states.

The underestimate for Trump could be for several reasons.
1. Higher turnout for young voters and their shift from traditional democratic bias to conservatives.
2. People don't answer polls or lie if they are voting for Trump.


About Polls:
People don't trust polls after most if not all polls predicted a large Clinton win over Trump in 2016.
Polsters have tried to improve since then.
The Washington Post polling team has "evaluated different survey methods, and only include those pollsters that apply the best ones."
But in 2020, the highest quality polls actually did slightly worse at predicting the election outcome than lower quality surveys. But on average (and over multiple elections), higher quality polls are generally more reliable and accurate."

In 2012 (Obama vs Romney) Obama did better than the polls.
In 2016 Polls gave Clinton a 4 to 6 point lead but she lost to Trump by 2 points.
In 2020 polls had Biden with an 8 point lead but he won by 4.4 points.
One explanation is that people who are going to vote for Trump either don't answer polls or lie.

See more on polling methods and accuracy below.

Polling:
Poll aggravators: They average polls from other sources.
270toWin and the Washington post both average multiple polls. Below is the results for battleground states
270toWin averaged 5 to 11 polls in each state. (varies by state)
Battleground States
Note: If you adjust by the amount Trump was undercounted in the 2020 polls, he would be ahead in all races.

Sorted by Electoral votes

The Blue Wall
The blue wall refers to the collection of states that had reliably voted Democrat in recent presidential elections through 2012. In a surprising victory, former President Donald Trump flipped three blue wall states in 2016: Pennsylvania, Michigan and Wisconsin. President Joe Biden won them back in 2020.
The three states account for 44 electoral votes, a large chunk of the 93 electoral votes that the seven battleground states make up.
Area Electoral
Votes
Oct 5 Oct 11 Oct 21 Oct 30Nov 4Wash Post Nov 4
Harris Trump 270toWin2024
Average
2 week
shift
National 49.6% 46.4% Harris 3.2%Harris 1.3%Harris 1.3%Harris 0.9%Harris 1.2%Harris +2
Pennsylvania 19 48.4% 47.9% Harris 0.5%Trump 0.1%Trump 0.3%Trump 0.6%TieHarris <1Trump +0.5
North Carolina 16 48.4% 49.2% Trump 0.8%Trump 1.6%Trump 0.4%Trump 1.4%Trump 1.3%Trump <1Harris +0.1
Georgia 16 48.4% 49.4% Trump 1%TieTrump 1.9%Trump 2%Trump 1.2%Trump +2Trump +0.3
Michigan 15 48.4% 47.4% Harris 1%Harris 0.1%Trump 0.4%Harris 1.9%Harris 1.8%Harris +2Harris +0.2
Arizona 11 47.9% 48.8% Trump 0.9%Trump 1.7%Trump 1.8%Trump 2.2%Trump 1.7%Trump +2Trump +0.4
Wisconsin 10 49.3% 48% Harris 1.3% Harris 0.3%Harris 0.5% Harris 0.7% Harris 1.1%Harris +2Harris +0.5
Nevada 6 49.6% 47.4% Harris 2.2% Trump 0.6% Trump 0.6%Trump 0.4% Trump 0.6%Harris +1Harris +0.8
Nebraska *
Distr. 2
Omaha
1 Harris was up 8-10% in the 2 most recent polls on November 4th
Iowa 6 Over the last weekend Iowa has become a swing state. They voted Republican in 2016 and 2020, but went Democratic in 2008 and 2012.
A Des Moines Register/Mediacom Iowa Poll on Nov. 2 shows. Harris leads Donald Trump in Iowa 47% to 44%,
270toWin still has Iowa for Trump 4.7% and Silver has Trump 4.4%
* Nebraska is split into 2 areas districts 1 & 3 with 4 electoral votes which is safely Republican and district 2 with 1 vote, which leans Democratic.

The above chart appears to be based on polls.

A copule of interesting scenarios:
Following is my simple calculation based on 2 scenarios.
Silver has a complicated model to predict the outcome of the election based on thousands of scenarios. See Below.

In the safe states Democrats have 225 electoral votes, Republicans have 219. Battleground states have 94.
So if Harris could flip Pennsylvania and hold Michigan, Wisconsin and Nebraska Dist 2 she would have 225+19+15+10=270 electoral votes (The mininum to win), Trump would have 219+16+16+11+6=268 electoral votes.

If Trump got Nebraska District 2 it would be a tie, 269 to 269..
In case of a tie each state would get 1 vote determined by the new house members in that state, so there are more states dominated by democrats, so Trump would win.

National Polls
In 2016 and 2020 the winner did not win the popular vote because of the way the electoral college works. Clinton had 48.2% of the popular vote and Trump had 46.1% in 2016. Democrats do very well in large states like New York and California, but the large margin doesn't help becasuse you get the same number of electoral votes whether you win by 30% or 3%.

See Presidential Elections for historic results by state.

In 2016 and 2020 the election winner did not win the popular vote because of the way the electoral college works. Clinton had 48.2% of the popular vote and Trump had 46.1% in 2016. Democrats do very well in large states like New York and California, but the large margin doesn't help becasuse you get the same number of electoral votes whether you win by 30% or 3%.


Outcome forecast


Nate Silver
Silver is an American statistician, writer, who analyzes baseball (players, teams ), basketball, and elections.
Around 2008 he started forecasting political elections at FiveThirtyEight.com. or 538

His recent election forecasting systems predicted the outcome of the 2012 and 2020 presidential elections with a high degree of accuracy. In 2016, his polls-only model gave Donald Trump, the ultimate winer, only a 28% chance of victory, although this was higher than any other forecasting competitors. He created his own election forecasting company Silver Buklletin in 2023 .

538
October 30: Harris 48.0% Trump 46.8%

The Economist
October 30: Trump 51% Harris 48%

Allan Lichtman's 2024 presidential election prediction. | ABC News
He predicts Harris will win.
Historian Allan Lichtman has accurately predicted the outcome of 9 out of the last 10 presidential elections, and he's done it without the use of polling data. Instead, he uses a set of 13 keys to make his predictions.

The keys include whether:

6 falses and white house party looses, so with 4 falses Harris wins.
Beting on the outcome PredictIt and Polymarket are platforms where you can bet on the results.
On October 28th the price of a Yes contract for Trump is 66 cents which will get you $1 and 34 cents for Harris which will get you $1.
Undecided Voters Could Make a Difference
A survey of Pennsylvania by Franklin & Marshall College in August says 15% are undecided.
Who are the Undecided Voters? | Franklin & Marshall College Poll
The Democrats in Pennsylvania are 95% sure of their vote. The Independents and Republicans are only 75% to 80% sure of their votes.

NY Times article on October 4 says 17% are undecided.
Donald J. Trump and Kamala Harris are starkly different presidential candidates. So why are so many voters - roughly 1 in 6 are still unsure of their choice?

Source: State of the Election | Vox Sept., 3

Oct 2 - A new national poll from Ipsos (A Public Opinion Specialist) found that former President Trump has the edge among undecided likely voters who do not support him or Vice President Harris. Ipsos found that when undecided likely voters are forced to pick a candidate, 53 percent of them chose Trump while 35 percent selected Harris. Another 15 percent skipped the question all together.


Ethnic Community inflence: Many Aribs in Michigan are refusing to vote for Harris out of anger over Gaza and Lebanon.

In Pennsylvania Ukrainian American are voting for Harris because Trumps position on support for Ukraine's war with Russia.
Polish people in Pennsylvania are not supporting Trump because of his relationship with Vladimir Putin, who is a feared danger for Poland.
Hispanics increase support of Harris after speakers remarks about Puerto Rico at Trump Madison Square Garden rally.

A 2020 report complicates simplistic narratives about race and the 2020 election |VOX

Why Latinos are turning to the Republican Party | CNN says,
"As Latinos settle and integrate, fewer are likely to think of themselves as immigrants, and may increasingly embrace an expanded sense of White identity in the way earlier light-skinned ethnic groups have over the course of US history."

The resons Democrats are loosing black support are more complex.
The support by younger backs is the thing that stands out. In the 18-29 age groups 62% lean Republican while 38% lean Democratic. The reasons Democrats are loosing black support are more complex.
The support by younger backs is the thing that stands out. In the 18-29 age groups 62% lean Republican while 38% lean Democratic.

Religion 2024

Gender Education 2016-2020:


In 2016 and 2020 the election winner did not win the popular vote because of the way the electoral college works. Clinton had 48.2% of the popular vote and Trump had 46.1% in 2016. Democrats do very well in large states like New York and California, but the large margin doesn't help becasuse you get the same number of electoral votes whether you win by 30% or 3%.
[One of my housemates when I was in grad school was doing his PhD thesis on the viability of the electoral college system and said it was the still the best system].

My speculation:
In my humble opinion I'm afraid that biases, misogyny and xenophobia may have more to do with the outcome than democracy, the economy, immigration, abortion, international relations and other real issues.
On the other hand I wonder if it will be like 1964 (Lyndon Johnson vs Barry Goldwater) where there was a big difference between the candidates in a lot of ways.
A History article "How Barry Goldwater Brought the Far Right to Center Stage in the 1964 Presidential Race" says,
"Goldwater was, without doubt, a divisive figure: Democratic detractors like Martin Luther King. Jr. and then-California Governor Pat Brown had compared his blunt, pull-no-punches rhetoric to Hitler. Within his own party, moderates responded with varying degrees of dismay or horror to his policies. On the eve of the 1964 GOP convention, the Republican governor of Pennsylvania, William Scranton, publicly released a letter he had written to Goldwater decrying the latter's "crazy quilt" of "dangerous positions," including his casual attitude toward the use of nuclear weapons."

LBJ won in a landside with 44 states and 486 electoral college votes..
There were only 6 states which went Republican, Alabama, Arizona, Georgia, Louisiana, Mississippi and South Carolina. because of, when it's usually around half.

in 2016 Trump won with 30 states and 279 electoral college votes.
Was 2016 and this year, as Yogi would say, Deja vu all over again?
Unfortunately it looks like No.

But this resulted in a pivotal change in American politics. The southern Democrats were the conservative power ever since they would not become Republicans because of Lincoln.
The South became Republican because of Goldwater.


Polling methods, bias and accuracy:
Accuracy:
In political polling, dozens of firms survey thousands of people in many different ways, from mail surveys, phone calls and online surveys.
Accuracy:

The Washington Post looks at over 20 Polls, most with with a sample size between 1,000 and 5,000
Some are: Reuters-Ipsos, NPR/PBS NewsHour/Marist, NBC News, ABC-Ipsos, Economist/YouGov, CBS News, NYT/Siena College, CNN, Wall Street Journal, ABC-Ipsos, CNBC, Quinnipiac University

Polls predicted a Clinton had a big advantage over Trump in 2016
In 2020 polls gave Biden a 4 point lead when actually won by 1.
In the 2022 midterms the democrats did better than the polls predicted.

Most sample sizes are around 1,000 up to 2,500 for NYT/Siena College.
Some history for Pennsylvanie from the Washington Post.
How the polls performed in Pennsylvania in previous elections
>
Election Final poll avg. (PA) Actual result (PA) Polls
underestimated
2020 Biden +4 Biden +1 Trump by 3
2016 Clinton +4 Trump <1 Trump by 5
2012 Obama +6 Obama +5 Negligible miss
If you added the undercounting margin for Trump in battleground states to the polls at the end of October, he would be ahead in all battleground states.
See Who is ahead in Harris vs. Trump 2024 | Washington Post

Polling Errors The Economist:

Polls have a history of being wrong:


Links:
2024 election Issues
Presidential Ratings.
Here's How The Presidential Polls Might Be Wrong | Forbes Oct. 2024
A guide to the 7 battleground states that could swing the election | Washington Post
What the Polls Are Really Saying | Politico
What 2020's Election Poll Errors Tell Us About the Accuracy of Issue Polling | Pew Reseaarch
Can we trust the polls this year? | Vox