We are not going back

Here is what I'll be watching for on election night. And a prediction. . . sort of.

Nov 05, 2024

You don’t need me to tell you that this election is very close. Every single polling average has this as a statistical tie. Every swing state is within the margin of error. Anyone making bold predictions that they know what will happen on election night should be ignored.

This thing can go either way. Just look at this crazy map based on the final NY Times/Siena swing state polls:

For the record, I highly doubt this will happen tomorrow night, but according to the final polls from NY Times/Siena in the swing states, this is how it would shape up.

For my final Substack before the election, I will put on my best Steve Kornacki/Nate Cohen costume, talk about what I’ll be watching for on election night, and give a prediction of sorts.

1. Did Ann Selzer do it again?

Polling Queen Ann Selzer might have done it again. Ann Selzer is THE platinum standard of pollsters for those who aren’t as familiar with her. In partnership with the Des Moines Register, Selzer consistently produces the most accurate polls of Iowa.

In 2016, her final poll found that Trump led Hillary Clinton by 7 points in Iowa. This shocking result, initially dismissed as an outlier, became a harbinger of doom for the Clinton campaign. History repeated itself in 2020, when she again produced an outlier poll showing Trump leading Biden by 7 points in Iowa. Trump won by 9 points.

When Selzer produces an outlier poll, she is almost always correct, and the polling consensus is wrong. Here is her recent track record in Iowa:

2022 Senate: Poll R+12; Result R+12
2020 President: Poll R+7; Result R+8
2020 Senate: Poll R +4; Result R+7
2016 President: Poll R +7; Result R+9
2014 Senate: Poll R+7; Result R+8
2012 President: D +5; Result D +6

On Saturday, she released her final Iowa poll for the 2024 election. In it, she found that Kamala Harris led Donald Trump 47% to 44%!

Obviously, this is a shocking result and has potentially exposed how wrong the majority of polls might have been during this cycle.

If this recent Selzer poll is correct, that means it picked up on some major demographic shifts that other polls have not:

Harris is winning senior voters and leading senior women voters 63% to 28%
Harris is leading all women in Iowa by 20 points and Trump’s support among men has shrunk to just 12%.

If Harris is really doing well with older voters and has a large lead with women voters, that bodes well for states like Wisconsin, Michigan, and Pennsylvania.

Now, it is possible that Selzer is wrong this time or that her findings are isolated to Iowa and don’t extend to Wisconsin. But I’ll be watching tomorrow night's results to see if what she found in Iowa is accurate and translates to other states in the region. If Harris does actually win Iowa, the Trump campaign will be in a lot of trouble.

2. Georgia and North Carolina should help us understand what kind of night we will have.

Georgia and North Carolina are the first swing states to close their polls. Georgia should be pretty fast at counting this cycle, so we should get some interesting data to help us understand what kind of night we face.

If it looks like Harris is performing equal to Biden’s 2020 numbers in Georgia, that will tell us we are perhaps headed for a good evening. To assess that I’ll be looking at three counties: Fulton County, Baldwin Country and Washington County. Fulton is where Atlanta is and Baldwin and Washington are more rural areas that Biden won in 2020. Seeing how Harris is performing in those counties will help us understand:

Her strength with white voters and if there is significant slippage among African American voters.

As a reminder, Harris does not need to win Georgia to win the election, but how that state is turning out could indicate whether we are in for a very close race or not.

3. The Red Mirage and Trump Victory Claim

Despite the good polling news and general momentum behind the Harris campaign, the most likely scenario is that this election will be very close. Similar to 2020, we are also likely to see an early “red mirage,” where the early vote count reported in the key swing states skews more Republican.

THIS DOES NOT MEAN THEY ARE WINNING THOSE STATES.

Urban centers tend to take much longer to report the vote, which gives the early impression that Trump has a big lead in these states. But we should not panic until a state is officially called.

Donald Trump is going to take advantage of this and declare victory tomorrow night. He will do so before any of those states have been officially. I want to think that he will wait until enough states show early leads that equal 270 electoral votes for him, but honestly, he might just claim victory whenever he feels like it.

This will be his attempt to create confusion and chaos in the coming days and weeks, which will help him discredit the election.

Don’t buy it, don’t panic, let the process play out.

Prediction. . . sort of.

I’ll be honest: Before the Ann Selzer poll, I was cautiously pessimistic about the election. I’ve been dubious about the polls for a while now, but I couldn’t quite get myself to believe they oversampled Trump's support since the historical evidence has been the opposite.

Although it was possible that the polls overcorrected and undersampled Harris' support, it was more likely that they were making the same errors as in 2016 and 2020, and Trump would overperform the polls again.

But the Selzer poll has changed my mind and shifted me to a cautiously optimistic attitude. Let me be very clear: Donald Trump can certainly win this race. I do not feel confident about any outcome and am not making any predictions. Every scenario is still in play here.

But here is what I think are the three most likely scenarios:

Harris wins the popular vote by around 2 points; Trump wins the electoral college narrowly.
Harris wins the popular vote by around 3 points; Harris wins the EC narrowly (Blue Wall plus Nebraska 2nd).
Harris wins the popular vote by 3.2-4 points; Harris wins a large majority in the EC (Blue Wall, plus GA, Nevada, and maybe NC).

The returns from Georgia and North Carolina should eliminate or support that third scenario.