PA : Trump Wins - The Election Day Math of why

Trump is going to win by 3 points. That’s my call and I am sticking to it, but why?

Public Polling

The public and private polling is dramatically rosier for Trump than it has ever been before it either of his other races. His image is better, his ballot score is higher, etc. RCP has him at +0.3 but as I stated in my earlier article I believe there will be a polling error in his favor. Meanwhile Harris’s image and ballot scores have been moving in the wrong direction since September.

This day RCP average in 2020 : Biden +2.6 (He won by 1.2), in 2016: Clinton +2.1 (She lost by 0.7), but if we back up just a week before a bunch of GOP “pollsters” flooded RCP Biden was up 4.3 points. That’s basically where the ‘high quality’ folks ended at. I think we’ll see a shift towards Trump. Ultimately its happened 100% of the time he’s been on the ballot but the n=2 so pick your side.

Again the increase in response bias of Republicans and non-college voters that’s occurred these last two weeks (which even the NYT admits), makes me believe we will see a polling error. Fundamentally I do not believe pollsters solved it and it will manifest with Trump outperforming the polls.

Advantage - Trump

Electoral Composition

This to me is the big issue people are missing. The PA electorate by registration is likely to be R+0.2 to as much as R+3.0. That has never happened before and is totally unprecedented. With PA likely to have 87% of the vote plus being partisan either registered Republicans or Democrats, it’s very important. Consider this table that assumes Harris and Trump win their own bases by the same % (something which all the data roughly says is best case for Harris, frequently Trump wins slightly higher Reg Dem percentages).

Voter registration has changed a lot in PA and I know Democrats say “oh so many of these people were already voting Trump” and they are right but that actually makes this pretty easy to figure out.

In 2020 the electorate was at the time D+3.6, but if we look at it by how people are registered NOW it’s D+0.6. So the electorate only needs to slide 0.61 points right to get the first ever GOP plurality PA presidential electorate in modern times.

Just look at Philadelphia as an example. Democrats probably need 700,000 voters to turn out in Philly to clear the margins they need. So far they have 180,000 returned ballots so between whatever mail comes in today and tomorrow plus Election Day turnout they need 520,000 votes. That’s a lot.

Plus they are at 76% return rate versus the state as a whole at 82%, which again is not the tea leaf you want to read if you are a Democrat. This will be the first election where more registered Republicans than Democrats vote. It may be narrow but every scenario out there starts with at worst case R+0.2 electorate.

Advantage - Trump

Rural Turnout

We do not yet know what PA rural turnout will look like but we have a few places we can look to see. The first is PA mail where Republicans adopting mail has resulted in massive increases, but we have so few election cycles with mail its hard to say what this signals. However we do have states like FL, NC, GA, and NV where most of the vote is in already and the signals are very clear YUGE rural turnout. They still need more but all signs point to a big rural turnout.

Advantage - Trump

Mail

So I know I said I wouldn’t read anything into the mail but… what the heck its here and we have about 2.1M data points on it.

Data as of the day before the election

So Republicans are basically going to get their 2020 mail numbers but Democrats going to be 550k+ off their marks.

The inescapable fact is that Harris will have a SMALLER raw voter margin out of 2024 by hundreds of thousands of votes than Fetterman had in 2022. That’s astounding. Now mail ballots moved 19 points to the right so far, and I don’t think its fair to say the whole universe will move 19 points to the right so Election Day will be more blue but the question is how much more blue.

The GOP has pushed people to vote early so the Dems are talking up “canibalizing.” I am sure there is some truth to it but the data doesn’t make it look massive. The GOP has 229,000 voters who voted by mail now and in-person in 2020. The Dems have 86,000 so net 143,000. Over potentially 5 million votes that net impact is moving Election Day 2.86 points left.

The big question is about the 686k mail voters who voted in 2020 but have not voted yet. The GOP equivalent is 262k. If we look at the vote history and apply turnout probabilities to it it’s likely that about 101k of those Democrats won’t vote as compared to 41k Republicans. That is roughly a NET loss of 60k votes for Democrats compared to 2020.

There’s a lot of guessing and intuition here so as always we’re gambling but fundamentally I’d rather be the GOP that’s gaining in registration, mail, etc than the other party.

Advantage - Trump

Conclusion

The turnout, electoral composition, mail, and honestly polling all point towards Trump so to me you have to read the data and feel like Trump is going to win. However politics is full of surprises (2016) so stranger things could happen (2016), but ultimately Trump by 3 with McCormick by a similar margin.

Disclaimer: I work for Dave McCormick + Stacy Garrity + other PA Republicans so yes I am a GOP hack but I do think the data is pretty clear.