How I got PA prediction right

Sorry it’s been awhile, but with the US Senate recount and then taking the kids to a much deserved Disney trip post-election, I haven’t had time to write much. That being said the first thing I wanted to answer is why my prediction of Trump +3, McCormick +1 ended up being right. (Not to brag tooo much)

Our Internal Polling

The immediate answer is our internal polling. The great folks at POS as well as our internal text-to-web tracker both showed this as the likely outcome. Our final surveys always showed Trump leading and Dave closing fast, in fact our final text tracker had Dave up one point on the Thursday before the election.

Why was that different than the public surveys? The answer is you guessed it… response bias. Our surveys worked hard to avoid oversampling liberal boomers and highly educated voters. The public surveys did not.

Without robust protections against these oversampling issues surveys are always going to be too Democratic especially in Presidential years.

My message to clients is always ignore the public polling and only pay attention to the polling that you pay for because that’s what we invest in doing right. Gene Ulm, our pollster, helped us to nail it right on the nose.

But it wasn’t just our polling as I laid out in my prediction post there were A LOT of other data points pointing to the GOP rolling up and down the ballot.

Philadelphia

My hot take before the election is Trump would break 20% in Philadelphia, which he did. I’ve always said if you can break 20 in Philly you can’t lose.

The warning signs for Democrats in Philadelphia have been going off for several cycles now. It was one of the few counties where Trump improved from 2016→2020 and even the voter registration data in the county has been relatively good for the GOP. Our internals frequently showed historic numbers for Republicans in the county.

And on top of all that, the vote by mail returns in Philly were some of the worst for Democrats in the entire state.

A 50,000 vote decline in Philadelphia for Democrats is no good, but the future news for them is worse. This decline has been balanced out so far by Democrats improving in the suburban areas. There’s a lot of evidence that things can get a lot worse for Democrats in Philadelphia BUT there’s also evidence that the suburbs are approaching the likely low point for Republicans. If they lose another 25k or so votes in Philadelphia in 2028 it is not clear to me where they can find those votes.

(Caveat: Miami Dade was moving hard left up through 2018, before its made a sharp right turn. Trends can and do change but the Democrats in PA NEED to reverse the Philadelphia problem quickly)

Note to the press: Democrats spent all Election Day knowingly LYING To you about Philadelphia turnout saying numbers were historically high, already over 2020, etc. It was all lies. You guys bought it. Do better next time.

Voter Registration

The second major telling point is the massive change in voter registration that has occurred. After my first major race I managed in PA in 2008 the GOP faced a massive 1.2 million voter registration disadvantage. The last 16 years have seen this number dwindle to under 300,000. This was public data and other folks like Patrick Ruffini had been talking about this for some time. The writing was indeed on the wall.

Vote By Mail

The last feature was the change in vote by mail. A lot of Democrats tried to say that the surge in the Republicans were a ‘mirage’ and ‘time shifting voters,’ and to some extent they were right, but what was missed by a lot was the ‘disappearing’ Democrats. We won’t know til the voter file posts but well over 600,000 mail voting Democrats from 2020 did not vote in the mail in 2024 and some real number will have not voted at all. This was a canary in the coal mine in voter intensity. It was clear that Democrats had a turnout problem and they did not in fact fix it on Election Day.

Will Pennsylvania become Ohio?

There’s been a lot of talk online that maybe PA is going to trend hard right. While it is possible, I think the reality is PA has largely been a swing state for 100 years and is likely to continue to be a swing state. Republicans wisely invested in massive voter registration, vote by mail, turnout and persuasion campaigns. We got outspent massively by Dems but this was one of the states where the spending gap was the smallest. We have to keep it up. While there’s a lot of bright spots in the Southeast (who ever thought we’d be saying that about Republicans in Philly again) and North East, Allegheny county looms as a real Republican problem to be addressed.

2026 is a different year and our voters will need to be convinced to turnout. We can make PA red, we can beat Josh Shapiro, and we can retake the state house but it will take a concerted effort the next 24 months to make that happen.