The Pennsylvanian Endgame

Road to November

Overview

It’s September.

As Labor Day rounds the corner and families across the country gather for their annual rituals, the endgame for 2024 is fast approaching. What started as a long-expected rematch between Biden and Trump ended up being turned on its head by possibly one of the wildest stretches of political twists and turns in the spring and summer of this year. With Vice President Harris taking over for the Democratic side of the aisle, both parties are campaigning vigorously up and down the ballot to claim one of the biggest prizes in the country - Pennsylvania and its 19 electoral votes.

Having said that, we still have quite a ways to go. For the mental well-being of almost everyone in the country and for you, dear readers, I’m going to explain how the timeline of the following ~9.5 weeks until Election Day will pan out as well as what happens on that day. And if it makes anyone feel better, I’ve been doing this for a few years now and I still feel like I’m flying blind. Electoral forecasting is tricky at best and with POTUS, congressional, Attorney General, Auditor, Treasurer, control of both houses of the legislature, and dozens of other races on the ballot, it’s going to be hard to keep up with everything.

Nevertheless, it’s important for everyone to be informed of the mechanics of how ballots will be cast in Pennsylvania this November. We have your typical in-person voting with standard polling station hours. And like most states, we have early voting, but there are some unique characteristics regarding it that you should be aware of. Counting has gotten easier over the years, but the lack of precanvassing per state law does not allow early votes to be processed accordingly - counties must do so on Election Day itself, which has presented challenges. In spite of that, the Department of State has pushed for ways to streamline the ballot counting process.

But while that situation unfolds, we have to look at the bigger picture. Namely, the endgame and the two major factors that lead to it - the things happening with the early vote and what occurs on Election Day. The patterns from both shape how the state goes. Buckle in and read along!

The Early Vote and the Firewall

As we speak, vote-by-mail (VBM) requests are pouring in - Pennsylvania’s version of early voting. Major counties are reporting solid numbers such as Allegheny (152,000+), Montgomery (110,000+), and Philadelphia (130,000)+. For those of you who have been following me long enough, you know that in the lead-up to Election Day, I typically start analyzing races based on those numbers. VBM ratios and splits are remarkably stable in Pennsylvania (depending on which metrics you’re looking at), thus making such a task doable. Those ballots are then reconciled with projected Election Day turnout and splits in order to forecast who’s going to win as well as a relatively good idea of how much they’ll win by.

Using this approach, I was able to determine that Fetterman (2022) and McCaffery (2023) would win and by comfortable margins, which they ended up doing. Bolstering this was the usage, among many other things, of a firewall metric. This concept has evolved as election cycles went on, but the initial version relied on a number-based benchmark being a threshold where, if a candidate crossed it, would virtually assure their victory when weighed against Election Day turnout. Driving this were some very eerie data-backed affirmations of this being viable. For example, in 2023, I estimated that McCaffery would need a ~400,000+ vote VBM firewall to be comfortable with winning his Supreme Court race. Not only did he hit it, but the percentage he exceeded it by was roughly around the margin he won by. It was a similar case for Fetterman.

I, however, have to stress that voting patterns change and this benchmark will also change - numbers and split wise. This year, we are facing two major races upballot instead of a marquee one; Harris v. Trump on the POTUS side and Casey v. McCormick for PA SEN. Subsequently, two different benchmarks/firewalls are going to have to be created. The factors involved in this November’s environment will go toward building them up, such as deltas between the two races, changed ratios/splits from years past, changes in the volume of VBM, etc. This is why I caution that if the GOP sees an uptick in VBM for them, it may not necessarily be a good thing - refer to the following section for why. On the flip side, I haven’t seen much in the way of meaningful gains for them on the mail-in front so far.

That said, we have until October 29th for Pennsylvania voters to request VBM. The earliest those ballots can go out is September 16th, but most counties are targeting more toward the end of September or early October. A few court challenges are wrapping up and some last-minute adjustments are being made, then they’ll be good to go. Once they start hitting voters’ mailboxes, we will begin getting updates of return rates and numbers as October goes on. You’ll see the reappearance of my handy maps, hotspots, splits, numbers to look out for, firewalls, and more. Typically, the last week of October or early November is when I start getting a really good idea of how certain races are going to end up, unless some happen to go down to the wire. But for everyone’s sanity, I truly hope we don’t get to that point.

Pennsylvania is Pennsylvania, though. On the POTUS level, whoever takes this state has a projected 90%+ chance to win the election. The stakes could not be higher for either major party. Because of that, it is so important to understand how the early vote works here and what to expect as we start receiving numbers. The firewall is not the only metric we’re going to look at or measure when it comes to calculating the odds and strength (or lack of) of candidates up and down the ballot. In fact, there’s often a misunderstanding about how the firewall works. It’s not just some random threshold where a candidate can take a victory lap. As mentioned earlier, it’s more of a numerical point propped up by a lot of other factors. And, as always, it is subject to change in its usage and/or what it can do analysis-wise.

In short, there’s a lot to consider about early voting here and what it can offer. All of that information will be collected over the next two months and put up against Election Day votes, which brings us to the second (and last) major component of our endgame - the coveted day itself. The day where dreams go to die, careers are secured or destroyed, and the fate of a nation hangs in the balance.

Election Day

As has been the case for election cycle after election cycle, the polls open at 7 AM sharp on November 5th and close at 8 PM. I expect to be able to track turnout in certain counties as the day progresses, Philadelphia chief among them. Those figures will be particularly important in determining how overall performances are holding up relative to 2020 and other elections as well as certain splits that are necessary for candidates for either party to achieve victory. For instance, if we see lagging urban turnout beyond what is considered the typical norm, it may bode poorly for Democrats. Inversely, if suburban turnout matches or exceeds 2020, Democrats may be able to afford a little bit of slippage in some urban precincts if the latter happens. Republicans will want to see 2020-level turnout *and* margins in the rural counties, particularly those in the T to stay competitive. If that starts to falter, their chances of victory will drastically diminish. There are a lot of different margins in certain areas that are possible in this November’s electorate, but one thing is for sure - as in 2022 and 2023, the Philadelphia collar counties and Allegheny will be the center of gravity. The benchmarks for either party in urban and rural areas can go up or down depending on what the aforementioned counties dictate turnout and split-wise.

Other than that, there is a misconception that I must dispel with Election Day projections, particularly as the votes start coming in after poll closing. Because of what happened in 2020, the winner of the state on the presidential level couldn’t be called for several days, bolstered by COVID-era delays and the need to count and process a drastic number of mail-in ballots for the first time following the passage of Act 77 by the Pennsylvania legislature. Consequently, some people are left with the impression that the same thing is going to happen again this cycle and thus create a space for bad-faith actors to stoke chaos.

Not so. Unlike 2020, most Democrats will not be casting ballots by mail. Instead of the 2 million votes and change of the roughly 2.6 million mail-in ballots cast for them that year, we are looking at ratios that more closely resemble that of 2022 and 2023 in addition to the raw totals themselves. Per current projections, we are likely on track to slightly exceed 2022’s VBM topline (a little over 1.2 million) - subject to change. It depends on the return rates and home stretch spikes, but that is the current target. Even if it ends up being more, of which I’m inclined to think so, the volume of mail to be processed and counted will be drastically less. You may recall that the last two election cycles went relatively smoothly and were called in a matter of hours. I expect the same to happen here unless certain races end up being decided by a point or less, in which case it may take another day or two, maybe more. Depends on the situation.

Additionally, be warned - there will be a very different pattern to how votes come in on Election Night. In 2020, there was a gradual red-to-blue shift as Republican-dominated Election Day votes came in first, followed by VBM in the reverse, which were overwhelmingly favorable to Democrats. Now, because of the lower VBM volume and the transition of a good portion of blue-leaning votes back to Election Day, I would expect a reverse shift to occur. Allegheny County has become remarkably good at mail processing and now typically drops VBM immediately upon poll closing, which will give Harris a wide lead in the state early on in the night. This will be followed by more mail drops in other counties and an initial trickle of Election Day votes, which are no longer as GOP-friendly as they were before.

Because of this new phenomenon (which was observed starting in 2022), Republicans should not expect to see a sea of red right off the bat. Instead, the overall vote will slowly get redder and redder with blue spikes in between as southeastern Pennsylvania’s initial batch of VBM starts dropping. With VBM still being quite blue in general bundled with a bluer Election Day vote, there will be a realignment of which party will mostly be playing catchup throughout the night. I suspect that bad-faith actors are going to find it difficult to make spurious claims with this pattern, but I have no doubt they’ll try anyway.

Could be a long night. Could be a short one. We’ll see!

Final Thoughts

Now that you understand the dynamics of the early vote and Election Day itself, my advice to you is to not panic. Stay calm and steady your course. The numbers will come when they come. One way or another, somebody’s going to win.

Of note: people have asked me who I think is favored to win the POTUS race at this juncture. Personally, it’s too early to be definitive about anything, given that we still have a lot of major events that can reshape the race over the next few months. On top of that, it’s going to be another month or so before we start seeing initial returns/splits from VBM and that could shake up projections as well.

Those disclaimers aside, my current feel based on what I see numbers-wise right now hovers around Harris+2 with Casey over in PA SEN winning by 7-8. Funnily enough, if you are a believer in the Washington State primary’s ability to forecast the environment of November elections, those results would be more or less in line with that, so keep that in mind.

In accordance with those results upballot, Dems would likely be favored to win the Attorney General race and the State House. Auditor and Treasurer would be more of a knife fight and the Senate likely stays in GOP hands barring a shakeup in SD-49 within Erie County. All congressional incumbents should survive in that kind of environment, though Cartwright in PA-08 and Perry in PA-10 will be races to watch.

But then again (again), nothing is set in stone. Maybe the general sense will be Trump+1 in a month. Perhaps Harris+5. We may see some fluctuations in the downballot races as well. Who knows? Embrace uncertainty.

Stay tuned for incoming real numbers… Until next time.