Okay, we're finally getting to the point where there is enough early vote in several key states to begin to watch for emerging trends. So I'll kick off my first early vote thread of 2024 with a look at Pennsylvania.
One quick note before we begin - for the sake of brevity I will generally refer to all ballots cast before election day as "early votes", regardless of how they were cast (mail, early in person, etc). I may talk about mode from time to time, but generally just combine them all.
If you look at the vote reported in PA so far, Dems have a solid advantage in terms of party registration, though the gap is smaller than it was in 2020 at this same point, but that doesn't tell the whole story. More on that in a moment.
You'll also note that there are more total ballots cast reported so far than were at this point in 2020. I believe that has more to do with timing/reporting than anything else, but I'll dig deeper on that when I have time.
Now let's look at the early vote in PA by modeled partisanship. It shows a wider Dem lead than at this point in 2020. Why is that? The answer is simple - the model believes that the unaffiliated voters are more Dem than they were in 2020.
With our data we can look more closely at these unaffiliated early voters. This chart shows the racial breakdown of unaffiliated early voters at this point in each year. Notice how they are far more likely to be voters of color this year? The white share is 6 pts below 2020.
Let's go back to the total early vote numbers though. We can look at how these voters cast their ballot in 2020. Here we can see that an impressive 37% of GOP early voters cast their vote on Election Day in person in 2020, as compared to 11% of Dems.
Looking at gender, women are accounting for a larger share of the early vote in PA than they did at this point in 2020, and in-line with 2022, suggesting the post-Dobbs environment continues (as we have seen elsewhere an in other data).
Now, looking at the racial breakdown of women early voters in PA, we see the biggest increases among women of color, especially Black women whose turnout is 248% of their turnout at this point in 2020, compared to 146% for white women.
I'm not going to wrap this in a tidy package and say what we're seeing is indicative of a win for one candidate or another. We can't do that. I will say I expected to see more of a partisan gap closing with Republicans spending more resources on converting early voters.
While Republicans have converted a number of election day voters to vote early, Democratic turnout, especially among voters of color, has negated those potential gains for the Trump campaign. Stay tuned, more states to come.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
One important thing to remember as we continue to digest the election results, especially around swings with Hispanics: some of what we saw was due to turnout differentials, not just vote switching. We still don't have vote history for most states, but here is one minor example.
In Delaware, Hispanic voters were 4% of the turnout this year. In 2020, by party registration, the Hispanics who voted were +46.9 Dem, but in 2024 they were +36.7 Dem. The total turnout from Dem Hispanics was roughly equal to 2020, but 35% more GOP Hispanics voted this year.
Not surprisingly, this surge in GOP Hispanic turnout was most pronounced among younger voters. Under age 30 Hispanics moved by 17 pts to the GOP, while over age 75 moved only 3 pts.
I've been looking at the universe of voters in GA who cast a ballot in 2020, but didn't vote in 2024 and are still registered in the state. There are just over 1 million such voters. Not surprisingly, they are modeled as more likely Dems.
One of the biggest drop-off groups in Georgia was younger Black men. 132,036 Black men under the age of 35 voted in 2020 and are still registered there, 58,151 of them didn't vote in this election (44%). For comparison, the dropoff among young Black women was 30%.
I've also started looking at the question of what impact the bomb threats had on turnout. We need to look at this at a more granular level, but the targeted counties had drop-off percentages that were basically in line with the state average of 22%.
I spent a little time this morning looking over voter registration data over the course of the election in various states, looking for signs in retrospect of a Trump win. Looking at PA, something jumps out that I think is instructive as to the path forward.
Much of the conversation around the future of the D emparty/the need to fix the brand with "working class" voters centers around white voters. But something different emerges in the new voter reg data that speaks to the need for Dems to invest in organizing communities of color.
Look at the party registration comparisons in PA, first among voters of color, then among white voters.
Our team worked with the folks at Hart Research, in a project for PSG Consulting that produced some very interesting results. Both surveys looked at different angles of campaign advertising/media consumption. Here are a few highlights:
The Hart survey looked at swing voters in battleground states, who found the negative ads about Trump most memorable, followed by the positive Harris ads. Trump ads were less memorable (perhaps suggesting his broader media strategy, podcasts, etc, was much more effective?).
Reinforcing that point, swing voters were just much more likely to have seen a Harris ad. Perhaps this also helps explain why the swing to Trump was much smaller in the battlegrounds, relative to the rest of the country.
We won't have the full picture of turnout differentials for a while (states take weeks to months to publish Election Day individual turnout history) but looking at county data a bit of a pattern is emerging.
Outside of the battlegrounds, similar to 2022, Dem turnout seems to have cratered, and there were likely GOP persuasion gains, resulting in those bigger swings in states like NY, NJ, MA, FL, etc
Inside the battlegrounds, the turnout variations are much smaller, but generally turnout in Dem counties is running just slightly behind 2020, while GOP turnout is running just ahead. Couple that with slight persuasion gains, and you have the small shifts that won it for Trump.
Facing several somewhat overwhelming questions and realities this morning. First, why the data that gave me optimism that Harris could win was just wrong, or at least my interpretation of it was.
It will take some time for all of the data necessary to answer these questions is available. But there are many bitter truths to acknowledge this morning.
Of course, analytics aside, the most bitter truth is that America chose a convicted felon who ran a campaign focused on fear and division, bullying and attacking the most marginalized among us.