Understanding the 2024 Early Vote – Split Ticket

In the run-up to the 2020 Presidential Election, Donald Trump repeatedly questioned the security of early voting (particularly mail-in voting) and encouraged Republicans to vote in-person or on election day. His base, distrustful of state boards of election and concerned about interference, largely followed suit. 

At the same time, Democrats strongly encouraged their supporters to vote by mail. This had the twin effect of banking votes while also protecting voters from COVID-19. Buttressed by concerns about potential meddling by Trump-appointed Postmaster General Louis DeJoy, Democrats participated enthusiastically and amassed a strong early vote advantage.

In all, Pew Research found that over 2/3rds of Americans reported having voted early in 2020 (i.e., mail-in or early in-person voting), and only 27% reported having voted in-person on election day. But there was a sharp skew by partisanship: 37% of Trump voters claimed to have voted on election day, compared to just 17% of self-identified Biden voters.

This disparity in voting habits significantly polarized the ballot return trends observed in the 2020 cycle, establishing a pattern with staying power. Subsequent elections appeared to confirm a continued Democratic preference for voting early, with Republicans more inclined to wait until election day or (at least) vote early in-person.

Since 2020, however, Trump and the GOP have come to embrace early voting—including the use of mail-in ballots. Democratic success in the high-turnout 2020 contest, combined with the ease of voting by mail, seems to have convinced Republican leadership that early voting can be an effective weapon. After all, a vote banked is better than a vote planned, as it allows campaign teams to focus on voters who haven’t yet voted instead of contacting voters who already plan to vote.

This cycle, Republican messaging has concentrated heavily on getting the base out—and out early. A superficial look at the numbers in select states suggests that the GOP’s efforts to coordinate vote timing may be paying off. In states with party registration turnout data, the Republicans have been outrunning their previous benchmarks significantly, even taking a perhaps-temporary lead in key swing states.

With misleading early vote analysis proliferating across Twitter, we decided to compile and compare key data from the last three election cycles from the five key states which report turnout by party registration: Florida, North Carolina, Pennsylvania, Arizona, and Nevada. Our goal is to objectively present the numbers we do have without drawing definitive conclusions. 

Before we dive into the data, it’s important to reiterate (as we wrote about in detail last week) that early voting data is not particularly predictive. Persuasion can never be learned from early vote, whether a state reports turnout by registration or not.

The best example we can give you is from our previously linked article. In Maryland, Biden won mail by 66 points and in-person early votes by 6 points, but lost election day by 18 points. Applying those same splits to the 2022 electorate would yield a Wes Moore victory by just over 10 points; in reality, he won by 33 points because the “missing” Democratic mail votes were cast in-person—shifting the margins to the left among those subgroups.

Thus, comparing early voting patterns observed across states to those recorded in past cycles is only useful to the extent it contextualizes the broader picture. With all of those caveats in mind, we can also say that Republicans are currently enjoying better early turnout than they got in 2020 in multiple key swing states. That does not mean that Trump will win. It does, however, suggest that Democrats will have to come out in force in the few remaining days or even on election day to win—which was not the case in 2020. 

Every election day voter Trump and his team can convince to vote early is another one that they do not have to worry about as the campaign draws to a close. While Democratic turnout behavior stands to make for a more evenly matched election day vote, it also raises important questions about the party’s ability to fully match GOP turnout in the end. A strong Democratic ground game may well provide the boost needed, but it does put more pressure on the party’s apparatus and voters.

Additionally, any major moves to early voting among core Republicans means fewer will turn out on election day. It’s improbable for Republicans to match their 2020 election day surge in terms of turnout or margin, especially with many Democrats simply reverting to voting in person.

Florida

We start with Florida, a former swing state that has stayed comfortably in the Republican column throughout the 2024 cycle. Donald Trump carried the state by just 3.5 points in 2020. This year, our model projects that Republicans will expand their margin in Florida relative to 2020. Split Ticket currently rates the state Likely Republican in our model.

In 2020, nearly 80% of Floridians voted before election day—whether by mail or in-person. Republicans ended up with a 13-point party registration advantage among Florida’s 4.3 million in-person early votes; Democrats, meanwhile, maintained a 14-point advantage by party registration among the state’s 4.7 million returned mail ballots.

As in other key states, Democrats started off 2020 with a substantial turnout advantage of 22 points by party registration. By election day, this figure had narrowed to just 1 point—with significant movement beginning around two weeks before the election after the start of in-person early voting.

So far this year, 6.2 million early votes have been cast. That’s more than half the total number of votes cast in Florida in 2020, which clocked in at 11.1 million. Since the beginning of in-person early voting, Republicans have expanded their overall advantage in terms of party registration to almost 12 points.

Daily reports do show independents and Democrats slowing GOP momentum and closing the gap, but it is almost assured that the final early electorate will be redder than it was in 2020. It is important to remember that a much larger portion of Florida’s total votes will be cast on election day this cycle compared to 2020 (for reference, although nearly 80% of Floridians voted early in 2020, this figure dropped to 61% in 2022), but it is hard to see Democrats drawing even under any circumstances, given how sharply Florida has lurched to the right.

The implications of this are minimal for the presidential race. We don’t see Florida being particularly competitive in the vast majority of scenarios. However, there is a competitive House seat in the state, with Whitney Fox looking to unseat Republican incumbent Anna Paulina Luna in Florida’s 13th district. Democratic turnout on election day will be critical to the outcome in Florida’s most competitive House race.

North Carolina

North Carolina was the only one of the seven swing states this cycle to have voted for Donald Trump twice, becoming the third of the swing states to enter the Leans Republican column this cycle. North Carolina is also known for mailing mail ballots early, although this year’s target date of September 6th was missed due to a (successful) lawsuit by RFK Jr. to remove his name from the ballot.

In 2020, Democrats took advantage of early mailing to amass a turnout lead that lasted up until election day—when 4.6 million votes had been received at a top-line margin of around D+6. North Carolina still boasts a Democratic registration advantage, but more registered Democrats cross over for Republican candidates at the federal candidates than the reverse.

This, coupled with a Republican election day surge, allowed Donald Trump to win the state even as he slipped elsewhere, including in neighboring Georgia—which he had won in 2016.

In 2024, aided by registration gains, Republicans now have a narrow but appreciable turnout edge in the state, leading by half a percentage point with 3.9 million votes tabulated, or 70% of the 2020 total votes cast. With early voting closed on Sunday and Monday, there is limited time for Democrats to overcome this turnout lead, so whether the state flips will likely come down to how many votes the party can get out on election day itself.

Pennsylvania

Out of the three tossup Rust Belt states, our model currently gives Donald Trump the best odds at winning Pennsylvania, where he is only a nominal underdog. Joe Biden carried the Keystone State by just over a point in 2020, with a total of 6.9 million votes being cast. Of these, approximately 2.6 million votes—or 38%—were cast by mail. In 2020, Democrats finished with over a 40-point registration advantage among mail-in votes cast.

This cycle, the Democratic early voting registration advantage has narrowed noticeably and currently sits at around a 24-point edge. The shift suggests that more Republicans are voting early than in 2020 and 2022—when Democrats finished with a 48-point registration advantage out of 1.1 million early votes cast. At the same time, early votes now make up a smaller share of the total votes cast. In 2022, for example, only 20% of the total votes were recorded before election day.

On top of the fact that a smaller number of participating voters can be expected to turn out before election day than in years past, it’s also important to remember that the election day vote will almost certainly end up being much less Republican than in 2020. It would therefore be unwise to use shifts in early vote party registration advantage to claim that Pennsylvania is anything other than a pure tossup.

Arizona

Arizona was won by Joe Biden in large part due to his early vote lead being great enough to stave off Trump’s gains during the remainder of the vote count. Currently, because of favorable fundamentals and polls, Split Ticket has the state rated as Leans Republican, a flip from 2020.

In 2020, Democrats started with a strong turnout advantage of about 14% three weeks before the election, which faded as the election drew nearer and more Republicans voted. By election day, the votes cast by party came to a 2-point Republican advantage, with 2.9 million votes cast, or about 85% of the final turnout.

This did not translate into a Republican advantage headed into election day; in Arizona in particular a sizable portion of Republicans no longer supported the party on the presidential ticket despite their registration. Additionally, non-affiliated voters also skewed towards Biden, meaning this tranche of early voters favored Biden, and a late Republican surge on election day was not enough to carry the state for Trump. 

This cycle, the picture appears rosier for the GOP. Although many of the registered Republicans could defect once again, it is certainly more favorable for Trump that Republicans currently have a steady 8-point lead on Democrats in the early vote turnout, with a large sample of just over 2 million early voters as of writing.

Democrats will either need to hope for mass GOP defections or otherwise supercharge turnout to bring their base out to vote. With November 1st being the last day to vote early, Democrats may cut into the GOP advantage with outstanding mail ballots, but the clock is ticking and remaining mail ballots don’t particularly lean toward either party.

Nevada

The emblem of Republican strength on the early-voting front is Nevada. Geographically similar to neighboring Arizona, Republicans here depend on suburban votes, but have as of now developed unprecedented leads in the early vote by muting the Democratic advantage in Clark and Washoe counties while simultaneously building a strong edge among the rural voters of the state.

As of writing, Nevada Republicans have racked up a 5-point lead with about 930,000 votes cast, or about two-thirds of the 2020 turnout. In order to reach a raw turnout parity with the GOP, Nevada Democrats will need to win the remaining third of votes by ten points, but a longer “landing strip” will require a less-lopsided margin, though a shorter runway will necessitate big wins with any remaining mail or in the in-person votes, which also close on the 1st of November.

Were this the end of the story, we would have no problems in saying that polls were decisively understating GOP strength. The twist is that as of 2021, Nevada is now a full VBM state, with every voter receiving a ballot in the mail, and we have very good evidence that ballot return patterns are changing this year. In each of the last few days, Democrats have closed the gap on the GOP in terms of daily returns, and they will now have three days of mail votes to cut into the statewide Republican lead. Additionally, there will be “day-of” mail drop-offs on election day, meaning that we could theoretically have over 250,000 mail votes left to come in Nevada.

Finally, Nevada has recently implemented automatic voter registration, which has seen a sharp spike in the number of registered independents in the state. This population skews young and nonwhite and is likely to tilt more Democratic than in years past, which complicates the math further. For these reasons, the early voting analysis rules that held in years past may be misleading in the Silver State if applied this time around.

This is why we are skeptical of those peddling analysis supportive of a GOP landslide based on Nevada’s early vote. Our model has the state at a 50-50 split, as of last night, and we see virtually no evidence in early voting to deviate from this. It is possible that Trump could win the state comfortably, but the early vote is not a good baseline through which to judge this.

Major Takeaways

It is crucial not to read too much into early vote returns. The inability to account for persuasion effects, coupled with systemic variation between states, makes it impossible to craft defensible predictions based purely on early vote comparisons. The numbers presented here should not be understood to foreshadow a red wave. They demonstrate only that the GOP has successfully embraced early voting, and that they have done a very good job getting registered Republicans to the polls.

This is a better strategy for Republicans in terms of targeting than the one they adopted in 2020. The reason is simple and purely mathematical: a vote cast is more valuable than a vote planned, especially as not all those who plan to vote will vote.

The upshot is that Democrats will have to rely more heavily on the election-day vote to win next week in the aforementioned states. Despite that, we want to emphasize that a strong Democratic performance on election day is perfectly possible. In 2022, for example, Democrats generally fared better in terms of the election-day vote than they had two years prior—an important instance of depolarization. Ultimately, the presidential race remains a pure tossup and early voting does not change that picture.

Data Credits: VoteHub, Michael Pruser, TargetSmart