Nearly 62 million Americans had cast a ballot by early afternoon on October 31, according to the University of Florida Election Lab, with five days left until Election Day, 2024. Early voting in-person and by mail has already surpassed the early vote totals for the 2012 and 2016 elections and may yet approach or surpass the share of early votes cast in 2020. “The meta-trend overall is a dip in the number of mail-in ballots that have arrived,” with more early voters choosing to vote in person, said Matt Carpenter, director of Family Research Council Action, on “Washington Watch” Wednesday.

“Voters really like that early, in-person option … going into the voting booth, filling in your bubbles, handing in your ballot,” Carpenter continued. “Voters are responding to that option, and they’ve been telling pollsters for a while now.”

While early votes are not tallied until Election Day, some states do compile data on the party registration of early voters. “We don’t know how these voters are voting, but you can draw some inferences that you don’t really get with standard, head-to-head polling at this stage,” Carpenter explained. “Typically, Republican voters prefer to vote in person, and Democrat voters tend to vote by mail.”

Based on early voting party registration information, “the Republicans are in a much better position this time around than they were in 2020,” Carpenter described. “The Democrat[ic] margin is shrinking a little bit here. We’re seeing Republican enthusiasm across the board.” In the 2020 election, 44.8% of early votes were cast by registered Democrats, and 30.5% were cast by registered Republicans. So far in 2024, 38.8% or early votes have been cast by registered Democrats, and 36.1% have been cast by registered Republicans.

Carpenter suggested this was “indicative of, maybe, Republicans getting some confidence in voting early and kind of getting the idea.” Namely, if “you vote early, you bank your vote, and then, the campaign can spend resources looking after low-propensity or no-propensity voters.”

More important than national trends, however, is the data coming out of seven swing states, where polling of the candidates has remained the closest throughout the entire cycle. These states happen to fall into three geographic regions, with Arizona and Nevada in the southwest, Georgia and North Carolina in the southeast, and Michigan, Pennsylvania, and Wisconsin in the Rust Belt. The results from these seven states will determine the results of the 2024 presidential election.

Let’s take these states in alphabetical order.

Arizona

So far in 2024, 1,826,892 early votes have been cast in Arizona, 75% of the 2020 total, according to the Election Lab’s latest data. These are all mail-in ballots, since Arizona does not offer in-person early voting. So far, 755,131 registered Republicans have voted (41.3% of ballots cast), compared to 616,680 registered Democrats (33.8% of ballots). In 2020, registered Democrats and Republicans each comprised 37% of the vote.

Counting only by party registration, Arizona Democrats held an advantage of 9,633 votes before Election Day in 2020. So far this year, Arizona Republicans hold an enormous advantage of 138,451 votes.

Georgia

So far in 2024, 3,482,352 early votes have been cast in Georgia, 87% of the 2020 total and perhaps on track to surpass the 2020 early vote tally. Only 214,689 of those votes are mail-in ballots, while 3,267,663 are in-person votes. By comparison, Georgia’s early vote in 2020 included 1,320,154 mail-in ballots and 2,694,763 in-person ballots. That is both a substantial decrease in mail-in ballots and a substantial increase in in-person early voting.

Georgia does not record the party registration of early votes. However, it does record age category, sex, race/ethnicity. So far, early voters in Georgia this cycle are most likely over 40 (75%), female (55.6%), and white (58.8%). Georgia also tracks the county of mail-in ballots, including the rates at which requested mail-in ballots are turned in. So far, four of the five counties with the lowest ballot return rates (under 60%) are urban, covering much of Atlanta, Macon, and Savannah, while the counties with the highest ballot return rates (as high as 90%!) are rural, Republican strongholds.

Michigan

So far in 2024, 2,390,559 early votes have been cast in Michigan, 84% of the 2020 total. Only 158,377 of those ballots were cast in person, compared to 2,390,559 ballots turned in through the mail.

Similar to Georgia, Michigan tracks other demographics, but not party registration, for early voters. Early voters in Michigan are more likely to be over 65 (46%) or 41-65 (36.1%) and female (55.1%). Breaking the mail-in data down by county, six of the 10 counties with the lowest return rates are urban, covering the three Detroit counties, Ann Arbor, Kalamazoo, and Saginaw. However, the return rate for Michigan counties occupied a narrower range, between 76% and 88%.

Nevada

So far in 2024, 872,011 early votes have been cast in Nevada, 78% of the 2020 total. These were roughly split between mail-in ballots (441,209) and in-person early votes (430,802). So far, 339,407 registered Republicans have voted (38.9%), compared to 297,255 registered Democrats (34.1%). In 2020, registered Democrats led registered Republicans 39.7% to 35.6% in the pre-Election Day vote.

This means that, on Election Day 2020, 45,103 more registered Democrats had voted than registered Republicans. This time around, 42,152 more registered Republicans have voted than registered Democrats.

North Carolina

So far in 2024, 3,621,936 early votes have been cast in North Carolina, 79% of the 2020 total. Only 191,078 are mail-in ballots, compared to 977,186 mail-in ballots cast in 2020. Meanwhile, 3,430,858 early votes have been cast in-person, compared to 3,620,531 in 2020. So far, 1,230,318 registered Republicans have voted (34.0%), compared to 1,183,577 registered Democrats (32.7%). In 2020, registered Democrats led registered Republicans 37% to 32% in the pre-Election Day vote.

This means that, on Election Day 2020, 260,112 more registered Democrats had voted than registered Republicans. However, so far in 2024, 46,741 more registered Republicans have voted than registered Democrats.

Pennsylvania

So far in 2024, 1,615,637 early votes have been cast in Pennsylvania, 61% of the 2020 total. All these votes are technically mail-in, since the state “doesn’t have typical in-person voting,” Carpenter explained. But voters “can go down to your precinct office, get an absentee ballot, and vote by mail there, in person,” and many have done just that. So far, 526,337 registered Republicans have voted (32.6%), compared to 913,074 registered Democrats (56.5%). By contrast, in 2020, registered Democrats led registered Republicans 64% to 23% in the early vote.

This means that, on Election Day 2020, 1,079,080 more registered Democrats had voted than registered Republicans. “Joe Biden had banked, I think, a 1.1 million-vote ‘firewall ‘going into Election Day that then Republicans had to overcome — just a monumental task,” Carpenter noted. So far in 2024, more registered Democrats have voted than registered Republicans, but it is a significantly smaller number: 386,737.

Wisconsin

So far in 2024, 1,224,779 early votes have been cast in Wisconsin, 63% of the 2020 total. In-person early votes (705,326) hold a modest lead over mail-in ballots (519,453), but there is no more information available about the Wisconsin early vote.

Summary

All told, 2024 early voting seems to be booming in the seven key swing states, especially in Georgia and Michigan, and less so in Pennsylvania and Wisconsin.

Additionally, in all the states where party registration is available (Arizona, Nevada, North Carolina, and Pennsylvania), Republicans are running far ahead of their mark in the 2020 election. In three of the states, they have replaced an early vote deficit with an early vote advantage, while in Pennsylvania they have slashed the Democratic early vote advantage by more than half.

(It’s worth stating again that the available data only provide information on whether a voter who cast a ballot was registered with a certain party, not which candidate they actually voted for. However, it can provide a reasonable approximation of where the candidates stand, as it’s reasonable to expect — and, in large numbers, will likely prove true — that voters affiliated with a particular party will vote for that party’s candidates.)