The first votes of the 2024 presidential election are rolling in, raising questions about what can be gleaned from the data in forecasting the possible outcome of the race.
Almost all states and Washington, D.C., have begun some mail-in or early in-person voting and releasing information on the number of ballots returned, the first actual data from the race itself. The states release how many ballots they have received, allowing comparisons to past elections, and many of them also share how many members of each party have cast ballots.
But some experts caution against extrapolating too much from the data given several unknown variables and the uniqueness of the last presidential election.
“It’s fair to make some observations, but it’s too early to make conclusions,” said Scott Tranter, the director of data science at Decision Desk HQ.
Mail-in and absentee voting has been a regular part of presidential races going back to the Civil War, but the practice has grown in more recent elections of the 21st Century. And Democrats have tended to be more naturally prone to voting early, while Republicans have been more likely to vote on Election Day.
The gap was particularly pronounced in 2020 amid the COVID-19 pandemic, with almost 60 percent of voters for President Biden voting by mail or absentee compared to about 30 percent of voters for former President Trump, according to Pew Research Center.
But so far, data being released in the main swing states that will likely decide the 2024 election have shown a major uptick in Republicans voting by mail or early in person.
Some caveats of any analyzing of this data include that the fundamentals of each election are different and even breakdowns by party don’t automatically show how people voted, hiding the possibility of cross-party voting.
Of the key swing states, Republicans have submitted more ballots in Arizona, Nevada and North Carolina, while Democrats have submitted more in Pennsylvania, according to the Election Lab at the University of Florida. Party breakdown data is not available in Georgia, Michigan and Wisconsin.
Nevada has seemed to gain some of the biggest attention for the advantage that Republicans have so far in votes cast already, as the state is a full vote-by-mail state, meaning all voters will receive a ballot in the mail if they choose to use that method.
Jon Ralston, an expert on Nevada politics and the editor of The Nevada Independent, has been tracking the daily intake of votes in the Silver State. As Republicans had submitted more ballots than Democrats for the first time since at least 2008, he said earlier this month that the totals could signal “serious danger” for Democrats in the state.
But pollster Nate Silver said Ralston is the only one whose analysis is more substance than “noise” and experts said observers should be wary about comparing this election to past ones.
Ken Miller, an assistant professor of political science at the University of Nevada, Las Vegas, said the public may be making too much of Republican improvement in early and mail-in voting, at least because of a shift in how Republican Party leaders have talked about the method, especially Trump.
Trump railed against mail-in voting throughout much of the 2020 cycle as susceptible to fraud and encouraged his voters to cast their ballots on Election Day. He ran on wanting to institute one-day voting, and to a certain extent has said he still supports that this year.
But Trump and his allies have made a concerted push to improve GOP performance in mail-in voting while not abandoning his false claims of voter fraud costing him the 2020 election. The “Swamp the Vote” initiative from the Trump campaign and the Republican National Committee has called for Republicans to overwhelm Democrats in turnout through mail-in and early voting.
The effort seems to have had at least some effect, given the tighter margins seen so far.
“Last presidential election, we had a presidential candidate who was encouraging his supporters not to vote early, not to vote by mail, and to vote on Election Day,” Miller said. “That’s pretty unusual. And so now we’ve returned to a sort of normal, but we’ve returned to a normal where voters are much more accustomed to [voting early].”
“Any comparisons people making to this election in 2020 are going to miss a key part of the story that Republicans naturally should have increased their early turnout,” he continued.
But questions remain about whether the change is more so a result of this shift in mindset about early voting among Republicans or a sign of charged GOP enthusiasm.
Tranter said both theories hold some water, as some data points support each one. He said data shows some new voters voting early from both parties, especially in Pennsylvania, but Democrats have also demonstrated enthusiasm among their base and showed signs of improving among white voters, who sometimes tend to vote on Election Day.
He said the number of new voters is measurable but isn’t “off the charts.”
“There’s good datapoints supporting both sides,” Tranter said.
But some said the data could be signs of what’s to come next week.
John Couvillon, a pollster and analyst who tracks early voting data, said the results may be a “flashing” sign for Democrats as their drop in mail-in voting compared to Republicans isn’t “being met with an equal and opposite increase” in early in-person voting.
He said the degree of the drop-off for Democrats compared to the 2020 race, which was a historic high-turnout election, suggests an enthusiasm deficit.
“Why there’s some statistical significance seeing a big drop-off of early voting is you can’t just assume that they’re all going to show up on Election Day,” Couvillon said. “Most will, of course, but not all.”
Of course, questions remain about the viability of comparing this year to 2020, when the pandemic produced an unusual political environment.
“What’s good is to compare it to a baseline, but we can’t compare it to 2020. With COVID, it was entirely distorted,” said Karl Rove on Fox News about early vote numbers last week.
Tranter said the DDHQ/The Hill forecast model does not take early voting data into account because of the swings in what comes in, making it difficult to be used.
“The main reason why it’s hard to use trends from early voting is very, very rarely do you have an apples-to-apples comparison,” he said.
In the absence of definitive conclusions, analysts said a wait-and-see approach is best.
“My advice would be the exact same as it would be for any polls that come out in the states, is that they’re just very murky snapshots,” Miller said. “And look, we’re going to know in a little over a week.”
