Unlike Pennsylvania, the state's new female voters are also slightly more likely to be registered Republicans than Democrats. But the Republican lead among new voters in Arizona has so far been largely driven by male voters.

Still, the number of new voters who have decided not to join a party is once again significant, and the way they vote could easily alter Republicans' apparent registration advantage among new voters who vote early.

A mixed picture emerges in the other swing states

A look at the remaining five swing states reveals a variety of patterns – and no clear conclusion.

In Michigan, there appears to be a large difference in the behavior of new male and female voters, although conclusions in Michigan are complicated by the fact that there is no party registration there and without this data it is difficult to predict the partisanship of Michigan voters I've seen big mistakes in the past. However, based on these estimates, modeling suggests that Democratic women slightly outperform Republican women among new voters. The same estimates suggest that the number of new Republican men nearly doubles the number of new Democratic men.

Wisconsin, like Michigan, appears to indicate a strong relationship between gender and partisanship among new voters – with new female voters leaning Democratic and new male voters leaning slightly Republican. However, the number of new voters who are likely to be unaffiliated requires extreme caution when attempting to read too much into these estimates.

In the other states with actual party registration data – North Carolina and Nevada – there is a new pattern: Voters who are not affiliated with a party are the largest group of new voters so far, among both men and women. How these independents vote is obviously crucial and also unknown – again highlighting the difficulty of drawing firm conclusions based on early voting data.

But one thing is clear: These voters could be decisive, as the number of votes cast by new voters in 2024 already exceeds the margin in many of the closest states in 2020. You are entering a polarized electorate and an election that many expect to close. And except in a few states where available data suggests a fuller story, the number of new voters who have no party affiliation — or the lack of party registration in key states — makes it difficult to know exactly how early voting is shaping up will impact this year's election results.

In such cases, it seems a futile exercise to peer into the bowels of the accumulated reports of early voting in the hope of turning them into a prediction of what is to come. Our pro recommendation: take a walk instead and enjoy the fall weather.