Battleground Polling Shows Ticket-Splitting Pattern

President Joe Biden trails in polling in Arizona. Rep. Ruben Gallego (D-ARiz.), right, is leading in a Senate race there. (Kenny Holston/The New York Times)
President Joe Biden trails in polling in Arizona. Rep. Ruben Gallego (D-ARiz.), right, is leading in a Senate race there. (Kenny Holston/The New York Times)
  • Oops!
    Something went wrong.
    Please try again later.
  • Oops!
    Something went wrong.
    Please try again later.

We have a new set of polls for you in the battleground states, including New York Times/Siena College polls of Michigan, Wisconsin, Georgia, Nevada, Arizona and the inaugural Times/Philadelphia Inquirer/Siena poll in Pennsylvania.

The results in the presidential race would have been surprising a year ago, but it’s hard to call them surprising anymore. Former President Donald Trump leads in five of the six states among likely voters, with President Joe Biden squeaking out a lead among likely voters in Michigan. (Among registered voters, Biden leads only in Wisconsin.) Trump’s strength is largely thanks to gains among young, Black and Hispanic voters.

Sign up for The Morning newsletter from the New York Times

What’s more surprising is the U.S. Senate results. This is the first time we’ve asked about Senate races this year, and the Democratic candidates led in all four of the states we tested: Pennsylvania, Wisconsin, Arizona and Nevada.

Not only do Democrats lead, but they also seem to do so in an entirely customary way, with ordinary levels of support from young and nonwhite voters, even as Biden struggles at the top of the ticket.

Nevada was ground zero for this striking ticket splitting. Trump led the poll by a staggering 12 points among registered voters, thanks to an eye-popping 9-point lead among Hispanic voters and a 13-point lead among those 18 to 29.

But in the Senate race, everything looks “normal.” Democratic Sen. Jacky Rosen led her likeliest Republican challenger by 2 points among registered voters, including a 46-27 lead among those 18 to 29 and a 46-28 lead among Hispanics.

Remarkably, 28% of Trump’s Hispanic supporters and 26% of his young supporters back Rosen.

This level of crossover voting has been extremely rare in the last few years, but it was pretty common before 2020. In fact, these results remind me a lot of the 2016 presidential election, when Trump surged in white working-class areas, Hillary Clinton surged in college-educated areas, and yet the Senate and House results by county still mostly followed the pre-2016 pattern.

With polls showing Trump making yet another demographic breakthrough, perhaps it shouldn’t be so surprising that ticket splitting is back as well.

What does ticket splitting mean for Biden?

There’s been understandable debate over whether Biden could really be doing so poorly among young, Black and Hispanic voters. After all, we’ve never seen anything like it before, and polls are hardly perfect.

But at least to me, the relatively “normal” down-ballot results strengthen the case that Trump’s breakthrough among young and nonwhite voters is probably real — which is to say, not an artifact of some kind of systemic polling error.

It’s consistent with other indicators (such as party registration, or recalled 2020 vote preference) suggesting that the polls are reaching the people who usually vote for Democrats — they just aren’t backing Biden.

And historically, huge polling errors tend to be systematic. Back in 2016, for instance, the polls missed Ron Johnson’s victory in the Wisconsin Senate race, not just Trump’s win for president in the state. Similarly, the 2020 polls overstated the prospects of Democratic candidates such as Sara Gideon, Gary Peters and Steve Bullock by every bit as much as they overstated those of Biden.

That said, there may be some good news for Biden here: These voters haven’t yet abandoned Democrats in full, and they might still be available to return to his side.

What is going on with likely voters in Michigan?

One of the most peculiar findings in the poll is the huge split between registered and likely voters in Michigan.

Overall, Trump led by 7 points among registered voters — the broader group of people who are registered to vote in the state.

But Biden had a 1-point lead among likely voters, the smaller group that represents the likely electorate of actual voters this November.

I find that gap to be more than a little hard to believe. It requires unlikely voters in Michigan to back Trump by about 30 points, even as Biden narrowly leads among those who actually show up and vote.

But it’s worth noting that the last Times/Siena poll of Michigan also had an unusually large gap between registered and likely voters, with Trump leading by 5 points among registered voters while Biden and Trump were tied among likely voters.

In this poll, Michigan voters who said they were only “somewhat” likely (or less likely than that) to vote backed Trump by 26 points, 54% to 28%.

Those without a record of voting in the 2020 election backed Trump by 34 points, 62% to 28%.

Arab and Muslim voters

When we did our last Times/Siena poll of the battleground states, we found signs of huge defections from Biden among a small sample of voters who were either Muslim or Arab.

We found it again.

Overall, Trump led, 57-25, among Middle East, North African or Muslim voters in the poll. Those who say they voted in the 2020 election reported backing Biden by a similar but opposite margin, 56-35.

When we asked Arab or Muslim voters who didn’t back Biden about their most important issue in the race, around 70% cited foreign policy or the war in the Gaza Strip.

c.2024 The New York Times Company