Tracking 2024: The Curious Gap Between the Presidential and Senate Polls

A ticket-splitting renaissance? (Probably not)

One of the more interesting facets of election polling this cycle has been the consistent difference in margins of support at the top of the ticket versus in those in key Senate races. 

Whether you’re looking at polling averages on 538 or RealClearPolitics, Democratic Senate candidates in key states appear to be performing better against their Republican counterparts than Vice President Kamala Harris seems to be doing against former President Donald Trump. 

These Democratic Senate candidates — be it Ruben Gallego of Arizona, Elissa Slotkin of Michigan, Jacky Rosen of Nevada, Bob Casey of Pennsylvania or Tammy Baldwin of Wisconsin — are all leading their opponents by at least a few percentage points as Harris and Trump remain within less than a point of each other.

Some have wondered whether this may herald a resurgence of ticket splitting, where voters cast their ballots for one party at the top of the ticket and the other down the ballot, which has only happened once in the past two presidential cycles (when Republican Sen. Susan Collins of Maine won re-election as Trump lost the state). But I think those who are swayed by this argument are making a relatively common mistake in the polling discourse: focusing on margins instead of shares.

For example, let’s take a look at our latest state-level survey results at the presidential and Senate level.

As you can see, the difference in margins between the presidential and Senate results are largely a function of the presence of undecided voters. While most of the Democratic candidates for Senate are polling right around Harris’ share of support against Trump, there’s a good bit of dropoff between Trump and Republican candidates. 

It’s not uncommon to see more undecided voters in surveys of Senate races. After all, the candidates, especially non-incumbents, are of a lower profile and therefore lesser known to voters. That means that the ultimate margins in these races will be largely driven by where these undecided voters gravitate. And on that front, our own data, illustrated by this next chart, very much points to Democratic advantages far short of what we’re currently capturing.

On the left side of the chart above, you see the share of Democratic voters who say they’d vote for Harris and the Democratic Senate candidate if the election were held today, while on the right it’s the share of Republican voters who would vote for Trump and the GOP Senate candidate.

What this chart tells us is that Republican Senate candidates have fewer committed voters from their party base than their Democratic opponents do. We should expect that to change as we get closer to Election Day, especially in Arizona and Nevada, where Republicans are roughly twice as likely as Democrats to be undecided at the moment.

So while none of this is to say that there’s no chance of a split-ticket outcome in 2024, it is a pretty low one that is obscured by a focus on polling margins. Paying closer attention to shares instead — and to shares of undecided voters specifically — as Election Day draws closer will give a clearer sense of where things are likely headed.

With that in mind, you can expect Senate races in the key battlegrounds of Arizona, Michigan, Nevada, Pennsylvania and Wisconsin to be closer than they currently appear.