No, it’s not suspicious that Democrats won Senate races in states Donald Trump won

Some social media users expressed incredulity when they learned that Democrats have won or are leading in U.S. Senate or governor races in states in which Kamala Harris lost to Donald Trump.

One Nov. 7 Instagram post shared a screengrab of an X post that contained a graphic showing the partisan breakdown of wins in a number of swing states.

Trump, the graphic showed, had won or was winning the battleground states of Arizona, Michigan, Nevada,  North Carolina and Wisconsin, but Democratic Senate candidates (or in North Carolina, the Democratic gubernatorial candidate) had won or were winning their contests in those states.

"What are the odds of this happening in 5 of the swing states?" words in the screengrab said. "Honest question. And not all Senate incumbents either." 

"Something stinks," the Instagram user wrote in response. "This doesn’t make any sense."

(Screengrab from Instagram)

The post has accurately described the results as they sat by midday Nov. 8, three days postelection. Although these results may not make sense at first glance, they happened for several reasons.

Voters have free will. Presidential and Senate votes tend to align; in most 2024 Senate races, the same party did win both contests in the same state. But there’s a long-standing phenomenon known as ticket-splitting, in which voters may choose a Democrat for one office and a Republican for another. 

Ticket splitting has become less common in recent election cycles as partisan polarization has increased. In 2016, every Senate race went to the same party as the presidential race did, and in 2020, only Maine voters split their ticket in voting for Republican Sen. Susan Collins and for Joe Biden for president. But the practice hasn’t disappeared entirely.

In gubernatorial races, states such as Kansas, Kentucky, New Hampshire, Nevada, North Carolina and Virginia have chosen a governor of a different party than the presidential choice in the most recent election. And, in the current Congress, more than two dozen U.S. House members represent districts that the other party won in 2020.

Candidates matter. Voters split tickets because elections are a choice between candidates, not just between parties. Incumbency can help candidates; so can a lack of controversy, or a bigger war chest for campaign ads or a better-run campaign.

Two of the Senate races on the list cited in the post had incumbents prevail (or had them on track to prevail, pending additional ballot counting): Tammy Baldwin in Wisconsin and Jacky Rosen of Nevada. Their achievements in office and voters’ familiarity may have helped these candidates. Some voters may have wanted to send a message on inflation in the presidential race but decided that the Senate has less to do with economic conditions than the president does.

Meanwhile, in North Carolina’s gubernatorial race, Democratic Attorney General Josh Stein faced Lt. Gov. Mark Robinson, a Republican candidate who faced a flurry of revelations about his past controversial comments on race and sex. Robinson won about 40% of the vote in a state in which generic Republican candidates for statewide office can usually count on 50% or better.

Featured Fact-check

Facebook posts

stated on November 21, 2024 in Facebook posts

“Taylor Swift has moved to Canada after leaving the U.S. following the big red wave.”

And in Arizona, Republican Kari Lake came to voters’ attention in 2022 when she lost a gubernatorial race built heavily on claims that elections are fraudulent, including the presidential race that Biden won. This year, she lost to Democratic Rep. Ruben Gallego, a retired U.S. Marine and a Latino in a campaign year in which Latino voters were pivotal within the electorate.

Don’t sleep on third parties and nonvotes. In the four Senate-race split decisions, the margins were narrow. In these four states, the Democratic Senate candidates won (or are winning) by 17,000 to 44,000 votes; in Nevada, Michigan and Wisconsin, Harris lost (or is losing) by between 29,000 and 82,000 votes. Arizona was wider at 155,000.

With such small margins, other factors can figure in election results if you run the math. 

In these states’ Senate races, third-party candidates received from 39,000 to 170,000 votes; in the presidential race, their haul ranged from 25,000 to 110,000.

Also, some people voted for president but not for a Senate candidate, and may have skipped other down-ballot races. Depending on the state, from 18,000 to 72,000 voters did this, according to PolitiFact’s calculations from The New York Times’ running election results page. If these Senate race-skippers consisted disproportionately of Trump voters, that by itself could have kept some of these Republican Senate candidates from winning.

Then there’s Nevada’s eccentric ballot option,"none of these candidates." In the presidential race, about 18,000 voters chose this option; in the Senate race, more than 40,000 did. The number of "none of these candidates" voters in the Senate race is more than twice as large as Rosen’s current lead.

A nontrivial share of voters chose to back a third-party candidate or decline to choose anyone in the U.S. Senate race. In some or all of these states, that could have spelled the difference between victory and defeat. 

Social media posts said "something stinks" that Trump won in several battleground states while Democrats won Senate or governor races in those states.

Although there’s a strong historical correlation between votes cast for president and senator, it’s not ironclad. Voters are free to choose candidates of different parties for different offices, and some fraction of voters does so.

In each of these races, the Democrats benefited from an advantage or two, such as incumbency or a politically damaged opponent, that may have put them over the top.

Finally, in the four Senate races cited, the margins were narrow, meaning that third-party voters or voters who chose to vote for president and not the Senate race could have easily spelled the difference between victory and defeat.

We rate the statement False.