Contrary to popular belief, there just weren’t that many people who voted for different parties in the presidential and Senate races this year. And where split-ticket voting did occur, it appeared to have more to do with the candidates on the ballot than the voters themselves. For example, Republican Sen. Susan Collins ran far ahead of President Trump in every corner of Maine, and Republican Sen. Steve Daines ran a bit behind him in almost every county in Montana. In other words, ticket-splitting didn’t vary that much within each state and was, in most cases, pretty minimal.
But some counties had more ticket-splitters than others, and those voters could be particularly important in Georgia. On Jan. 5, control of the U.S. Senate will be decided by a pair of runoff elections in the Peach State, and Democrats Jon Ossoff and Raphael Warnock will likely need to win some votes from people who didn’t vote for them last time, as Democrats got fewer votes than Republicans in both races on Nov. 3.1 Both Democrats in the runoffs also ran behind Joe Biden, who won Georgia 49.5 percent to 49.3 percent.
How Trump’s pardons compare to other presidents | FiveThirtyEight Politics Podcast
There weren’t too many of these split-ticket voters (Biden outran Ossoff by only 1.6 percentage points statewide), but as you can see in the map below, we have a sense of where Ossoff and Warnock need to make up ground: Suburban or well-educated areas, such as the Atlanta metropolitan area, including Fulton (where Biden did 2.8 points better than Ossoff), Cobb (2.4 points better) and Forsyth (2.1 points better) counties. (It’s easiest to use Ossoff’s vote share as the comparison point because multiple Democratic candidates were on the ballot in the other Senate race, so Warnock advanced to the runoff with less than 33 percent of the vote. However, the numbers are still similar when you compare Biden with the total Democratic vote share in the special election.)
The greater Atlanta area was not the only spot where Biden outperformed Ossoff, though. He also did so in well-educated Clarke and Oconee counties to the east — perhaps owing to the University of Georgia’s presence there. (The pattern also shows up in a handful of rural counties, such as Early County in the southwest, but this is probably less important. These counties are so small that just a few Biden-but-not-Ossoff voters can produce what looks like a big swing. So the most efficient strategy for Ossoff and Warnock is probably to focus on squeezing more votes out of metro Atlanta. The top six counties where Biden outperformed Ossoff the most in terms of raw votes were all located there.)
|county||Biden Votes Over Ossoff|
And a precinct-level analysis, such as the one done by Lenny Bronner at The Washington Post, reveals that upper-class white neighborhoods in these counties, such as Buckhead, drove Biden’s overperformance. Bronner found that the Atlanta-area precincts where Biden outperformed Ossoff had a median household income that was $41,000 higher than the average precinct in the 10 counties he categorized as part of the city’s metro area. The number of adults with a bachelor’s degree was also 37 percent higher, and the share of white people was 21 percentage points higher.
And this trend extended far beyond Georgia. For example, Biden did 1.6 percentage points better than Democratic Senate candidate Cal Cunningham in North Carolina. But according to analyst J. Miles Coleman of Sabato’s Crystal Ball, he outran Cunningham by 6.1 points in North Carolina state Senate District 37, which covers a wealthy white neighborhood of Charlotte colloquially referred to as “the Wedge.”
Or look at Iowa. Democratic Senate candidate Theresa Greenfield did 0.3 percentage points better than Biden statewide, but Biden did 2.2 points better than Greenfield in Dallas County, a fast-growing exurban county just to the west of Des Moines that has the state’s highest median household income. That was Biden’s best performance relative to Greenfield anywhere in the state (although, notably, he also outran Greenfield in Des Moines’s Polk County and Iowa State University’s Story County).
Similarly, Biden’s vote share was significantly higher than the Democratic Senate candidate’s vote share in the counties that contain and surround cities such as Chicago, the Twin Cities in Minnesota, the Research Triangle in North Carolina, and a host of other metropolitan areas nationwide.
The fact that ticket-splitting was higher in upper-class white communities jibes with research by political scientist Ashley Jardina, who studies the individual voting behavior of white voters (though not specifically in the suburbs). Jardina told FiveThirtyEight that Trump’s culture-war appeals and handling of the COVID-19 pandemic have turned off many wealthy, white, traditionally Republican voters, which led some of them to not vote for Trump but to continue to support the GOP in down-ballot races.
If anti-Trump Republicans were indeed the reason Biden outran Democratic Senate candidates in the nation’s suburbs, it could be a bad omen for Democrats in future elections. If these high-profile new Biden voters are still voting Republican further down the ballot, it suggests they are not yet fully ready to leave the GOP behind — and could return to the fold in the 2022 midterms or in the 2024 presidential election, if Republicans nominate a candidate who’s closer to George W. Bush than Donald Trump in temperament. This could move the suburbs, that newfound source of Democratic strength, back to the right.
Or not. America’s suburbs are also getting more racially diverse and better-educated, two traits that correlate with voting Democratic. And despite all the ink spilled over Biden-Republican ticket-splitters (perhaps because many members of the media share their well-educated white demographic profile), it’s important to remember that there simply aren’t that many of them: Throughout this article, we’ve only been talking about differences of a few percentage points. The difference between Biden getting 46 percent of the vote and a Democratic Senate candidate getting 44 percent2 really isn’t all that significant — except, of course, in a close election like the Georgia runoffs. But in a race that close, any bloc, no matter how small, is an important constituency.