It is hard to know what role negative feelings about black politicians played in the likely gubernatorial losses of Democrats Stacey Abrams in Georgia and Andrew Gillum in Florida. Neither race has been officially called, but the Democrats trail in both, which were among the handful of races nationwide that seemed to most capture the interest of Democratic activists. It’s also not clear how much the Georgia race was affected by controversial election practices implemented by Republican Brian Kemp, who was serving as secretary of state as he ran for governor, making him Georgia’s chief elections officer.
Here’s what we can say, however: First, Abrams, considering the conservative lean of Georgia, had one of the best performances among Democrats’ 36 gubernatorial candidates, many of whom are white, male or both. Second, Gillum’s performance was basically average, compared with other Democratic gubernatorial candidates. And finally, neither Abrams nor Gillum was able to really shift the default partisan dynamics in their states, unlike, say, Democrat Laura Kelly in Kansas or Republican Charlie Baker in Massachusetts, who both won in states where their party labels alone could have really hurt them.
To look more closely at how these two candidates did, we compared the results in all 36 states that had gubernatorial elections last week to the partisan lean in those states (essentially a measure of how liberal and conservative states are, across election cycles and different races).1 This method is useful because it creates a kind of apples-to-apples measure: How did Abrams and Gillum do compared with other 2018 Democratic gubernatorial candidates, considering the differences in the politics of various states and the broader pro-Democratic dynamic of the 2018 election overall.
Overall, in 25 of the 36 races, Democrats did better than the party’s standing according to partisan lean in those states. This is more evidence that 2018 was a favorable year for Democrats overall.
The biggest overperformances in governors races though, came from fairly moderate Republican incumbents in blue states who were heavily favored pre-election: Massachusetts’ Charlie Baker, Maryland’s Larry Hogan and Vermont’s Phil Scott. The best Democratic performance was by Laura Kelly, who defeated Kris Kobach in Kansas. She outperformed her state’s partisan lean by almost 28 percentage points. (Kansas leans about 23 percentage points to the GOP, and Kelly won by about 5 percentage points.) This is not particularly surprising: Pre-election polls had shown Kelly could win, numerous prominent Republicans in the state endorsed her over Kobach, Kobach has a controversial reputation, and some Republican voters in the state had turned against the outgoing GOP administration.
|state||name||Margin||Partisan Lean||Diff||Diff including national vote*|
|NM||Michelle Lujan Grisham||+14||D+7||D+7||0|
Abrams ranked sixth among Democratic candidates in terms of overperformance — at about 10 percentage points. (Georgia is about 12 points more conservative than the nation overall, while Abrams is down by less than 2 percentage points in her race against Kemp.) Gillum, by this measure, did the 18th best of the 36 Democratic candidates, right in the middle. He ran about 5 points ahead of Florida’s 5-point Republican lean. He trails Republican Ron DeSantis by less than 1 point.2
So that data (and the comparison of Abrams and Gillum to the 34 other Democratic gubernatorial candidates) tells a pretty positive story about them. But here’s another way to look at the data: How did Abrams and Gillum do compared to how Democrats did overall on Election Day? Democrats currently lead in the national House popular vote by about 7 percentage points.3 Translating that to Florida and Georgia, that would suggest the national environment in Florida on Election Day 2018 probably narrowly favored the Democrats (by about 2 percentage points), while in Georgia it would favor the GOP (by about 5 percentage points.) By this measure (which you can also see in the last column in the table above), Abrams still overperformed, but the gap is about 3 percentage points, not the 10 points suggested by only comparing her margin to partisan lean. Gillum, by this measure, probably did 2-3 points worse than you would have expected, since he is just barely trailing in his race.
So Abrams did well and Gillum OK, but neither had the big overperformance of Kansas’ Kelly. And it’s worth thinking about why. Abrams, Gillum and Kelly are particularly worth comparing, because all three ran against Republican candidates who cast themselves in the mold of Trump.
Kelly could simply be a superior candidate, although she is not the compelling speaker Abrams and Gillum are. Outgoing GOP governors Nathan Deal in Georgia and Rick Scott in Florida are more popular than Sam Brownback, the Republican who ran Kansas most of the last eight years before he stepped down earlier this year for a job in the Trump administration. So Kobach probably had the biggest drag from his party. We think this is a really big factor.
Kelly’s win, like those of Republican moderates in the Northeast, wasn’t out of left field, because of something past election cycles have shown: Party matters in gubernatorial races, but voters are typically much more likely to back a candidate from the opposite party for governor as opposed to Congress or the presidency.
But Gillum and Abrams in particular were testing something that complicates the potential of cross-party voting in a governor’s race: Is it harder to get any kind of crossover vote, even in a governor’s race, if you are a black Democratic candidate running in a state with a high minority population where politics are divided along racial lines as well as partisan ones?
There’s no doubt that some whites in Southern states in particular have negative attitudes about blacks. But a more subtle factor to consider in Southern states is that when politics and race are tightly aligned, it may be harder to cross party lines. Abrams and Kelly were not that different ideologically, but few Republican-leaning elites endorsed Abrams. Again, maybe that was because Kobach is uniquely controversial. But Georgia Republicans leery of Kemp would have had to make a bigger leap — culturally, politically, and racially — to align with Abrams, compared to backing Kelly in Kansas. This is a big part of why many states in the South, including Georgia, are among the least elastic in the country (they have very few swing voters).
All this doesn’t mean that Stacey Evans, the white Democrat who ran against Abrams in the Georgia Democratic primary for governor, would have won the general election. Evans might have lost too, just differently — a worse performance around Atlanta and among minority voters, where Abrams was strong, but perhaps a better one in more rural, white areas.
Here’s the good news for Abrams, Gillum and Democrats hoping the party lifts up more black candidates in the future: Abrams in particular did much better than a lot of other white Democratic candidates in outperforming partisan lean. Wisconsin’s Tony Evers won, Iowa’s Fred Hubbell lost, but both ran fairly narrowly ahead of where Democrats generally stand in those states.
In other words, we don’t want to dismiss liberal claims of voter suppression in Georgia or the historical barriers blacks have faced in getting elected to statewide office in America. That said, there is one clear obvious factor that limited Beto O’Rourke in Texas, Abrams and Gillum, three candidates whom liberal activists were very excited about but didn’t win. Their states are generally Republican-leaning — and opted not to go the way of Kansas and break that historical pattern for them.
CORRECTION (Nov. 15, 2018, 4 p.m.): A previous version of the first graphic in this article gave the incorrect party designation for partisan lean and vote margin when a data point was hovered over.