On Tuesday night, Brian Kemp—the Georgia secretary of state and self-proclaimed “politically incorrect conservative” who owns a truck, talks about standing for the national anthem and happily supports President Trump—beat Lieutenant Gov. Casey Cagle in the state’s gubernatorial runoff, officially becoming the GOP nominee. Kemp will face off against Democrat Stacey Abrams, the Sanders-backed former Minority Leader of the Georgia House who is the first black woman in American history to win a major party gubernatorial nomination.
Kemp and Abrams are interesting characters—they’d be worth writing about even if they were running in an un-competitive state. But they’re running in Georgia (a typically red state that Democrats hope to turn purple in 2020 or 2024) in what looks like a good year for the Democrats.
This race has the potential to be very interesting. So it’s worth taking stock of some of the basics—exactly what Georgia looks like politically and demographically and what the early outlook in this race is.
Georgia—Turning Blue or Not?
The conventional wisdom on Georgia seems to be that it’s a light-red state that’s going to turn blue sometime soon—that the demographics of the state are changing, and Donald Trump’s 5 point win there in 2016 is a sign that it’s slipping away from the GOP.
Certain parts of that story may end up being right, but the current reality is more complicated. And when you pick at the details, you start to see problems for Abrams in this race.
First, Georgia did take a leftward step in 2016, but it wasn’t inexorable demographic forces that pushed it in that direction.
This graphic shows Georgia’s partisan lean in every presidential election since 2000 (calculated by subtracting the statewide presidential two-party vote share from the national popular two-party vote share). Trump’s 2016 performance did represent a drop for Republicans. He won the state by about 5 points while losing the national popular vote by 2 points. Mitt Romney, on the other hand, won the state by 8 points while losing the popular vote by 4 points. John McCain only won the state by 5 points, but the national environment was much worse for Republicans in 2008 than 2016 (McCain lost the popular vote by 7 points). And George W. Bush won the state by 17 points and 12 points in 2004 and 2000 respectively.
Those numbers don’t look great for Trump. After adjusting for the national climate, Trump’s showing in Georgia looks like a real decline from the stable-ish equilibrium set up by recent Republicans.
Part of this movement is demographic change—Georgia’s nonwhite population has been growing. But the specific conditions of 2016 (the urban-rural and education divide among whites) was a big factor.
This graphic shows the number of two-party votes cast for each candidate in each large cities, small towns, rural areas, and other “CBSA Divisions.” (these divisions group suburbs and some exurbs in with the cities—more info here). You’ll notice that the “Mega City” category (which, in Georgia, is basically Atlanta) usually casts over half of the state’s major party votes. That’s good news for Trump Era Democrats. Trump lost ground in many well-educated, suburban areas around the country—and a notable amount of Georgia fits that description.
We can get more granular by looking at data from the Rob Griffin, Ruy Teixeira, and John Halpin at the Center for American Progress. They found that most demographic groups in Georgia voted similarly in 2012 and 2016 with one important exception: college-educated whites. Trump still won the group in 2016, but his margin was about 15 points lower than Romney’s (the Asian vote also shifted strongly against Republicans, but they’re still a relatively small group in Georgia so they didn’t shift the overall result as much).
That leaves Georgia Democrats in a better position than they’ve been in before—they improved their performance with a major group without losing much ground with other groups and got within 5 points of winning the state.
But that narrative doesn’t match the conventional wisdom, and it’s not clear that Abrams can will be able to turn Clinton’s gains into a winning coalition.
Despite the leftward shift of 2016, Georgia is still a light-red state and there’s still a decent baseline advantage for the GOP.
Georgia is also less “swing-y” than other states, and that helps Republicans. Both parties have large bases, but the overall math of the state usually ends up favoring the GOP.
This graphic shows the two-party vote of every state in the 2016 election along with the state’s “elasticity.” There’s more than one way to define this term, but the basic idea is to measure how much a state does (or doesn’t) swing over time. Mathematically, it’s the standard deviation of the two-party presidential vote since 2000.
But basically if a state’s elasticity is higher, that means it was all over the map in recent presidential elections; lower elasticity indicates greater consistency.
Georgia is one of the more inelastic states in that graphic. It varies more than some states (e.g. Massachusetts) but it’s been reasonably consistent despite the wild ups and downs of the last 18 years of U.S. politics.
That’s partially because there’s some racial polarization in Georgia voting. Republicans typically win white voters (especially blue-collar whites) by a large margin, Democrats win black voters by a large margin and the final totals end up favoring Republicans. According to the CAP numbers, Obama and Clinton won about 91 percent and 87 percent of the black vote, respectively. But they won 16 percent and 17 percent of the white, non-college educated vote. The state isn’t completely racially polarized (Republicans have a few black voters, and college-educated whites are much more split than non-college educated whites), but right now the numbers add up in way that gives the GOP a consistent advantage.
Democrats could break through this advantage by turning out nonwhite voters, holding onto the white voters they won over in 2016, and persuading some new white voters. But that’s no small task. The demographics and political history of the state seem to say that these factors still favor the GOP in 2018.
So What’s the Bottom Line?
There isn’t much polling in this race. But Sabato, Inside Elections, RealClearPolitics, and Governing all rate the race as “Likely Republican” while Cook has it at “Leans Republican.” Taken together, these ratings don’t rule out an Abrams comeback, but they do suggest a strong initial edge for Kemp. And that makes sense. Georgia is a red state, and Abrams needs multiple things to go right for her, while Kemp could underperform Trump and still win. Georgia might be more competitive in a future presidential election, but Kemp starts with the the advantage.
