Georgia Democrats are on a roll. Dormant as a serious political force for years, the state has now entered center stage as a national battleground. Polls show highly competitive races at every level in Georgia, including the state’s two Senate races. Our friends at Sabato’s Crystal Ball recently moved the special Senate election to Leans Republican while the Cook Political Report moved it to Tossup at the Presidential level.
Even with this newfound momentum, Democrats in the Peach State are worried. Under Georgia law, if no candidate gets 50% in a statewide contest, the top two finishers face off again in a runoff election a month later. Since 2008, the prospect of a runoff in any election has kept Democrats awake at night. That year, Democrat Jim Martin narrowly finished second in the November race for Senate, but lost by 15% in the runoff a month later to incumbent Republican Saxby Chambliss.
Largely because of that race, Democrats and election observers have viewed runoffs as an obstacle to any potential Democratic win. But an examination of recent runoffs in southern states points to them being very winnable for Democrats – and their 2008 failure being an anomaly. Since 2008, there have been numerous runoff contests in Louisiana, Georgia, and Mississippi. Racial and partisan turnout data from these elections do not paint a grim picture for Democrats.
Starting with Louisiana, the state has seen four high-profile statewide runoffs since 2008. The 2014 Senate election, the 2015 Gubernatorial election, the 2016 Senate election and the 2019 Gubernatorial election. Across these four elections, Democrats increased their vote share from the primary to the runoff in every single one. On average, this increase was 5.5%.
In Mississippi, the state had a runoff election to fill a vacant seat in the U.S. Senate in 2018. In the November election, Mike Espy was the only Democrat, and he received 40.9%. In the runoff he increased this by 5.5% to receive 46.4%. Georgia also had a runoff that year, with the race for Secretary of State seeing no candidate get a majority in November. John Barrow, the Democrat, did see his vote share decline, albeit only by .6%. That’s hardly a massive drop in support.
Runoffs and Racial Voting
Of course, the assumption around runoffs is not that voters suddenly like Democrats less after a month. The main reason runoffs are seen as difficult for Democrats to win is a misconception about racial voting patterns. Many party operatives and observers believe that Black voters tend to turnout at much lower rates for runoffs than they do for November general elections. This is critical in these states, as the Demcoratic coalition in each is heavily reliant on Black voters. But again, this presumption is almost entirely incorrect.
In all four of the major runoffs in Louisiana, Black turnout was strong. On average, in the four general elections, Black voters made up 28.3% of the electorate. But in the runoffs, the average share of the electorate that was African-American increased to 29.7%. The story in Georgia is slightly more complicated. The Black share of the electorate did drop from general to runoff in 2018, but only slightly. Black voters made up 28.9% of the electorate in November, and 27.3% of the electorate in the December runoff. Again, hardly a drop that points to December runoffs being unwinnable for Democrats.
The state of Mississippi gives no data that is comparable to what is provided in Louisiana and Georgia. However, for the special Senate Election held in 2018, it is possible to look at how each county’s total votes changed from the general to the runoff. Below is a map that shows this as well as a map that shows what percentage of each county in Mississippi is Black.
As can be clearly seen, the biggest drops in turnout came largely from heavily white counties. Even wealthy DeSoto County (just south of Memphis), saw a large drop in turnout. By and large, an increase in votes cast was correlated with a county having a higher percentage of African-American residents. This is the opposite of what the conventional wisdom would suggest should have happened.
The Bigger Picture
This phenomenon is not unique to runoff elections. Contrary to popular belief, Black turnout is fairly robust. According to the United States Election project, Black turnout in midterm elections since 1986 has ranged from 33% in 1990 to 51.3% in 2018. White turnout is slightly higher, but has a virtually identical range. It has gone from 39.8% in 1986 to 55.2% in 2018.
The biggest reason for the survival of the belief that Black turnout is volatile is Barack Obama. In 2008 and 2012, Black turnout soared to unprecedented levels. Peaking at nearly 70% in 2008, Black voters turned out at a higher rate than white voters in both of Obama’s elections. Something that had not been done before-and has not been done since. When turnout rates returned to more historically average levels in 2010, 2014, and 2016, Democrats jumped to hasty conclusions.
The conventional wisdom became that Black turnout was perpetually on the cusp of falling off a cliff unless circumstances were ideal. On top of that, pundits saw two low turnout midterms, 2010 and 2014, and concluded low turnout would always benefit Republicans. But both of these conclusions were erroneous. Looking at changes in Black turnout from 2014-18 provides a much more accurate picture. Black turnout in 2014 was 36.4%. It jumped to 59.9% in 2016 before clocking in at 51.3% in 2018. White turnout followed a similar pattern. It was 40.8% in 2014, 64.7% in 2016, and 55.2% in 2018.
Additionally, while it is true that Democrats did poorly in low-turnout elections during Obama’s tenure, there is good reason to think this pattern would not hold today. The nature of each party’s coalition was different in 2008. Back then, Republicans typically ran up crushing margins in wealthy white suburbs. John McCain won big in Gwinnett and Cobb counties that year, for example. The Democratic coalition in states like Georgia, in addition to being reliant on Black voters, was more dependent on ancestrally Democratic working-class Whites. This meant the GOP had a base that was more consistently high turnout, giving them an advantage in runoff elections.
These days, Democrats are the ones running up huge margins in the Atlanta suburbs while Republicans rely on working-class whites. Their turnout of course, being more variable than their college-educated counterparts. Any advantage Republicans had in 2008 would not exist in a 2020 runoff.
Of course, these dynamics make it obvious why Martin fell flat in that 2008 runoff. The aforementioned spike in Black turnout driven by Obama’s candidacy helped him in November but couldn’t be maintained in the runoff. The state does not give turnout data for the runoff, but in Georgia that year, Black voters made up 30% of the electorate. This is a huge increase from 2004, when Black voters only made up 25.3% of the electorate.
Furthermore, the national dynamics at play in that 2008 election were unique. Democrats after November held a 58-40 advantage in the Senate. A win would have given them a near filibuster proof majority, with the ongoing recount in Minnesota the potential 60th seat. It is possible voters were wary of potentially giving Democrats even more power after the party had just scored a crushing win a month earlier. Such a dynamic would be highly unlikely to be at play in 2020.
In fact, a runoff in Georgia could be the deciding factor in which party holds a Senate majority, creating incentives for voters of both parties. This is especially true because Georgia’s second Senate race, the special election to fill the remainder of Johnny Isakson’s term, is almost certain to go to a runoff. While David Perdue or Jon Ossoff could win outright in November, the balance of power in the Senate will not be set until after the December runoff in Georgia. This could create a scenario where the seat is critical, and turnout that is much higher than previous runoffs.
A runoff in Georgia does not spell doom for Democrats, nor should it make Republicans sleep easy. Until now, analysts have baked in the assumption that a runoff in either Senate race would present a massive challenge for Democrats, but the data does not bear this out. In fact, it shows a potential runoff presents just as many issues for the GOP as it does Democrats.