Every poll that has appeared recently has shown Brian Kemp ahead in the gubernatorial election. They have, with the same consistency, shown his lead to be within the margin of error so although Kemp leads, the contest is a statistical dead heat. Kemp’s advantage is in keeping with consistent GOP success over the last two decades but the narrowness of the margin is striking. In both 2014 and 2016 Republicans won top-of-the-ticket contests by 200,000 votes, a decrease from the high water mark a decade earlier but nonetheless comfortable. A Stacey Abrams victory would make her the first non-incumbent Democrat to win a statewide, constitutional office in 20 years.
What are Abrams’ prospects for doing better than the polls that show her losing narrowly? A relatively simple model with just a couple of moving parts describes Georgia’s electoral dynamics. Two key components show little sign of movement. African Americans constitute the Democratic Party’s core constituency and will cast more than 90 percent of their ballots for Abrams. She might exceed 95 percent support and approach Obama’s 98 percent. On the other side, white males will give more than three quarters of their votes to Kemp. Donald Trump attracted 80 percent of white males in 2016.
This leaves white women holding the balance of power. They were ten points more willing to vote for Hillary Clinton and Jason Carter than were white men. There are distinctions among white women. In 2016 college-educated white women were almost twice as likely to vote for Clinton as their sisters without degrees. Younger, single white women are also more receptive to Democratic overtures.
To overcome the Kemp lead reported in the polls, Abrams will have to do better with white women than any non-incumbent Democrat in 20 years. If white men cast 20 percent or less of their votes for Abrams, she will have to get about twice that large a share of the white female vote. She will not need a majority, but she would need a substantial minority of the votes from white women.
With white women one of the moving parts in Georgia’s electoral system, the other item is the share of the vote coming from African Americans. Four years ago, Michelle Nunn’s consultants told her that to win she would need for African Americans to cast at least 30 percent of the vote – the relative level of participation achieved in both of the Obama campaigns and slightly higher than the 28+ percent in 2010. In 2014 blacks came close to Nunn’s target, accounting for almost 29 percent of those who went to the polls.
If Abrams succeeds in inspiring African American turnout to achieve historic levels vis-à-vis all voters so that blacks cast more than 30 percent of all ballots that would allow her to win with less white female support. Of course, if blacks cast less than 30 percent of all votes then she will have to get more white support. Abrams has been working to expand the minority electorate for years. Will those efforts pay off in 2018?
Abrams and her supporters have launched multiple attacks on Kemp alleging that as secretary of state he has acted to suppress black turnout. The implementation of the exact match law, a proposal to close seven Randolph County precincts and the 2017 voter purge have all entered the conversation and sparked legal challenges. Democrats warn that these actions disproportionately burden their supporters and could cost them victory in a close election. It is possible, however, that these actions may actually hold the key to an Abrams’ triumph. Essential to her success is mobilizing individuals who typically do not vote in midterm elections. Claiming that her opponent is bent on keeping minorities from voting may provide the stimulus needed to get some who would otherwise sit out the election to go to the polls. The threat that something is about to be taken away heightens the attractiveness of the item in question.
The polls that have attracted so much attention are another factor to consider when thinking about how the election may play out. The art in polling involves how the analyst weights the sample. The goal is to have a sample that matches the electorate that will cast ballots in the upcoming election. Thus the sample needs to be weighted accurately in terms of its racial, gender, age composition and so forth to reflect those who cast ballots.
Polling done for the Atlanta Journal-Constitution has weighted black turnout to equal what it was in the last most recent comparable election, the 2014 midterms. That weighting has blacks casting 28.7 percent of all ballots. Two other recent polls have gone with the Obama turnout figures and assumed that blacks will cast 30 percent of the vote. If Abrams achieves her goal of bringing into the electorate individuals who have not voted in the past or who voted in presidential elections but not mid-terms and if this increment includes disproportionately minority voters, then the weightings being used in publicly-released polls may underestimate her support. The failure to anticipate higher levels of minority turnout may account for several surprises in Democratic primaries earlier this year which resulted in Andrew Gillum’s nomination for governor in Florida and the upset of senior, white congressmen by minority women in Boston and New York.
In past years analyses of the national electorate have shown that as the election approaches, disaffected Republicans return to the fold. This year that willingness to put aside disagreements and rally to the GOP nominee has drawn impetus from the Kavanaugh hearings. During the past year, polling showed Democrats enjoying a sizable enthusiasm gap. More Democrats than Republicans expressed great interest in the upcoming elections and an intention to vote. GOP outrage over what they believed to be unfair attacks on the Supreme Court nominee has erased some of the Democratic enthusiasm advantage. A reinvigorated Republican electorate could make the Kemp advantage in polling a reality at the ballot box.
The closeness of the polls and the novelty of an African-American woman atop the Democratic ticket have made Georgia’s gubernatorial election one of the most watched in the nation. The millions spent by the campaigns have alerted the electorate to the differences in the candidates’ stands. All of this may produce a record turnout. If Republicans are as motivated to vote as Abrams’ Democrats, it may be difficult for her to attain the turnout advantage that seems to be critical for her campaign.
Charles S. Bullock, III, is the Richard Russell Professor of Political Science at the University of Georgia. The third edition of his co-authored Georgia Politics in a State of Change will be published in early 2019.