Minimizing and Curing Deficient Ballots Could Be Key in GA Runoffs

5 min readDec 18, 2020

By the Socio-Economic Research & Tech Team

Methodology: Analysis of Georgia Secretary of State data

Photo by Element5 Digital on Unsplash

Main Points:

  • While only a minuscule percentage of Georgia ballots were either rejected or designated as “spoiled” in the general election, if the upcoming Senate runoffs are as close as anticipated, deficient ballots could be the difference between victory and defeat.
  • In the general election, younger Georgians and people of color were overrepresented among voters who cast deficient ballots and those who did not cure their ballots.
  • Only 0.19% of vote-by-mail ballots submitted in the Senate runoff election thus far have been rejected or marked spoiled, but a racial gap persists, suggesting an opportunity for targeted outreach.

In the 2020 general election, Georgia embodied the adage “Every vote counts,” with Biden edging out Trump by fewer than 13K votes out of 5 million. Unfortunately, not everyone who attempted to vote in Georgia successfully did so. Analysis of Georgia Secretary of State (SOS) absentee records suggests that close to 4400 voters had their ballots either rejected (~4100) or marked as spoiled (~250). Although this number represented less than 0.1% of total ballots cast in the state, rejected and spoiled ballots could play an important role in the Georgia Senate runoffs on January 5 if the margins are as razor-thin as the presidential election or even closer.

Of the nearly 1.25 million vote-by-mail ballots returned for the 2021 Senate runoff election through December 17, only 2383, or 0.19%, had been rejected or marked as spoiled. At the same point in time during the general election, 0.12% of returned VBM ballots had met the same fate.

In this post, we take a deeper look at deficient ballots from the general election to explore what implications they might have for the upcoming runoffs.

Reasons for ballot rejection and spoiling

Ballot rejections fell into five categories, with late ballots and signature issues making up over 98% of rejections (Table 1):

Moving from rejected ballots to ones designated as spoiled, the latter do not have standardized explanations. However, manual inspection of the reasons given indicates that the vast majority of spoiled ballots were caused by voter error (83.33%) or ballot damage (7.26%).

In our analysis moving forward, we combine rejected and spoiled ballots, referring to them collectively as deficient ballots.

The geography and demographics of deficient ballots

Unsurprisingly, the counties with the highest number of deficient ballots are also some of the largest in Georgia: Gwinnett (782), DeKalb (476), Cobb (421), Chatham (290), and Fulton (212) counties make up the top five. However, if we look at counties by the percentage of votes marked deficient, those larger counties are much further down the list, while smaller ones like Taylor (0.60%), Clay (0.49%), Dougherty (0.42%), Washington (0.34%), and Liberty (0.27%) (Figure 1) lead the pack.

When we drill into the demographics of voters with deficient ballots, a clear pattern emerges: the voters affected were younger (Table 2) and more diverse (Table 3) than the overall Georgia electorate; however, there wasn’t a big difference in terms of gender (Table 4). Perhaps most striking is that Black and Asian voters — two populations that were key to Biden’s victory in Georgia — were overrepresented in the population of voters with deficient ballots. For example, Asian voters’ share of voters with deficient ballots was 2.5x higher than their share of the electorate.

New voters seem particularly prone to ballot rejection and/or spoiling: Voters with deficient ballots in the general election were 11 percentage points more likely to have registered to vote in the last two years compared to the electorate at large (41.38% vs. 29.62%, respectively).

The importance of ballot curing

Thankfully, a deficient ballot does not have to be final. Georgia, like a few other states, allows voters to cure certain deficiencies on their ballots. We estimate that more than 5300 Georgia voters cured their ballots during the 2020 election. Combining data from individuals who cured their ballots with data from individuals whose ballots were deficient gives us the most complete picture of the rejection process. Specifically, we observe that:

  • For gender: Female voters were more likely to have their ballots designated as deficient at any point in time compared with male voters. However, they were also more likely to cure their ballots than their male counterparts. As a result, the female/male share of voters with ballots ultimately marked as deficient is similar to the makeup of the electorate.
  • For generation: Perhaps unexpectedly, it was Generation Z, Boomers, and the Silent Generation and older that were much more likely to have their ballots marked deficient at any point in time. However, older voters made the effort to cure their ballots at higher rates than their younger counterparts. The end result is that Generation Z made up a larger portion of voters whose ballots were ultimately marked deficient compared to their share of the electorate.
  • For race: Black and Asian voters were much more likely to have their ballots marked deficient at any point in time compared to voters of other races. On the other hand, white voters were significantly less likely to have their ballots marked deficient. However, Black voters appear to have taken greater initiative to cure their ballots than voters of other races. Despite those efforts, Black voters were still heavily overrepresented in the population of voters whose ballots were ultimately marked deficient.

Race gap in Senate runoff deficient ballots

When we examine the small sample of voters who have had their ballot marked deficient in the runoffs, we find that Black voters make up a larger proportion of voters with deficient ballots compared to their share of voters who have returned mail-in ballots thus far. That is, while Black voters made up 30% of voters who had returned their runoff ballot as of December 17, they represented 43% of voters whose ballots had been rejected or marked as spoiled.

With more than two weeks left until Election Day, voters still have time to fix their ballots.

Follow us on LinkedIn, Twitter, and YouTube.




A powerhouse of technology, data science, and creative innovation in support of progressive causes.