Georgia election audit finds close match in hand and machine counts

Audit reviewed 231,000 ballot sample in secretary of state’s race



An election audit completed Friday showed similar results between a hand count and machine tally of ballots in the secretary of state’s race, a result that Georgia election officials said confirmed the outcome.

The audit examined a randomized sample of 231,000 ballots cast in the secretary of state’s race and found few discrepancies in Republican Brad Raffensperger’s reelection over Democrat Bee Nguyen.

Across the state, the audit counted 21 more votes for Raffensperger and 18 fewer for Nguyen.

“This audit shows that our system works and that our county election officials conducted a secure, accurate election,” Raffensperger said.



Georgia law requires an audit of one race after each general election, and Raffensperger decided to check his own win, which had the largest margin of victory in any statewide race, at 9.3 percentage points.

Raffensperger declined to audit the much closer race for U.S. Senate, which is heading to a runoff between Democratic U.S. Sen. Raphael Warnock and Republican Herschel Walker.

Almost all batches of ballots examined were within two votes of results recorded on election night, according to data released by the secretary of state’s office. Of 328 batches, eight of them were off by more than two ballots.

Discrepancies were within the expected margin of error and likely caused by human error during the hand-counting process, according to the secretary of state’s office.

Credit: Georgia secretary of state

Credit: Georgia secretary of state

The largest inconsistency occurred in Clay County, where auditors counted 11 more votes each for Raffensperger and Nguyen, as well as four more votes for Libertarian Ted Metz.

The county elections superintendent investigated the issue and found it was likely caused by test ballots mixed in with actual ballots from the election, according to the secretary of state’s office. Test ballots are supposed to have an identifier, such as a watermark, that distinguishes them from official ballots, but these ballots didn’t.

The audit included all 159 counties in Georgia and was the state’s second post-election ballot review, starting on Thursday and ending Friday.

Georgia’s first election audit took place after the presidential election in 2020, when all 5 million ballots cast were recounted to determine that Democrat Joe Biden had in fact defeated Republican Donald Trump by about 12,000 votes.

This year’s audit covered a sample 6% of 3.9 million ballots cast to ascertain with a greater than 95% confidence level that the outcome of the race was correct.

Critics of Georgia’s voting machines, manufactured by Dominion Voting Systems, say the audit wouldn’t necessarily detect problems if ballots had been manipulated.

But state election officials say paper ballots provide a record to verify computer tallies, and there’s no evidence that voting equipment has ever been tampered with during an election.

The audit was conducted with software from VotingWorks, a nonprofit company that the state also used in its audit two years ago.

Audit data is available for download from the secretary of state’s website at