Paper ballots recounted to check election results in Georgia

A polling assistant cleans a voting machine at the Paulding County Municipal Building on Oct. 14 in Dallas, Ga. Paulding and six other counties were the first to test the new machines that will be in all precincts in the state for the March 24 presidential primary election. (Photo/Rebecca Wright for the Atlanta Journal-Constitution)

A polling assistant cleans a voting machine at the Paulding County Municipal Building on Oct. 14 in Dallas, Ga. Paulding and six other counties were the first to test the new machines that will be in all precincts in the state for the March 24 presidential primary election. (Photo/Rebecca Wright for the Atlanta Journal-Constitution)

A recount of ballots printed out by Georgia’s new voting system confirmed the accuracy of electronically counted election results, state election officials said Wednesday.

But critics say the state’s audit proved nothing, and they believe ballots created by computers remain vulnerable to tampering and inaccuracies.

Election workers on Tuesday reviewed a sample of paper ballots printed by touchscreens during last week's election in Bartow County, one of six counties that tested the state's $107 million voting system. Voters in the rest of the state will switch to the new system starting with the March 24 presidential primary.

“An important part of the new voting system is the ability to audit with the use of paper ballots. This feature provides the confidence voters deserve,” Secretary of State Brad Raffensperger said.

During the audit, four teams of two election workers each pulled a random sample of 80 ballots out of 1,550 cast in Cartersville. The teams read the printed-out text on the ballots and tallied the results in the race for mayor and a referendum on Sunday morning alcohol sales.

The review showed there were 66 votes for Cartersville Mayor Matthew Santini and 12 votes for challenger Nicole Butler, a result that confirmed Santini's victory on Election Day with 85% of ballots cast, said Mark Lindeman of Verified Voting, a national election integrity organization that helped with the audit. The probability that Santini didn't actually win based on the audit sample is about 1 in 1 billion. The audit also confirmed the outcome of the alcohol sales question.

“It doesn’t take a statistics degree to figure out there was a preponderance of votes for Santini,” said Lindeman, Verified Voting’s director of science and technology policy. “There was strong evidence that the reported outcomes were correct based on the ballots.”

Election workers also conducted a full hand recount of all 480 paper ballots in the Cartersville East precinct. That count showed a one-vote difference from the electronic tally, likely because of human error during the audit, Lindeman said.

Rhonda Martin, an election integrity advocate from Fulton County, said audits of computer-printed ballots fail to confirm election results because many voters don’t check their ballots for accuracy. She wants Georgia to use hand-marked paper ballots, which would create a voter-verified source document for audits and recounts.

"It shakes the foundation of doing an audit when it's based on these documents," said Martin, a plaintiff in a federal lawsuit seeking paper ballots filled out by hand. "We really care about the intent of the voter, and that's not part of that process."

Martin said the public should have been able to observe the audit, but many people weren’t able to because the secretary of state’s office didn’t announce it until minutes before it began Tuesday.

Audits of paper ballots will be required throughout Georgia starting with the November 2020 election, according to a state law passed this year. The State Election Board is considering standards for conducting audits and recounts.