Election audits match machine tallies in most counties, ‘minimal’ discrepancies in others

Lincoln County Auditor Sheri Lund (left) oversees election workers conducting the county's first-ever post-election audit on June 20, 2024. (Makenzie Huber/South Dakota Searchlight)

Post-election audits across South Dakota matched primary election night results from tabulator machines in most counties, according to statewide data released Wednesday by the Secretary of State’s Office.

Post-election audits are a routine practice in most states to verify whether equipment used to count votes worked properly and yielded the correct result, but the review was South Dakota’s first. The Legislature passed a law during the 2023 legislative session requiring the audits.

Of the state’s 66 counties, 52 reported no discrepancies between their post-election audit and tabulator results. Ten other counties did not have to conduct a post-election audit because they held a recount of a race. Three counties — Mellette, Tripp and Lincoln — reported one discrepancy each.

Minnehaha County reported 52 discrepancies in the full election recount that Auditor Leah Anderson chose to conduct, but she said the list of discrepancies “looks minimal.”

Voters reject machine-counting ban in all three counties where it was on the ballot

The state’s top elections official, Secretary of State Monae Johnson, pledged during her 2022 campaign to implement post-election audits. She described the first audits as a success.

“The transparency of this process, conducted by audit board members from multiple parties, and the accuracy of the audit results should reinforce every voter’s belief that their vote mattered and was counted accurately,” Johnson said in a news release.

County auditors are required to hand count ballots in 5% of their county’s voting precincts, though most counties reviewed more than the required amount. Some counties conducted a full hand count of their results, such as McPherson and Minnehaha counties.

Auditors across the state hope the primary post-election audit will prepare their election workers for the general election in November.

Some auditing boards had to recount races several times, one county re-audited its election results, and paperwork slowed the process down in at least one county, auditors said.

The Secretary of State’s Office report did not specify how long the post-election audits took for each county or how much the audits are costing the state or counties. The Secretary of State’s Office has not yet responded to requests for comment from South Dakota Searchlight.

Minnehaha auditor says paperwork slowed process; commissioner pushes back

Anderson told Minnehaha County commissioners ahead of the audit that she expected to hand count the county’s full 13,189 ballots in five hours. It took 12 hours.

It took about an hour for Moody County to count 104 ballots for its two audited races. Lincoln County took nearly three hours to review three races on 592 ballots.

What slowed the process down, Anderson said, was over 1,593 pieces of paper that audit board members had to print and sign their names on after each race was counted in a precinct.

“I wish there was a way to condense some of the paperwork,” Anderson said. “The paperwork isn’t doing much for an audit.”

Minnehaha County Commissioner Joe Kippley said Anderson’s full recount wasn’t necessary, but did return “what you’d expect if everything is running the way you want it to be.” He added that the level of time and work to run a full recount “could have been predicted.”

“I don’t know if you can make a ‘volume of paperwork’ critique when you made the choice to do 20 times the work you were asked to do,” Kippley said. “When it’s your first year running this new process and state statute says ‘do a 5% audit,’ there should be some self-checking. Maybe it’s 5% for a reason.”

Lincoln County’s 592 audited ballots represented about 10% of ballots cast in the primary. Auditor Sheri Lund said her auditing boards only had to sign 20 pieces of paperwork.

“I don’t think our paperwork was overwhelming,” Lund said.

Post-election audit meant to reinforce confidence in elections, auditors say

In McPherson County, the post-election auditing board reported that it was four ballots short in one precinct for the District 23 House race and had an extra ballot in another precinct compared to the ballots scanned by the tabulator on election night. The county conducted a full recount of its ballots.

Auditor Lindley Howard and the McPherson County state’s attorney asked a judge to unseal the ballots for a re-audit of the results. In Howard’s re-audit, the ballots and votes matched the tabulated results.

“We saw way more human errors through the audit process than machine discrepancies. All our discrepancies were human errors,” Howard said. “I think we proved this time that the tabulators work and I don’t want to do a 100% audit again.”

Some of Moody County’s audit workers are skeptical of the tabulators, which can be a good thing, Moody County Auditor Tawny Heinemann said. Being involved with the audit let them work with and understand the process better, she added.

“They see what marks could look like and why a ballot wasn’t counted and if it wasn’t counted correctly,” Heinemann said. “Being able to see that the numbers matched and what some of the checks and balances are helps a lot.”

Anderson doesn’t believe a 5% audit is enough. She plans to audit about 25% of votes in November.

She worries that the state’s post-election audit process isn’t clear enough where she should stop, though, saying that she would continue to audit precincts if a race is off by one vote until she gets “clean results.”

Kippley said he doesn’t think a “clean results” audit is needed. The audit doesn’t have any influence on the election results, unless an auditing board finds a discrepancy that changes an outcome in an election. Only minimal discrepancies were found in audits across the state.

If a race was close enough, he said, it would go to a recount board anyway, which would have more authority to review ballots.

According to the state post-election audit guide, counties are instructed to conduct a post-election audit and report their findings. They do not have to continue auditing until they get within a certain margin of error.

Analyzing voter intent

Anderson’s audit boards found a discrepancy in 52 of Minnehaha County’s 570 races — most off by one vote. The Minnehaha audit also found that the tabulator machine counted a handful of ballots cast by handicapped voters that weren’t properly stamped by precinct workers.

The largest discrepancy was three votes, which wasn’t enough to change an outcome. Audit boards are only supposed to consider marks the tabulator would count for the audit, not voter intent, according to the post-election audit guide.

Secretary of State Johnson said in a press release that the discrepancies reported by Minnehaha County and others “appear to be due” to auditing boards considering voter intent.

Jessica Pollema, president of South Dakota Canvassing Group, which advocates for the replacement of machine tabulators with hand counting, said the state’s guidelines were made to have auditing boards “think like a machine.”

“That totally defies the purpose of a high end count post-election audit,” said Pollema, who served as the superintendent of the Minnehaha County post-election audit board. Pollema spoke on the McPherson outcome at a recent Minnehaha County Commission meeting.

Linda Montgomery, a Lincoln County resident who served on her county’s post-election audit board, told Minnehaha County commissioners she was instructed to “think like a machine” as well.

Montgomery’s audit board in Lincoln County had a difference of one vote from the tabulator, Lund said. She suspects the difference was due to human error rather than the machine. Another post-election audit board in Lincoln County had to recount a race five times before agreeing on a number.

An audit of the machines is meant to make sure they’re doing their job correctly, Lund said. By law, tabulators and resolution boards are not supposed to consider voter intent . Considering voter intent would cause an “apples to oranges” comparison. It could also open the door to bias, Lund added.

“We give you instructions for marking a ballot. In order for that vote to count, you need to mark it the way we tell you. If you mark it any other way, you’re supposed to ask for a new ballot, like instructions say,” Lund said. “We’re not going to take it and say, ‘Hey, what did you mean by this?’ or ‘I think they thought this way.’”

At least one ballot in Minnehaha County was rejected as an overcount (when a voter marks a ballot in favor of too many candidates) on election day because the voter crossed out a vote for a candidate rather than asking for a new ballot. Anderson used the example to explain that the ballot should have been diverted to a resolution board to identify the ballot as “mismarked” rather than an overcount.

Anderson said she plans to bring the findings to the state and to the election systems vendor for South Dakota, Election Systems & Software.

Kippley said Anderson’s full recount proved his faith in tabulator machines.

“She did a 100% audit in the June primary and it proved my point,” Kippley said. “It came out with basically the exact same results.”

SUBSCRIBE: GET THE MORNING HEADLINES DELIVERED TO YOUR INBOX

DONATE: SUPPORT NEWS YOU TRUST

The post Election audits match machine tallies in most counties, ‘minimal’ discrepancies in others appeared first on South Dakota Searchlight .