For the first time in at least several decades, Park County poll workers counted some local ballots by hand in a Thursday audit — and came up with the exact same results as those tabulated by machine on election night.

From roughly 9 a.m. to 1:30 p.m. Thursday, a dozen election judges manually reviewed 524 ballots that had been cast at the Park County Fairgrounds on Aug. 18. The group tallied how many votes were cast for which candidates across seven of the more than two dozen races that appeared on Powell area ballots.

When the results were read aloud in the county clerk’s office early Thursday afternoon, the group collectively cheered the news that their hand-counted tallies were a perfect match for those tabulated by the machine on election night.

Judge Vicki Taylor expressed hope that the identical totals would show the public that the machine counts are correct.

“That’s what we’re hoping for,” responded Park County Clerk Colleen Renner.

“Well, you need to convince the people of that,” offered judge Paul McConnell, “and that’s the tough part.”

The push for hand counts

There’s been a growing movement for hand counts since 2020, when President Donald Trump claimed that widespread fraud led to his loss to former President Joe Biden.

For years, a local group led by state Sen. Dan Laursen, former Park County Republican Party Chairman Larry French and Tim Lasseter of Cody has pushed the county commission to stop using their electronic tabulators and switch to a hand count. The current leaders of the local Republican Party have also issued calls for hand counting, but the county commissioners have said state law requires ballots to be counted by machine.

Advocates’ efforts to change the statutes have generally floundered in the Wyoming Legislature, but this past winter, lawmakers did authorize a one-time hand count audit. Senate File 113 requires counties to manually count certain races that appear on about 5% of the total ballots cast in each election. One of the ES&S DS200 tabulators that was deployed at the Powell fairgrounds processed 5.2% of the more than 10,061 ballots cast across the county and was chosen for the local hand count comparison.

Counties were required to check the results of the U.S. Senate race, the race for superintendent of public instruction and all legislative races; in this case, that included portions of Senate District 19 and house districts 25, 26 and 50. The Park County Clerk’s Office also decided to review one of the contested county races, and the treasurer’s race was chosen at random.

Last week’s hand count was in addition to a separate mandatory audit in which counties must check a certain number of scanned ballot images and make sure the machine correctly logged the voter’s choice. That audit — which was conducted on Aug. 20 — also confirmed Park County’s results.

Thousands of votes to count

The election judges wound up hand-counting 2,547 individual marks on Thursday — tallying the number of votes for the candidates on the ballot, for candidates written-in, for too many candidates (overvotes) and blank sections of the ballot (undervotes). There were a total of 268,759 choices logged across the county’s primary election, which means Thursday’s audit checked not quite 1% of the total.

If the group’s pace was representative of how long a full hand count would take, the county could need as many as 4,800 man-hours to manually count the entire election. That’s the equivalent of 120 people putting in a full workweek.

However, Thursday’s process wasn’t built for speed or for a full hand count, and First Deputy Park County Clerk Hans Odde said he believed at least one other county worked at a faster clip; hand count advocates have said their trial runs have resulted in relatively speedy processing.

Lasseter and French observed a portion of Thursday’s audit, and Lasseter remarked that he believed the process had been set up to take as long as possible.

Odde told the Tribune that the state’s county clerks collaborated on developing a uniform process for the hand count and generally based them on techniques taught by hand count advocate Rick Weibel.

Regardless of how fine-tuned the process is, there’s no disputing that the machines are multiple times faster than people. For example, all of the county’s votes were tallied and publicly released by 8:45 p.m. on election night — less than two hours after the polls closed.

Human error

The workers started off the morning by making sure they had 524 ballots. Their initial tally came to 522, but after doing it again, the judges realized they’d skipped over two ballots in their initial count.

The poll workers generally split into groups of four to count ballots, with two people audibly confirming each vote cast and two people independently counting the votes. The groups’ tallies generally matched the machine’s on the first try, though the final batch of Democratic ballots — which was processed a little differently — had to be counted three times before the judges’ count agreed with the official results.

In a Friday interview, election judge Richard Jones said he hopes people are satisfied that the machines work, as he called hand counting slow, expensive and error prone.

“I worked in both the casino industry and the banking industry, and they have never ever once counted a piece of money by hand,” he said, adding, “all the problems I’ve ever seen with machines and stuff like that have been human error or programming errors or something else. It’s not been the machines …”

Jones noted that there were several errors within the hand-count that had to be caught and corrected.

“It’s just not the way to do it,” he said, adding, “Humans are terribly imperfect, and machines basically aren’t.”

Still supporting a hand count

As for election judge McConnell, he’d still prefer that Wyoming count all of its ballots by hand.

Around the state, “a lot of people want a hand count, and they do not trust any machine at all,” he said in an interview. 

McConnell believes that distrust is a part of the reason for lackluster participation in local elections.

“If you really look at the number of people that are legally available to come vote, it’s under 50% that vote,” he said, “and that’s a shame. That makes it easy to cheat if there’s a cheater.”

McConnell also has concerns about the machines, believing they can be hacked and the results changed.

“I don’t think that’s happening in Wyoming, but it could be,” he said.

On the flip side, McConnell, like Jones, praised Park County’s hand count operation on Thursday, saying the elections office did a great job doing what state lawmakers required.

“We came up exactly with what it was supposed to,” McConnell said of the matching vote tallies between the people and machine, “so I thought that was fantastic.”

And for those who have doubts about the process, “I hope it makes them a little bit more confident there are things to make elections safe,” he said.

Next steps

Senate File 113 only requires counties to conduct the hand county audit this year, though lawmakers “could come back and codify it into law. Who knows?” Odde said.

He expects Secretary of State Chuck Gray and county clerks will share thoughts about the recent process with the Legislature’s Joint Corporations Committee in October.

Excluding the extensive amount of staff time put into the count, the audit came at a relatively small expense to the county: Paying the 12 judges for five hours of work cost $600, plus a bit more for lunchtime pizza from Domino’s, Odde said.

Clerks will conduct another hand audit following November’s general election. Odde said he expects they’ll review a similar amount of ballots for the general election, but “hopefully a few more, because that would mean that we would have a better turnout than 10,000.”