Election season does not end on Election Day for local registrars, especially if a municipality is subjected to an audit of its results by the state.

In late November, Secretary of the State Denise Merrill randomly selected 38 polling places from throughout the state to be audited. Among those chosen were New London’s Science & Technology Magnet High School, Waterford’s Oswegatchie School and Norwich’s Rose City Senior Center. Norwich, Waterford and New London registrars of voters traveled to Hartford on either Monday or Tuesday this week for the audit, which scrutinized three races, including state House, state Senate and U.S. Congress.

The registrars all said there were almost no irregularities, and the ones that did show up would have altered races by only one or two votes. The Secretary of the State's Office, University of Connecticut and State Elections Enforcement Commission will analyze the results of the audits.

SOTS spokesperson Gabe Rosenberg said the purpose of the audit is to ensure that the machine count accurately reflects ballots marked by the voters.

“The totals from election night are compared to the totals determined in the audit,” Rosenberg said Tuesday. “To date, the tabulators have been accurate and the audits have validated the results from Election Day.”

Per state statute, 5% of polling places that use optical scan machines are subject to audit. In the past, this would mean a somewhat expensive hand count, but in recent years, election officials have had the choice of traveling to Hartford and having results checked by a machine.

“We responded to the audit requirement as soon as we heard we were included, and we took the ballots up to Hartford for an audit using their high-speed scanning equipment,” Norwich Democratic Registrar Dianne Daniels said. “There are a limited number of spaces, so early requests for auditing in Hartford are essential.”

Waterford Democratic Registrar Julie Watson Jones explained the motivations for executing the audit in Hartford.

“What the audit is doing is testing the machines. Normally, if we did it here in town, we would have to bring back poll workers and do a hand count,” Watson Jones said. “The machine is going to do a better job of replicating what our machine did. In Hartford, we looked at batches of 50 ballots, and they’re all brought up on a giant screen. Both registrars look at each ballot one by one on this giant screen, and anything that questions whether it went through our machine, it would highlight in red.”

Norwich Republican Registrar Dianne Slopak said the audit, which took about three hours on Monday, would have taken days and a more robust staff, including poll workers, if it had been done in Norwich.

“This is the third time that we’ve been able to go to Hartford for the audit,” Slopak said. “It saves the city a tremendous amount of money because Dianne (Daniels) and I are on salary, so we’re already being paid for the day, and it’s just a matter of us going to Hartford. If we were here, we would’ve had to pay a whole lot of people.”

She described the audit process similarly to Watson Jones.

“The machine scans the ballots, it puts them on the screen, and we have visuals of every single ballot, and we can judge whether we think they were tabulated correctly,” Slopak said. She said Norwich brought in about 2,500 votes.

New London Democratic Registrar Bill Giesing said it was the first time New London had gone to Hartford for a machine audit.

“So we went up to Hartford, and we went to one of the stations set up for towns. We separated our ballots into piles and fed them through approximately 50 at a time,” Giesing said. “We then had to view the screen to make sure the ballots were read properly.”

He said he and New London Republican Registrar Rob Pero were pleased with the audit outcome.

“All the numbers came in great,” Giesing said.

Watson Jones said that of the 2,131 votes Waterford brought in for the audit, 2,130 were counted. The machine in Hartford didn’t pick up the checkmarks one voter had used instead of filling in the circles on the ballot.

In an election defined by absentee voting, none of the three New London County towns audited had absentee votes examined. At the moment, an audit only includes absentee votes if they are counted at specific polling places. In Waterford, New London and Norwich, all absentee ballots are counted centrally.

“Centrally counted absentees have never been counted. This is the first year where absentees have been so big, so they wouldn’t have been able to change it that quickly,” Watson Jones said. “When you’re talking about setting up new statutes, it’s a process. That’s just not the way it’s set up right now. In the future I see that probably changing, but right now, these audits were done specifically to test these machines.”

The registrars all recognized the import of election accountability. Still, this year’s election season called for longer hours for election officials in addition to learning and implementing new processes. So, is it annoying to have to go through an audit after an already taxing several months?

“More than you could possibly know,” Slopak said.

She noted the frequent selection of Norwich polling places for the audit.

“Theoretically, this is a totally random audit system in this state, and we come up just about every other year. Something’s fishy with that, if you ask me,” Slopak said. “I’ve been a registrar for 11 years now, and I believe we’ve been audited five or six times. That’s crazy. The small towns are very rarely chosen. And we’ve never had an issue, all this time, we have never had an issue with an audit, so it’s kind of mind boggling to me. In my opinion, if you haven’t had any trouble with an audit in a couple of years, leave us alone!”

About 1,675,000 votes were cast in Connecticut’s 2016 general election compared to more than 2,330,000 this year. A record number of people — 665,597 — voted by absentee ballot this year.

s.spinella@theday.com