Pennsylvania’s primary results are still unofficial. To be approved, counties go through several processes to make sure everything worked well.
A big step is auditing votes to make sure they were counted correctly on primary night. Counties finished the process Friday, the last major hurdle before Monday’s certification deadline.
Results go through two checks. The first is a 2% audit. The second is a newer method called a risk-limiting audit. Commonly called an RLA, it is considered the gold standard tabulation audit recommended by the Department of Homeland Security and the National Academies of Sciences, Engineering and Medicine.
Mark Lindeman, policy director with the election technology and security nonprofit Verified Voting, helped the state design its RLA system.
“An RLA is a statistical sample of votes from all over the state and it’s simply designed to make sure that the winner really got more votes,” Lindeman said.
Pennsylvania is one of six states that require a statewide RLA, according to Verified Voting.
Lindeman said the law requiring the 2% audit is vague in how it should be carried out, as opposed to the RLA, which has a detailed methodology that “produces results that ring like a bell.”
Jonathan Marks, deputy elections secretary for the Department of State, said the audit is designed to be transparent and ensure results are correct. The process is done in public view, from the process of randomly selecting the race and ballots to be audited to counties conducting the audit.
The state requires all members of a ballot review board be duly sworn in. Centre County goes a step further and has the local Democratic and Republican parties send representatives to observe the process.
“I’m not deluding myself into believing that we’re going to convince every skeptic, but I think providing this transparency and undertaking these post-election processes prior to certification will give some individuals confidence where they may have had some doubt,” Marks said.
This special type of audit adapts to every election, so in races with small margins, lots of ballots are checked, while in races with large margins elections officials have less work to do.
“The math is complicated, but the idea is simple,” Lindeman said. “If you check vote counts from all over the state and they all can come back fine, you have pretty strong assurance that the vote counts are good.”
Centre, Columbia, and Dauphin election directors all reported RLA results with no discrepancies between the hand-audit and the machine count. The Department of State is scheduled to publish a statewide report early next week.
Read more from our partners, WITF.
