Votebeat is a nonprofit news organization that covers voting access and election administration in the United States Sign up for Votebeat Arizona’s free newsletter here.

A Pinal County supervisor who lost his bid for sheriff is calling on the Arizona Attorney General’s Office to investigate the county’s primary election results, claiming he has identified a suspicious pattern that suggests inaccuracies in the results.

The candidate, Republican Kevin Cavanaugh, presented his analysis to Votebeat, saying he believed the data provided “strong evidence that fraud occurred.”

But two independent analysts who reviewed the results said that was not necessarily evidence of malpractice or error. And officials in the southeastern Maricopa County district are confident the election went smoothly.

A spokesman for the attorney general’s office confirmed receipt of a complaint from Cavanaugh, but said the prosecutor’s office could not confirm whether it would investigate the case.

Cavanaugh said he would raise his concerns about the results when supervisors meet on Monday to certify the election results.

Voting deviated from usual pattern, says supervisor

The pattern Cavanaugh said he has identified concerns the similarities between early voting results and Election Day voting in at least a handful of local elections. There is usually a difference between how these two groups of voters vote: In recent elections, for example, Republicans in Arizona were more likely to vote in person than Democrats. Cavanaugh said in the local elections he studied, the results for both groups were nearly identical, which he believes is a sign that something went wrong.

But two Arizona-based election data analysts who reviewed Cavanaugh’s analysis at Votebeat’s request – Republican Benny White and Democrat Sam Almy – say the voting patterns he identified do not necessarily indicate fraud or error. Many factors contribute to how and why voters vote, they said, and voter behavior has become increasingly erratic in recent elections.

“It’s certainly a curious thing and very interesting historically,” that this pattern contradicts other recent voting patterns, White said. “But to me that’s all – it’s just a curiosity.”

District Attorney Kent Volkmer, also a Republican, agreed with Cavanaugh that the pattern was odd, but Volkmer also said that after checking with election officials about how the election was conducted and what checks and balances were in place, he had confidence in the outcome.

“They were able to explain to me to my complete satisfaction that nothing unusual happened in this election, that nothing was fishy and that there was nothing suspicious about it,” said Volkmer.

Republican Dana Lewis, who oversees elections in the county, said she also believes the results are accurate for many reasons, including that pre- and post-election testing and a manual count conducted by her office showed that the machines were programmed to work accurately. She also pointed to confirmation from observers from all major political parties that everything went well.

She said she and her team would not be distracted by the efforts of those who promote the hackneyed and ridiculous narrative that poll workers in any way influence the outcome of an election – and this in a facility designed for transparency and whose staff are committed to the accuracy of the results.

District Attorney sees benefit in expanded hand counting

Lewis said this election in the county went much better than the midterm elections, which further increases her confidence in the outcome.

In November 2022, the county initially failed to count hundreds of votes after staff made several errors while counting ballots. A countywide recount uncovered the problems and updated the final results. The county has since moved to a new building, developed new processes to stay more organized, and hired new staff.

Cavanaugh has long been a critic of the county’s Board of Elections and Lewis. For example, earlier this year he filed a court brief saying he was preparing a defamation suit against Lewis, claiming the county had failed to properly investigate the whistleblower’s claims. The district attorney later said he believed the whistleblower’s claims were without merit.

Cavanaugh’s team’s analysis of recent results shows that in races for county sheriff, district attorney, tax assessor and two supervisor districts, Election Day and pre-election votes for those candidates were within one percentage point of each other. In his race, for example, he received 31.0% of the pre-election vote and 30.1% of the Election Day vote.

He called this “odd” considering that looking at primary election results in the past has often found a larger difference between a candidate’s share of the early vote and their share of the vote on Election Day.

Volkmer said he noticed this pattern when the results of his race came in, and they surprised even him. He lost his primary to a Cavanaugh ally, Brad Miller.

“It never changed from initial decline to early maturity to late maturity,” said Volkmer. “I was personally surprised that I couldn’t make up any ground.”

Still, Volkmer said he and his campaign team accepted the result and that the pattern may have something to do with low voter turnout.

Cavanaugh said he is interested in asking a court to order an expanded manual count of those specific races – but not before supervisors certify the results on Monday.

Arizona law requires a hand count of a certain number of ballots after each election. But that law only requires the counting of certain precincts and only a certain number or percentage of ballots. The Arizona Court of Appeals ruled in October that counties cannot expand their hand counts in ways not provided for in state law.

Volkmer said he believes an expanded manual count could increase voters’ confidence in the results.

Democratic Party Chairwoman Lisa Sanor, who observed the machine test and vote counting on election night, said she was confident in the results and did not notice anything suspicious that evening.

But after speaking with Cavanaugh, she agreed that a manual count of the races he identified would settle the matter.

“For transparency, right?” she said. “To suppress it.”

Checks and balances include tests and live cameras

Lewis said she was disappointed and frustrated by Cavanaugh’s claims. She said both the county’s and the secretary of state’s logic and accuracy tests showed the machines were 100 percent accurate.

The county counts votes using machines that are not connected to the internet and is monitored 24/7 by security cameras that are broadcast live on the county’s website. There are strict rules about who can enter and leave the counting room, and party observers are present at all times.

Almy and White cautioned against drawing conclusions from a single pattern in a data set. Other data, such as voting behavior, could help explain the unusual pattern, White said. However, Arizona counties are not currently granting access to that database.

Additionally, Almy said, high-profile and national elections are more likely to experience differences in vote counts that sometimes occur between early voting and voting on Election Day. In local elections with low voter turnout, Almy said, voter behavior is more likely to be explained by looking at individual candidates and campaigns.

“A lot of it depends on where they are on the ballot,” Almy said.

And while such trends might have been more predictable in the past, since 2020, misinformation about the insecurity of mail-in voting combined with COVID-19 has confounded voting behavior in key national elections, both White and Almy noted.

“It’s easy to jump to conclusions from a single set of numbers,” White said. “But that’s not necessarily the whole story.”

Jen Fifield is a reporter for Votebeat based in Arizona. You can reach Jen at [email protected].