Arizona Republican voter turnout was just 0.35% off

Earlier this year, the Arizona Republic set out to project how many state residents will vote in the November election.

Projection was the basis a tool that helps readers know how many votes are left to count in the presidential election – and it proved to be on target.

The Republic’s projection was that 3.37 million Arizonans (or 3,377,525 to be exact) would turn out to vote this year, about 77% of the eligible electorate. The Arizona Secretary of State’s office reported this week that 3,389,319 people voted in the presidential contest.

The difference between these figures is about 0.35%.

Not everyone who voted actually voted for one of the presidential candidates; of the total votes cast, the projection of the Republic was about 1.5% lower than the actual number.

“That’s impressive!” Paul Bentz, senior vice president of research at Arizona-based political consulting firm HighGround, said before the election told the Republic he expected between 76 percent and 78 percent of eligible voters to show up.

More than 3,420,000 Arizona voters — or about 78.5 percent of the state’s eligible electorate — cast ballots in this year’s general election, according to the data. data from the state secretariat.

Most of those who attended voted in the presidential race: Fewer than 40,000 Arizonans did not choose a presidential candidate, figures from the secretary of state’s office show.

With more people submitting their ballots late — combined with Arizona’s batch of withdrawals on Election Day — coming up with an accurate projection was “a little harder” this year than in the past, Bentz said.

“Having done this for 20 years … we often say there’s no such thing as an election day in Arizona, there’s an election month,” Bentz said. “A lot of people made it almost more like an election day than we’re used to.”

How the Republic estimated the turnout

The Republic’s data team modeled and analyzed historical election data to arrive at an estimate of voter turnout, which was rounded down to the nearest 10,000 votes.

Reporters used a combination of several approaches — from the simple to the complex — to generate predictions, then averaged them to create a rounded turnout projection.

Members of the data team worked with Christopher Weber, director of the Arizona Voter Project at the University of Arizonato model voter turnout in presidential elections. Weber has a PhD in political science and a background in statistics.

With Weber’s guidance, a reporter used data collected by L2, a research company that combines voter registration information from the Arizona Secretary of State’s Office with demographic information about registered voters across the state, to conduct several analyzes of regression.

Statistical methods allowed for a better understanding of the relationship between voting in previous presidential general elections and factors such as an individual’s age, voting history, and party affiliation.

“If you’re going to predict whether someone votes or not, I can’t think of a better measure than who voted in the last election,” Weber told The Republic. “You’ve built a regression model that takes information from past elections to predict the present. And because there’s a remarkable degree of stability in whether people vote, especially now, your indicator is probably the best you’ve got.”

The reporters used cross-validation, a standard method in the machine learning industry, to ensure that their models would perform well on new data and generate more accurate predictions.

Contact Sahana Jayaraman at [email protected]. Follow X @SahanaJayaraman