The sheer scarcity of polling locations was one problem. Where they were located was another.
Eighteen of 60 locations — nearly a third — had fewer than 1,000 voters all day, with several servicing just a few dozen voters each. The report, obtained by the New Times on Thursday after filing a public-records request, details the vote counts in each of the 60 polling places, all of which were open to any would-be voter. (See the report below.)
County officials slashed the number of polling locations from 200 in the 2012 Preference Election, calling it a cost-cutting move. This drastic step was not followed by careful regard to geographic population difference, the report shows.
The super-long wait times in Maricopa County caused frustration, outrage, at least one protest, and calls for official investigations. Waiting an hour or more was typical for many voters in metro Phoenix; some waited as long as five hours, with hundreds still in line long after the media reported that Hillary Clinton and Donald Trump had won by double-digit margins. People in line by the official closing time of 7 p.m. were allowed to vote; some didn't go home until after midnight.
County Recorder Helen Purcell accepted blame for the problem, but also committed a public-relations faux pas when she told a TV reporter that voters were to blame “for getting in line.” Governor Doug Ducey called the bungle “unacceptable” and called for a new law that would allow independents to vote in Arizona's Presidential Preference Election as they can in all other major elections.
Phoenix Mayor Greg Stanton called for a federal investigation on Wednesday, and Arizona Secretary of State Michele Reagan called for bipartisan public hearings.
The new report shows a sporadic voting pattern among the county's few polling stations.
For much of Tuesday, lines were relatively short for large parts of the day at the South Mountain Community Center near Central and Southern avenues, and the Cooperative Extension at 43rd Street and Broadway Road, which only had 809 and 872 voters, respectively.
While voters waited in line on Tuesday for Arizona's Presidential Preference election at the Fountain Hills Community Center, which serviced 1,207 people, the report shows it was a slow day at the Fort McDowell Indian Community Recreation Center's assigned polling spot – a short drive away. Twenty people voted all day there, the report shows.
Putting three in Gila Bend wasn't the best idea — there were only 140 people total at all three, all day. It would have been better to have had two more in the Mesa-Gilbert area, which got socked with too many voters and too few locations. The three stations in Gilbert received a total of 9,013 voters.
The report shows other interesting differences: In Goodyear, Peoria, and north Phoenix, two or more locations with far fewer voters were just minutes away by car. The county's website is high-tech, with a button that showed voters which three polling locations were nearest to a user's mobile device – but it needed information about the lines like the “Wait Times for Disneyland” app.
The wait-time problem probably grew worse almost everywhere, though, around 4 p.m., when the Valley's biggest procrastinators got off work and in line.
Officials have given several reasons for the voting fail, and much blame has been heaped on Purcell, a seven-term politician. She's taking back the blame-the-voters comment. Her spokeswoman, Elizabeth Bartholomew, says it was “wildly taken out of context.” (Though it doesn't sound out of context in Channel 10's video, which can be seen below.)
Blame's been flying around, that's for sure. In a release on Wednesday afternoon, Reagan said she had “no explanation why county election officials decided to reduce the number of polling places...”
Yet, Reagan and Purcell gave a news conference together on Monday about the upcoming election, and neither gave the slightest hint or warning of the fiasco to come. They did, however, point out that turnout was expected to be higher than normal.
And that's another weird thing about the voting tieups: Bartholomew says county officials estimated that 10 percent of county voters would show up at the 60 polling locations. That's about 120,000 people out of the county's 1.2 million eligible voters. But only 6.7 percent showed up, she says.
Nearly all of the rest voted early by mail.
In-person turnout on Tuesday “was perceived as high simply because there were fewer polling places,” she says.
Reagan, conceivably, could have asked for an explanation of the reduction in polls before Tuesday – after all, she's the state's top election official. Purcell's office points out that the five-member county Board of Supervisors all voted to approve the county's election plan, which was enacted to save money.
A warning given at one Board of Supervisors meeting about the need for more polling stations wasn't discussed by officials, and the problem was scheduled by Supervisor Andy Kunasek to be talked about after the election, Bartholomew says.
County officials say a large number of independent voters standing in line without the ability to vote in the election gummed up the works. Only Republicans, Democrats, and Green Party members were allowed to vote on Tuesday. But independent voters or those who the system didn't recognize were given provisional ballots, some of which were still being counted on Thursday night.
It's unclear exactly what impact these voters had on the length of lines, especially when there were so few polling locations overall. But some stations processed far more provisional ballots than others, the report shows, which may have led to more delays at those stations.
For instance, while the polling station with the most voters on Tuesday, the Southeast Regional Library in Gilbert, gave out 554 provisional ballots out of 3,189 total voters who showed up, or 17 percent, at some other locations, 40 to 50 percent of the people who showed up needed a provisional ballot.
At the Tolleson Parks and Recreation Center, for example, 789 people out of 1,394 voters required provisional ballots.
All told, 23,264 provisional ballots were given out at the polling locations out of 80,081 total ballots, for a total of 29 percent. Less than 10 percent of ballots in the 2012 general election were provisionals.