RICHLAND COUNTY, SC — Richland County’s elections office used turnout from previous elections to help decide the number of voting machines distributed last month, two poll managers and a machine technician said.
That might have been one of many miscalculations by the Elections & Voter Registration office – but so far not publicly acknowledged – that prompted machine shortages that created hours-long lines and disenfranchised uncounted others.
State law requires one machine for each 250 registered voters. The law has no specific provision for using turnout as a gauge.
“It was definitely based on turnout,” Ward 11 poll manager Michael Sullivan told The State newspaper of a conversation in October with a key staffer in the elections office when he complained that too few machines were designated for his Ben Arnold Boys and Girls Club polling place in Rosewood.
“That was a novel interpretation of the law I’d never heard of,” said Sullivan, who has served four years as a poll worker in Richland County.
State Elections Commission director Marci Andino said county elections officials may use turnout or absentee ballots to decide how to allocate machines to precincts with histories of low or high voter turnout. But officials may not cut the countywide number.
“You never want to reduce the (countywide) total,” Andino said. “You want to use one machine per 250 voters and then you can use turnout as a factor ... to make adjustments as needed.”
Other poll managers and technicians reached by the newspaper either said they did not hear that turnout was a factor or they did not recall it being mentioned. Most of the 17 technicians who worked on Election Day, Nov. 6, declined to speak to a reporter.
Elections director Lillian McBride, in her only public accounting of the fiasco, said someone in her office reduced the calculation of 864 machines she had determined in June would be needed. That person has yet to be publicly identified. She did not mention turnout.
Steve Hamm, the Columbia attorney hired by the county Board of Elections & Voter Registration, said he has no information that McBride or her staff took prior turnout into account in deciding the numbers of machines.
Board chairwoman Liz Crum agreed, as did Sen. Darrell Jackson, D-Richland, who has been among the most outspoken legislators saying her mistakes should not cost McBride her job.
Brandon precinct manager Al Roblee said he was told on the Friday before the vote of a “formula” that adjusted for turnout in the 2010 gubernatorial election.
“Somewhere in the process there had been a formula worked out,” said Roblee, whose precinct’s voters cast their ballots at Annie Burnside Elementary School.
The formula was “based on the 2010 election and the number of voters and the number of votes cast per machine,” he said, adding he never saw the formula but discussed it with elections office precinct coordinator Becky Brown during his requests for more machines for the Brandon precinct.
Even if using turnout had been proper, the calculation would have been inaccurate, Roblee said.
“That’s a fallacy in the (formula). They used the wrong year,” he said. “It should have been 2008.” That’s because, unlike 2008 and this year, 2010 was not a presidential election year when turnout is highest. Turnout figures for 2010 show that 41,000 fewer Richland County voters cast ballots than in 2012.
Andino said that twice in pre-election meetings, she warned county elections officials to steer clear of comparisons with the 2010 election.
“That’s exactly what I told them this year: not to look at 2010 but to look at 2008 – go back four years, not two, because we had a lower turnout, particularly in absentees in 2010,” Andino said.
Andino said she offered that advice most recently at a Sept. 13 workshop that four Richland County election staffers registered to attend. Those staffers include Brown and Cheryl Goodwin, a former State Elections Commission employee who now oversees training for machine technicians and the preparation of machines for elections in Richland County.
Roblee said he’s been Brandon poll manager since 2008. He said his conversation with Brown on Nov. 2 at elections headquarters was about his request for 14 machines instead of the eight that had been delivered.
“That’s all you need, based on the formula,” Roblee said Brown told him. “What she told me is that Cheryl (Goodwin) used a formula based on the 2010 election. I have no idea what the formula looked like.”
Efforts to reach Brown and Goodwin Friday were unsuccessful.
Roblee, asked by a reporter why he did not challenge the formula and cite the law, said he knew the elections office would not budge. “They weren’t going to make a lot of changes,” Roblee said.
Sullivan, of Ward 11, said he learned of the turnout calculation when Brown responded to a Sept. 27 email he sent her seeking more than the three assigned machines. His precinct received five machines in April for the City Council elections, when turnout hovered around 12 percent. The turnout countywide last month was 65.5 percent.
Recounting a face-to-face conversation on Oct. 2 with Brown during a training session for poll managers, Sullivan said, “To the best of my recollection, she said, ‘We’re basing it (the number of machines) on 250 people who actually voted.’ ”
“I can’t recall if they were basing it on 2008 or some average (with other elections),” said Sullivan, who has vowed never again to work Richland County polls as long as McBride remains director.
More than six weeks after the election and following a legislative delegation hearing and a preliminary report for an attorney hired to get to the bottom of the mess, Sullivan said, “I still can’t figure out how they allocated machines.”
Machine technician Andy Clark said he learned of the turnout calculation during an Oct. 18 training session for technicians. Goodwin conducted the mandated training in a fourth-floor conference room at the county administration building, which houses the elections office.
“I heard her say, ‘We compare the voter turnout in previous elections to determine the number of machines,’ ” Clark said.
“She was referring to ... voter turnout as a gauge not only as to the allocation of machines but as to staffing polls,” according to Clark, who said he has worked as a “rover,” the term for technician, during three elections this year. The Oct. 18 session was one of several sessions held for technicians in preparation for election season.
Clark was assigned to troubleshoot machines in four precincts in Lower Richland.
“I would never want to work for anyone who wouldn’t honor their accountability,” Clark said of McBride, Goodwin and Brown.
Roosevelt Martain said he attended the same training session. He said he does not recall any mention of turnout and doubts the number of machines being sent to precincts was discussed.
“I don’t know why that would come up,” Martain said, citing that technicians review proper machine maintenance and troubleshooting. Rovers are assigned to zones across the county for which they are responsible. Martain said his zone included seven precincts, largely in northern Columbia.
Martain, who has 12 years of experience as a technician, blamed the election meltdown mostly on the size of the turnout.
“To me,” Martain said, “it was just a national presidential election.”
But the 2012 meltdown was not just another election to 70-year-old Rallie Sigler, who lives in the WildeWood subdivision in Northeast Richland.
“It’s just a sinking feeling to know that something you treasure so much was sort of seemingly handled in such a flippant manner,” Sigler said. “It was treated in a way that sort of demeaned it.”
More Monday on what went wrong
The Richland County Elections & Voter Registration Board on Monday will hear an update of attorney Steve Hamm’s interim report on what went wrong Election Day.
The board will meet at 4 p.m. on the fourth floor of the county administration building, 2020 Hampton St.
Reach LeBlanc at (803) 771-8664.