Vote results further delayed
Officials vow to fix machine issues by Saturday
By Mike Cherney - The Sun News    21 January 2008

More than a day after the polls closed, Horry County still could not release full results in a Republican presidential primary plagued by county and state election commission mistakes.

Officials vowed Sunday to ensure that the problems that marred the Republican primary would not be an issue for the Democratic presidential primary, slated for Saturday.

Four of the county's 118 precincts had yet to be counted late Sunday night, said Sandy Martin, the county's elections director. She expected final results to be released before noon today.

The S.C. State Election Commission estimated that 20,000 voters, or 15 percent, turned out in the county. That's down from 2000, the year the last Republican primary was held, when 24 percent of registered voters cast ballots.

As soon as the polling began Saturday, voters started complaining about malfunctioning voting machines and emergency paper ballots that were running out. Even though the machines were eventually fixed, another problem was discovered by election officials once voting was over.

The machines' pre-programmed poll-closing time was incorrect, and technicians had to follow a multi-step procedure to manually shut them off before counting could begin. The closing date for the machines was set for Jan. 26, the date of the Democratic primary, instead of Saturday night.

Poll workers are not trained to manually close the voting machines, so they had to drive them to the elections office in Conway where technicians closed them down. Once closed, the machines print out the results.

The state election commission is responsible for sending the opening and closing data to the counties for their machines, said Garry Baum, a spokesman for the S.C. State Election Commission. He did not think a similar issue had happened in the past and did not know if it had happened elsewhere in the state on Saturday.

"All elections are prepared for a one-day election, and in this case we were preparing for two separate ones," he said. "I wouldn't use the word 'confusion,' but it was the state elections commission's responsibility."

Martin said that without the programming error by the state, the county would have been able to tabulate all the results on Saturday night. Less than half of the precincts were counted that night. Another 25 remained to be counted on Sunday.

But in the commotion of the previous evening, all the voting machines had gotten mixed up, so election officials did not know which machine came from which precinct. To match a precinct to a machine, officials needed to open up the case that holds the machine and locate the machine's serial number.

Four machines - one for each outstanding precinct - remain to be tabulated, Martin said late Sunday.

The initial problem with voting on Saturday stemmed from a routine test by the county to ensure the machines were operating properly before the vote. But the machines were not recalibrated correctly after the test for the actual election, Martin said.

As a result, some machines in about 80 percent of the precincts were not working at the start of the vote. One or two were affected in some precincts, but in others, all the machines were down, Martin said.

The machines were working across the county by about 4 p.m. after technicians reset the disabled machines. While the machines were down, 4,000 people voted on paper ballots instead, Martin said.

Some voters said Saturday they were turned away, or told to come back later by poll workers who ran out of official paper ballots. Martin said she was unaware of anyone who was not able to vote as a result of the malfunctions.

If official paper ballots run out, poll workers are allowed to let people vote on any piece paper of paper they can find. But Edward Munns, a poll manager in Myrtle Beach, said that issue was not addressed during a training seminar for poll workers earlier this month.

Poll workers had to attend one of 15 training seminars in the weeks before the election, Martin said. Each were three hours long, and she said it was possible that the paper ballot issue was not addressed in all of them.

Munns said he had to improvise after his polling station at a fire station on 38th Avenue North ran out of paper ballots by handing out sheets of yellow legal paper to serve as ballots.

"I wasn't going to turn anybody away, they were going to get to vote one way or another," he said. "In some places, they turned people away, and I wasn't going to do that in our poll."

During standard testing before an election, test votes are logged onto each machine, said Ken Fields, a spokesman for Election Systems & Software, which manufactures the iVotronic machines used by Horry County. The machine has a security feature that requires it to be reset after any voting, even during a test.

The feature is intended to prevent actual votes from being erased during a real election, Fields said. He said there would have been no problem if proper procedures had been followed.

"Overall, those machines across the country have recorded millions of votes and have have performed really well," he said. "It's terribly important that the called-for procedures be followed in order to prepare for election day."

Fields said the company, which has sold its voting machines to municipalities in 46 states, will work with the county to ensure the proper procedures are followed in the future.

Half of the county's 613 voting machines were used in the election, Martin said, and the other half will be used for the Democratic primary. Martin said those machines will be completely reset, and staff from the state commission will be sent to help. The machines were first used in 2006.

As the counting continued into Sunday, one political science professor who specializes in elections said the human errors that occurred on Saturday should not have happened.

"People knew that the Republican primary was going to be a high-profile election, and this was not a new implementation of a voting system," said Thad Hall, a professor at the University of Utah.

Hall said it was unlikely the glitches affected the outcome of the election, especially since most people who vote in primaries feel strongly for their particular candidate and would come back to vote if they were told to do so.

"This is actually the best-case scenario for screwing up an election," Hall said. "There is only one race on the ballot. Imagine if this had happened in November."