I reported for The City Paper yesterday on the Davidson County Election Commission's decision not to use electronic poll books (EPBs) in this fall's general elections.
The commission had deployed the EPBs to 60 voting precincts for the August primaries as a trial run, with plans to use them in all 160 of the county's precincts in November. Now, after evidence surfaced that some voters, including three elected officials, received Republican ballots in error, all precincts will be using paper poll books. The EPBs will only be used for informational purposes, such as looking up a voter's correct precinct.
Elections Administrator Albert Tieche has maintained all along that while there were some issues with the EPBs, the results of the county's elections were not in question. He endeavored to prove that by way of a lengthy presentation of precinct-by-precinct data last night, during which he further explained what went wrong.
From the City Paper
In a presentation to the commission, Elections Administrator Albert Tieche illustrated how voters in three precincts may have received Republican ballots in error but were given the correct ballots before they actually voted. The problem, he said, arose when poll workers failed to follow protocol to make sure that correction was properly recorded.
In the case of Sheriff Daron Hall, Tieche said that if Hall had alerted a poll worker that he had been given a Republican ballot in error, the problem could have been fixed before he voted.
A spokeswoman for Election Systems & Software (ES&S), the company that manufactures the electronic poll books, testified that the commission did not instruct the company to make the Republican ballot the default. In an effort to decrease the amount of steps in the process, however, the setup of the poll books made it easier for a poll worker to erroneously select an application for a Republican ballot.
“This was a door left open to human error,” Tieche said. “Some of our humans went through that door.”
Basically, when a voter shows up to the polls for a primary, they are supposed to be asked by a poll worker which ballot they want to vote on — Republican, Democrat or General. At that point they are to be issued a ballot application, which they bring to the voting machine. A poll worker at the voting machine is then supposed to load the ballot listed on the application.
In the three precincts where problems occurred, Tieche said, the issues began when the first poll worker either failed to ask which ballot a voter wanted, or mistakenly selected the Republican ballot. In either case, Tieche (and the ES&S spokeswoman) said that the setup of the electronic poll books made it easier for the poll worker to select an application for a Republican ballot. Both say that setup has been changed.
In most cases, Tieche showed, voters alerted the poll worker at the voting machine when they realized they had received a Republican ballot in error, and at that point, the poll worker corrected the mistake and loaded the correct ballot. The problem, Tieche said, was that those poll workers should have sent the voter back to the first poll worker to get a correct ballot application. When they didn't do that, the result was that voters did get to cast the correct ballot, but a discrepancy was created between the ballots cast and the voter participation history — that is, the commission's collection of data showing which primary voters voted in.
For instance, according to Tieche's data, in one House District 58 precinct, there were no Democratic ballot applications issued, but 89 Democratic ballots cast. Those voters did indeed get to vote in the Democratic primary, but that is not currently reflected in their voter participation history. The commission did discuss ways to give those voters a chance to have the record corrected.
Over at the Metro Council meeting, Joey Garrison reports, the council voted to withhold funds for the additional EPBs, but will revisit the issue at its next meeting.