Election results don't add up

Review shows discrepancies in voting

By JOHN MARTIN Evansville Courier & Press November 23, 2004

At several Vanderburgh County polling places, the number of people signing in to vote Nov. 2 exceeded the vote totals, while at other places votes outnumbered signatures.

Election officials say there were no such discrepancies in the May primary election, when Vanderburgh County's new touch-screen voting machines were used for the first time, but when voter turnout was minuscule compared to the general election.  
No local election outcomes have been challenged, but the voting process and the precise count continue to be a source of debate.

Democratic county commissioners continued Monday to press for an independent audit of election results. Others argued a post-election evaluation conducted by manufacturer Election Systems & Software has shown how and why Election Day problems occurred, and anything else would involve unnecessary expense.

County Clerk Marsha Abell said several scenarios can explain why numbers of signatures and votes at some polling places don't match.

Long lines of voters were reported throughout Vanderburgh County. Some people said they waited at least two hours to cast a ballot. She said that because of those long lines, there were instances of voters signing in but then leaving before voting. In other cases, voter confusion about a final step in the electronic voting process caused some votes to be canceled, Abell said. To lock in a vote on the iVotronic machines, voters had to hit a flashing red "VOTE" button at the top of the screen and then touch a section on the screen that said they did not wish to go back and change their choices.

There were instances of voters walking away before going through that final step, Abell said. And other voters, Abell said, may have managed to vote without first signing the poll books.

Workers in the Election Office are doing a precinct-by-precinct study of votes cast compared to signatures. Most of the county's 139 precincts have been checked. Workers are paying particular attention to precincts where the discrepancy of votes and signatures is more than 10.

At the Knight 1 precinct in eastern Vanderburgh County, there were 25 more votes than the poll book said there should have been, while at the Center 14 precinct, there were 12 fewer votes than signatures.

The ES&S post-election examination of the machines' performance was overseen by Abell and Evansville residents Tom Massey, a Republican, and Tammy Barnett, a Democrat.

After a vote is cast, the iVotronic machine is supposed to automatically reset, and a poll worker is supposed to activate it for the next voter by ing a cartridge.

The ES&S review showed some machines locked up as a result of poll workers putting cartridges into machines too quickly, before the machines had reset from the previous voters, Abell said. Those lockups contributed to long lines at some polling places.

Massey, an attorney and a former Vanderburgh County Election Board member, said ES&S opened each of the 532 iVotronic machines used for the election.

He concluded that each vote properly cast on the machines was counted and that "it would be a waste of taxpayer money to pay for further audit of the 2004 general election."

Barnett, however, said "it is premature to rule out" having an independent audit, and the county should ask questions about what kind of data the process would show.

Barnett suggested the county appoint a community task force to study the election process and recommend improvements.

County Commissioners President Catherine Fanello said she still wants an independent audit. The county has been quoted a price of $37,500.

Massey said he questioned whether the county has legal authority to conduct such a review. Fanello asked the commissioners' legal staff to take up that question with the Indiana Election Commission.

"I really think it's imperative to make sure the equipment is working properly for $2.9 million," Fanello said.