Michael Duck, Maryland Gazette

Sep. 5, 2003

ROCKVILLE -- Members of Rockville's elections board derided a Johns Hopkins study criticizing the state's electronic voting machines during its meeting last week, calling it "fantasy" and its methodology "pathetic."

"The research was very, very poorly done," said Doris Ecelbarger, chairwoman of the city's Board of Supervisors of Elections.

"At one point, it struck me that this man had never voted," board member Linda Aksamit said of Aviel D. Rubin, the Johns Hopkins professor who led the study.

Before the meeting, board members had reviewed the Hopkins report as well as the response from the machine's manufacturer and programmer, Ohio-based Diebold Election Systems. Many drew from their own experience with the machines as election judges.

The board's support for the machines would seem to restore the city's plan to use the state's Diebold voting machines in the mayoral and council elections in November. But until a state-ordered review of the machines is completed this month, county and state officials will not speak with city officials about the issue.

City Clerk Claire Funkhouser advised the board to wait for the state report before recommending anything to the City Council. With the election just two months away, the board has less than a month to make an official recommendation.

Funkhouser, whose office coordinates the city election in conjunction with the board, seemed surprised that board members all reached the same conclusions.

"I didn't know that everybody was this clear," she said at the board's meeting, adding that she had expected more discussion.

Board member David Celeste said the city's biggest problem in this election will be the public perception, spawned by media coverage of the Hopkins report, that the machines are not secure.

Board member Margaretta Tutson also said defeated candidates might blame "faulty" machines for election results.

But she noted that elections officials often hear such complaints, regardless of technology.

Rockville resident Stanley Klein already voiced his misgivings about the city's use of computerized voting machines at the City Council's Aug. 4 meeting.

Klein is a member of the Institute of Electrical and Electronics Engineers and part of a team working to develop security standards for voting machines. He told the council that the Rockville election might be a good practice run for someone trying to corrupt a state election.

He pointed out that nobody ever realizes someone is stuffing a ballot box unless there is a huge discrepancy in the results.

"If they slice a few votes here and a few votes there, nobody ever knows," he said.

Celeste disagreed. Speaking from his experience as an election judge, he said, judges can see how many votes were cast on each machine at the end of the election day.

If, as the Hopkins report suggests, a voter created a fraudulent "smart card" allowing him or her to vote multiple times, it would be obvious, Celeste said. If one voting machine tallied thousands of votes while others at the same poling place counted only a few hundred, a judge would know immediately there was a problem.

"I never thought I'd be in a corner defending these machines," Celeste said, adding that he hated them after his experience as an election judge in last year's county primary. After receiving better training for the general election, he has become an ardent fan of the machines and a vocal detractor of the Hopkins report.

The Hopkins report charges that "voters can trivially cast multiple ballots with no built-in traceability, administrative functions can be performed by regular voters, and the threats posed by insiders such as poll workers, software developers, and even janitors, is even greater."

In reaching some of its conclusions, the Hopkins report assumes election officials might run the Diebold voting software on ordinary personal computers, and that official election results could be transmitted using a regular Internet service provider.

Board members said those assumptions are incorrect. The software runs on specially designed voting machines without a keyboard or ports a voter could use to access the software's computer files. Similarly, results are never transmitted over the Internet. Only unofficial results may be sent over a phone line via a modem, and no results ever go through a third-party Internet service provider.

"The authors of this study had no clue," Celeste said. He said they used "pathetic" methodology and some of their assumptions were based on "fantasy."

Similar objections to the Hopkins report were raised when the Hopkins report was first released.

After Florida's problems with voting machines in the 2000 presidential election, Gov. Parris N. Glendening (D) ordered Maryland to update its voting systems. Montgomery was one of four Maryland counties to use the new touch-screen machines in last year's elections.

In late July, the state approved a $55.6 million contract for 11,000 new voting machines. The Hopkins study was released just days later, and Diebold quickly released a point-by-point rebuttal of the Hopkins report.

On Aug. 6, Gov. Robert L. Ehrlich Jr. (R) ordered Science Application International Corp. to review the machines. It is that report that Rockville officials are awaiting.

Should the state report find serious problems with the machines, the city could be in trouble. Funkhouser said that the county has already sold its old voting machines to California, and a state regulation requires municipalities to use the Diebold machines.

Funkhouser said municipal and county officials have asked the state to suggest alternatives as part of the report. Gaithersburg and TaKoma Park also have upcoming elections and face the same problem.