Analysis: Flawed ballots more likely on touch screens

BY JEREMY MILARSKY

South Florida Sun-Sentinel    16 January 2005

FORT LAUDERDALE, Fla. - (KRT) - Florida's touch-screen voting machines performed better in the Nov. 2 presidential election than they did in the March primary, but were still outmatched by older voting devices that use pencil and paper ballots, according to a South Florida Sun-Sentinel analysis.

Voters using the ATM-style voting machines in November were 50 percent more likely to cast a flawed ballot or have an unregistered vote in the presidential race, compared to voting machines employing simple paper ballots.

"I'm not surprised at all," said U.S. Rep. Robert Wexler, D-Fla., who lost a federal lawsuit last year that demanded touch-screen machines generate a paper receipt.

Fifteen of Florida's 67 counties use touch-screen voting machines, including Broward, Palm Beach and Miami-Dade. The pros and cons of the newer technology have been debated around the country, with the secretary of state in Ohio last week announcing touch-screen machines would no longer be used in elections.

Generally, Florida has praised the performance of touch-screen devices, touting them as the future of voting and the solution to the punch-card machines that added to the confusion of the controversial 2000 presidential election.

While optical-scan machines once again seemed to outperform touch screens in November, both obtained an error rate of less than .5 percent. In the 2000 presidential election, the undervote rate in Florida counties using punch cards was 1.5 percent; the rate for optical-scan counties was 0.3 percent, according to a University of Florida study.

The November 2004 evaluation was based on undervotes cast on each system, or those instances in which the voter skipped the presidential race or a choice was not tallied for whatever reason including machine and software error. Overvotes are those in which more than one choice was recorded for a single-candidate race.

A spokeswoman for Florida Secretary of State Glenda Hood, the state's top election official, declined to comment on the newspaper's study. However, she downplayed the overall numbers, saying the difference between the two voting systems was not "significant."

"I believe it is incorrect to state that touch-screen machines were outperformed by optical-scan machines due to a minor difference in the undervote rates," state spokeswoman Jenny Nash wrote in an e-mail message to the newspaper. "An undervote ... is the prerogative of the voter and not an error."

Of 5 million votes analyzed from the November election, there were 18,555 examples of flawed or unregistered presidential ballots. Of 2.7 million votes on touch screens reviewed, 11,824 ballots had no vote registered for president. Of 2.3 million votes on optical-scan machines, 6,731 ballots were not recorded or flawed.

The presidential race was chosen for study because it appears as the first race on the ballot and is considered by many election experts as the least likely race that voters would deliberately not cast a ballot. The study did not examine votes cast absentee or those cast during the early voting period.

The results of the November election study in Florida mirror a similar newspaper examination of the March 2004 election that indicated voters using touch-screen machines were six times more likely to cast flawed ballots than voters using the optical-scan machines.

A state study examining undervotes in the Nov. 2 election is due on the governor's desk by Jan. 31, Nash said.

Broward county officials in 2001 spent more than $17 million on the new touch-screen voting system, more than three times the price of a voting system using pencil and paper.

Ion Sancho, election supervisor for Leon County and an advocate of a paper-ballot system, said he was not surprised by the study's results.

Sancho, like many other critics of the new touch-screen technology, has concerns that voters cannot independently verify their choices after casting a ballot on an ATM-style machine. No system in Florida gives voters a receipt with their choices on it.

Without such a system, Sancho, Wexler and others said there's no way to be sure undervotes aren't being recorded by touch-screen machines because of a machine error or because voters are struggling to properly use the devices.

County commissioners in Broward and Palm Beach counties have discussed whether to install paper printers on their touch-screen machines, but state officials have not approved any such machines.