Conflicting studies hamper search for better voting machines |

Conflicting studies hamper search for better voting machines

ROBERT TANNER, Assiciated Press

NEW YORK – Voting sounds so simple.

One person, one vote. The candidate with the most votes wins. For president, the Electoral College balances out states big and small.

The tangled reality became apparent in last fall’s overtime election. A few people – enough to sway a close race – invariably make mistakes and spoil their votes. The problem is, some machines seem to encourage more errors than others.

The answer, clearly, is to find machines that confuse the fewest voters. But that isn’t so simple, as state legislatures and Congress are learning in their search for a more trustworthy election system.

It’s hard to pick one voting system over another because little analysis of voting technology exists, and the reports that are available often conflict.

Modernization is the new watchword. But some studies have found low-tech voting methods – mechanical levers and paper ballots – work better than state-of-the-art systems.

Optical scan technology, where voters darken circles as in standardized school tests, has support in Florida, Illinois, Arizona and more. But a Georgia study of optical scan voting found places with large minority populations had much higher numbers of uncounted votes than the rest of the state.

Some think computers are the answer. But the latest ATM-style electronic machines turned out errors nearly as often as the punch-cards at the center of the Florida recount, according to a statistical review of the past four presidential elections.

”There isn’t good comparative data,” said Susan King Roth, an Ohio State University design professor who has studied election equipment problems. ”You almost need something like a Consumer Reports for voting systems.”

Votes can go uncounted for a number of reasons. Machines break down, ballots confuse voters, poll workers add wrong. But experts say the most frequent cause is that people in the polling booths make mistakes.

There are overvotes (when a person casts two votes for one race, or a machine counts two votes because of a stray mark on the ballot) and undervotes (when a person accidentally erases a vote, or starts but doesn’t complete a vote as with Florida’s dimpled chads).

To make things murkier, some people intentionally leave votes blank.

Nationally, uncounted votes are estimated at about 2 percent in presidential years – more than the difference that separated Al Gore from George W. Bush in the popular vote nationwide.

Florida estimated its uncounted votes at 2.9 percent, much more then Bush’s slim margin of victory in that decisive state. Georgia’s was 3.5 percent.

Even smaller numbers of uncounted votes are unacceptably high, experts say.

”Talk to people in the civil engineering business, who design drainage projects and dams,” said Michael Traugott, a research scientist at the University of Michigan’s Center for Political Studies. ”They have this concept of designing for a 50-year-flood or a 100-year-flood. This was a once-in-a-lifetime election.”

And the system couldn’t handle it, turning what should have been a simple election into a series of complex arguments before state and federal courts.

In Georgia, officials took a postelection look at the results in their state and found some troubling trends.

Secretary of State Cathy Cox discovered precincts with the largest minority populations had a higher rate of uncounted votes regardless of the technology used, but that optical scan machines made the problem worse.

”We believe the data we have makes a compelling argument that further deployment of optiscan systems would be bad policy, and could perhaps even be considered a decision that disenfranchises minority voters,” Cox told a U.S. Senate committee studying election reform earlier this month.

In contrast, a study of national election returns from the past four presidential elections drew very different conclusions. Academics believe the joint study by the California Institute of Technology and the Massachusetts Institute of Technology was the first of its kind.

Yes, punch-cards were found to produce the highest number of uncounted votes. But electronic voting systems were just as bad – though researchers insisted the technology showed the most promise.

Optical scan, the Caltech-MIT study found, was one of the better technologies, with uncounted votes making up 2.3 percent of all ballots cast with that system. Votomatic punch-cards came in at 2.9 percent, and electronic voting was at 3 percent. The best were paper ballots (2 percent) and mechanical lever machines (1.6 percent).

The study has been criticized, however, by analysts and industry officials who found electronic voting to be as reliable as optical scan.

Policy-makers, in Congress and in the states, need guidance. Experts agree legislators should look at which equipment work best, how people interact with voting machines and what steps election administrators can take to educate voters before ballots are cast.

”The promise of December is that hey, higher technology solves our problem,” said Chris Thomas, Michigan’s election director. ”The reality is it might help, but it’s not going to solve them. You’re going to find it’s the human element.”

On the Net:

Caltech-MIT study:

Georgia study:

Federal Election Commission:

AP-WS-03-18-01 1331EST

Support Local Journalism

Support Local Journalism

Readers around the Lake Tahoe Basin and beyond make the Tahoe Tribune's work possible. Your financial contribution supports our efforts to deliver quality, locally relevant journalism.

Now more than ever, your support is critical to help us keep our community informed about the evolving coronavirus pandemic and the impact it is having locally. Every contribution, however large or small, will make a difference.

Your donation will help us continue to cover COVID-19 and our other vital local news.