By WARREN CORNWALL
Herald Writer
(c) 2000, The Daily Herald Company
All votes are not created equal in Washington state.
Counties that use a punch-card ballot, made infamous by the Florida debate over hanging chads, were twice as likely to reject votes for president and more often rejected ones for the U.S. Senate than counties using optical scanners.
Whether in sparsely populated places like Island County, or more urban spots like Spokane and Bellingham, counties that used punch cards consistently rejected a higher percentage of votes in this fall’s general election, according to a computer analysis by The Herald.
The results suggest that when it comes to casting and counting votes, technology does matter.
The machinery behind elections has come under intense scrutiny amid the closest presidential race in modern times. Much of the dispute over Florida’s election results revolved around whether punch-card technology was more confounding to voters.
To see whether such concerns were borne out in Washington, The Herald analyzed detailed election results from all 39 counties. Here’s what we found:
“What it suggests is that punch-card voting, of all technologies we have, is probably the most flawed,” said Herb Asher, a professor emeritus of political science at Ohio State University who has studied ballot errors with punch cards and reviewed The Herald’s findings.
Those warnings have led Bellingham lawmaker Doug Ericksen to call for legislation requiring counties to replace their punch-card systems. Ericksen, a Republican in the state House, said The Herald’s findings bolstered his argument that punch-card systems are inferior and need to be replaced.
Ericksen said his plan would help strengthen voter confidence shaken by the furor in Florida, and ensure votes are counted as accurately as possible.
“I think it’s an issue that needs to be addressed. It will be better for the state of Washington, it will restore confidence in the system that we use to vote,” said Ericksen, who won his seat in 1998 by 108 votes following a recount.
Election officials from punch-card counties, however, show little appetite for replacing the costly machines, even as some admit they probably lead to more voter mistakes.
Punch-card systems may not be as “user friendly,” said Secretary of State-elect Sam Reed. But that doesn’t mean they should be scrapped.
“This is still a very good system. There are problems with optical scan as well,” said Reed, who now oversees elections with punch cards as Thurston County auditor. “I think the counties are capable of deciding for themselves.”
The difference in how often votes were rejected is small, less than 1 percent in both races. But it amounts to thousands of votes statewide.
Gorton might have benefited if punch-card counties had the same rejection rate as optical-scan counties, based on the voting results. It likely wouldn’t have reversed the election, however.
If the rejection rate was the same, an analysis of county-by-county results showed Gorton might have narrowed the gap by roughly 700 votes. That assumes the additional ballots would have gone to the two candidates in the same proportion as the overall results in the county.
Those findings suggest one candidate might have benefited more, said John Wilkerson, a political science professor at the University of Washington’s Center for American Politics and Public Policy. But he cautioned voting mistakes may be concentrated in one neighborhood, or among people with lower education levels, making it difficult to know who would have gotten the votes.
The politics of punch-card counties didn’t escape Ericksen’s notice.
“In Washington state, it seems to me that the majority of punch cards are in your rural more conservative areas,” said Ericksen, who lives in Whatcom County, where punch cards are used.
Election officials and experts in voting technology say the different rejection rates found by The Herald likely stem from a combination of how people vote and how the votes are counted.
Voting with punch-card ballots is usually more complicated than with optical-scan ballots, making it more likely a person will mistakenly not vote for a race or will vote too many times, said Scott Konopasek, Snohomish County elections manager. The county uses optical-scan equipment. Konopasek earlier managed elections in Salt Lake County, Utah, which used punch cards.
The problem may be aggravated for increasingly popular absentee ballots. Voters with an optical-scan ballot vote by putting a mark next to their candidate’s name.
“There’s no further thinking than, ‘Here’s the candidate, here’s the mark,’” said David Elliott, assistant director of elections with the Washington Secretary of State’s Office.
With a punch card, they usually must look at a voter’s guide, then search a numbered punch card and poke out a corresponding hole with something like a bent paperclip. At the kitchen table, with none of the voting-booth equipment, that may lead to more mistakes, Konopasek said.
It’s also easier to decipher what a voter meant to do on an optical-scan ballot, said Snohomish County Auditor Bob Terwilliger.
With punch cards, election officials must try to divine the voter’s elusive intent from dents or holes. With optical-scan ballots, they can read check marks or X’s that didn’t register, or see if a person tried to erase one vote and filled in another, Terwilliger said.
He has even seen ballots where people wrote a note alongside the race, explaining how they meant to vote.
A certain percentage of the skipped votes and overvotes are also intentional, said Konopasek, as voters decide not to weigh in on a particular race, or vote for every candidate.
Then there are the chads.
If voters fail to punch a chad out of the ballot card, the tiny piece of paper can remain lodged in the ballot and fail to register as a vote when passed through the machine.
In Washington, election workers try to solve the problem by checking each ballot and brushing away hanging chads before the card goes to a counting machine. Reed credited that for the tiny change in voting results between the first count and the recount conducted on his race and the Senate race.
Some election watchers had earlier predicted punch-card counties would see a bigger change in votes, as a second trip through a voting machine dislodged chads.
A Herald analysis showed optical-scan counties actually had slightly more variation from the first to second count. The differences between the two systems were so small they were statistically unimportant, said Michael Brett, an assistant professor of civil and environmental engineering at the University of Washington who reviewed the findings.
The difference in rejection rates, however, passed statistical muster, Brett said.
Still, some cautioned about drawing too strong a conclusion from the numbers.
Differences in rejection rates can also be affected by how long voters have used a particular voting system, how complicated a particular ballot is, or whether a voting area is well lit, said Richard Smolka, editor of Election Administration Reports, a publication geared toward election managers.
“You never rule anything out,” he said.
The punch-card ballot has no shortage of detractors at the moment.
“There’s a serious problem. More people are going to mess up with punch cards than other kind of equipment, and that’s not fair,” said Robert Richie, executive director of the Center for Voting and Democracy, a nonprofit, nonpartisan election watchdog group in Takoma Park, Md.
In 1988, fears of inaccurate vote counts prompted the National Bureau of Standards, a precursor of the federal National Institute of Standards and Technology, to urge an end to the use of prescored punch-card ballots.
“It’s just user unfriendly,” said Roy Saltman, the now-retired engineer who wrote the bureau’s report. “Election administrators that think it works fine are ignoring the facts in front of their faces.”
Fourteen of Washington’s punch-card counties use the prescored ballots, while two – Yakima and Franklin – use punch cards where the hole is made with a machine similar to a hand-held hole puncher.
Despite the criticism, local election administrators are reluctant to endorse costly changes to a system they feel has worked well for years.
Island County bought a new punch-card system in 1999, after election officials decided optical scanners were too expensive. The county paid $35,000, rather than the more than $250,000 it would have cost for the scanners, auditor Suzanne Sinclair said.
“I have no objection to replacing the system if the state wants to cough up the money for it. Otherwise I would foresee using it for a few more years just to get our money’s worth,” she said.
Konopasek also cautioned against ditching a fairly reliable system in favor of optical scanners at a time when other devices like touch-screen computers may soon gain wide acceptance.
“Those are emerging technologies, and if you were to throw out punch cards in the state of Washington, you’d only have two options. One is to go to paper ballots, and the other is optical scan,” he said.
The UW’s Wilkerson said more votes would likely be added by investing the money in get-out-the-vote efforts, rather than the new equipment.
Saltman, however, said cost concerns need to be weighed against the price of ensuring everyone has an equal chance at voting and having their votes counted.
“How do we value making sure our votes are counted in a democracy?” he said. “I don’t think our votes are valued enough.”
Talk to us
> Give us your news tips.
> Send us a letter to the editor.
> More Herald contact information.