New York Times
December 24, 2004
COLUMBUS, Ohio, Dec. 23 - William Shambora, 53, is the kind of diligent voter who once assumed that his ballot always counted. He got a rude awakening this year.
Mr. Shambora, an economics professor at Ohio University, moved during the summer but failed to notify the Athens County Board of Elections until the day before the presidential election. An official told him to use a provisional ballot.
But under Ohio law, provisional ballots are valid only when cast from a voter's correct precinct. Mr. Shambora was given a ballot for the wrong precinct, a fact he did not learn until after the election. Two weeks later, the board discarded his vote, adding him to a list of more than 300 provisional ballots that were rejected in that heavily Democratic county.
"It seems like such a confused system," said Mr. Shambora, a John Kerry supporter who blames himself for the mistake. "Maybe if enough people's votes had counted, the election might have turned out differently."
From seven-hour lines that drove voters away to malfunctioning machines to poorly trained poll workers who directed people to the wrong polling places to uneven policies about the use of provisional ballots, Ohio has become this year's example for every ailment in the United States' electoral process.
With a state recount expected to be completed next week, few experts think the problems were enough to overturn President Bush's victory here. And many of the shortcomings have plagued elections for decades.
But with the 36-day Florida recount of 2000 proving that every vote counts and with the two major parties near parity, the electoral system is being scrutinized more closely than ever. Election lawyers and academics say Ohio is providing a roadmap to a second generation of issues about the way the nation votes.
Congressional passage of the Help America Vote Act of 2002 - which mandated the provisional ballot as a failsafe and provided states money to update voting technology - was considered a landmark overhaul that would help prevent another Florida.
But an array of voting rights groups contend that Ohio has underscored shortcomings in the law, including one of its centerpieces, the provisional ballot. Now those groups are pushing for a re-examination not only of the law, but also of other voting issues, including the role of partisan secretaries of state in overseeing elections, electronic voting and the elimination of the Electoral College.
"We're in an environment where people believe that even the tiniest number of votes can have a huge impact," said Doug Chapin, director of Electionline.org, a nonpartisan clearinghouse for voting information.
Ohio is emblematic of that attitude.
In the two weeks since Mr. Bush was certified the winner here by 118,000 votes out of 5.7 million cast, watchdog groups have filed lawsuits contesting the outcome and questioning the counting of provisional ballots. The state has nearly completed a recount, at the request of the Green and Independent Parties. Liberal Democrats have demanded investigations into whether there was voter fraud, tampering and intimidation in urban districts.
"This has fundamentally shocked people's sense of whether any election can be accurately counted," said Daniel Hoffheimer, counsel to Mr. Kerry's Ohio campaign.
It is far from clear that Republicans in Congress will have any appetite to revisit voting issues, and many Republicans here argue that the system suffered only minor glitches, even with high voter turnout. "There are no error-free elections," said Secretary of State J. Kenneth Blackwell, a Republican whom Democrats have accused of worsening the state's voting problems in the way he interpreted state law.
But Mr. Blackwell acknowledged that the election spotlighted the state's outdated voting system, with 68 of 88 counties still relying on punch cards. In an interview, he called for updating voting machines, and also for early voting, multiple-day voting and other changes that he said would shorten lines and encourage people to vote.
"I don't think it's wrong to have high expectations," he said.
Certainly there were problems on Election Day.
In Franklin County, a computing error initially awarded nearly 4,000 extra votes to President Bush. In Mahoning County, improperly calibrated touch screens resulted in an unknown number of votes incorrectly going to President Bush before the problem was caught.
And most recently, election challengers in various Ohio counties have said that the tabulators used to count punch cards may have been tampered with before the recount.
But experts in election law say little clear evidence of fraud has emerged. Democratic officials have joined Republicans in arguing that any conspiracy to deny Mr. Kerry votes would have required Democratic complicity, because each of Ohio's 88 county election boards has two Democrats and two Republicans.
Yet there were widespread problems, many of which point to defects in the election rules, experts say.
"I think the problems weren't sufficient to cast doubt on the results," said Edward B. Foley, director of the Election Law Program at Ohio State University's law school. "But I do think there were more problems than usual in Ohio."
Provisional ballots are a prime example. In 2002, Congress authorized using the ballots in federal elections for voters whose names do not appear on registration rolls. The ballots are sealed and held until after an election, so a voter's eligibility can be checked. Valid ballots are then counted, others discarded.
But Congress largely left it to the states to promulgate rules for provisional ballots, resulting in a hodgepodge of policies. In Ohio, Mr. Blackwell, who was co-chairman of Mr. Bush's state campaign, ruled that provisional ballots would be counted only when cast from a voter's proper precinct. (At least 26 other states followed the same practice.) Democrats challenged the ruling, but a federal court upheld Mr. Blackwell.
Rules for reviewing provisional ballots also vary widely within the state. Some counties checked voter registration records dating back several years to validate ballots; others searched only recent records. Cuyahoga County, a Democratic bastion that includes Cleveland, did not check older records, and its rejection rate for provisional ballots was about 35 percent. The state average was 23 percent.
Mr. Blackwell says that despite the complaints, Ohio had one of the country's highest acceptance rates for provisional ballots: 77 percent of its 155,000 provisional ballots were counted, the highest in a 16-state survey by Electionline.org. Illinois and Pennsylvania, which went for Mr. Kerry, accepted only about half of their provisional ballots.
Perhaps the most visible of Ohio's problems were its long lines. Christopher McQuoid reached his polling place in Columbus at 4:30 p.m., congratulating himself for beating the after-work rush. By 7:30, he was getting impatient. And when he finally voted at 9:30, there were 150 people in line behind him.
"I was lucky," said Mr. McQuoid, a radio announcer. "I had the day off."
But how many people decided not to vote because of long lines, and was it enough to make a difference? No one has been able to say with authority. Much attention has focused on whether elections officials served one constituency better than another.
Among the 464 complaints about long lines in Ohio collected by the Election Protection Coalition, a loose alliance of voting rights advocates and legal organizations, nearly 400 came from Columbus and Cleveland, where a huge proportion of the state's Democratic voters live.
"It's possible that it made a difference in the outcome but unlikely," said Dan Tokaji, an assistant professor of law at Ohio State, where academics plan a voter survey to test whether large numbers were discouraged.
In Columbus, Franklin County election officials reduced the number of electronic voting machines assigned to downtown precincts and added them in the suburbs. They used a formula based not on the number of registered voters, but on past turnout in each precinct and on the number of so-called active voters - a smaller universe.
By contrast, the state's most populous county, Cuyahoga, allocated machines based on the total number of voters, a move that the county's election director, Michael Vu, said helped stave off even bigger lines.
In the Columbus area, the result was that suburban precincts that supported Mr. Bush tended to have more machines per registered voter than center-city precincts that supported Mr. Kerry - 4.6 machines per 1,000 voters in Mr. Bush's 50 strongest precincts, compared with 3.9 in Mr. Kerry's 50 best. Mr. McQuoid's precinct, a Kerry stronghold, lost one of the four machines it had in 2000, despite an increase in registration.
"Somebody came up with a very sophisticated plan for machine distribution which, either by accident or design, greatly enhanced the president," said Robert Fitrakis of Columbus, who is part of a group that has contested the election results in court.
Matthew Damschroder, a Republican who is the director of elections in Franklin County, said the urban precincts lost machines because many of their voters had not voted recently and because those precincts historically had had low turnout.
Indeed, election results show that a much higher suburban turnout on Nov. 2 meant that machines in Bush areas were more heavily used on average, although whether that was because their voters were less easily discouraged by long lines or simply more efficient in voting is unclear.
"Most of the precincts that stayed open late because of long lines were in the suburbs," said William Anthony Jr., a Democrat who is chairman of the Franklin County election board.
Another area of contention is the large number of ballots - 96,000 by recent counts - that registered no vote for president. Known as "residual" or "lost" votes, they involve cases where no candidate for president appeared to have been selected or where multiple candidates were chosen, rendering the ballot invalid for that race.
The problem was pronounced in minority areas, typically Kerry strongholds. In Cleveland ZIP codes where at least 85 percent of the population is black, precinct results show that one in 31 ballots registered no vote for president, more than twice the rate of largely white ZIP codes, where one in 75 registered no vote for president.
Experts say punch cards contributed to the problem, because the ballots, which require voters to punch a hole through a heavy-stock paper, are prone to partial perforations, or the buildup of chads. Election officials say that nearly 77,000 of the 96,000 residual ballots in Ohio were punch cards.
But Mr. Foley, the election expert at Ohio State, noted that some people consciously withhold their votes for president and that 77,000 residual punch cards is in keeping with failure rates for punch cards nationwide.
Mr. Blackwell said Ohio's residual votes actually declined this year from 2000. Of the 4.8 million votes cast in 2000, about 90,000 - 1.9 percent - registered no vote for president. This year, 96,000 of 5.7 million votes cast - 1.7 percent - did so.
Mr. Blackwell favors changing to a system that uses an optical scanner to read a paper ballot, which, he said, meets federal requirements, is less expensive than other machines and can handle more voters. But he said groups who say that just about every electronic voting system can be hacked are not helping things.
"There is still evidence out there that we need to transform the machinery," he said. "But it will be harder to do now."
When the recount is completed next week, no one expects the questions about the election to die, with several groups poised to challenge the recount.
"I think the majority of Democrats feel that the election was more or less accurate," said Dan Trevas, the spokesman for the Ohio Democratic Party. "But others are suspicious. Irregularities that are normally overlooked have become the focal point of attention this year. I just can't see those people walking away satisfied."
James Dao reportedfrom Columbus for this article, Ford Fessenden from New York and Tom Zeller Jr. from Cleveland.