Page 2 of 3

Recounting Ohio

Was Ohio stolen? You might not like the answer.

Except it doesn't. Talk with Eaton and she is quick to volunteer that Barbian never used the phrase "cheat sheet"—those were her own words, dashed down in a rush after a lawyer advised her she had witnessed illegal activity and should testify at the Conyers hearings. Eaton says that no one took Barbian's cheat sheet advice seriously and adds that "I still don't know if there was fraud," though she does find his visit suspicious. And although Eaton is angry that she was fired—and has retained legal counsel in the matter—she does not believe her whistleblowing was the only cause. She claims Schwartze had long wanted to get rid of her "because I stood up to her." Schwartze declined to comment.

None of the skeptics hint at this more nuanced version of Eaton's story. What's more, the Conyers report says Triad "essentially admitted" to providing cheat sheets to Ohio counties. That's news to Triad. Barbian didn't respond to my phone messages, but Triad president Brett Rapp insists that "no tampering whatsoever took place." Skeptics note that Rapp is a contributor to the Republican Party. But figures listed in the Conyers report show that his donations have averaged less than $350 a year since 1998—hardly a high roller.

The Terrorist Threat That Never Was

Now to Warren County, where officials locked down the building used to count votes and told a Cincinnati Enquirer reporter that there'd been a terrorist threat. The skeptics are right that the FBI denied issuing any such warning. But it's not true that votes were counted in secret, say both Susan Johnson, the Republican Board of Elections director, and Sharon Fisher, the Democratic deputy director. Not only were Johnson and Fisher present, so were the four Board of Elections members (two Democrats, two Republicans) plus an observer from each party. The only person shut out, Johnson says, was the reporter, "but reporters have never been allowed into our counting room before."

Fitrakis responds that the goal of the lockdown was actually "to divert ballots to an unauthorized warehouse where [Republicans] could manipulate the vote." He claims to have maps that reveal where the warehouse is. But what, exactly, would that location prove?

Vote-Hopping and Other Oddities

What about Miami County, site of the 13,000 mystery votes? Roger Kearney, a contract employee who manages the Miami County website, says he understands what aroused the skeptics' suspicion. The problem, he says, is that Miami County considers a precinct to be "reporting" as soon as a single vote is reported. Thus, when election officials were distributing results on election night, their last two tallies of the night said that 100 percent of precincts were reporting. "The reason the last report had 13,000 more votes," he says, "is that those votes hadn't been counted yet, but they were there in the system."

Kearney adds that he tried, twice, to explain this error to Fitrakis, who had published the allegation in the Free Press. "I'm a Democrat," notes Kearney, "and I told him I'd be glad to find fraud here and turn the election around, but that didn't happen." As for the "Hussein-like" turnout, the county's Board of Elections director, Steve Quillen, acknowledges that 98.55 and 94.27 percent turnouts were in fact reported, but only in 2 of the county's 82 precincts. What's more, Quillen caught the errors and corrected them before the official tally was announced a week later—though, according to Vanity Fair's fact checkers, he never pointed that out to the magazine.

Other dubious results also appear to have been caught and corrected, including 4,258 votes reported in a precinct near Columbus that had only 800 registered voters and a negative 25 million votes initially reported in Mahoning County. Mahoning County's Board of Elections tech specialist, Chris Rakocy, also offers an innocent explanation for the problem of "vote hopping," in which electronic voting machines mistakenly registered a vote cast for Kerry as a vote for Bush. "We had that calibration problem on 18 of 1,148 machines on Election Day," says Rakocy, adding that the errors were quickly corrected. He says that even on faulty machines, voters could check their ballots and correct any mistaken entries, just as they would when withdrawing money from an ATM.

The GOP's Closet Gay-Rights Voters

If voting machines were hacked, skeptics argue, that could explain some improbable results in three Bush strongholds near Cincinnati. In Warren, Butler, and Clermont counties, Kerry got 132,684 fewer votes than Bush did. But Kerry also got 11,923 fewer votes than C. Ellen Connally, the Democratic candidate for Ohio chief justice. It is "beyond plausible," argues the Free Press, that Connally, an African American supporter of gay rights, would do better than the top of the Democratic ticket, especially in three Bible Belt counties that overwhelmingly approved a gay-marriage ban on the same ballot. Kerry's true count must have been suppressed. "Take Ohio without those three counties and Kerry would have carried the state," argues attorney Cliff Arnebeck, a Fitrakis ally.

Not so fast, replies Michael O'Grady, the legal counsel to the Ohio Democratic Party. O'Grady, who helped advise Connally's campaign, agrees that her results in those counties do "stand out." But he credits the 8 to 10 percentage boost that female candidates often get from voters simply because they are female. And, he adds, many Ohioans didn't know that Connally supported gay rights or even that she was black—her campaign deliberately downplayed those facts.

Exit-Poll Enigmas

The discrepancy between exit polls and the official results is a key part of the skeptics' argument: Kerry was projected to win nationwide by a close but comfortable 3 percent, and in Ohio by 6.5 percent. But the skeptics betray a poor grasp of exit polling, starting with their claim that exit polls are invariably accurate within tenths of a percentage point. In truth, the exit polls were wrong by much more than that in the 1988 and 1992 presidential elections.

Warren Mitofsky and Joe Lenski, the pollsters who oversaw the 2004 exit polls, concluded that one source of their incorrect forecast was an apparent tendency for some pro-Bush voters to shun exit pollsters' questions. "Preposterous," claims Mark Crispin Miller, who also sees trickery in the adjusting of exit polls after the election, though that is utterly routine. And is it really so strange to imagine that Bush supporters—who tend to distrust the supposedly liberal news media—might not answer questions from pollsters bearing the logos of CBS, CNN, and the other news organizations financing the polling operation?

Besides, how do skeptics explain New Hampshire? The state conducted a hand recount of precincts that critics found suspicious; the recount confirmed the official tally, as Ralph Nader's campaign, which paid for the exercise, admitted. Apparently one reason Bush did better than expected in those precincts was an influx of conservative Catholics who relocated from neighboring Massachusetts—the kind of anomaly that can confound even persuasive-sounding assumptions about voters.

Page 2 of 3