The 2004 Presidential election took place a full two days after Halloween. But the way things have gone in the weeks since, you'd think the ghouls must have hung around for the voting.

Across the country, Americans have been coming out of the woodwork with voting stories of the weird, the unbelievable, and the downright evil. There were, for instance, the numerous voters whose names vanished from voting lists; the Florida counties with three-quarters registered Democrats, which mysteriously voted three-to-one for Bush; and the eerie precinct in Ohio where 13,342 registered voters somehow cast 18,472 ballots.

All and all, over 30,000 voting irregularities were reported for this election. All evidence, say supporters of John Kerry, that President Bush's victory is a sham, the product of Republican lies, deception, and dirty tricks.

Take the "mismatched" counties, for example--wins for Bush, even though the majority of voters are registered Democrats. Pretty odd, right? Not really, say a number of analysts. First, points out Slate assistant editor Josh Levin, even if 90% of Florida voters had stayed true to their registrations--the nationwide average--the Republicans would only have lost about 600,000 votes, a number unlikely to have changed overall results. Furthermore, according to Cornell political scientist Walter Mebane the Florida irregularities really aren't that unusual. Said Mebane in Salon magazine: "Those counties have trended strongly Republican over the past twelve years." The discrepancy between registrations and votes, he says, resulted from "old-style southern Democrats who have not bothered to change their registration." Indeed, the 2000 election saw similarly wonky results in these constituencies.

But what about the fact that weird voting only took place in counties with optical-scan machines? Doesn't that suggest a conspiracy? Actually, there's a fairly mundane explanation for that too, according to MIT political scientist Stephen Ansolabehere told Salon. The optical scan machines, he said, tend to be used in "rural areas or places with low population density, and those counties happen to be more Republican."

Okay, maybe that makes sense. But what about the Ohio precinct with six thousand more votes than voters? Some analysts who counted votes across the entire county in question think there may have been as many as 93,000 voter-less ballots. Perhaps enough to call into question the neck-and-neck results in Ohio. Explain that one.

No problem, says Kimberly Bartlett, a spokeswoman for one of the counties with reported phantom voting. The apparently impossible numbers are simply the result of a glitch in the software the county uses to show election results on their website. Inexplicably, the computer lumps absentee ballot counts from several voting precincts together and assigns them to one precinct, making it look like there were more votes there than voters. But in fact, Bartlett says, if you look at votes versus registered voters for the county as a whole, the numbers do add up. Correcting for the software anomaly, Fairview Park--the precinct that appears to have 18,000 ballots and 13,000 voters--actually only shows 8,421 presidential votes cast. Well within reason.

Another challenge that's been raised to the legitimacy of Bush's win involves "residual ballots" in Ohio. Unlike many other states, Ohio largely shunned electronic voting machines because of security fears. So most voters were using the good old punch-card systems, the same ones that proved so controversial in Florida in 2000. The paper voting system leads to a significant number of ballots being cast but not counted for various reasons: A voter might have chosen too many candidates, or the vote-reading system might have mistakenly decided that the person did so. And let's not forget the infamous hanging chads, another reason why many ballots are tossed out.

In 2000, a Harvard report suggested that voting systems in Ohio resulted in about 20 of every 1000 ballots being thrown out. Using that stat, journalist Greg Palast has argued that there could be as many as 110,000 uncounted ballots now kicking around the state (although one Cleveland newspaper reported the number at only 93,000). A re-examination of these ballots might reveal wrongful disqualifications, yielding further genuine votes.

Another source of extra votes in Ohio might be "provisional" ballots. Because of the fiascoes of the 2000 elections where many legitimate voters were turned away from polls because their names had disappeared from the voting list, Congress this year allowed unregistered voters to cast back-up ballots. The idea was that such votes could later be counted if the person was indeed found to be an eligible. But according to the Seattle Times, Congress failed to specify how such votes should be evaluated, thus leaving Ohio's 155,000 provisional ballots in limbo. Taken together with spoiled ballots, this means a possible 265,000 votes have gone uncounted in the state. Bush defeated Kerry in Ohio by about 136,000 votes.

This may sound like a possible election-breaker, but in fact a recount would be unlikely to change the presidency. Firstly, the reality is that many of these ballots would still be thrown out--ones spoiled for a legitimate reason, or cast by people rightfully struck from voting lists. Even if half were deemed legit--an unlikely scenario--these would have to go 100% to Kerry in order to change the overall result. Even if 75% of uncounted ballots were miraculously admissible, Kerry would still have to garner nearly 70% of these votes in order to win. According to Salon, such numbers are unlikely given that Bush won 4 of 10 votes even in Ohio's most Democratic counties.

Were there irregularities in this election? Certainly. Are they worrying? Again, yes. Several congressmen have already called on the General Accounting Office to study what went wrong, and improve America's unwieldy voting systems. Did the above-mentioned problems cost John Kerry the presidency? Likely not. All indications are that Kerry lost because Americans simply didn't vote for him. As Walter Mebane--the political scientist who argued vehemently that Gore actually beat Bush in 2000--puts it, "If this is evidence that [Republicans] stole this election, they've been stealing elections for a long time."

Posted 11-15-2004 12:57 AM by Doug Casey