Florida E-Vote Fraud? Unlikely 

By Kim Zetter  Wired News

02:00 AM Nov. 10, 2004 PT

Since the election, liberal blogs and discussion sites have been raising a ruckus over an analysis of Florida voting results that's been spreading on the internet, which shows President Bush received a much higher number of votes than the number of registered Republicans in many counties.

The analysis has led lawyers for the Democratic Party to look into allegations of fraud, and was cited in a letter sent by Democrats in the House of Representatives last week to the Government Accountability Office requesting an election investigation.

But academics at several universities, who received a flood of e-mail urging them to look into the matter, say the results are due to high numbers of Democrats in rural areas voting across party lines, and to independent voters who chose Bush in higher proportions to Sen. John Kerry.

The academics say the intense scrutiny has been good for democracy and has highlighted the need for instituting mandatory election audits that would help catch anomalies with voting machines and restore voter confidence in results. But Stanford University professor of government Jonathan Wand said the analysis can be harmful if done improperly.

"It's important that when allegations are made that people bring to bear the correct evidence and statistical analysis to actually back it up," Wand said. "What is destructive is when the allegations are made and they are misconceived or implausible. That's not helping anything. But the general process of people paying attention is a very good one."

It all began last week when a voter posted a chart online showing that in 47 of 67 Florida counties, Bush received more votes than there are registered Republicans. In 15 counties, Bush received two times more votes than the number of registered Republicans, and in four counties, he received more than three times the number of registered Republicans. In Dixie County, for example, 9,676 people registered to vote. Nearly 78 percent of them were registered Democrats while 15 percent were registered Republicans. Yet Bush received 4,433 votes, while Kerry received only 1,959. Overall, Bush received about 20,000 more votes than there are registered Republicans in the state.

Activist Sam Parry called the numbers "so statistically stunning" that they bordered on the unbelievable.

The counties with suspicious results all used optical-scan machines, made by various manufacturers such as Diebold Election Systems and Election Systems & Software. Ironically, in counties using touch-screen voting machines which activists predicted would be the source of fraud this year Bush's votes were more proportionate to the number of registered Republicans.

The optical-scan results could be explained if many so-called third-party voters such as those registered for the Green Party or those who registered independent voted for Bush. But Bush's votes exceeded the number of Republicans and third-party registrants in many counties, meaning that a large number of Democrats voted for Bush.

Activists and progressive commentators reject the notion that large numbers of Democrats voted for Bush, because the Democrats put a lot of effort into increasing turnout for their man, Kerry. And early exit polls showed that Bush would receive only about 14 percent of the Democratic vote in Florida (a figure that lines up with the 13 percent Bush received in 2000). The polls also showed Kerry winning Florida's independent votes 57 percent to 41 percent.

The discussion in online forums and on activist news sites suggests that the optical-scan machines may have been rigged or possessed glitches that gave Kerry votes to Bush. Parry wrote, "If Bush's totals weren't artificially enhanced, they would represent one of the most remarkable electoral achievements in U.S. history."

Activists have been calling for a manual recount of optical-scan paper ballots, which they say would show whether the computer-counted tallies match the paper ballots. They are urging presidential candidate Ralph Nader to request a recount in Florida, as he recently did in New Hampshire.

Although the 20,000 non-Republican votes for Bush would not change the election outcome Bush won Florida by about 350,000 votes activists say the optical-scan anomaly, combined with other problems in the state (such as touch-screen machines that registered votes for Bush when voters pressed the screen for Kerry) should be examined.

But according to academics, the internet pundits are reading the data out of context. Demographic figures and vote trends over several years show the numbers to be consistent with previous elections. According to University of California at Berkeley political scientist Henry Brady, the Republican vote share has been going up in Florida's rural optical-scan counties for years.

"If you look at a trend line, what you got in 2004 is what you'd expect. There's nothing odd going on there," Brady said. "We're not unwilling to entertain the idea that screw-ups occurred. But to just use 2004 as your basis is not the best way to do analysis." 
 Three professors of government also examined the numbers after being pressured by many people, including lawyers for the Democratic Party, and concluded the same thing.

Walter Mebane of Cornell University, Jasjeet Sekhon of Harvard University and Jonathan Wand of Stanford did extensive analysis of the 2000 election, and concluded that Vice President Al Gore should have won the presidential election. But this time Bush won.

"I don't believe Bush really won in 2000," Sekhon said. "So people think our bias should be that Bush didn't win again. But that's not what we've found from this data."

Sekhon said most of the Florida counties in question are small, rural Florida Panhandle counties. The area has a high percentage of registered Democrats, but they're primarily elderly white voters who have been increasingly voting Republican without changing their party affiliation. Indeed, results for the optical-scan counties in 2000 showed the same disproportionate number of votes going to Bush.

Activists say this means the fraud was going on for years, but Sekhon points out that much of Kerry's campaign money in Florida came from cities, whereas Bush's money came from rural areas like the optical-scan counties. Activists say the results still should not have happened because the Democrats had done extensive voter registration in Florida.

But Sekhon said this assumes that Republican and Democratic voters are the same. Although Democratic and Republican registrations increased this year, the two parties went after different types of voters, which affected the turnout and results. Republicans went after occasional voters voters who went to polls inconsistently while Democrats went after people who had never voted. Of the two groups, the latter were less likely to cast ballots on Election Day, because they were less committed to voting.

"It's better to target marginal and sometime voters, rather than people who have never voted," Sekhon said.

Furthermore, although Democratic registrants increased more than Republican registrants this year, the largest increase in registration numbers was among independent voters, who tend to vote Republican. Sekhon said despite the increased registration numbers for Democrats, Democratic registration in Florida had gone down overall in the last decade by about 480,000, a fact activists ignored. Republican registration has also gone down, but by a lower amount. And independent registration has gone up by about 680,000 which offsets the Republican , since independent voters tend to lean Republican in Florida.1

Walter Mendane of Cornell said just because they didn't find fraud in Florida "doesn't mean that nothing went wrong in this election. It just means this particular thing is not what went wrong. It may be that if one looks at the election results overall, you may still find other surprising things."

The academics said they don't expect the rumors to disappear. Sekhon said because state laws vary about which votes get counted and how states audit elections, voters are mistrustful, especially when election officials fail to follow state laws, as in Florida recently where the secretary of state ruled that counties using touch-screen machines wouldn't have to recount votes if anyone contested an election a violation of state law.

Sekhon said it's reasonable for people to think that something was amiss in this election.

"We know something really bad happened in 2000," Sekhon said, "so why is it unreasonable to think that in 2004 the same thing happened?"

Questions about election results won't go away, he said, until the government mandates a system for routinely auditing elections nationwide as Canada does. Although California has a mandatory 1 percent manual recount law which often catches glitches that occurred with optical-scan machines and other equipment other states do not. The rumors in Florida could disappear, for example, if election officials recounted a percentage of the optical-scan ballots. But unless Nader or someone else requests it, it won't happen.

- - -