Thursday, November 18, 2004

Scholars on the votes, Ohio undervotes

Scholars on the votes, Ohio undervotes
Keith Olbermann
November 18, 2004

SECURE UNDISCLOSED LOCATION— We return to Academic Dueling In Our Time, already in progress.

A UC Berkeley sociology professor, director of his school’s Survey Research Center, is scheduled to conduct a news conference at 1 p.m. ET today at which his “research team” will report that “irregularities associated with electronic voting machines may have awarded 130,000-260,000 or more excess votes” to President Bush in Florida.

The advance word of the news conference gives little detail, but suggests Professor Michael Hout might be treading out onto thin ice. His study is said to show “an unexplained discrepancy between votes for President Bush in counties where electronic machines were used versus counties using traditional voting methods.”

The Berkeley group may have new material, but if not, it could be pinioned by the fact that some of the apparent variations between optical scanning and other voting methods in Florida, might also be explained by— or, even better explained by— historical voting patterns in Florida’s Dixiecrat counties of the north, and the Panhandle.

Regardless, this is now shaping up as the BCS of presidential election analysis. A joint report out of the CalTech and MIT voting project— suggesting that the much-decried exit polling of election night really wasn’t outside the margin of error at all when analyzed on a state-by-state basis— had already been countered by a Penn professor’s report using the exit polling for Florida, Ohio, and Pennsylvania.

Now it’s not just CalTech and MIT versus Penn— but also UC Berkeley versus CalTech and MIT.

Stay tuned for the halftime show.

And stay tuned for the latest disaster from Ohio.

For 40 years, the Dayton Daily News reports this morning, Shirley Wightman has worked at polling places on election days. Two weeks ago, she says, turnout was high - 611 voters - and she and her colleagues paid careful attention to their punch-card, chad-filled, voting stations in Washington Township, Ohio.

“We checked the machines periodically,” Ms. Wightman told the paper, “and I could see nothing wrong with them.”

Yet when the votes were tallied, 168 of the 611 voters had made no choice for president. Unless these were the famed undecideds we heard so much about in the closing weeks of the campaigns, something went terribly wrong. 27 and a half percent of the voters in that “Washington X” precinct in Montgomery County officially didn’t have a presidential preference.

This was the high point of the Daily News’ investigative analysis of the still-unofficial voting results in its county— or more properly, perhaps, the low point. The paper discovered that of the 284, 650 votes in Montgomery, a total of 5,693 registered no valid vote for president. And the percentages were significantly higher in the 231 precincts that wound up voting for Kerry (2.8%) than did the 354 that wound up voting for Bush (1.6%).

Besides Washington X, a second County precinct exceeded 27% ‘undercount,’ as the election professionals, such as they are, call it. Washington X, Kettering 3-A, and five of the other top ten ‘undercount’ precincts by percentage wound up supporting Bush.

Since, as the papers note, political scientists suggested that the poor and the lesser-educated are presumed to have more trouble with punch card voting, there are several logical disconnects here. Given the outcomes in those two precincts, Washington X and Kettering 3-A, were those mostly Bush voters who managed to blank out more than a quarter of their own ballots, or did the precincts wind up voting for Bush because more than a quarter of the ballots had no valid presidential vote?

What happened in the voting precincts in Moraine, Ohio? 2,557 votes were cast at seven sites there. The President won the city by 2%. The number of ballots without a valid presidential vote was 5.6%.

What do the state undercounts in Ohio look like? Did they reduce Bush’s margin of victory? Did they eliminate votes for Kerry? What the hell happened?

The least likely explanations are that these people couldn’t make up their minds, or screwed up only the presidential part of their ballots.

“It is very difficult to believe that a quarter of the people would not vote for president, especially in a year like this,” University of Virginia political scientist Larry Sabato— an old friend of Countdown— told the Daily News. “If I were the election officers in those areas I would be doing some very extensive checks of those machines.”

As the Ohio recount nears, the number of hotspots continues to multiply. You are aware of the remarkable late night voting lines throughout the state, and the mysterious Glitch of Youngstown which initially registered negative 25,000,000 votes. There is the Gahanna machine which gave one presidential candidate 4,000 extra votes in a community of 600. And the farcical “walling off” of the vote counting in Warren County, because the county head of security was told face-to-face of an FBI terrorism warning there - except the FBI says it didn’t issue any terrorism warnings there.

The Associated Press today carries a report of 2,600 ballots in nine precincts around Sandusky, Ohio that were counted twice— as that paper puts it— “likely because of worker error.” The Clyde precinct showed a voter turnout of 131%, to the dismay of the head of the elections board, Barb Tuckerman.

Ms. Tuckerman, in one of the great quotes of the election, told the News-Messenger of Fremont, Ohio: “I knew there was something amiss.”

Tell me about it, Barb.

What do you think? E-mail me at