Will Voting Machine Integrity Decide the Senate?

Control over the agenda of the United States Senate remained up in the air on the day after Election Day, as the margins for the vote counts in Senate races in Virginia and Montana remained razor thin on Wednesday morning.

According to the Associated Press vote count, with all but one Montana precinct reporting, challenger Democrat Jim Tester led incumbent Republican Conrad Burns by 1,729 votes -- about two one-hundredths of the state's population.

And in Virginia, with 99% reporting, challenger Democrat Jim Webb held an edge over incumbent Republican George Allen by just 8,047 votes, or just fewer than four one-hundredths of the total number of votes cast in that race. A recount will take place in Virginia, under whose state law a recount is automatically triggered at margins of 0.5% and below; and voters should expect a similar recount in Montana.

If the two AP tallies are verified by their states' respective secretaries of state, Democrats will have seized a one-vote majority in the US Senate, having won over seats in six of the seven highly-contested Republican strongholds - a feat NBC commentator Tim Russert likened to an "inside straight." Democrats also took control of at least 33 former Republican seats in the House of Representatives, far more than the 15 the party needed to wrest control there.

But how much of the "inside-ness" of this well-played hand may have been provided by technology, or the lack thereof? This question may be asked and re-asked in the coming days, and perhaps weeks, as courts may end up deciding once again the validity of elections with national importance.

In tests three weeks ago of Virginia's eSlate electronic voting machines, which are manufactured by Hart InterCivic, the Democrat Senate candidate's name appeared as "James H. 'Jim"' in machines being delivered to three counties. Omitted was not only the all-important surname "Webb" but the even more critically important (D) denoting the candidate's party affiliation. George F. Allen's full name appeared on these electronic ballots, though his (R) was omitted.

Yet the only truly technical glitch thus far reported nationwide concerning Hart InterCivic centered around its eScan paper ballot readers in Lancaster County, Pennsylvania, where dozens of machines were plagued by paper jams.

In recent weeks, a white paper compiled by ProjectVote.org (PDF available here) warned of possible problems with Hart InterCivic equipment based on its past record, which included the inadvertent addition of at least 100,000 votes to a Tarrant County, Texas, primary race last March where only 58,000 votes were actually cast.

According to the Virginia State Board of Elections, eSlate equipment was used yesterday in the Alexandria, Charlottesville, and Falls Church localities, which according to the US Census Bureau database, would collectively represent about 2.5% of the state's overall population.

Just 3% of households in Falls Church have median incomes below the poverty level, while over one-fourth of households in Charlottesville qualify as below-poverty. The statewide median is 9.9%, and Alexandria falls just below that number at 8.9%. If eSlate voting equipment does come under scrutiny in Virginia, no single segment of the state's economy will hold exclusive claim to the disenfranchisement banner.

Voters in at least eight Virginia counties, according to the Richmond Times-Dispatch, were given inaccurate instructions concerning their local voting precincts. State voters generally arrived at the wrong precincts and were redirected to the proper ones. In at least one case, the culprit appears to have been campaign volunteers who passed out pamphlets including precinct voting station addresses, in neighborhoods that lay outside those precincts' boundaries.

But allegations of fraud on the part of both parties were reportedly received by the FBI, which will be investigating all of them, no doubt under increased public scrutiny.

Meanwhile, in Yellowstone County, Montana, unfamiliarity with the new voting equipment is being cited as the reason why ballots there were being recounted by hand until late into the morning. Yellowstone is the state's largest county, with just under 15% of Montana's population. According to the Billings Gazette there, incumbent Burns leads challenger Tester by just under 2% in that county, though Tester leads the unofficial count statewide.

Precincts throughout Billings recently deployed the AutoMark electronic system, manufactured by ES&S. Its hallmark has been a design which is geared toward use by the disabled. But as ProjectVote.org reports, in ten noteworthy incidents in the past six years, AutoMark equipment was discovered to be recording votes different from those their users selected. Three of those incidents alone took place in Broward County, Florida, which selected AutoMark in response to the disastrous vote count in that state in 2000.

Voting tallies have also been reported slowed down in Flathead and Gallatin counties, which have also been considered typical Montana Republican strongholds. However, since most Montana counties proudly use paper ballots, it's likely the tallying there is simply slow by design.