6:50 p.m. The reliability of voting equipment in the United Sates has not improved, with reports of failures coming in at around the same fairly high rate as in 2008—with Virginia potentially emerging as one of the worst trouble spots.
At least 1,000 reports of equipment breakdowns at polling places nationwide have been made so far today, Pamela Smith, president of the nonpartisan group Verified Voting told me a short while ago. The group tracks such reports as it studies the integrity of the U.S. voting system. “Virginia is the big issue—some very long lines in Virginia and equipment malfunctions, and things that seem rather anomalous,” she said.
Smith added that at some point later this evening, she plans to post her findings on the Verified Voting website. Virginia uses direct recording electronic voting technology, or DRE, with no paper backups. With luck it won’t end up being the deciding swing state with a close margin.
“So far no jurisdiction has really been immune to reports in terms of equipment failures. We’ve seen a lot of reports that ‘instead of the old six-to-eight machines we used to have, there are two-to-four working.’ And we’ve seen reports of equipment down for hours at a time,’“ she added.
It will take hours or more to really calculate the total number of reports and put it in perspective, but she added: “I would say it’s about the same as 2008 in terms of volume so far.”
6:19 p.m. Touch-screen technology is great, until it disobeys you and chooses another candidate for president.
Today a video was posted on Reddit showing a voter in Pennsylvania trying to vote for Barack Obama on a direct recording electronic voting machine—commonly called a DRE machine. The machine highlights Mitt Romney’s name instead. The polling site removed the machine from service, according to a report on MSNBC.
Pennsylvania is a key swing state and their DRE systems, like those in Colorado and Virginia, don’t have paper backups. If the race is extremely close, glitches like this one, involving only small numbers of votes, could not only decide the race but prevent anyone from auditing the results. In 2008, 1,800 reports of problems with such machines were reported according to this analysis.
4:44 p.m. Uh- oh, Ohio.
Ohio is a crucial battleground state in today’s election, and its electronic voting machines could yet be at the center of a huge controversy.
Lawyers for Ohio’s Republican secretary of state, Jon Husted, were in court today over a lawsuit filed on Monday by Democrats concerning last-minute software changes to the state’s electronic-voting machines (see “The States With the Riskiest Voting Technology.”)
Husted’s aides had previously said that a contract inked back in September simply eases reporting of results on election night. But the complaint includes the allegation that, at the request of Husted’s office, the voting machine contractor ES&S “…has installed a ‘back door’ into such hardware and software that enables persons who are not under the supervision and control of defendant Husted, and who are not under the supervision and control of Ohio’s boards of elections, to access the recording and tabulation of votes using facilities not under the control of defendant Husted or Ohio’s boards of elections.” Court filings are available here, including raw notes from today’s hearing.
It is far from clear right now whether there is a real issue here in terms of voting integrity or worsened security. But if Ohio turns out to be extremely close, we could be in for a drawn-out legal battle over electronic voting machine integrity.
4:15 p.m. As we predicted last week, Facebook is running one of the biggest get-out-the-vote efforts in history (see “How Facebook’s Plans Could Affect the Election”). It wants you to “tell friends you’re voting.” Those are four very important words, and they may produce hundreds of thousands of additional votes today.
If you’ve logged into Facebook today, you’ve probably seen that message, or something similar. What you may not know is how effective such messages can be at increasing real-world turnout. Social pressure is very powerful, even online, as recent research has proved.
On Election Day in 2010, similar Facebook reminders, which included faces of friend’s who’d indicated they’d voted, wound up inspiring an additional 340,000 people to go to the polls nationwide, according to a paper published this year in the journal Nature by Facebook researchers and academics. The researchers spent more than a year checking the real-world voting behavior of the 61 million people who had logged in and thus seen the message that Election Day (see “How Facebook Drove Voters to the Polls”).
Explaining the nuances of how this social peer-pressure works, study co-author James Fowler, a political scientist at the University of California, San Diego, says two forces were very powerful. One was the inclusion of faces of friends who’d voted (versions of the message without faces didn’t have any effect on voting). And the second was additional posts by friends saying they’d voted. These second-order posts, he found, were even more influential than the original nudge from Facebook.
We can’t be sure yet whether Facebook is doing another study, though it stands to reason that they will be. Today I’ve looked at lots of the Facebook messages from friends and colleagues (and a few perfect strangers) and they all look the same to me, so if there’s an A/B study going on, I can’t see evidence of it. If there are differences in the choices of which friends’ faces they’re showing, it would be hard to tell, of course.
The key thing is—the messages do include the faces as well as language designed to inspire voting. And though Facebook is clear in saying this is a nonpartisan effort, I’d wager this benefits Barack Obama, if only because there are more registered Democrats than Republicans in the United States—and because Facebook membership skews younger than the electorate. Facebook has 160 million U.S. members while there are about 140 million U.S. registered voters.
The extent of the overlap is not clear, but political analysts have told me that, for example, about half of Ohio’s voting-age population is on Facebook.
2:30 p.m. In the past two days, New Jersey has faced confusion over voting practices after the Republican Lt. Gov. Kim Guadagno announced Saturday that registered voters displaced by Hurricane Sandy could vote by e-mail or fax–a unique domestic adaptation of how many military and overseas ballots are already submitted.
The announcement was soon followed by anecdotal reports of nonworking e-mail addresses to obtain and submit those ballots, prompting a great deal of news coverage, such as an All Things D post that concluded: “New Jersey has become an unwilling test case to see if Internet voting can work in the country that invented the Internet.”
The stories aren’t quite what they seem. What New Jersey decided to let Sandy victims do isn’t exactly Internet voting–as in, clicking an icon on a Web interface as if you were standing in front of an electronic voting machine at a polling place. Voting through a Web interface is something security researches have made clear is inherently insecure (see “Why You Can’t Vote Online”).
Rather, New Jersey allows you to receive a traditional ballot by fax or e-mail from your county, fill out that ballot, and then fax or e-mail the ballot back. In addition, state officials have made clear that any voter who does this, must also mail a paper copy of that e-mailed or faxed ballot for later verification. Of course, the practice of faxed or e-mailed ballots raises the possibility that someone else could submit one under your name. But the paper trail also helps address a key security problem.
New Jersey also isn’t unique in allowing this. All told, 33 states allow the e-mailing and/or faxing of ballot for some, or all, voters–almost all of these directed at the needs of military or overseas voters. Other states include Arizona, North Dakota, and the swing state of North Carolina.
Is this a concern? Nationally, little data is kept about how many votes are cast this way and the fate of those votes, says David Jefferson, a researcher and voting technology expert at Lawrence Livermore National Laboratory in California. “We have absolutely not even a remote idea of how many votes will be cast over the Internet [by e-mail and faxed ballots], nor do we know what data will be released, e.g. how many votes are cast that way in each state, jurisdiction, or precinct, how many were e-mail vs. fax, and how many were rejected for some reason versus how many counted,” he says.
In a close race, the fate and integrity of e-mailed or faxed ballots could become an issue. But Charles Stewart, an MIT poltical scientist and voting technology expert, says that in the range of possible technology problems, this one isn’t the highest. “I’m less alarmed by e-mailing and faxing, because we’ve already taken the leap and saying we can do absentee ballots by mail. It’s hard to argue that it’s a different kind of concern than traditional mailed absentee ballots.”
What’s clear, though, from the problems some New Jersey residents have had in getting working e-mail addresses, is this: “States need to figure out this stuff before disasters happen,” he added.