While the 2000 election made hanging chads memorable, the real culprits behind that electoral fiasco lay elsewhere. Topping the list of factors that led to the four to six million lost votes: registration issues, confusing ballot layouts, and polling place shortcomings. Variations on the theme are rife. Consider Florida’s 2000 Palm Beach County election, in which a confusing ballot layout led 19,000 people to vote for two presidents, nullifying their votes completely. Or a poll worker in upscale Brookline, Massachusetts, who entered an election hall backroom with 200 ballots and a ballot marking pen, and exited 20 minutes later, alone and unsupervised in the interim.

Voting horror stories? Voting technology expert Ted Selker has them “up the wazoo,” as he puts it. Co-director of the Caltech-MIT Voting Technology Project (VTP) and associate professor at the MIT Media Lab, Selker has been part of a team that launched a nationwide review of voting procedures following the 2000 election. Their 2001 report, “Voting: What Is, What Could Be,” contributed to the writing of the 2002 Help America Vote Act, which established the bipartisan Election Assistance Commission; the report also has served as a guidepost for ongoing voting research and technology.

Registration, the VTP discovered, was the number one problem that caused lost votes — two to three million went astray in 2000. “We found two to three times as many errors were due to registration than to any other problem,” says Selker, citing scenarios such as people mistakenly thinking they’re registered if they file a change-of-address report, or people going to the wrong polling location, where they’re allowed to cast a vote for president, but not for any of the other races.


Segueing from the number one problem of election 2000 to its number one lesson, Selker points to the critical need for oversight: “There’s no election technology, period, that can’t be circumvented by people who are not supervised. When you have people supervising each other, you have a way of stopping one of them from changing things. You have a better election.” In virtually all the country’s 17,000 voting jurisdictions, people now know they must pay attention to voting issues and that they’re being scrutinized. “We had 1 to 2 percent fewer lost votes in 2004 compared to 2000 …” — due to a combination of improved oversight, training, equipment, and polling place — “… and that’s completely because people are aware of the issues.”

Another key lesson is the need to perform run-throughs of all election processes. Poll workers must be experienced in all facets of the process; they shouldn’t arrive at their polling locations, as they did in Cambridge, Massachusetts, in 2006, without keys to turn on voting machines for disabled people. (No disabled people were able to vote independently in Cambridge that morning.) Polling officials also should take into account that it takes half an hour to boot and set up these voting machines — a reality Cambridge poll workers failed to factor.

As obvious as it sounds, voters should know where the voting booths are located, as well as the most expeditious way to get there; in one Boston location Selker visited in 2004, voters waited in one artificially long line for nearly an hour because they didn’t know, and no signage indicated, that three of the four precincts in that polling place had no line.


“All jurisdictions should require audits of voting preparation,” concludes Selker, who speaks at gunshot speed as he makes suggestions for the 2008 elections. One critical recommendation is eliminating folds in ballots used in optical scan machines — equipment that optically records which selections on the paper have been marked. Folded ballots cause misreads at the folds; they smudge the ballots; they misalign in printing and don’t stack properly — forcing poll workers to continuously open ballot boxes to restack them, often without supervision. The upshot: a greater possibility of errors and fraud.

Another recommendation is targeted to the four in ten voters nationally who use direct record electronic (DRE) — or computerized — machines. Selker strongly endorses the use of headphones, which gives voters simultaneous audio verification and can eliminate at least 0.5 percent of votes mistakenly cast for the candidate next to the one they meant to select, which happens with all voting systems. Adding this additional feedback mechanism would do away with approximately one million lost votes each election, as well as improve voting for the 14 percent of voters with reading disabilities. Selker also strongly recommends bringing marked sample ballots into voting booths to help people remember the choices they want to make and to check their work.

While voting technologies hold enormous promise — to make voting easier, more convenient, accessible, secure and accurate — improved voting is ultimately not a technology issue. “Technology can reduce mistakes, but better voting practices don’t depend on whether the technology has a power supply. It depends on people,” says Selker. Technology’s job, he asserts, is to improve society. “This is what the VTP is all about. It’s a huge new direction for MIT, which traditionally has focused on technology, technology, technology. Now we have to understand that technology is about helping people do things they want to do more easily, about being part of solutions that really matter.”