Campaigning for Computerized Voting

Despite staunch opposition from the computer science community, ATM-style electronic voting may offer the best hope for escaping the mess inflicted by paper-based balloting systems.

By Simson Garfinkel

September 3, 2003

Over the last two decades, geeks have rarely passed on an opportunity to replace a perfectly good mechanical device with a computerized system. Got one of those old-fashioned cash registers? Replace it with a PC and a touch screen. Got a hotel with perfectly good door locks and metal keys? Rip them out and replace them with computerized locks and swipe-cards. Wherever you look, pinball is out, video games are in.

But there is a rising chorus of geeks-a chorus led by some very high-profile computer science professors and researchers-who say that one machine should never be computerized: the voting machine. These computer professionals say that accurately counted free elections are the bedrock of democracy. Voting, they claim, is too important to be done on a computer. The irony is delicious-it's sort of like group of doctors arguing for the return of leeches because the President of the United States is too important to be treated by modern medicine.

Specifically, the computer scientists are opposed to that new generation of voting machines that resemble automatic teller machines. These systems are called "direct recording electronic" (or DRE) voting machines because people vote on the touch screen and the votes are recorded directly on the computer's hard drive, without any paper being harmed in the process.

There are a lot of reasons to like these DRE machines. Because the voting is done on a large touch screen, they can use big fonts that are easier for the elderly to read. The machine can be programmed to reject attempted votes that are patently wrong, like voting both "yes" and "no" on a referendum question. The machines can be equipped with speech synthesizers, allowing people who are blind or illiterate to vote on a truly secret ballot for the first time in their lives. They can even confirm the voter's choices on a second screen-which means that there would be no more elderly Jewish voters in Palm Beach accidentally casting their ballots for Pat Buchanan.

Nevertheless, most computer professionals are opposed to the DRE machines. One reason is that there is fundamentally no way to audit them: If 600 people vote at a DRE on Election Day and the machine says that 310 voted for the Democratic candidate, who is to say that the number 310 is true? Perhaps only 280 voted Democratic, but the machine was programmed to randomly flip 5 percent of the Republican votes to Democrat before recording them on the computer's hard drive. To make this sort of programmatic tampering harder to detect, perhaps the program was devised so that the flipping would only happen on the first Tuesday in November. On other days-presumably the days when election officials tested the voting machine-no vote flipping would take place. To make it even harder to detect, perhaps the flipping occurs only when the machine discerns that the vote is close; this would avoid the embarrassment of having polls predict one outcome, and having the machines tally another.

This sort of election-stealing logic would be easy to code into the voting machine's operating system. The logic could be written by a lone programmer-perhaps an activist hacker with a grudge-without the knowledge of the voting machine company. The logic could be so well hidden that not even a careful review of the machine's source code would find it. This isn't as far-fetched as it might sound: Unauthorized features called "Easter eggs" are routinely hidden in commercial software, even software shipped by Microsoft.

Story continues below

I keep writing "most computer professionals" because I recently met one who isn't opposed to DREs: In fact, he's positively enthusiastic about them. And this man isn't just anybody; he's Ted Selker, an award-winning inventor with many patents, formerly with IBM Research, currently a professor at the MIT Media Lab, and member of several panels and commissions that looked at the issue of voting following the debacle of the 2000 presidential election.

I met Selker a few days after he had attended a meeting of computer scientists and election officials in Colorado. He was livid. He had just spent two days listening to the experts of the field talk about all of the failings with DREs and how these systems could be used to steal an election.