Was There E-Voting Fraud?

By Debra D'Agostino  |  Posted 11-08-2006

David Dill is a computer science professor at Stanford University and the founder of Verified Voting, a non-profit organization that encourages voter-verified paper ballots rather than paperless systems. Dill spoke with Online Editor Debra D'Agostino about vote flipping, poll security and lessons learned from yesterday's election. What follows is an edited transcript of his remarks.

CIO Insight: While there were widespread reports of problems with voting machines, most of them were said to be caused by human error. Does this mean the machines are safer than you thought they were?

Dill: It was inevitable that things would go wrong. We have a lot of new equipment in this election that people aren't familiar with. There have been a lot of issues just because workers didn't know how to operate the machines. The poll workers try to get the machines up and running and discover they can't, so the voters begin to line up. You call the state office, but the lines are busy because everyone is having the same problem. The end result is that voters get turned away.

For more news and analysis, see the Midterm Election Report

But of course the e-voting manufacturers are going to blame the problems on human error. Otherwise they would have to accept responsibility for the problem, and they're not about to do that. When you talk about user error there are two important points to keep in mind. First, when you are dealing with a system that is inherently complex, you are going to have errors. You can't just blame the user for that, even if it's the user's fault. A lot of this voting equipment isn't very well designed for use by the minimally trained poll worker. Second, it really doesn't matter to voters if the error is caused by the manufacturer or the poll worker. Either they have confidence that their votes are counted, or they don't. I don't think that this election improved voter confidence in these systems.

Keep in mind also that it's a little early to get a clear idea of how bad the problems are in this election. In the 2004 election, there was a lot of happy talk about how smoothly the whole process went—it wasn't until days later that the truth began to emerge. It's like the fog of war. I hope the press won't be reporting how peachy everything is this time around.

Have you heard any official reports of tampering?

Except for the guy who broke the machine, I have never seen smoking-gun proof of election fraud with electronic voting. But my concern is not that there's a lot of fraud—it's that there's no way to tell if fraud is happening or not.

Look at this problem of vote flipping. Vote flipping is when the voter, typically on a touchscreen system, enters his or her vote, but when the confirmation screen comes up, it's different from what the voter entered. This happened all over the country in the 2004 elections on four different types of machines (from Sequoia, ES&S, Diebold and Danaher), with documented cases in California on machines that were pulled out for parallel testing. There were also complaints of vote flipping across the nation during yesterday's election. The problem is usually blamed on screen calibration errors, but I don't understand how these machines can go out of calibration so easily. And without an investigation into the problem, there's no way to tell.

Sounds like you suspect we'll be seeing more problems cropping up over the next few days.

It depends on whether there is a close race, and if someone raises an issue in that race. The right thing to do in terms of the voter flipping issue is for the EAC to call for an investigation, since it happened on a nationwide basis. Failing that, one of the states should take the lead and do a thorough inquiry.

The bottom line here is that we should be thinking more carefully about what to do when something does go wrong. There is a very big difference between not being able to boot up machines and not having the necessary emergency ballots so that citizens can vote.

So is that the most important lesson to be learned from all this?

Absolutely. The chances of voters being turned away from precincts with paperless voting are much higher than those that use paper ballots with an optical scan. Having paper as a backup means you never have to turn voters away—as long as you have enough ballots. That's the direction we should be headed in.