By John Gideon on 8/21/2006, 5:57pm PT  

Guest Blogged by John Gideon

Tonight, Dobbs reporter and stand-in Kitty Pilgrim interviews Professor Avi Rubin about his new book and about the dangers of Direct Recording Electronic voting machines.

Video in Streaming Flash format...

The video is also available on Rubin's new blog.

The text-transcript of tonight's segment on Lou Dobbs Tonight follows in full...

PILGRIM: More than half of all American voters will cast their ballots on electronic voting machines this November. But the accuracy and the integrity of these machines increasingly are in question tonight. Well, my next guest has written a provocative new book on the threat that e-voting machines pose to our very democracy. Avi Rubin is a professor of computer science at Johns Hopkins University and he's also the author of "Brave New Ballot: The Battle to Safeguard Democracy in the Age of Electronic Voting." Thanks very much for joining us tonight.

AVIEL RUBIN, AUTHOR: Good evening. PILGRIM: Aviel, you know, you point out some very, very compelling e-voting issues. And let's just bring our viewers up to speed on some of your worries. The first one is there's no way for voters to verify that their votes are recorded correctly. There's no way to count the votes in a publicly viewable fashion. And meaningful recounts are impossible.

For the American voter, this does not bode well for confidence in the system. What makes you think this with such conviction?

RUBIN: Well I'm speaking here about the electronic voting machines that are being used in many places. And these are the direct recording electronic or DREs. And the problem is that these are simply computers. And they've been programmed by people, computer software often has bugs in it. And it's almost the ideal platform for somebody who wants to rig the election or tamper with the election in some undetectable fashion.

Now the problem is that when a voter goes up to one of these machines and votes and makes their choices, when they leave the polls, they have no idea how that machine recorded the votes that they cast inside of it. There's no way for them, for example, to have confidence that the data that's now inside of that voting machine corresponds to how they voted.

Worse, if the election becomes controversial or if it's very close and a recount is needed and many states actually have laws giving the candidates the right to have a recount, there's no way to get an independent count of those votes.

If you ask a computer to add a bunch of number, it will give you a particular result. And even if you ask it to do that 100 more times if the data is incorrect or corrupted, it is going to give you that same incorrect result. So that's one of my primary concerns.

PILGRIM: You know, and recounts are not unusual in elections if there's a particularly close election. Let's go to some other points that you make. And this is about the security of the machines. Your basic problem is that the machines must be trusted not to fail. And you say in the Diebold machines we found gross design and programming errors. And we cannot determine the quality of vendors' machines because their code is proprietary.

I know you have vast expertise in this area, especially in checking the security of these machines. Tell us why you're really concerned about the design of these machines at this result?

RUBIN: Well, everybody who has worked in the software field knows that software's very complex and very difficult to get right. So when you're building something like a voting machine, in the case, for example, of the Diebold voting machine, which my research team examined, it was 50,000 lines of C++ code. There's a lot of room for error there. And it's really crucial that that software be designed using formal software engineering processes.

And the software that we looked at from the Diebold voting machine was really poorly written. We found all kinds of security errors that I outline in the book and the use of cryptography for example, that's used to encrypt information on the machine was using outdated and broken ciphers.

And so I'm very concerned in the first hand that we're even using this kind of technology, but furthermore, I'm concerned with the fact that we're taking the plunge and using it, we're not even using well designed systems.

PILGRIM: Now, some states, in fact, 27 states have instituted laws or mandates that a paper trail be available on a vote so that people can check if the machines record it properly. But other states have not. There are many court cases that now take on this issue. And let's listen to a sound bite from one of the plaintiff attorneys in one of these cases. This is what he says about what's going on.


LOWELL FINDLEY, PLAINTIFF ATTORNEY: Sometimes litigation is the only way to get the attention of the officials and, in the end, sometimes the only way that the change can be brought about, by asking a judge to order it.


PILGRIM: Now Aviel, we've looked at some very flawed elections and controversial elections in Georgia, Maryland and Ohio and other places around the country. Is litigation the way to go? Why are not state officials re-examining their decision to purchase these machines?

RUBIN: I think we have to attack this from all angles. And in the short term I've supported some of this litigation. I believe that when the election officials are not responding to what the security experts are saying, litigation might be the only way. I think in the long term it has to be through legislation.

I think that we need laws that will require us to have transparent systems where the votes can be publicly observed, the vote counting can be publicly observed and recounts are possible. But if we're not going to have legislation like we don't have in my state of Maryland, then I think for the public to bring lawsuits challenging the use of these machines is not only reasonable, but I think it's the right thing to do.

PILGRIM: Thanks very much, Aviel Rubin, author of "Brave New Ballot: The Battle to Safeguard Democracy in the Age of Electronic Voting." Thanks very much for being with us Aviel.

RUBIN: Thanks a lot, Kitty.