A Debate on Electronic Voting: A Tool To Improve Elections or Rig Elections?
Friday, February 27th, 2004
With days to go before Super Tuesday, Electronic voting critic Bev Harris, author of Black Box Voting, debates the head of Common Cause Georgia, Bill Bozarth, on whether electronic voting could threaten the future of democracy. [includes transcript]
Millions of American voters are projected to use electronic voting machines in the upcoming presidential election. Many of these machines will get their first test when voters head to the polls in 10 states holding Democratic presidential primaries on March 2nd in what is known as Super Tuesday.
But concerns over security flaws in voting machines have sparked wide-ranging public debate. Last week, a Sacramento judge heard a request for an injunction against Secretary of State Kevin Shelley and registrars of 18 counties using electronic voting technology.
The lawsuit was filed by publicist-turned-investigative reporter Bev Harris in an attempt to force counties not to use electronic voting machines during next Tuesday's primaries.
In 2002, Harris uncovered a public Internet site that posted the source code of major electronic voting machine manufacturer Diebold, Inc., and found that the company had failed to meet voting security standards.
Since then separate analyses of the code have claimed to find alarming security flaws in Diebold's software. Voters in California, Georgia and Ohio will use electronic voting systems for the first time during the Super Tuesday vote.
Bev Harris, publicist turned investigative reporter who has written the book Black Box Voting: Ballot-Tampering in the 21st Century. <http://www.blackboxvoting.org> She recently filed a lawsuit in California in an attempt to force counties not to use electronic voting machines during next Tuesday's primaries.
Bill Bozarth, executive director of Common Cause <http://www.commoncause.org> Georgia. He has defended <http://www.commoncause.org/states/georgia/evs.htm> the use of electronic voting systems in Georgia.
AMY GOODMAN: Bev Harris joins us in our studio today, author of "Black Box Voting: Ballot Tampering in the 21st Century." We're also joined on the phone by Bill Bozarth, head of Common Cause, Georgia, who has come out in support of electronic voting machines. We welcome you both to Democracy Now!.
BEV HARRIS: Hi.
AMY GOODMAN: Why don't you explain the lawsuit that you are involved with and explain your concerns about electronic voting, Bev Harris?
BEV HARRIS: Yes, what we’re concerned about is that they don't have adequate security procedures or auditing procedures in place. And therefore with these machines, especially the ones don't even produce a paper ballot, there's really no way to verify that the machine recorded the vote accurately. Added to that we know that the machines have security flaws that are significant that could enable tampering. And added to that, we know that there have been programmers involved in the company who have criminal backgrounds. So when you put all of those things together, we wanting security and auditing procedures to make sure that our vote is counted as we cast it.
AMY GOODMAN: Can you talk, Bill Bozarth, about Common Cause's decision to support electronic voting?
BILL BOZARTH: Yes. We have looked this into Georgia very thoroughly, first hearing some of the concerns last year of many voices, like Miss Harris and others. What we have tried to do is to do a thorough investigation of the electronic voting process in this state, and I can really only speak with any authority whatsoever on what we have done in Georgia. We can find no evidence in the 2002 election, when these machines were first used statewide, that even though there may be some reasons for suspicion, we find no grounds whatsoever to conclude that there was any tampering with the voting process in 2002, and we have also looked very carefully at the security procedures surrounding the process. They are very thorough. And while theoretically something could happen, we can find no evidence that it has. So, our position has been that the overwhelming advantages of electronic voting and the terms of the accuracy that we have achieved, compared, to prior systems is certainly a positive step forward, and we are very reluctant to endorse a policy that would take us to a point where we have lost that edge. There are some security measures that have been recommended; that the Georgia people are implementing that came out of the study in Maryland. We think those are the most important things, and they have agreed to do that for the 2004 election. That's important, but we just don't simply agree with some of the skeptics have the degree of concern that warrants measures of going back to where we were.
AMY GOODMAN: Bev Harris?
BEV HARRIS: First of all, let's talk about the security procedures in Georgia and whether or not there's evidence that there was a breach. I believe that Georgia in 2002 showed the biggest security breach in the history of the United States electoral system. We now know and we have evidence that one man programmed the commands for the computers that controlled a million votes that nobody looked at what he programmed. After the election, they overwrote those programs with something else, so no one will ever know what he programmed. As for whether the new systems are more accurate, in fact, Cal Tech and MIT did a report on this and they found that the only system that is less accurate, in other words, that loses more votes than punch cards, is the touch-screen system.
AMY GOODMAN: What do you mean, loses more votes?
BEV HARRIS: What this study looked at is how many votes that voters cast that did not get registered for whatever reason. They found that the touch screen is actually the only system that performs worse than punch cards, which makes us wonder why in the world we are all charging toward this.
AMY GOODMAN: Why don't they register?
BEV HARRIS: Because of software flaws or because of various problems with the touch screens where they're not sensitive enough, they're not calibrated correctly. Things like that. Many of the talking points that we're hearing are simply not accurate. For example, he is saying, “well, we don't want to go back where we were before.” Again, that's not what we're asking for. We are asking for a paper ballot and auditing methods where we can verify that the vote was counted correctly. That's not saying that we are going back to what we did 100 years ago. We're saying let's audit the system. Very simple recommendations, not expensive, not time consuming.
AMY GOODMAN: How exactly to you audit it?
BEV HARRIS: A couple of things: One, you want do a random, robust spot check of the paper ballot against the machine to make sure that the machine tally is correct and there wasn't a software flaw.
AMY GOODMAN: How do you do that paper –
BEV HARRIS: You can select it by random precincts and compare the paper ballot -- in other words, tally up the paper ballot in perhaps 5% of precincts and compare -- that are randomly chosen and compare it with what the machines came up with. If there's a discrepancy, you expand the size of audit, because there should not be a discrepancy.
AMY GOODMAN: But you’re saying that paper would vote on paper ballots and a computer.
BEV HARRIS: If they vote on a computer, the computer can easily kick out a ballot. We have printers. We keep hearing all of these ridiculous things like the printer will jam. Well, no, it won't. In fact, the printer that's used inside the Diebold machines is similar to the one that's used in Wal-Mart. It's a very robust printer. By the way, only about 300 ballots are cast on the touch screen in a day's voting. We don't hear people saying, Wal-Mart can only print 300 receipts and the printer jams and they run out of paper. That's just simply not how it works. We’re simply saying that we need a printed, physical record that the voter has verified what presents their vote. We need a random spot check that’s robust enough, about 5%, to check, and we also need to run reports that the votes have not changed as they travel through the system. That's simple that it takes 60 seconds and costs nothing.
AMY GOODMAN: Bill Bozarth, your response.
BILL BOZARTH: I didn't hear everything that she said. I have a little trouble on the line. Couple of points in response. Number one, I think she is -- I would like to see the proof that somebody actually did something to those systems in Georgia in 2002 to change votes. I have never seen any conclusive information on that. I believe there's been a lot of talk around some changes that were made to the code in order to solve some technical problems that were not taken through the verification process. The decision was made in Georgia that either they had to live with the one or two percent failure and screen freezes or apply the patch for that in order to correct the problem. They applied the patch without the total testing process, and only tested it afterwards. That's certainly something we would have preferred not to do, but we understand why the decision was made. The verification process of the same code, which was kept intact, proved there was nothing in there that would have affected the results. So, there still has not been, to my satisfaction, any sort of case made that anything has happened so far. In response to the subject of voter verifying paper ballots, the common cause –
AMY GOODMAN: Let me put that point to Bev Harris, the issue of what actual proof do you have. There are a lot of rumors that are going around.
BEV HARRIS: We actually have proof that has been verified by three independent sources. It wasn't just one program modification. It was eight. Eight times someone put program modifications into the Georgia system. I actually have six of thOSE program modification patches myself.
AMY GOODMAN: What do you mean when you say program modifications?
BEV HARRIS: What they did was—we had a programmer in Vancouver, Canada who was writing alterations and changes to the program that was certified and approved. He then put them on a website. They were downloaded by technicians under the instructions of Diebold and then they were placed on the machines. Nobody looked at them. When they say, we looked at this and we know there's not a problem, that simply isn’t true. Nobody has gone back and reverse engineered all eight of those patches.
AMY GOODMAN: What about the proposal by congress member, Russ Holt, to simply get a paper receipt?
BEV HARRIS: Right. And we want to call it a paper ballot because a ballot has legal standing where a receipt does not. He has a bill in congress. We should be passing this bill. We need one more thing with that bill, and that is more robust auditing procedures. We haven't really decided what we're doing with this paper once we get it, and we need to do a little bit better with that. Those things are fairly simple and we can hammer them out.
AMY GOODMAN: What is the argument against having a paper receipt or paper ballot?
BEV HARRIS: Really, they haven't come up with anything that holds water. There are some talking points, such as the blind can't read a paper ballot, but what the visually impaired citizens that I talk with say they would like to have a paper ballot that is compatible with a standard book reader. They could vote at the polling place, but they could also vote in their home if they can't find someone to drive them to the polling place, which is a common problem. This would give them the full right to vote absentee or at the polling place.
AMY GOODMAN: Let me put this question to Bill Bozarth, executive director of Common Cause, Georgia. Do you support the bill that congress member, Russ Holt, has put forward that would give the voter a paper receipt or a paper ballot?
BILL BOZARTH: Just for a point of clarification, we have a bill in our current state senate in our general assembly today that would require that in Georgia in 2004. It's not clear that it will pass but there's legislation both at the national and the local level trying to impose that requirement. But my position on paper ballots is that if we can find the right way to do it, there is certainly no harm in it. Some of the things that have got to be straightened out before those that's totally clear as to how that can be implemented, involves a legal position that says in the case of a difference of opinion, what is the official ballot. We need to sort that out. Despite Miss Harris's playing down the possibility of mechanical failure, anybody that's worked with printers over the years thinks it worth the contention that putting 28,000 printers in the polling process in addition to the other equipment is going to be a logistical problem and present some challenges. That's not a reason not to do it, but we cannot minimize the impact in terms of poll workers who do these jobs once every couple of years. Having the expertise, collective expertise, to deal with that extra aspect of it. So, that's a challenge, but that's not a reason not to do it. What we feel in Georgia -- it's my position in Georgia is once there's a system that has been approved and verified as a working system, the various national agencies, which to my knowledge has not been done yet on any verified paper ballot process or implementation, I think Georgia should embrace that. We are not going to be able to do it in time for 2004. There's not enough time to do it right. I think in the future as that settles down. That's a reasonable position to take. I don't think in 2004 we're going to do it. Let me just make one more comment. In terms of being able to trust the process. You know, we have to trust the process that we have today. One could make the contention that a box full of paper, be it optical scan ballots or punch card ballots or written by hand, all have the vulnerability of somebody opening the box and replacing the ballots where they actually get counted. That could be done today with a variety of systems. Electronic voting is not uniquely vulnerable to this. We ultimately have to trust the people that do the process to make it work. In Georgia, from all of the work that I have done, all of the interaction that I have had with the polling officials, while certainly over the years, things have happened, overall, that's a process that I think we can all trust and continue to approach any of these decisions on that basis, not on the basis that we have to guard against any mischief whatsoever because we simply cannot do that.
AMY GOODMAN: What about the concerns of -- well, a number of those who have looked, like former Diebold systems manager, Rob Baylor, who said that the company failed to adequately test the troubled equipment and balked when he warned them of widespread problems with the machines. Looking at a Salon.com piece by Farhad Manjoo, "Will the Election Be Hacked?" He says last summer computer scientists at Johns Hopkins University and Rice University found major security flaws in the Diebold machines, concluding that the Georgia system falls far below even the most minimal security standards. In January, experts at Raba Technologies, a consulting firm in Maryland, discovered additional failures in that state's Diebold systems. Internal Diebold emails shows that company engineers knew about the problems and in some instances chose to ignore them. Your response. Yes.
BILL BOZARTH: I have read those, and I have tried to get, you know, dialogues going with the various people to get into more detail about what did happen in Georgia. Again, I think it's -- I was in the information technology business for a number of years. I understand some of the things. I don't pretend to be an expert in voting rights and software, but I understand the software development process and that one can look at these and draw potential conclusions about the quality of the code. I know that a number of people have criticized the way in which the code was written and so forth, but ultimately, when you have an election system, our election systems are based really on two major software groups. One is the underlying operating system, which is Microsoft Window's product, and the code that rides on top of that which is a package of software known as Gems, which Diebold acquired. All of this gets cobbled together into something that works. The comments about what happened in that process and trying to get it tested and verified, all of that becomes somewhat historical when you get a system and go through a verification process prior to the election. The verification process is overseen by a number of groups from the national level, the national association of the directors, the local technology group in Georgia, which has responsibilities for the validity of this. The local county officials and anybody that wants to observe the validation process. It’s a lot goes into it after the code gets cobbled together. To me it’s much more significant and believing if it works or not, as opposed to exploring the history of how it came together. This particular employee, as I understand, left Diebold, and there are all kinds of stories floating around that really don't add up to reasons to mistrust what happened in Georgia. One of the stories is that there was one of the files they found was called robgeorgia. People were making implications that that meant somebody was trying to rob the election, and in fact that, was the gentleman's name. There's interesting things to read about, but if you go and look at what we implemented, I’m satisfied what we have done so far and the procedures that we have surrounding it or -- are that we have an honest election. Certainly more honest than what we’ve had in the past. Not perfect. Not to say that we cannot continue to find ways to tighten it up. I think that anybody that wants to say that Georgia had an election fraud of some kind in 2002 simply hasn't made the case to me yet.
AMY GOODMAN: Bev Harris, your response.
BEV HARRIS: Let me talk about the certification of the machines. He is talking about a logic and accuracy test. That's what the officials trot out everytime to tell us that the machines count accurately. I went to observe one of these in California. This system—by the way, they use very similar systems, but this system that I’m going to tell about is Sequoia, which is actually one that’s being considered for New York State. This was the most ridiculous bamboozling process I’ve ever seen. You go in there and you don't actually touch the machine. The machine basically runs a self-test. It has a videogame-looking thing, as it does this, which rolls votes by looking like the bars, bells, and cherries on a Las Vegas slot machine. It makes a Pac Man noise. Not only are you not allowed to videotape this ridiculous test, but you have to leave the room for an hour and then you come back and they hand you a piece of paper (no one knows where it came from). It has numbers on it, and they say, “See? The machine counted correctly.” We had computer programmers and observers and citizens there laughing out loud at the ridiculousness of this system. This doesn't prove anything. Also, even if you had a valid test before or after the election, that doesn't say anything about remote access, because people wouldn't remote access during the test. It doesn't say anything about triggered events where you would have a program launch based on a trigger, which could be as simple as casting a vote. And I notice that the gentleman didn't answer any questions about the four independent reports that -- some of which you brought up, all of them identified significant security flaws. This is not just a technician who was installing software in a historical sense. The most recent report was in January this year. I have the release notes from the (inaudible) program. They have not fixed the problem.
AMY GOODMAN: You also have the now much quoted CEO Of Diebold, Walden O’Dell, who bragged, quote, he's “committed to helping Ohio deliver its electoral votes to the president next year.” A big supporter of President Bush in Ohio. Does that concern you at all, Bill Bozarth?
BILL BOZARTH: Yeah. I have written on that. First of all, let me frame the common cause organization a little bit here. We are in the business of viewing the governance process skeptically. There are certainly a lot of flaws in our political system that we address day and in day out. My inclination is not to be defensive of the establishment at all. I simply weighed in on this because it was such an important thing. And I still maintain my position that while I’m skeptical of people trying to get overdue influence in the process, there's so many ways to do that through the influence of money and campaign contributions and lobbying that nobody really needs to steal an election in order to gain that -- this disproportionate power. As to what Mr. O’Dell said, that having made that public statement that he somehow was able to -- the organization that he ran -- to create a process by which machines in any one jurisdiction, let's say Georgia, could somehow be programmed so that they would be able to change the outcome of the election. Now, keep in mind that almost every polling location has a unique ballot. It would have to be a fairly grand scheme to go in and not only be able to change the outcome in a general sense, but with unique ballots around each of our 159 counties and multiple ballot pipes all across those counties, you have literally thousands of combinations of voting -- voting models that would have to be somehow anticipated and then changed in a way that it wouldn't be obviously detectable. To say that Wally O’Dell could somehow orchestrate that within Diebold and be able to implement it strains credibility beyond imagination. While Wally O’Dell shouldn't get awards for being politically astute and it's clear that was a very stupid statement for him to make, but to conclude that because he made the statement that somehow the series of events could take place to result in a stolen election, I simply cannot follow that logic.
AMY GOODMAN: Bev Harris, your response.
BEV HARRIS: If he doesn't believe that the conflict of interest that Wally O’Dell cited is a problem, what about the fact that the senior vice president for research and development, including under Diebold up until April, 2002, and at that time Georgia had already bought the machines, was convicted of 23 counts of embezzlement and spent four years in prison for sophisticated computer crimes, abusing a position of trust. This is a high-ranking official who had access to every voting system in the Diebold program, including the optical scans, touch screens and the Gems program. We have breakdowns at multiple levels here. All we're asking for is to be able to verify that the vote was counted correctly. When you use only a computerized system, and you throw away the paper ballot, that is exactly equivalent in a bookkeeping system to saying I’m only going to use my QuickBooks and we're going to throw away the invoices and canceled checks because we don't need them. The only person who can look at my QuickBooks system is my bookkeeper, and I’m just going to trust this person. That's foolish. There are a certain percentage of people who have character flaws. We’ve had elections problems for—there’s a tremendous amount of money at stake in terms of manipulating elections. We just don’t want to leave these doors open. Let’s close the doors so that it’s more difficult.
AMY GOODMAN: Is there a model that you would accept, that you think is more reliable?
BEV HARRIS: Yes. I mean, it’s probably cheaper and more reliable to just go back to the paper but there are advantages to these machines. One of the things we must do, and we have to do this right away (and by the way, all the Diebold machines have a printer in them already), we have to get a report of the results at the polling place before this has ever been transmitted anywhere, and post it at the polling place for 24 hours and compare the number at the polling place with what shows up at the county and is reported to the state for the precinct. Amazingly, this is not done. And if you don't do that -- basically, we are talking about digital data a cartridge or memory card. It shouldn't change when it goes from the polling place to the central office. Since we have identified several points of attack in between the polling place and the central count, we should be checking that.
AMY GOODMAN: Well, we have to leave it there. I want to thank you both very much for being with us. Bev Harris is author of, "Black Box Voting: Ballot Tampering in the 21st Century." On the line with us from Georgia, Bill Bozarth, executive director of Common Cause, Georgia. This is an issue that we will continue to cover. This is Democracy Now!.