Thursday, Jul 24, 2008
Georgia used Diebold's touch-screen machines in 2002, right?
Yes.
And Georgia also had some wacky results, right?
They did. They had six upsets. The most famous one is Max Cleland [the Democratic senator and the incumbent]. That's because he was quite far ahead in the polls and an 11-point shift happened overnight and [Republican] Saxby Chambliss won instead. And the other upset that surprised people was Sonny Purdue, who was the first Republican governor elected in 134 years.
Do you think those elections were legitimate elections?
Well, I think that it was an illegal election in that they had no idea what software was on the machines at the time. Georgia was a situation where they had changed the software not once or twice but seven or eight times so it went through so many permutations without even being examined by anyone, and nobody has any idea what the machines did. [Harris says she confirmed these preelection changes to Diebold's software in conversations with Georgia voting officials, but Diebold denies that any changes were made. In February, Joseph Richardson, a spokesman for the company, told Salon: "We have analyzed that situation and have no indication of that happening at all."]
I do find this suspicious -- they have since scrubbed clean the flash memory and gotten rid of the small cards that store the results from each touch-screen machine. They've overwritten it with a whole new thing. What's amazing is you keep paper ballots for 22 months, and they're an awful lot bulkier than these credit card-size memory cards, but for some reason they felt compelled to get rid of them all. They have also overwritten all of the GEMS programs in the counting machines. They've gone through and overwritten everything in the state.
OK, so we should talk about how Diebold responded to your posting these memos.
As soon ... a few days after we posted them they sent us a cease-and-desist letter -- interestingly authenticating the memos and laying claim to them, telling us that they were copyrighted. So they claimed copyright and they told us to take them off the Web.
Right. By claiming copyright they're saying they own them, so that seems to indicate they are authentic memos.
Exactly.
So what's your response to their copyright claim?
Well, I don't believe you can protect intent to break the law by slapping a copyright on it. And the memos that we posted show that the law has been broken. If you can protect intent to break the law, all anybody would need to do is take their bank robbery plans and put a copyright on it, and then say nobody can look at them because they're copyrighted.
Do you really think that their memos show intent to break the law?
Oh yes, yes. The Ken Clark memo is absolutely clear. It says they have been aware of these security flaws for years and they have chosen not to correct it. He says something to the effect of, find out what it will take to make this problem go away. [Referring to a voting equipment certifier, Clark tells a colleague to "find out what it is going to take to make them happy."] He says if you don't mention [a problem] you may "skate through" certification. And talking about doing "end runs" is not a good thing either.
And what's disturbing is the very same thing that these memos are talking about -- overwriting the audit log -- in the presentation in which they sold their machines to the state of Georgia they specifically bring up the audit log and say that no human can change it. This shows they made fraudulent claims, frankly.
There's a thing called a Qui Tam suit which citizens can file if they feel that federal money has been spent based on fraudulent claims. I haven't done it because it gives you a gag order and I refuse to be gagged even for billions of dollars, but these things are wide open for such a thing. If you go and look at the sales documents, they made one claim after the next.
So because the memos show what you say is clear intent to break the law, that's why you don't think that they have a valid copyright claim.
Well, the other issue is an overriding public interest. We are told that we are to depend on these systems in 37 states and yet they are admitting that they are easy to tamper with.
Are you going to respond to them?
Well, these memos are on the Web in so many locations that we took them off and put a link to someone else who put them up. So that fulfills our requirement under the law.
But do you know if it's possible for you to face any --
-- any retaliation? It's certainly possible that they will try retaliation, and if so I will use the full extent of the law available to me for full discovery of everything. And I think that going through discovery will become a very uncomfortable process and perhaps put some people in jail ... Not on our side, by the way.
At this point activists are now taking these memos from various places on the Web into their state attorneys general and asking for an investigation, and since Diebold has now authenticated them it's no longer, "I found this on the Web," it's, "I found this on the Web and Diebold says they wrote them."
When Diebold is put to greater scrutiny, won't the elections officials say, "We won't go with Diebold, but we'll use touch-screen systems from this company or this company?"
Well, I think that won't fly in the long run because the same illness is afflicting all of them, and that is that they are not auditable and secret. The solution is pretty simple and obvious, and that is to get properly auditable machines. A lot of the security stuff goes away -- the most bulletproof system that I know anyone has come up with is one that is a touch screen but then prints a ballot that the voter verifies.
Whatever the software is doing, if you have something with a really bulletproof audit -- the voter verifying the paper, and the computer tally -- if those two things match, you've got a pretty good confidence level.
If Diebold, ES&S and Sequoia want to come up with a nice paper trail, voter-verified paper trail that's a touch screen, I'm supporting them. But right now they're fighting it tooth and nail.
How are they fighting it?
For one thing they had a meeting on Aug. 22 -- the voting machine manufacturers and the Election Center [a nonprofit management division of the National Association of State Election Directors, which handles part of the voting-machine certification process] and a lobbyist. The whole purpose of this meeting was to try to get the public to figure out how to accept machines without a paper trail.
How did you find out about this meeting?
Actually, this is kind of funny. My publisher found out about this. It was a teleconference and he just called in under his own name and nobody asked him where he was from, and he sat in on the whole meeting. [Harris' publisher, David Allen, posted notes on the meeting on his Web site.]
The meeting had quite a few things of concern in it. They were being told that as an industry they had to come up with $200,000 in seven days in order to come up with a P.R. campaign to whitewash their P.R. problem, as they put it.
So apparently they feel they have a problem?
Yeah, they do. And in this particular meeting, one of the things they discuss is, they say, "Now we need to make sure the press never finds out this because we don't want them to know we have a problem." [According to David Allen, Harris Miller, the president of the Information Technology Association of America, said, "We just didn't want a document floating around saying the election industry is in trouble, so they decided to put together a lobbying campaign."]
Was there anything discussed about addressing the problem?
Absolutely, what they want to do is not fix the problem, but they agreed to fix the perception of the problem.
Did they indicate what they thought would be a problem with printing paper ballots?
No. It was a foregone conclusion that we don't want paper.
But they say that they would try to convince the public that having no paper is fine?
Right.
It's rather confusing why they're fighting this ...
Yes, actually I find it a little bit suspicious frankly.
What do you mean by that?
Well -- it just seems like, OK, most of us who've ever run a business before, you know what the public wants. Diebold could have early on become a hero by saying, "You know what, this is a problem, but here's what we're going to do. We're going to make sure that you guys have what you want, we're going to get you this paper ballot." And instead there's this huge amount of money being expended to avoid it. It's such a simple solution -- it's too much fighting over something that's so simple and that is pretty much agreed on by all of the tech experts anyway.
The last thing I wanted to talk to you about is the California recall.
Hey, you Californians. What in Sam Hill are you doing?
Well -- as you know, the other day the 9th Circuit Court ruled that the election should be put on hold because punch-card systems are being used in six counties. Do you have any opinion on that -- on whether it's a good idea to hold off on the election because of the punch-card systems? Isn't it better to have punch cards than touch screens?
Well, here's my opinion on that. First of all I don't understand why you guys are doing this election, but be that as it may. There's a study by MIT and Caltech from 2001, and it found that optical scans lose about 3 percent of the vote, punch cards lose about 4.1 percent, and touch screens lose 5.7 percent. [Harris' numbers are a bit off. The Caltech MIT study, which was one of the most thorough investigations into what went wrong in the 2000 election, analyzed "residual votes" -- "uncounted, unmarked and spoiled ballots" -- caused by different types of voting machines. For the presidential race, 2.5 percent of all votes cast on punch-card machines were residual votes; the rate was slightly lower, 2.3 percent, for touch-screen machines. But in gubernatorial and senatorial races, punch-card machines had a 4.7 percent error rate, while touch-screen machines had an alarming 5.9 percent error. The study's 95-page report is available here.]
If you're going from punch cards to optical-scan ballots, that is an upgrade, but if you're going from punch cards to touch screens, that makes no sense. According to the research, the one system that is currently being sold that is less accurate than a punch card is a touch screen. The court decision doesn't make a lot of sense to me. It sounds to me that, as is so typical with this, you have people who really don't understand the issues and don't understand much about how the computer programs work forming decisions based on a combination of what politicians and vendor P.R. people say.
But one of the problems with optical-scan ballots is that you have to print up a lot of paper -- and, you know, if this election is postponed until March, a lot of the counties are going to have huge bills because they have to print new ballots.
Oh, goodness! I hadn't thought of that. Huge, huge bills, completely wasted.
So isn't that an argument for touch-screen voting?
I think the touch screens, if they had a paper trail so that we could do a proper audit, they would be my choice. The thing is if you speak Chinese, they can print something in Chinese. There would be no reason for all these combinations of ballots that folks have. It's kind of a nightmare which would be solved with the touch screens that can print.
Yes, I imagine that's one of the main selling points for touch-screen machines.
I would think so. It's just that they're not auditable. I'm not opposed to it, and I think it has tremendous advantages, but it just needs to be auditable. That's a deal-breaker -- it has to be auditable. And why I've been so down on Diebold is because they're the poster child for why it has to be auditable.
Get Salon in your mailbox!
