Ok, what’s the poop on the Diebold voting machines. I’ve heard a few times that the CEO of Diebold is a major Bush campaign contributor and that the voting machines have some sort of anti-tampering software that has been disabled.
(I first read that in Rolling Stone in a interview, but sorry, I can’t remember who it was. Then I’ve heard it several times in the past month, and even here in the thread: …Election…What’s changed?)
Obviously, one becomes very concerned when a voting machine can be tampered with and leave no paper trail.
Can one ask to vote on paper instead of the machine? Is that option available?
Can the machines be secretly skewed?
How widely will they be in use in November?
The whole idea of not having a paper trail to follow seems really scary to me. Salon has been doing a series of in depth articles on the situation. (go to the site and search for Diebold) I’m not reassured by the company’s replies. Some poster will probably point out that Salon is primarily a liberal magazine but I don’t know why everyone, liberal and conservative alike, wouldn’t want a voting system that will produce a hard copy that the voter can use to verify his or her vote and that can be recounted in the case of disputes.
I’m not overly concerned whether the CEO is a Bush supporter or a Nader supporter. If his company deliberately engaged in an attempt to subvert elections, not only would they be opening themselves up to significant legal and finanacial liability, but they would also destroy their market as soon as this got out. In fact, I am pretty sure that the current rumors alone are sufficient to depress demand for the product.
But it does seem idiotic for a company that produces ATMs and the like not to include a similar level of security and auditability into voting machines.
I agree, Salon’s coverage has been excellent. I personally discount the fact that the head of Diebold is partison Republican – CEO’s often are. There’s no need to invoke that particular Bogeyman when there are so many others. The whole system is poorly designed, doesn’t meet minimal security standards, and it unauditable. The last is the worst – those who do believe that the CEO may have done something and the surprising results in the Georgia elections were rigged, they have nothing to base their claim on because there’s no way to go back and double-check the results.
If you thought that 2000 was bad for elections, just keep waiting…
But national hand-wringing is too late. We need to address this issue on the local level. Cities and counties are already deploying these systems, and the decisions are being made by well-intended bureaucrats and public servants who are completely unqualified to analyze the security of specific products. They are purchasing a bill of goods they don’t understand, and we will be stuck with these systems for years after they’re deployed. We need to block the deployment of these systems until they can be vetted by qualified professionals, and we need to disqualify all the vendors who are trying to prevent that level of analysis, whether by claiming their source code is proprietary and not subject to review (as Diebold does) or by vendors who intentionally circumvent the version control systems to slip in changes after systems are in lock down (as Diebold did in Georgia).
I’ve been voting since the elections of 1980, and I’ve only voted on those cardboard paper punch ballots. But, back in the olden days, some time previous to my experience, they used mechanical voting machines. You’d go in and actually pull a lever that would increment a counter for that candidate. These things didn’t leave a paper trail, and apparently worked well enough.
The big difference between these and the Diebold machines is that you could open up the front cover of a mechanical device, pull the lever a few times, and see for yourself that it’s working OK.
And I think that’s a big difference. The vote was recorded mechanically and you could easily verify the machine was working. To change a vote, you’d have to physically do something to the machine. With the electronic systems, you have to have faith that the vote is being recorded correctly and that no one has anyway to change it. Plus, it’s not like the pollworkers will have enough computer saavy to know if the electronic machines are working correctly. We all know computers can behave unpredictably; how often have you had a post completely eaten on this very message board?
We are much more relaxed about who our government is over here in Australia and are much more relaxed about voting. We are a a fairly laid back people. Every election I have to pull myself out of a hangover stupor on the couch drive or walk up to the nearest school, church or wherever the hell you have to go and get a bit of paper so damn big you have to fold it over 3 or 4 times, cram into a tiny booth and fill in a million boxes or just 9, i dont know anyone who has filled in all the boxes yet.
man a machine and a few buttons would make my voting life bliss, whatever gets me in and out of that place quickest.
This case is being used to argue that such high-priority source code should be made open-source. (I won’t take a side in GQ, but I will present both arguments fairly as I understand them.)
The pro argument runs like this: Trusting the source code to an electronic voting machine to any one group is the equivalent of trusting the voting process itself to that group, at least as it pertains to the regions that will use that source code. Since the idea of a private company holding the strings on the voting process that directly is absurd, the source code should be open to public review and audit. That way, all interested parties will be able to verify its correctness, security, and complete and utter lack of bias. Any modifications to the source code under such a scheme would be rigorously reviewed by the people who are knowledgeable in such matters and, more than likely, have a direct personal interest in the quality of such code. Both bugs and malicious modifications would be weeded out.
The con argument runs like this: It is possible that crackers could sneak problem-laden code past the review process, possibly tainting the final product come poll day. In addition, the Diebold corporation has a right to preserve its trade secrets as they were used in the source code.
The rebuttal could run thusly: There are many prominent and widely-known open-source projects of high importance to the world economy (such as the Linux OS, the Apache webserver, and the various BSD OSes). None of them have been compromised by crackers in this way. And if the Diebold corporation feels justified in using secret procedures in programs that directly influence the public weal, why should Diebold be trusted to write such code? Openness is the watchword of democracy.
I expressed my dismay that the machines are too easily hackable, that there’s no paper trail. and asked what they are doing about them. Maybe you all might want to do the same.
You can check out some internal Diebold emails that people found on the internet (yeah, that’s right folks, they didn’t protect their own network security…and we expect them to keep our votes safe). They talk about so many bugs in the system that it’s scary.
When the emails were first leaked, Diebold threatened to sue any server that held the files, including my university. Sadly, the administrators here have balls the size of grape seeds so they took away internet access for the poor guy who was hosting the files. Later Diebold realized it had no support so it just gave up.
We definitely need open-source software. Forget that proprietary crap. For all we know, the software might have a secret “back door” that will allow the CEO of Diebold (or whoever knows about it) to tamper with the vote tallies. Remember what happened in Florida last time; a few hundred votes more or less could have altered the outcome of the election.
I don’t like the idea of paper receipts that the voters take with them. Others suggested that the voter could use it to prove they voted as instructed and then collect a payoff. My objection is that there’s no way to verify the overall results if the voters take the paper receipts away.
Instead, I’d have the voter make the selections on a touchscreen machine that then prints out a receipt showing the votes placed. Once he verifies that the selections were as intended (to avoid a repeat of the Florida situation), he would drop the receipt in a slot where the votes would be scanned and tabulated, and then the paper dropped in a box. Then the election monitors could open a sample of boxes to verify the paper against the electronic count.
Another thing that was pointed out (perhaps in Salon) is that it seems odd that all the other systems Diebold builds (ATMs, building access systems, alarm monitoring, prepaid meal cards, etc.) have audit trails built in, but the voting systems they designed did not.
