VoteTrustUSA: The Long Road to a Reliable Voting System

[Wilms]

The Long Road to a Reliable Voting System

By Warren Stewart, VoteTrustUSA

April 01, 2006

Among their many activities, the National Institute of Standards and Technology (NIST) chairs the Election Assistance Commission's Technical Guidelines Development Committee (TGDC). The TGDC is responsible for the development of the nation's Voluntary Voting Systems Guidelines (VVSG), which although optional for the states, are pretty much mandatory for e-voting vendors who are expected to build to these standards. This is one reason why Diebold's use of interpreted code which is in violation of these standards, has come under so much scrutiny.

As previously reported one aspect of the VVSG itself that is severely deficient is its Hardware Reliability spec that allows almost 10% of e-voting systems to fail in any 15-hour Election Day. Such failures, especially of touch screen machines, have resulted in voter disenfranchisement, possibly even affecting the outcome of elections.

Suggestions for improving this standard (which dates back to 1990) made by experts such as Dr. Stanley A. Klein, Dr. Rebecca Mercuri and Alfred DuPlessis (a bona fide Reliability Engineer) as early as 2002, appear to have fallen on deaf ears in both the IEEE Voting Systems Standards Committee as well as the EAC itself as recently as last year when this issue was raised in public comments on the 2005 VVSG.

snip

Howard Stanislevic, a computer network engineer and research consultant for the VoteTrustUSA E-Voter Education Project said of the recent opinion by NIST, "I am very happy that the voting systems Reliablity issue was raised by Dr. Goldfine, who stated that there is now a consensus within NIST that this standard should be improved. The issue now is by how much, at what cost, and most importantly, when? Clearly, the current standard is inadequate."

snip/links

http://www.votetrustusa.org/index.php?option=com_content&task=view&id=1151&Itemid=26

[mopinko]

1. make them meet the same standards as

wall street. my hubby works in IT infrastructure at an sec regulated exchange. suggesting these standards for reliablity and accuracy would get him laughed out of his job.

[Bill Bored]

3. Been there; done that. This standard is a joke. And not just on Apr. 1! nt

[AuntiBush]

2. Question:

Is there any way possible we can request in writing proof of how our vote was registered, after votes have been tallied?

[Wilms]

4. Not if you want to keep the balloting secret..

Currently, the law says balloting should be secret. While many people seem to agree with that, the acceptance of that notion is not universal.

[Land Shark]

5. it's not easy but there might be this way:

1. polling place issued 1000 ballots with serial #'s on them (current practice) and then voter removes stub w/ serial number on it before depositing completed ballot. (current practice)

2. Instead of removing stub, the 1000 ballots would be mixed in a box and the voter could randomly take a ballot. The serial number on that ballot is now their tracking number if they WRITE it down (stub stays on/or there is no stub). However, now the serial number is not trackable to the voter.

3.the serial number is scanned along with the voted ballots and the recorded votes can then be tracked and recorded together with the serial number without being able to trace it back to the individual voter.

[Wilms]

7. That would work, I think, except it won't prevent vote fraud and coercion.

Though mail-in doesn't prevent that, either.

[Febble]

9. I think that's brilliant

I think we should do that in the UK too. All you need is for the blank ballot to be selected randomly, with stub, BEFORE voting. Stub is then receipt.

cool.

To Wilms: yes, coercion and fraud are a problem for mail-in - our UK HBPBs are a good system, but the one weakness is postal voting - and indeed there were incidents of fraud during the last election when postal voting was made more widely available. Mail in is not a panacea. There's more than one way to skin a cat.

[Wilms]

10. I agree that mail-in is no panecea.

In fact, I'm a bit skeptical of it.

My comment was not an endorsement/excusing of mail-in. It was only to point out LS's concept was no worse.

[Febble]

11. Exactly.

I think coercion is a real issue.

There is a real difficulty balancing secrecy of the ballot with transparency in the count, and both of them are important to the prevention of election fraud. That's why I think LS's suggestion is so good.

Cheers

Lizzie

[mopinko]

6. sometimes i think this is where we went wrong.

seriously. i understand the concerns. but i would rather have a few whinos get a drink than the mess we are in now.

[Wilms]

8. I wonder the same thing.

Though I'm also concerned that a husband-beater may force her spouse to vote a certain way.

For a state-wide, or federal election, my guess is that "vote fraud" is a minor concern. But for a local election, I imagine a handful of coerced/bought votes could be very problematic.