Dave Zweifel (Madison Capital Times): High-tech voting must be fraud-free

[Dudley_DUright]

Some insist it's nothing more than paranoia run amok, but there's a debate mounting over new voting machine technology and how it could be used to fix elections.

Although it's been going on for months, the debate reached a new level last week at a national convention in Denver of clerks, recorders and election officials.

Some top computer scientists told the conventioneers that what happened in Florida in 2000 could pale in comparison to the vulnerabilities of high-tech voting machines that counties throughout the country are purchasing to replace punch-cards and other paper ballots.

"What we know is that the machines can't be trusted, it's an unlocked bank vault, a disaster waiting to happen," a Stanford University computer science professor was quoted as saying in the Denver Post.

more...

http://www.madison.com/captimes/opinion/column/zweifel/57736.php

[gristy]

1. That would be Dr. David Dill from Stanford

His classes have started (he had been on sabatical last year), and he still had time to go to Denver last week. Bless him.

[kcwayne]

2. The manual system is an unlocked bank vault too

I am a computer scientist as well, and don't share the idea that an automated system by definition is a disaster waiting to happen. This is an unpopular view on this forum, but I would like to offer a few non-technical viewpoints anyway.

When they started re-counting votes in New Mexico (2000), they were running into the same problem as in Florida. There is no voting precinct that could pass any form of audit or generate the same counts twice in the entire country.

The difference between an automated system and the manual one is that the manual system is a fraudulent exercise, the scope the fraud being unknowable. Since it moves slowly and is of such large scale, you delude yourself into thinking it can be accurate and fair because it takes too many dishonest parties to sway the results. (But check this against the dead voters in Cook county 1960, or Florida 2000).

An automated system has the potential to be tampered with by a small group, and so paranoia causes people to be very afraid of all of the hidden demons in the machine that will possibly sway the vote. Since one believes that one can't "see" the vote processing in the same way that one "sees" the manually conducted processes, trust in the system is an extremely hard sell. But the reality is you cannot see or monitor manual systems any better than automated ones, and the potential exists to build a secure automated system that is practical with substantially better security than a manual system.

Think this is BS?? Go to your ATM machine and withdraw $100. What is the probability that you'll get the $100 and your account will be debited correctly? How closely do you audit this, and how often have you ever found an error? How often have you been offered a fake ATM card that will allow you to dip into someone else's account? Now consider that there are millions upon millions of these transactions every single day that vastly eclipse the number of votes that must be counted in an election. If these systems were porous, and leaked money improperly, would banks and consumers continue to use them? The point is the ATM is an excellent model of a secure system that is certainly not a disaster waiting to happen, or it would not be here.

The biggest argument against automated systems is traceability. This is a technical problem, not a conceptual one. There are ways to build secure and tracable systems, the ATM example being one. No voting system proposed that I am aware of has these attributes, but it is not an impossible implementation as the techno-phobes are suggesting.

By way of disclosure, I have no financial interests or associations in any form with parties that would benefit from promoting automated voting machines. I work in the video technology area for a consumer electronics company.

[Fozzledick]

3. Comparison of Black Box Voting to ATM is invalid

Every ATM transaction produces

1: Paper hardcopy of the transaction

2: An audit trail of the individual transaction (which is reported to the user monthly)

This is exactly what BBV is lacking.

Would you trust your money to a system that gave you no record of your deposits and produced no audit trail of individual transactions? ("We're sorry you don't agree with our balance for your account, but we have no record of the deposits you claim."):wow:

[kcwayne]

4. As I said in my post

"The biggest argument against automated systems is traceability. This is a technical problem, not a conceptual one. There are ways to build secure and tracable systems, the ATM example being one. No voting system proposed that I am aware of has these attributes, but it is not an impossible implementation as the techno-phobes are suggesting."

There are specific technologies and methods that could be designed into a system to address the issue of traceability. For example, suppose your voter registration card was a smart card and your votes were encoded on the card. You could check your vote on a reader. Audits could be performed by reading the card and checking the vote database to verify that the transactions in the database and on your card matched. There are a whole series of other things that would need to be done to insure accountability of the tallying system.The discussion is too involved for me to spend the time posting a solution on a bulletin board when readers don't even throroughly read short posts.

[Upfront]

5. Question

Are you a Republican?

[kcwayne]

6. Absolutely not

Are you a Luddite?

[punpirate]

8. Read the Caltech/MIT study group report on this.... n/t

[kcwayne]

10. Do you have a URL?

[gristy]

7. false foundation of argument

But the reality is you cannot see or monitor manual systems any better than automated ones,

You seem to be basing your argument on this statement. This is certainly not a conclusion which can be drawn based on your argument. You have nothing here.

OK, are you up for a challenge? You write your best possible "procedure" which minimizes the chance of fraud in an election with electronic voting machines, and I'll show you where the fraud can occur - and not onesie-twosie fraud. I'll show you how you can shift several percentage points of voters AND you will never know.

Then I'll write up a procedure which minimizes the chance of fraud in an election which uses paper ballots. You will be hardpressed to show that fraud on the order of that attainable with electronic voting will be achievable without a wide and tight-lipped conspiracy in place. The difficulty in establishing such conspiracies, and not being found out, is what makes paper more secure.

[kcwayne]

9. You are on

I don't have a design worked out completely, but I have been thinking about this more lately, and think I can pull together my final thoughts in a couple of days. I will definitely get back to you on this.

[Loyal]

11. Great story!

Eye-opening :kick: