ha ha ha.

Nov 09, 2005 10:09

All the propositions were defeated. I admit I voted in favor of the communistic Prop 79, but I never really thought it would pass. Oh well ( Read more... )

Leave a comment

arturis November 9 2005, 22:03:37 UTC
See, what you actually want is accountability and redundancy, which can be accomplished without paper.

I agree that you should get a voting receipt with a random verification number on it so you can associate yourself to your vote. And I think you need two separate machines, one with the voter registration databases keeping track of who voted when and the other with the actual vote counting which generates arbitrary numbers to assign to each vote. Then with your receipt and the records stating that you did vote at that polling place at that time, you can verify your vote. Ideally the vote verification process could be automated after the fact and web-ified such that anyone can check on their vote by logging in with their receipt and some personal information. The system can check the public info against your personal info, namely that you attended a certain polling place at a certain time. The system would also know if the receipt you hold was issued from the same polling place at roughly the same time and could authorize you based on that. You could check your vote online and issue a complaint if something ended up wrong.

Anyway, I think purely or mostly electronic systems can work. I just thing Diebold didn't try to do it right.

Reply

anadamous November 9 2005, 23:45:34 UTC
You can accomplish redundancy with two different databases, but only paper gives you an absolute canonical record. I mean, if the computers differ, which vote do you take? Do you just throw the vote out? This will be especially important for the first couple of years of using the machines. Hopefully after a couple of years of trials,

I also disagree that you should be able to log in with the receipt and some personal information. There should be no connection between the voter and his record, other than the ID number which he happens to have. The ID shouldn't know when it was issued.

I'm sensitive to thatdumbjerk's point about printer paper jams, but office printers are designed to take a lot of different types of paper and are loaded by random idiots. How often does an ATM paper receipt jam? I think it would be a matter of using equipment and paper that's standardized for not-jamming.

Reply

arturis November 10 2005, 02:27:27 UTC
You can accomplish redundancy with two different databases, but only paper gives you an absolute canonical record. I mean, if the computers differ, which vote do you take?

What? If a computer prints out the paper, what makes it more authoritative? What you're basically saying is that with a paper ballot the user could verify the accuracy of thier vote before submitting it fully and then the paper ballot wouldn't change over time so it would perfectly preserve the correct vote. But fundamentally that system is based on the verification process by the user, so if the electronic system allows that same verification to take place at any time after the vote is cast, it doesn't matter that electronic information is more easily changed.

I also disagree that you should be able to log in with the receipt and some personal information. There should be no connection between the voter and his record, other than the ID number which he happens to have. The ID shouldn't know when it was issued.

This is obviously a failure of my explaination. What I want is one machine that records the time that you checked in to the polling place, and who you are (as the current regristration system works), and another machine which records a vote and correlates it to a number that it assigns and prints on a receipt. When you log in, two things are checked. The first is that your personal information matches the logs in the first machine, that you were at the polling place in question at the time recorded. The second is that your receipt matches the same information about polling place and time.

The receipt wouldn't need to have anything printed on it except the vote number, but the database of recorded votes would have each vote, its number, the time it was cast, and the polling place. If you want to make the information less trackable, you could record the time accurate to the hour rather than the minute, so it would be harder to connect votes to people.

But let me tell you why this information is necessary. You need to have a system where someone can challenge the record of their vote and the proof that they were counted can be supplied. But you don't want a system where the receipt is effectively a 'bearer' instrument and anyone who gets ahold of the receipt can look up the voting record for it and claim that they were the one who cast that vote and demand that the vote be 'corrected' to what they want it to be.

You want a system that is anonymous enough that people can't be connected to their vote by anyone but themselves. So that connection of person to vote is handled by the receipt, which only one person can have. With the personal information prerequisite, say address and full name, then someone who picks up a discarded receipt will have little chance of being able to use it for deviousness. The system still doesn't keep enough infomation to create a direct link between people and their votes, so it can't be used to learn about how someone has voted.

Then, of course, if someone issues a complaint, the web log system should be designed NOT to record any of their personal information. Instead it should only say that vote number X is disputed.

Reply

anadamous November 10 2005, 17:40:18 UTC
I don't much care about the paper ballot, honestly. I just think it would make people feel better, at least for the first few years, and I just wanted to show that it would be easy to do. All we need are two redundant databases that each independently store the votes, so that the two databases can be compared with one another.

There should also be a third machine that registers that you have voted, and counts the number of votes for the polling-place. This machine does not know your voter number, however.

I disagree about the vote-challenging process. This is how it should work: every single vote is online. You get an anonymous, generic-looking receipt with your number. You could easily fake this receipt if you wanted to; it's for your information, and not to show to others. You can go online and check your vote, to make sure it was registered correctly. If you find an error, you can complain. But here's the catch: you don't have to prove it was your vote. When you make a complaint, you have to give your full contact information. Each voter can only make one complaint - it doesn't ensure that they're complaining about their own ballot, but it does ensure that each individual can only complain once. And your complaint is just, "There was an error on this ballot." You don't say who you meant to vote for or anything. (This should reduce the possibility of voter intimidation.) Which ballot you're complaining about is stored independently of the record of the fact that you complained, just like your voter record is stored independently of your vote. Then, if the number of complaints reaches a certain threshold, a recount is ordered. If it reaches another threshold, you must hold another election.

Reply


Leave a comment

Up