On Sat, 10 Aug 2002 17:06:26 -0400, you wrote:
Go look up discussions on google about cryptographic protocols for internet voting. It just ain't possible without the most strict, obscene, biometric, draconian, "is a person", non-anonymous methods you ever saw.
Sure it is. The measures, if any, taken to insure that the "person" being granted a "digital voter registration card" is a "qualified voter" can be as lax or as stringent as the issuer may require. There is no reason that they would need be more stringent than current process, which, in the US, prohibit voter registration staff from requiring verification of identity. See the "Motor Voter" law.
The point to democracy, in the industrial/agricultural political sense, is one man, one vote. One *anonymous* vote.
Except in Chicago, etc., etc.
On the net, paradoxically, that is completely impossible. Votes can be sold.
No different from the current arrangement. Voting in many jurisdictions can be done today by mail. How would a digital vote, using cryptographic protocols to insure anonymity, and authenticity (the registered person who was issued the digital voter registration has digitally signed the vote) be less likely to be "sold" than a mailed in vote? And pardon the political comment, but almost all votes are sold now, as in the United States the democratic custom has declined to using votes essentially to transfer wealth from earners to voting blocs.
If you fix it so that you can't sell votes without forgoing your identity -- and thus your freedom -- and physically showing up somewhere to vote, or at least proving that you have a device that identifies you as a voter in the most immediate terms possible, you can sell your vote, anonymously, on the net, for whatever the market will bear, and *that* person can *re*sell your vote, and so on, just like it was voting rights to a share of stock.
It is quite simpler to do such fraud with mail in votes, or even "buy me a drink and I'll vote however you'd like", or "yes, this is my pictureless voter registration card, and I'm here to vote".
That bit of cryptographic mobiosity is probably down at the semantic level of consistency versus completeness. Somewhere, Goedel and Russell are laughing.
A laugh a day keeps the economists away.
The net result, of course, of any kind of truly anonymous internet voting, is anarchocapitalism, where people sell their voting control over assets, including political "assets", over and over in secondary markets, on a continuing basis, in real-time. No political small-d democrat (or small-r republican, or small-l libertarian, whatever) I've ever heard of would call that a "true" democracy.
The "sold vote" boogeyman". You need to submit evidence that "anonymous" "internet" voting is more likely to be fraudulent than paper, voter-present by mail voting. You have submitted none, and the "cryptography" word is insufficient to scare me off. The "bogus digital voter registration" boogeyman. You may also wish to show how digital voter registration cards would be more likely to be bogus than "Motor Voter, no-id required" registration cards. Good luck. The "crypto" boogeyman. I challenge you to show that current, published crypto voting protocols cannot accomplish the following: 1. one digital sig, one vote, the first one, and the others are discarded 2. no dig signature, no vote 3. no dig voter registration, no dig sig 4. anonymity, i.e., no connectibility between the voter's choice and his identity. 5. auditability, i.e., connection between each voting "lever throw" and a dig sig for the current vote. Next, the "internet" boogeyman. It's just a pipe/wire/whatever. Bits. Don't be afraid. If the bits are properly signed, no problem and whether "internet" bits or voter-machine-punched-paper-tape-bits is irrelevant.