(today's SMBC)

@DVK-in-exile I imagine that in the case of fraud happening where someone votes who is not dead but simply didn't actually vote is far more common, and also far less detectable.

@Bobson Democracy works best when the people trust it, and if they need a box of paper to feel they aren't being oppressed by "the man", then a box of paper is necessary. If they really want to prevent fraud they would encrypt each vote, the public key being provided during the voting period, then have them tallied by a master device using an unshared private key. They would also have to enforce voter ID.

@GorchestopherH So you're advocating that each voting machine encrypts every vote with the master device's public key? That's strong protection against tampering, but how can you prove that when the voter chose "Candidate A" than the data which was encrypted was actually "Candidate A" and not "Candidate B"? If the machine encrypts the wrong data, you'll get the wrong data when you count.

Also, how do you prevent someone from digitally stuffing the ballot box by making the machine encrypt extra votes that no one ever actually cast?

You might also be interested in this form of voting system, which is the reverse of the form of VVPT I suggested before. Here, the voter marks a ballot, then feeds it into the machine, and confirms that the machine has correctly read their ballot.

But there's still the paper that can be looked at for auditing purposes.

(The two concepts are not mutually exclusive, although the two specific implementations are mirrors)