Counting votes

If the interlocks are broken, it does not matter whether someone has been monitoring who went in and out of the radiation area. When it is time to button up and turn on the beam, a search team goes in to make sure nobody was left behind. (And sometimes the safety officer hides a dummy in the experimental hall, to keep the search team on their toes.)

If the package seal is broken on the gauze pads, the burden of proof that the contents are safe is on the person providing the material. The surgeon needs a clear chain of possession and observation with trusted agents.

The default assumptions change when security is broken. If a cracker gets into your computer system, you assume that he has everyone’s password, and left nasty files behind. I can assure you that getting that cleaned up is a lot of work.

If ballots turn up outside of the regular procedures designed to prevent fraud, or are brought in from outside those procedures, the default assumption is that they have been tampered with. The burden of proof is on the election commissioner to prove that they are not. If she fails to provide clear chain of possession and observation by the required neutral observers, the votes are tainted. If the ballots are not collected and processed in the prescribed way by the prescribed deadline, the votes are tainted.

You cannot readily estimate the effect of tampering, and all votes are questionable. You can’t just scale the numbers and say "Instead of 150 votes for the Whigs and 50 for the Tories we’ll count 75 for the Whigs and 25 for the Tories," because tampering is intended to change the balance. On the other hand, this is not adequate to prove that the vote was tampered with, so absent incriminating evidence nobody is going to go to prison.

Holding new elections in Broward for the local elections seems like a no-brainer. The state-wide elections are another matter—the partisans know how many votes they need to arrange for, and will carefully make sure they have them. It may be that a new state-wide election is needed.

I have been dubious of early voting--when you hold ballots for that long guaranteeing neutral observation and possession is hard--and I consider e-voting an invitation to disaster. "There are lots of very smart people doing fascinating work on cryptographic voting protocols. We should be funding and encouraging them, and doing all our elections with paper ballots until everyone currently working in that field has retired." Messes like this do not encourage me.