I didn’t want to hijack BrainGlutton’s thread, so I started a new one. Attached below is an election process proposal. What are the pros and cons of this scheme? Is it sufficiently robust? Does it have shortcomings that our current system does not? Is it markedly better than our current system?
I can think of one big shortcoming and a few holes in the process, but I don’t want to spoil the fun…
Ballots are cast using whatever high-tech gizmo is available. Each voter receives a paper ballot-receipt that includes a sufficiently random ballot serial number and the details of the votes cast on that ballot.
After polling closes, votes are counted digitally and the results are published in detail on the internet. Details include:
a. A table of all registered voters and whether they cast a ballot or not. This table is compiled as a result of validating voter registration and not from the process of casting a ballot.
b. A table detailing each ballot cast that is indexed by ballot serial number. Of course, no reference to the individual casting the ballot is includes (or possible for that matter).
All registered voters are responsible for verifying their status in the registered voter table (e.g. whether they voted or not).
Registered voters that cast a ballot are responsible for using their ballot serial number to verify that their votes were cast correctly.
Registered voters are responsible for notifying election officials if either (3) or (4) is incorrect. Election officials investigate the results if they receive a sufficient number of notifications.
Registered voters certainly have the right not to perform (3), (4), or (5).
Interested third-parties can use the published results to validate:
a. The sums for each candidate.
b. That the sums do not exceed the number of ballots cast.
Good point. I should have been more specific with (2). Paper copies of the results would be available to the public at the polling place or similar location. I just mentioned internet publishing because I thought it was a more rigourous requirement. I don’t expect most people would validate that their ballot was counted correctly, except in possibly close elections.
Thanks, I completely missed this problem. Anything more specific than an “I Voted” sticker can be a liability. Is there a way to mitigate the risk? Perhaps the ballot receipt only contains the ballot serial number and the comprehensive results are published after a week or so. Would this eliminate the majority of coercion?
This is one of the holes that I alluded to. Perhaps absentee ballots could include a ballot serial number that is used if the ballot is filed? This probably wouldn’t compromise privacy any more than the ballot itself.
What’s to keep a person from claiming their ballot was ‘wrong’ when, after the fact, the total results suggest that changing one’s vote could change the outcome? Let’s say I voted for a third-party candidate but it turns out the margin between Dems and Republicans was very close, much closer than I thought it would be when I voted. The possibility of changing one’s vote hours after the fact because it was wrong opens the door for this sort of abuse, i think.
Good question - I didn’t explain this in my OP. Basically each ballot is validated before the ballot receipt is printed. Once the receipt is printed, the ballot is cast and there is no way to change it. A registered voter can notify election officials only if there is a discrepency between the ballot receipt and the published ballot results. Once the receipt is printed, it is no longer possible to complain that you accidentally voted for the wrong person. Presumably the high-tech voting gizmo is sufficiently robust that it lets each voter verify their choices before the final cast.
Thanks for the heads-up Polerius. Sorry to (subconciously) steal your idea. From the comments in your thread it seems that it is dificult to provide transparency and complete anonimity at the same time. The ballot receipt could encode the votes cast, but sooner or later someone could look them up on the published results. Jurph suggests one method, but it seems limited to a single contest.
Old System:
Oppressed immigrant is paid $5 to vote for Candidate A.
Oppressed immigrant goes into the poll and votes for whomever he wants.
Oppressed immigrant come out and says “I voted for A!” when interrogated by the Boss and gets his $5.
New System:
Oppressed immigrant is paid $5 to vote for A
Oppressed immigrant goes into the poll and votes for whomever he wants and gets a reciept.
Oppressed immigrant comes out and says “I voted for A!” when interrogated by the Boss.
Boss says “Well see about that!” and uses the reciept to check the vote.
How about this:
Voter enters the polling place and does whatever they need to do to veryify they are eligible to vote.
Poll worker validates electronic token and hands it to voter.
Voter goes to the electronical voting machine and inserts their token.
Machine verifies the token is valid, and presents the voter with a ballot.
Voter votes.
Machine prints out hard copy version of ballot.
a) Voter confirms paper ballot is correct. Machine records vote. Machine invalidates token. Machine issues verified paper ballot to voter.
b) Voter alerts that paper ballot is incorrect. Machine instructs voter to tell poll workers. Poll workers tell machine to destroy paper ballot. Voter goes back to step 3.
Voter places hard copy in ballot box.
Poll workers collect invalidated tokens from machines for reuse.
Votes are tabulated electronically.
Paper ballots are counted at various locations for random spot-checks of system integrity, and anywhere irregularities are suspected. For the purposes of recounts, the paper ballot is the legal vote.
Judging by your comments about the printed receipt matching the published results, do I understand correctly that the ballot reciept would have a recap of all the votes the person made? I think that removes too much secrecy.
In my OP I suggested that the ballot receipt have the detailed choices in text. Since this may compromise secrecy, an alternative would be to encode the choices using a digital algorithm that is only readable by computer. This prevents basic coercion and still allows for verification of the ballot. However, as rasied earlier the big secrecy issue is that eventually the results are published by ballot serial number and so coercion is still possible.