Hacker News Reader

Comment by helloguillecl | Hacker News Reader

I cannot imagine a system more transparent than the one implemented in the country where I come from (and some others as well):

After closing the polls, the paper votes are counted manually (in from of candidates representatives and public in general) by the poll station "vocales".

"Vocales" are citizens (4 in total) randomly selected for each poll, who must operate the poll during the entire day. They must report the results of the counting directly to the central system and sign a certificate, so everything is traceable down to the polling box, and it is very difficult for a candidate to conspire against a fair count. Results are normally available from 1 to 3 hours after closing the polls.

I'm completely against e-voting as I feel that transparency in every election is way more important than efficiency, and I think that it cannot get more transparent than this.

View on HN

It sounds like this is worse than what I've experienced in multiple US jurisdictions. Any of the random nominees can unilaterally discard my vote without detection. In the US, I physically insert my ballot in a sealed counting machine that the poll workers cannot tamper with without detection. Nobody who has access to my ballot knows which vote is mine, so nobody can censor me without detection.

If the 4 people counting the ballots are chosen truly at random after the voting is done, and they count the votes independently from each other, and there is some punishment (even a mild one, like a moderate fine) for disagreeing with the majority count, why would a counter discard your votes ? they have 0 gaurantess that others would discard the same number of votes so why risk being the minority report and getting punished ? bonus points if there is a reward for agreeing with the majority.

If the randomness gaurantees hold and everybody knows they hold, I have trouble imagining a better system. It has a blockchain-y feel to it, you have no reason to trust any single rando, but a bunch of random non-colluding people all acting consistently is a very strong indicator that they are actually telling the truth.

Yes if they all count everyone, then they have to collude to discard my vote.

The better system is what you described, plus a box that I insert my ballot in that counts it. The electronic machines used in SF are basically the same as an additional random poll worker, with the added benefit that the machine's count can only be manipulated by physically inserting ballots into the machine.

You can have electronic voting machines print a paper record which can be counted in case of issues.

we could even have generated guid on the paper records that you can look up in a database to make sure your vote is being recorded correctly, without exposing any identifying information.

I have this in San Francisco, although I've done the "lookup" part only by email.

But how do I know that no one is correlating those GUIDs with my personal information? And more important: how do you convince a common citizen that their votes are not being tracked when there's a unique identifier right there?

The way I see it, it's a system that only works in elections when it's not needed. Otherwise, all you need to suppress people's votes is some thugs saying "we have men on the inside - if you vote for someone else, we'll know". It doesn't need to be true, just plausible enough to make people think twice before voting.

It's very difficult to provide an easy to use vote receipt that can't also be used for tampering (threats or bribes for the wrong/right votes).

(Maybe I am reading too much meaning into "recorded correctly")

This is a solved problem: https://news.ycombinator.com/item?id=31675726

I am no expert and so will not argue stridently, but are they really easy to use?

https://en.wikipedia.org/wiki/Scantegrity looks sort of complicated.

For the voter who just wants to vote and leaves it to others to do any auditing, they simply take their ballot into the booth and fill in the bubble next to the option they want to vote for. It looks just like normal optical scan voting except filling the oval with the provided pen reveals a code instead of completely blackening the oval.

For the voter who wants to check later that their vote was included correctly in the total they have to also to note the codes that are revealed when they fill the bubbles.

Voters that wish to check that the have not received a rigged ballot can ask for two ballots, and then pick one of the two at random and reveal all the codes before going into the booth and using the other to actually vote with. They can afterwards look up all the codes on that first valid to verify they are all valid. This actually just shows that the ballot they picked to verify with instead of vote with was not rigged, but someone trying to slip in rigged ballots has a 50/50 chance of being caught each time a voter asks to do such a check.

Reply on news.ycombinator.com