Such a scheme has a huge flaw: it allows extortion/coercion. Suppose I am a mob boss and I wish to become (insert position here). I send my goons to inform you that voting for me is a wise decision because I can offer you protection from the vicious thugs in the area. It would be a shame if the vicious thugs burned your house down, wouldn't it?
So, you'd be forced to reveal your bitvote address to me under threat of violence. Once you vote, I just check that you voted for me. And if you didn't...
A real cryptographic voting system goes to great lengths to prevent this scenario. In such a system, a voter can prove to himself that his vote counted toward the proper party, but crucially the voter cannot prove such to anyone else. This prevents coercion.
Of course, real cryptographic voting systems have a bunch of other nice properties, but this is a really major one that most people overlook.
Example of a cryptographic voting system: Helios (https://vote.heliosvoting.org/). Also see the tech docs (http://documentation.heliosvoting.org/verification-specs/hel...). Helios is an online system, so it's not ready for prime-time voting --- it would need to be adapted into an in-person scheme, which cryptographers have already created --- but it does demonstrate the concepts involved.
I think there is a simple way to counter-act vote buying/coercion: if the voter has a mechanism to cheat the cheater. If I can accept a hundred bucks to vote A, then walk around the corner and change my vote to B in secret, it creates sufficient disincentive to buy or coerce votes.
It's not a huge flaw, since our current voting system (in any country you care to name) allows exactly that, and your scenario is not observed in the wild, since it is illegal.
The typical US voting system allows coercion? When I voted, I went into a publicly visible booth, recorded my vote, pressed submit, and went about my business. I suppose I could have been forced to take a picture with a cellphone, but I could have just selected the option they wanted, taken the picture, and then swapped back to the candidate I wanted. Is there something I'm missing here? Any way you slice it, even if the current system allows coercion, if we're talking about cryptographic voting systems, we might as well use ones that don't suffer the same issue.
> your scenario is not observed in the wild, since it is illegal.
I don't mean to be snide, but I've witnessed a great deal of illegal things, so "since it is illegal" is not that convincing.
Even if you discount the threat of violence because of the possibility of police intervention (e.g. you report being threatened), consider the notion of vote buying: a rich person just says they'll pay you lots of money to vote for them. This isn't necessarily something people would want to report, given the possibility of financial benefit, especially if the person is down on their luck.
At any rate, if electoral fraud is something that is "not observed in the wild", people sure do make a big fuss over it (see http://en.wikipedia.org/wiki/Electoral_fraud for some references).
Ah! Of course. You're right. I've never used an absentee ballot, so it completely slipped my mind.
This is actually why Helios voting (which is end-to-end auditable) doesn't prevent coercion either: your voting is done 'at range' (online), so a coercer can just stand over your shoulder. Just about the only surefire way to prevent coercion is to somehow physically give someone security while voting.
Okay, well, anyway, my point still stands; even if coercion is possible in the current-day low-tech system, if we're going to discuss a cryptographic replacement, we might as well restrict ourselves to the baseline "best-studied" ones so far, virtually all of which include coercion-defeating mechanisms.
At a minimum, if a physical presence were the norm with cryptographic voting and there was an option to be absent (like today), then we would be in an identical scenario with regard to coercion while gaining the end-to-end auditability of the cryptographic scheme. It seems like a net positive, although many critics of cryptographic voting see the increase in complexity as unworkable.
You're missing the fact that it's very possible to use coercion to discourage certain groups from voting at all. Just target the groups that are likely to vote for your opponent, and you've essentially used coercion to manipulate the election.
This is a problem which would be solved, somewhat ironically, by the US adopting mandatory voting. Shift a really nominal fine onto "not voting" (like $30), and it will become politically impossible to get away with marginalizing groups from voting.
Because people might ultimately think "well my vote probably wouldn't have mattered" when the line is 8 hours long on a Tuesday, but they will definitely care when they get fined $30 - but they're not going to simply argue the fine should go away, because it will be a little reminder of "oh yeah, and you didn't get a say in your government either!"
If I were Australian, I would have a problem with it, because I am what you might call a principled abstainer. But apart from personal issues regarding perceived rights violations, there are several obvious advantages and disadvantages, similar to any aspect of election design. I'm not saying compulsory voting is game-breaking, but merely that it's a compromise that one could argue is not desirable.
You're under no obligations to actually vote in Australia, just to show up. You can freely get your name ticked off, put a line through all the candidates and write "try harder" and walk out.
Fair enough. It's hardly the worst idea I've ever hear of, nor the most upsetting to be personally (the fact that "modern" Western governments still reserve the authority to perform military conscription is far worse). I would still prefer not showing up for paying the fine.
When I lived in Ecuador long ago, voting was mandatory. The people with your views ran a spirited campaign with posters that said "VOTA NULO", vote null. I'm not sure exactly how it worked there, but in some places there's an explicit "vote for nobody" choice, and those choices are counted and reported: http://es.wikipedia.org/wiki/Voto_nulo
Actually, people can and do sell votes this way in places in Italy. How they do so is extensively documented. It's one of the reasons that cameras are supposed to be strictly forbidden in voting booths - so they can't take a picture verifying how they voted.
>it allows extortion/coercion. Suppose I am a mob boss and I wish to become (insert position here). I send my goons to inform you that voting for me is a wise decision because I can offer you protection from the vicious thugs in the area. It would be a shame if the vicious thugs burned your house down, wouldn't it?
How is this different than what currently happens?
Our current voting system has an emphasis on the secret ballot.
There are records of which polling place a person voted at,
and whether they showed up to vote, but (in theory) there's no way to tell which of the thousands of ballots from that polling place match to which individual voter.
But there is also no way for us the verify a valid and safe vote. Especially with our lively die bold machines. I'll take the thugs, give me a fair vote.
People from several parties, in a public, observed location. It's not hard; it works here in the UK, and it worked in the US before all this voting machine nonsense started.
So, you'd be forced to reveal your bitvote address to me under threat of violence. Once you vote, I just check that you voted for me. And if you didn't...
A real cryptographic voting system goes to great lengths to prevent this scenario. In such a system, a voter can prove to himself that his vote counted toward the proper party, but crucially the voter cannot prove such to anyone else. This prevents coercion.
Of course, real cryptographic voting systems have a bunch of other nice properties, but this is a really major one that most people overlook.
Example of a cryptographic voting system: Helios (https://vote.heliosvoting.org/). Also see the tech docs (http://documentation.heliosvoting.org/verification-specs/hel...). Helios is an online system, so it's not ready for prime-time voting --- it would need to be adapted into an in-person scheme, which cryptographers have already created --- but it does demonstrate the concepts involved.
This is an area of active research, BTW.