Hacker News Reader

Comment by apo | Hacker News Reader

I'm generally pessimistic about the desirability and utility of block chain voting systems.

Spend any amount of time around Bitcoin beginners and the one of the recurring themes will be lost private keys.

And "lost" includes countless thefts through network-based attacks aided and abetted by people conditioned for years to take computer security lightly.

If voting identities are controlled by private keys, then voters will lose their voting rights left and right.

Worse, just because an election is "on a block chain" doesn't mean that voter fraud isn't being committed. For example, buying and selling cryptographic voter identities could become a very lucrative industry.

View on HN

In addition to all that, pretty sure Ethereum isn't anonymous. Since the public keys are presumably registered with the government can not only tell who you voted for, stuffing the ballot box becomes trivially easy by just registering a bunch of fake public keys for which you control the private keys. Maybe some of these problems could be solved by generating the key at the ballot box & just throwing away the private key, but then I fail to see what addition a blockchain provides as you still need to verify the public keys are valid first.

The keys aren’t known to an individual. They are used as a way for citizens to confirm that their vote was counted.

An example- go to the poll. Cast a vote and get a print out with a private key.

Go home and check the blockchain to make sure your vote is recorded. If your vote is missing, report somewhere official with your sheet of paper.

There’s lots of variations, but anonymity is only given up if your vote is missing. Then you just vote again. Smart contracts make sure your original lost vote can’t be applied.

The US needs something like this because currently our digital voting leave no record and there’s no way for an individual to confirm their vote was counted accurately.

That’s intentional in the US, otherwise people would sell votes or get coerced to vote.

It also means that you can prove to others that you voted the "right" thing. With all its implications.

> Cast a vote and get a print out with a private key. Go home and check the blockchain to make sure your vote is recorded.

Apart from the problem that this enables vote buying or coercion, it also:

(a) is a good use for a public database, but putting that public database on the blockchain is just a useless complication of things; and

(b) only addresses part of the problem: yes, any voter's vote is counted; but is everything that's counted an actual person's unique vote? That is, this doesn't protect against ballot stuffing by multiple voting or by simply faking database entries.

Overall, not a terribly good approach, and even if it were, no need to "blockchain" it.

a) it’s really inefficient, but necessary to make sure every vote is counted. It’s not a database per se, just a record.

b) I don’t think vote stuffing is solved by blockchain as that requires more work at polling site, audit of voting software, etc. but that’s an existing problem that is a higher risk by citizens not knowing if their vote is counted. A blockchain voting solution doesn’t solve all problems, just some of them.

There are likely better approaches. I was just responding to the question of how this could help and why useful since someone upstream was worried about anonymity.

> it’s really inefficient, but necessary to make sure every vote is counted. It’s not a database per se, just a record.

I don't see why it would be necessary to have a blockchain. It would work just as well for the election authority to publish the official database of all votes and everybody ensuring that their vote is in that database.

> There are likely better approaches. I was just responding to the question of how this could help and why useful since someone upstream was worried about anonymity.

Right. The approach with a personal verification key is a good one.

What if I don’t trust the election authority? Having a public ledger as each vote is recorded is really powerful (or a smart contract for all votes to be countable whenever polls close).

It would also reduce the ability to pad with votes after the fact. Would help for analysis of any abnormalities (eg, five times population voting, stuff like that).

I think it becomes more useful the less trustworthy the government (eg, Russia).

Maybe we can skip the farce of voting and move government selection directly to proof of stake?

This looks like a premature assessment. The problem with voting systems are many, and so are the benefits and ways to apply blockchain tech. There are other examples in different comments already, but the most glaring prematurity here sounds like some sort of comparison being made to crypto currency. I honestly don’t understand why you’re doing that or the point you’re making.

For one other example so far unmentioned, mere distribution of public keys (not reliance on them) along with open source decentralized blockchain is grounds for verification far beyond what we have now, which lingers near the zero mark. Why would we complain about that?

The simple answer is to combine blockchain with legacy methods of verifying that votes aren't sold or double voting occurs. So that probably still means going to a physical polling location for the forseeable future. The use of blockchain tech wouldn't initially replace that, it would just provide a 100% accurate paper and open trail of the vote.

Reply on news.ycombinator.com