4 ms·

Why not paper ballots? The trust is greater (you can't hack a piece of paper), it's idiot proof, can be counted electronically and hand counted in the case of d

by Thlom 10y ago

Why not paper ballots? The trust is greater (you can't hack a piece of paper), it's idiot proof, can be counted electronically and hand counted in the case of dispute or even as part of procedure. And much easier to expand to more polling stations as you don't need any voting machines on site. Just paper and a secure box to put it in.

cbanek 10y ago
Of course you can hack a piece of paper. You can add more pieces, add misleading pieces, fake pieces, etc. And even with a secure box, there's always the risk of conspiracy to remove votes. Do I think this kind of thing happens? No. But encryption could be used to make the trust of voting even better, if done correctly. This would allow secrecy of votes, the ability to verify the validity of a vote without seeing who they voted for, and the ability to bundle the votes together and hash them, for example. Paper scales great, until you need to collect and retain it.
- Klathmon 10y ago
  But those "paper hacks" take man-power, physical resources, time, and a lot of people. To do any of those across a state you are looking at at LEAST a few hundred people. Nation wide? Thousands. In a perfect world. With electronic voting, all it takes is one person at some point in the chain, or at best a handful of people.
  - cbanek 10y ago
    Why not have everyone in the chain sign? The point is you'd have to have a secure chain from the voter all the way to registration. I don't see how the number of people in the validation chain would change. The only real problem is a technological problem, like a bug or improper use of encryption. Especially if you're only trying to validate the results, and not trying to have secrecy. Each person signs / records the hash of the votes they're validating is. Combine those hashes together and sign that to validate at higher levels, or use something more fancy. It would be obvious if something got changed due to malice or fraud. What would this one person do exactly to sabotage the system?
    - Klathmon 10y ago
      One person can design a button which votes wrong 1/30 times. One person can exploit a bug in the encryption implementation that allows you to access metadata about who voted for what. One person can design the UI of the voting booth to vote incorrectly in some cases. (and if you don't believe this is possible, take a look at some of the "underhanded c contest" [0] winners and how insanely simple some of those exploits are, and how powerful they can be). One person could write malware to search and phone home with any private voting keys that it can find and on election day that person can vote many times. We as a species can't even write a secure SSL implementation, and you think that we would be able to write a bulletproof voting cryptosystem that can be used by every single person? But even aside from that, what happens when a person loses their key? Are they disenfranchised for life? If there's a way to regenerate keys, what is stopping someone from regenerating a few extra every now and then and voting with them? How do people without a PC vote? Remember, voting isn't like a website or application. 100% of people need to have the ability to vote. If your system disenfranchises even one person who should legally be able to vote, it can't work. Paper is near perfect already. It takes a monumental amount of resources (and one hell of a literal paper trail) to sway a paper vote in any meaningful way. It's easy to count (you already vote in small counties and districts, just count them there and pass the numbers up the chain), and it's fairly quick (the US counts the votes for the presidential election by the end of the night! That's crazy!) With an electronic system, you need people to audit it. That means that only developers would be able to vet the voting system. Your "average joe" wouldn't be able to verify if his vote was cast, he would need to rely on others. And what percentage of developers know encryption well enough to be able to safely audit a cryptosystem? 10%? 1%? That's a hell of a lot of power to be putting in a small group of people's hands. With paper, anyone can verify. Any language, any "economic class", the illiterate, anyone. You can walk down and sit next to the ballot box in your district, and watch that thing all damn day. You can count along with everyone else at the end of the day, and you can sleep well knowing your vote was counted, as you watched it, and you can do the math yourself and make sure of it. Paper works. [0] http://www.underhanded-c.org/ http://www.underhanded-c.org/
      - cbanek 10y ago
        Any system has flaws, that's for sure. I'm not saying either that everyone should have a key, or how that key would be distributed. I haven't even thought of what the system would look like. I don't think this system would need to be online either. But yes, I think such a system could be developed. And yes, I think the results could be audited. We have code running nuclear facilities, fighter jets, and other sensitive parts of our government. In terms of how easy it is, just look at how hard it was to get recounts of votes, not to mention the 'hanging chad' situation, where people were arguing over what the paper votes were. Even the UI on the paper is very important to the whole experience to not confuse the person. Right now your average joe isn't able to verify if their vote is cast on paper I believe. You have the number on your ballot, as least in nevada on a receipt, but validating that (or at least in a meaningful way, seeing the paper) is almost impossible. For local elections and non federal elections, I think the system would be easier to mess with, due to the fewer number of people involved in the system. Overall I'm not saying paper is bad, or that tech would be easier - voting is incredibly complicated. But I think it's entirely possible to design a good system. Personally I think the mail-in ballot systems are the best, but of course I'd like to have some way of checking on that, which isn't easy.
- angry_octet 10y ago
  Paper systems are pretty hack resistant. Duplicating ballot papers requires large printing presses and a system for distributing the ballots. If it was a problem then standard secure printing techniques could be used. Vote tallies from the polling booths can't be faked, so ballot boxes would have to be stuffed in real time. Stealing votes is also difficult and has low marginal gain in close elections. Ballot collection and retention scales great. We have trucks and warehouses as part of the regular economy. The cryptographic properties you mention seem nice, but no-one has implemented them. Paper is also great for anonymity, cost and consistency.

by Thlom 10y ago

cbanek 10y ago

Of course you can hack a piece of paper. You can add more pieces, add misleading pieces, fake pieces, etc. And even with a secure box, there's always the risk of conspiracy to remove votes. Do I think this kind of thing happens? No. But encryption could be used to make the trust of voting even better, if done correctly. This would allow secrecy of votes, the ability to verify the validity of a vote without seeing who they voted for, and the ability to bundle the votes together and hash them, for example. Paper scales great, until you need to collect and retain it.

Klathmon 10y ago
But those "paper hacks" take man-power, physical resources, time, and a lot of people. To do any of those across a state you are looking at at LEAST a few hundred people. Nation wide? Thousands. In a perfect world. With electronic voting, all it takes is one person at some point in the chain, or at best a handful of people.
- cbanek 10y ago
  Why not have everyone in the chain sign? The point is you'd have to have a secure chain from the voter all the way to registration. I don't see how the number of people in the validation chain would change. The only real problem is a technological problem, like a bug or improper use of encryption. Especially if you're only trying to validate the results, and not trying to have secrecy. Each person signs / records the hash of the votes they're validating is. Combine those hashes together and sign that to validate at higher levels, or use something more fancy. It would be obvious if something got changed due to malice or fraud. What would this one person do exactly to sabotage the system?
  - Klathmon 10y ago
    One person can design a button which votes wrong 1/30 times. One person can exploit a bug in the encryption implementation that allows you to access metadata about who voted for what. One person can design the UI of the voting booth to vote incorrectly in some cases. (and if you don't believe this is possible, take a look at some of the "underhanded c contest" [0] winners and how insanely simple some of those exploits are, and how powerful they can be). One person could write malware to search and phone home with any private voting keys that it can find and on election day that person can vote many times. We as a species can't even write a secure SSL implementation, and you think that we would be able to write a bulletproof voting cryptosystem that can be used by every single person? But even aside from that, what happens when a person loses their key? Are they disenfranchised for life? If there's a way to regenerate keys, what is stopping someone from regenerating a few extra every now and then and voting with them? How do people without a PC vote? Remember, voting isn't like a website or application. 100% of people need to have the ability to vote. If your system disenfranchises even one person who should legally be able to vote, it can't work. Paper is near perfect already. It takes a monumental amount of resources (and one hell of a literal paper trail) to sway a paper vote in any meaningful way. It's easy to count (you already vote in small counties and districts, just count them there and pass the numbers up the chain), and it's fairly quick (the US counts the votes for the presidential election by the end of the night! That's crazy!) With an electronic system, you need people to audit it. That means that only developers would be able to vet the voting system. Your "average joe" wouldn't be able to verify if his vote was cast, he would need to rely on others. And what percentage of developers know encryption well enough to be able to safely audit a cryptosystem? 10%? 1%? That's a hell of a lot of power to be putting in a small group of people's hands. With paper, anyone can verify. Any language, any "economic class", the illiterate, anyone. You can walk down and sit next to the ballot box in your district, and watch that thing all damn day. You can count along with everyone else at the end of the day, and you can sleep well knowing your vote was counted, as you watched it, and you can do the math yourself and make sure of it. Paper works. [0] http://www.underhanded-c.org/ http://www.underhanded-c.org/
    - cbanek 10y ago
      Any system has flaws, that's for sure. I'm not saying either that everyone should have a key, or how that key would be distributed. I haven't even thought of what the system would look like. I don't think this system would need to be online either. But yes, I think such a system could be developed. And yes, I think the results could be audited. We have code running nuclear facilities, fighter jets, and other sensitive parts of our government. In terms of how easy it is, just look at how hard it was to get recounts of votes, not to mention the 'hanging chad' situation, where people were arguing over what the paper votes were. Even the UI on the paper is very important to the whole experience to not confuse the person. Right now your average joe isn't able to verify if their vote is cast on paper I believe. You have the number on your ballot, as least in nevada on a receipt, but validating that (or at least in a meaningful way, seeing the paper) is almost impossible. For local elections and non federal elections, I think the system would be easier to mess with, due to the fewer number of people involved in the system. Overall I'm not saying paper is bad, or that tech would be easier - voting is incredibly complicated. But I think it's entirely possible to design a good system. Personally I think the mail-in ballot systems are the best, but of course I'd like to have some way of checking on that, which isn't easy.
angry_octet 10y ago
Paper systems are pretty hack resistant. Duplicating ballot papers requires large printing presses and a system for distributing the ballots. If it was a problem then standard secure printing techniques could be used. Vote tallies from the polling booths can't be faked, so ballot boxes would have to be stuffed in real time. Stealing votes is also difficult and has low marginal gain in close elections. Ballot collection and retention scales great. We have trucks and warehouses as part of the regular economy. The cryptographic properties you mention seem nice, but no-one has implemented them. Paper is also great for anonymity, cost and consistency.