5 ms·

Beyond the security implications (which are huge and ought to be a deal breaker on their own), this seems to have a huge problem in that it's not private. Your

Beyond the security implications (which are huge and ought to be a deal breaker on their own), this seems to have a huge problem in that it's not private. Your abusive spouse/boss/dealer could tell you that they want to see you use the app to vote their way, and there's not much to prevent it. While going to a local precinct is inconvenient, no one knows what you do once you walk in that booth. That's important to avoid coercion.

lukifer 7y ago
While I'm in total agreement, and think that this is likely a bad idea, doesn't the same problem exist with mail-in voting? An abusive party can hover over your shoulder while filling it out, and then walk you to the mailbox.
- slipheen 7y ago
  Absolutely. IMO this exacerbates the existing problem. That said, if the app supported changing your vote later on, it could be preferable to the mail-in solution.
- anon463637 7y ago
  There's no solving infrequent social problems with technology, only socio-politically. I'm sure it does happen to a few people, but it sounds extremely rare. At least a mail-in ballot usually has a "receipt," (that serial number thing that's torn off at the top/bottom) although it does little to show or prove a vote was counted, especially with the subjectiveness of interpreting handwritten marks. The bigger problem is electronic voting has no real records making it far easier to manipulate because there's nothing permanent to recount. I suggest that to solve the receipt issue, privacy, speed of counting and accuracy to the best degree available, it's best to: 0. Make voting day a national holiday. 1. Mail every voter a durable, physical RFID token (signed by a closely-held private key) that has an unique code that is not recorded who it is given to. If they haven't received one before voting in-person, they can receive a random token. 2. The voter is first checked to make sure they haven't yet voted by keeping a database of "has voted"... completely independent of actual votes. 3. The voter either drops-off the RFID in a container in a booth for their vote preference, or they mail it sealed in two envelopes (outer mailing info, inner vote preference). 4. Votes are counted both by weight and by RFID scanning (whole containers of votes are scanned in batch) for redundancy. 5. Voters can check online in real-time if their vote was counted by searching for the RFID code they used. 6. Recounts are a matter of mass-scanning large containers (with appropriate chain-of-custody) of votes. 7. Tokens are reused to reduce waste.
  - lukifer 7y ago
    +1 for election day as a national holiday; I'd even go so far as to make voting mandatory (with support for blank protest ballots, health/circumstance exemptions, etc). That said, while the RFID system you describe seems achievable (and clearly an improvement over closed-source easily-hackable voting machines), I think it misses an important component of a democratic process: voter trust. Even if the tech is fully auditable, requiring the average voter to trust an elite technological caste does reduce trust, compared to pen and paper, which are auditable by the vast majority of citizens. Also, the RFID can't be the only mechanism; having a fixed address can be used for authorizing identity (as with mail-in ballots), but is not and should not be a requirement to vote. At any rate, I agree that technology is no silver bullet; solutions should be sociopolitical first, and technology is merely a tool to that end. I do think there are worthwhile innovations to consider; if anything, I'd like to disrupt polling moreso than voting, which I think has a surprising political influence both during and between election cycles, and yet which we outsource to private media companies with their own biases and incentives. If we could crack the problem of distributed identity that is resistant to Sybil attacks, we could have ongoing/persistent voting, liquid voting, public choice economics, etc, enabling more fluid feedback loops with our representatives (and maybe someday, even "pass-through" representative direct democracy).
jmg8766 7y ago
I never thought about how you could be coerced into voting a specific way like that and it does seem to be a big problem to me. But is that already possible with mail-in voting?
sornaensis 7y ago
How is this different from voting by mail, then? e.g. basically everyone in Oregon votes by mail only.
- zacherates 7y ago
  ... more relevant than Oregon: "Washington State votes by mail." [1] [1] https://www.sos.wa.gov/elections/faq_vote_by_mail.aspx https://www.sos.wa.gov/elections/faq_vote_by_mail.aspx
- filoleg 7y ago
  >How is this different from voting by mail, then? WA also has vote by mail, but this kind of online voting is very different. To vote by mail, you have to receive the ballot at your address, physically mark your choices, and then mail it back in. The potential for fraud or abuse of this system is fairly limited. To vote with this WA online voting system, you need to know the name and the date of birth, that's it. Aside from illegality of voting fraud, the potential for abuse is just humongous, all you need is just a dataset of names and DOB.
devonkim 7y ago
Voter intimidation at a local or institutional level is absolutely massive but giving people the _choice_ of voting in various ways hopefully negates the bad. I'd love to understand the data and anecdotal stories on this as well if any is available. I suspect that those in such abusive situations don't even get to vote whatsoever because they're physically captive / bound, but that is 100% conjecture on my part. Nobody advocating for this voting technology wants to encourage any form of intimidation and distortion of individuals' wills like you've described.
Analemma_ 7y ago
Washington already has voting by mail, and the solution to this problem there is that you can vote by mail but also travel to a poll station on election day and that vote overrides your mail ballot. I assume the same would be true here.
all_blue_chucks 7y ago
Seattle has long ago decided that advantages of remote voting outweigh the disadvantages. If your boss doesn't give you time off to go vote you are essentially being coerced not to vote anyway. The only difference here is whether the post office or a TLS connection are more trustworthy to deliver your vote.
- deogeo 7y ago
  Not just TLS connection, but the whole stack of voting software vendor, OS vendor, and phone hardware. Edit ("posting too fast"): For paper votes, no computers need be involved at all - they can be counted and summed by hand.
  - all_blue_chucks 7y ago
    That is already a factor. Even with paper ballots they eventually get entered into a computer for tallying. Which has a whole stack that must be trusted.
  - cwkoss 7y ago
    Also your phone's keyboard transmits predictive text data re: write in candidates.
conanbatt 7y ago
I think low voting participation is a much more serious issue than coerced votes. If 10% of people more voted, there is no chance in hell you get a 0.1% increase in coerced voting.

lukifer 7y ago

While I'm in total agreement, and think that this is likely a bad idea, doesn't the same problem exist with mail-in voting? An abusive party can hover over your shoulder while filling it out, and then walk you to the mailbox.

slipheen 7y ago
Absolutely. IMO this exacerbates the existing problem. That said, if the app supported changing your vote later on, it could be preferable to the mail-in solution.
anon463637 7y ago
There's no solving infrequent social problems with technology, only socio-politically. I'm sure it does happen to a few people, but it sounds extremely rare. At least a mail-in ballot usually has a "receipt," (that serial number thing that's torn off at the top/bottom) although it does little to show or prove a vote was counted, especially with the subjectiveness of interpreting handwritten marks. The bigger problem is electronic voting has no real records making it far easier to manipulate because there's nothing permanent to recount. I suggest that to solve the receipt issue, privacy, speed of counting and accuracy to the best degree available, it's best to: 0. Make voting day a national holiday. 1. Mail every voter a durable, physical RFID token (signed by a closely-held private key) that has an unique code that is not recorded who it is given to. If they haven't received one before voting in-person, they can receive a random token. 2. The voter is first checked to make sure they haven't yet voted by keeping a database of "has voted"... completely independent of actual votes. 3. The voter either drops-off the RFID in a container in a booth for their vote preference, or they mail it sealed in two envelopes (outer mailing info, inner vote preference). 4. Votes are counted both by weight and by RFID scanning (whole containers of votes are scanned in batch) for redundancy. 5. Voters can check online in real-time if their vote was counted by searching for the RFID code they used. 6. Recounts are a matter of mass-scanning large containers (with appropriate chain-of-custody) of votes. 7. Tokens are reused to reduce waste.
- lukifer 7y ago
  +1 for election day as a national holiday; I'd even go so far as to make voting mandatory (with support for blank protest ballots, health/circumstance exemptions, etc). That said, while the RFID system you describe seems achievable (and clearly an improvement over closed-source easily-hackable voting machines), I think it misses an important component of a democratic process: voter trust. Even if the tech is fully auditable, requiring the average voter to trust an elite technological caste does reduce trust, compared to pen and paper, which are auditable by the vast majority of citizens. Also, the RFID can't be the only mechanism; having a fixed address can be used for authorizing identity (as with mail-in ballots), but is not and should not be a requirement to vote. At any rate, I agree that technology is no silver bullet; solutions should be sociopolitical first, and technology is merely a tool to that end. I do think there are worthwhile innovations to consider; if anything, I'd like to disrupt polling moreso than voting, which I think has a surprising political influence both during and between election cycles, and yet which we outsource to private media companies with their own biases and incentives. If we could crack the problem of distributed identity that is resistant to Sybil attacks, we could have ongoing/persistent voting, liquid voting, public choice economics, etc, enabling more fluid feedback loops with our representatives (and maybe someday, even "pass-through" representative direct democracy).

jmg8766 7y ago

I never thought about how you could be coerced into voting a specific way like that and it does seem to be a big problem to me. But is that already possible with mail-in voting?

sornaensis 7y ago

How is this different from voting by mail, then? e.g. basically everyone in Oregon votes by mail only.

zacherates 7y ago
... more relevant than Oregon: "Washington State votes by mail." [1] [1] https://www.sos.wa.gov/elections/faq_vote_by_mail.aspx https://www.sos.wa.gov/elections/faq_vote_by_mail.aspx
filoleg 7y ago
>How is this different from voting by mail, then? WA also has vote by mail, but this kind of online voting is very different. To vote by mail, you have to receive the ballot at your address, physically mark your choices, and then mail it back in. The potential for fraud or abuse of this system is fairly limited. To vote with this WA online voting system, you need to know the name and the date of birth, that's it. Aside from illegality of voting fraud, the potential for abuse is just humongous, all you need is just a dataset of names and DOB.

devonkim 7y ago

Voter intimidation at a local or institutional level is absolutely massive but giving people the _choice_ of voting in various ways hopefully negates the bad. I'd love to understand the data and anecdotal stories on this as well if any is available. I suspect that those in such abusive situations don't even get to vote whatsoever because they're physically captive / bound, but that is 100% conjecture on my part. Nobody advocating for this voting technology wants to encourage any form of intimidation and distortion of individuals' wills like you've described.

Analemma_ 7y ago

Washington already has voting by mail, and the solution to this problem there is that you can vote by mail but also travel to a poll station on election day and that vote overrides your mail ballot. I assume the same would be true here.

all_blue_chucks 7y ago

Seattle has long ago decided that advantages of remote voting outweigh the disadvantages. If your boss doesn't give you time off to go vote you are essentially being coerced not to vote anyway. The only difference here is whether the post office or a TLS connection are more trustworthy to deliver your vote.

deogeo 7y ago
Not just TLS connection, but the whole stack of voting software vendor, OS vendor, and phone hardware. Edit ("posting too fast"): For paper votes, no computers need be involved at all - they can be counted and summed by hand.
- all_blue_chucks 7y ago
  That is already a factor. Even with paper ballots they eventually get entered into a computer for tallying. Which has a whole stack that must be trusted.
- cwkoss 7y ago
  Also your phone's keyboard transmits predictive text data re: write in candidates.

conanbatt 7y ago

I think low voting participation is a much more serious issue than coerced votes. If 10% of people more voted, there is no chance in hell you get a 0.1% increase in coerced voting.