Wow, great to hear and a great opportunity for improvement. If a sensible solution comes to light then hopefully it will have a domino effect on other states. Bonus points for wanting transparency and a system that's open source. Let's just see what actually happens though.
"Bonus points for wanting transparency and a system that's open source."
Transparency is not a bonus, publicly verifiable elections are a requirement for a healthy democracy to ensure power is legit. Open source does not really matter. Even if it's open source, it's hard to assure the public that "that open source software you downloaded from github" was actually running in a counting machine and nothing else (i.e. a custom made virus) at the moment of counting.
Interesting. Good to see they seek information about systems that "are based on voters marking paper ballots".
Contrary to popular belief, in state or nation-wide elections paper is a safer way to vote than electronics because the former is simpler and thus easier to verify by the general public.
Most people don't realize how hard it is to beat the transparency and simplicity of a physical ballot box. It can be inspected by the public at the start of the day so anyone interested can verify it's empty and people can keep an eye on it during the day and see if anyone tampers with it. Then at the end of the day it's important the votes are manually tallied right after the ballot box is opened. This makes the process transparent to most people in society and assures the highest level of scrutiny and confidence that the vote and vote count are legit.
It's hard to replace this with anything more complex i.e. electronics.
Topic submitter here (and current President of the SF Elections Commission).
As a software developer and open source contributor myself, I'd really like to see open source happen. But if no vendors respond to the RFI / RFP with a credible open source proposal, it will make things a lot harder.
Although I'm not a fan of it [0], there has been a commission to research the feasibility of electronic voting in the Netherlands. Outcome was it would not be cheaper than the current paper voting system and some fundamental parts still have to be on paper [1].
Currently they are researching the costs of implementing the proposed voting and counting machines (with human readable paper separating both machines). Also there was a preference for the software to be open source.
You could get in contact with some people of the commission:
https://twitter.com/cryptoron (part of the commission, now leads the costs investigation of the new proposed machines and software)
https://twitter.com/ArjenKamphuis (part of the commission and part of the hacker movement that got eVoting machines banned in the Netherlands in 2007 [2]).
[0] I don't think electronics outweigh the security of a paper based voting system anytime soon. Because of it's simplicity and transparency it's publicly verifiable in practice. Professor Halderman, who did a lot of research in electronic voting, won't expect secure electronic voting systems that can be used at large scale within the next 10 years, if ever: https://www.youtube.com/watch?v=JY_pHvhE4os#t=58m55s
Is this a chance for Score Voting or Approval Voting [1]?
Edit: It looks like they are going with instant-runoff, which is more complicated and has higher levels of regret. Does anyone know why they have selected this versus the easier Score Voting option?
> Does anyone know why they have selected this versus the easier Score Voting option?
Instant-runoff voting (aka ranked-choice voting) is the law in San Francisco. It was passed by the voters back in 2002.
By the way, systems like approval and score voting suffer from the problem of voters having to vote against their favorite candidate to support a "lesser of two evils."
Think of an approval election with Donald Trump, Hillary Clinton, and Elizabeth Warren (where Warren is your favorite and Clinton is the lesser of two evils). Do you "approve" of Clinton, which would help defeat Warren? Or do you "approve" only Warren, at the risk of getting Trump?
11 comments
[ 3.8 ms ] story [ 44.1 ms ] threadTransparency is not a bonus, publicly verifiable elections are a requirement for a healthy democracy to ensure power is legit. Open source does not really matter. Even if it's open source, it's hard to assure the public that "that open source software you downloaded from github" was actually running in a counting machine and nothing else (i.e. a custom made virus) at the moment of counting.
Contrary to popular belief, in state or nation-wide elections paper is a safer way to vote than electronics because the former is simpler and thus easier to verify by the general public.
Most people don't realize how hard it is to beat the transparency and simplicity of a physical ballot box. It can be inspected by the public at the start of the day so anyone interested can verify it's empty and people can keep an eye on it during the day and see if anyone tampers with it. Then at the end of the day it's important the votes are manually tallied right after the ballot box is opened. This makes the process transparent to most people in society and assures the highest level of scrutiny and confidence that the vote and vote count are legit.
It's hard to replace this with anything more complex i.e. electronics.
For a recent discussion see:
https://news.ycombinator.com/item?id=10076919
As a software developer and open source contributor myself, I'd really like to see open source happen. But if no vendors respond to the RFI / RFP with a credible open source proposal, it will make things a lot harder.
Basically, SF would have to contract to build its own system. I made this point at last week's Commission meeting here: https://www.youtube.com/watch?v=SOOoXSOtR-s&t=1h04m55s
Go back to paper ballots counted by hand.
Currently they are researching the costs of implementing the proposed voting and counting machines (with human readable paper separating both machines). Also there was a preference for the software to be open source.
You could get in contact with some people of the commission:
https://twitter.com/cryptoron (part of the commission, now leads the costs investigation of the new proposed machines and software)
https://twitter.com/ArjenKamphuis (part of the commission and part of the hacker movement that got eVoting machines banned in the Netherlands in 2007 [2]).
[0] I don't think electronics outweigh the security of a paper based voting system anytime soon. Because of it's simplicity and transparency it's publicly verifiable in practice. Professor Halderman, who did a lot of research in electronic voting, won't expect secure electronic voting systems that can be used at large scale within the next 10 years, if ever: https://www.youtube.com/watch?v=JY_pHvhE4os#t=58m55s
[1] Unofficial abstract of the commission: https://www.kiesraad.nl/sites/default/files/every-vote-count...
[2] http://wijvertrouwenstemcomputersniet.nl/English
Edit: It looks like they are going with instant-runoff, which is more complicated and has higher levels of regret. Does anyone know why they have selected this versus the easier Score Voting option?
[1] http://www.electology.org/#!approval-voting/cc04
Instant-runoff voting (aka ranked-choice voting) is the law in San Francisco. It was passed by the voters back in 2002.
By the way, systems like approval and score voting suffer from the problem of voters having to vote against their favorite candidate to support a "lesser of two evils."
Think of an approval election with Donald Trump, Hillary Clinton, and Elizabeth Warren (where Warren is your favorite and Clinton is the lesser of two evils). Do you "approve" of Clinton, which would help defeat Warren? Or do you "approve" only Warren, at the risk of getting Trump?