It is never easy to resolve conflicting requirements (e.g. 'Only eligible voters shall vote', 'voters shall not be connected in any way to their vote' - the implied requirement here being the existance of an 'air gap' between electronic roll and voting system). Voter-verified audit trail requirements contain such conflicts, especially where legislation requires that voters are prohibited from retaining prrof of their vote (vote-buying risk).
Brazil solved this by never allowing the voter to control the receipt.
Special legislation needs to be debated and produced to deal with the following nightmare scenario:
Manual recount (possibly error-prone) does not agree with electronic count(possibly subversion-prone) over some relevant dimension.
Good for you - jbn. I see you have a functioning bullshit-meter. The world needs more people like you asking the right questions.
You make some very valid points.
At the heart of your argument is that electronic voting is unnecessary, even if characterised by 'perfect' integrity. In my humble opinion you are correct. Paper and pencil systems can possess virtues such as relative simplicity, portability, low cost, ease of use and inherently implement a voter-verified audit trail.
Those who cannot use these systems can still get by if they are prepared to let someone help them complete the ballot.
I will argue that those Florida punch-card systems are an improvement over the paper ballot because they share the above characteristics (assuming the voter gets the chance to inspect the punched card before it's admission to the 'ballot box')plus they enhance the democracy by reducing the chances that an under-educated voter will unintentionally render his or her ballot invalid.
However, whichever method is used:
1) There must be a consistent density of polling places per unit of population across the voting region.
2) Polling place staff need to be comprehensively trainied, and the OIC should be a seasoned hand both with the electoral process and the voting method(s) employed at their polling place.
After the 2001 ACT election, the number of informal ballot papers was around 5%. The number of informal electronic ballots was closer to 1%. Some of this improvement can be attributed to the fact that the software prevents many of the common mistakes electors make with paper ballots, effectively returning democratic access to those who have been inadvertently disenfrachised. One may take the view that people who don't understand how to vote properly don't deserve to, but I don't - their rights are just as important as anyone elses and their voice just as valid.
Similarly, delivering voting instructions in a dozen written language has improved access for other voters.
Ergo, there are qualitative benefits which can enhance access and therefore the quality of the democracy (Which I'll briefly define here as a measure of information symmetry and actual numbers of effective voters).
Acknowledging the risks and specific problems inherent in any mechanism for collecting and counting votes for a democratic assembly is a prerequisite to ensuring that the democratic process is well-served by it.
Since electronic voting introduces new risks into the process, these need to be fully understood and provably mitigated against in order to gain the beneficial effects without undermining the process as a whole. A voter-verified audit trail can then bring personal satisfaction to each voter as to the integrity of the process. eVACS was implemented without such a facility because it was not funded, but such an extension is programmatically straightforward - most of the work would be in elucidating the detailed requirements from the relevant electoral body as to how the scheme would be managed.
And all this is going to cost, yeah?
Is it worth it?
It depends whom you ask. The blind people I've spoken to say 'hell yes!' for obvious reasons. Die-hard (and sighted) libertarians say 'hell no!' because they perceive themselves to be coerced into subsidising someone else.
In summary
Do we really, really, really need electronic voting? No, of course we don't. Democracy can get along fine without it. The Florida debacle wasn't really about the machines functionality, more about preventive maintenance and staff training.
Can electronic voting significantly enhance the 'quality' of a democracy? Yes it can, by promoting equal access to the democratic process.
Can electronic voting significantly reduce the 'quality' of a democracy? Yes again, if it is not done properly and transparently. Election subversion is an ever-present danger and needs be heavily and provably mitigated against in any system.
Does eVACS preclude the use of a voter-verified scheme?
While it is true that the eVACS system does not provide a voter-verified audit trail, it is not quite accurate to say it is not an improvement over proprietary solutions.
Non-proprietary code by itself is an improvement.
External auditing and a compilation process provably free of trojans performed on digitally signed code, producing digitally signed binaries is an improvement.
The capability of a voter-verified audit trail would be another improvement. What was not mentioned in the wired article was the fact that paper ballots were available for those who wanted them, so I didn't feel the need to chuck my job in protest.
As background, the main driver behind the eVACS system was to gain a faster, more accurate count using the complex hare-clark algorithm. The actual vote-collection software was mostly driven by a desire to allow blind and ilterate electors to vote in secret like anyone else.
Electronic vote collection was never intended to replace paper - merely offer an alternative voting method.
I would find it highly disturbing if American voters were provided only with a machine incapable of voter-verification and nothing else...
Slashdot Mirror
It is never easy to resolve conflicting requirements (e.g. 'Only eligible voters shall vote', 'voters shall not be connected in any way to their vote' - the implied requirement here being the existance of an 'air gap' between electronic roll and voting system).
Voter-verified audit trail requirements contain such conflicts, especially where legislation requires that voters are prohibited from retaining prrof of their vote (vote-buying risk).
Brazil solved this by never allowing the voter to control the receipt.
Special legislation needs to be debated and produced to deal with the following nightmare scenario:
Manual recount (possibly error-prone) does not agree with electronic count(possibly subversion-prone) over some relevant dimension.
Regards
Matt Quinn
Good for you - jbn. I see you have a functioning bullshit-meter. The world needs more people like you asking the right questions.
You make some very valid points.
At the heart of your argument is that electronic voting is unnecessary, even if characterised by 'perfect' integrity. In my humble opinion you are correct. Paper and pencil systems can possess virtues such as relative simplicity, portability, low cost, ease of use and inherently implement a voter-verified audit trail.
Those who cannot use these systems can still get by if they are prepared to let someone help them complete the ballot.
I will argue that those Florida punch-card systems are an improvement over the paper ballot because they share the above characteristics (assuming the voter gets the chance to inspect the punched card before it's admission to the 'ballot box')plus they enhance the democracy by reducing the chances that an under-educated voter will unintentionally render his or her ballot invalid.
However, whichever method is used:
1) There must be a consistent density of polling places per unit of population across the voting region.
2) Polling place staff need to be comprehensively trainied, and the OIC should be a seasoned hand both with the electoral process and the voting method(s) employed at their polling place.
After the 2001 ACT election, the number of informal ballot papers was around 5%. The number of informal electronic ballots was closer to 1%. Some of this improvement can be attributed to the fact that the software prevents many of the common mistakes electors make with paper ballots, effectively returning democratic access to those who have been inadvertently disenfrachised.
One may take the view that people who don't understand how to vote properly don't deserve to, but I don't - their rights are just as important as anyone elses and their voice just as valid.
Similarly, delivering voting instructions in a dozen written language has improved access for other voters.
Ergo, there are qualitative benefits which can enhance access and therefore the quality of the democracy (Which I'll briefly define here as a measure of information symmetry and actual numbers of effective voters).
Acknowledging the risks and specific problems inherent in any mechanism for collecting and counting votes for a democratic assembly is a prerequisite to ensuring that the democratic process is well-served by it.
Since electronic voting introduces new risks into the process, these need to be fully understood and provably mitigated against in order to gain the beneficial effects without undermining the process as a whole. A voter-verified audit trail can then bring personal satisfaction to each voter as to the integrity of the process. eVACS was implemented without such a facility because it was not funded, but such an extension is programmatically straightforward - most of the work would be in elucidating the detailed requirements from the relevant electoral body as to how the scheme would be managed.
And all this is going to cost, yeah?
Is it worth it?
It depends whom you ask. The blind people I've spoken to say 'hell yes!' for obvious reasons.
Die-hard (and sighted) libertarians say 'hell no!' because they perceive themselves to be coerced into subsidising someone else.
In summary
Do we really, really, really need electronic voting? No, of course we don't. Democracy can get along fine without it. The Florida debacle wasn't really about the machines functionality, more about preventive maintenance and staff training.
Can electronic voting significantly enhance the 'quality' of a democracy? Yes it can, by promoting equal access to the democratic process.
Can electronic voting significantly reduce the 'quality' of a democracy? Yes again, if it is not done properly and transparently. Election subversion is an ever-present danger and needs be heavily and provably mitigated against in any system.
Does eVACS preclude the use of a voter-verified scheme?
While it is true that the eVACS system does not provide a voter-verified audit trail, it is not quite accurate to say it is not an improvement over proprietary solutions. Non-proprietary code by itself is an improvement. External auditing and a compilation process provably free of trojans performed on digitally signed code, producing digitally signed binaries is an improvement. The capability of a voter-verified audit trail would be another improvement. What was not mentioned in the wired article was the fact that paper ballots were available for those who wanted them, so I didn't feel the need to chuck my job in protest. As background, the main driver behind the eVACS system was to gain a faster, more accurate count using the complex hare-clark algorithm. The actual vote-collection software was mostly driven by a desire to allow blind and ilterate electors to vote in secret like anyone else. Electronic vote collection was never intended to replace paper - merely offer an alternative voting method. I would find it highly disturbing if American voters were provided only with a machine incapable of voter-verification and nothing else...