Voting Machine Makers Claim Names of Entities That Own Them Are Trade Secrets
techdirt.com
techdirt.com
It's bizarre how much of a wild west the US electoral procedures are. You'd think this would be extremely strongly regulated, but instead it's a free-for-all where the real interests are kept as opaque as possible.
That opaque bit is the crux of the matter. By design, doing that is completely unnecessary! We do not make votes personally identifiable, and that opens the door for the rest of the process to be public, trusted, etc...
After a bit of research, done back when this crap started, I arrived at these pillars for trustworthy elections:
Freedom: Basically, people can vote or not. This puts the burden of garnering votes on those wanting to win elections. It's not a given votes will be cast.
Anonymonity: This is all about reinforcing freedom. No vote record = people free to vote their minds with far fewer overall worries than there would be otherwise. If we eliminate this one, electronic votes can make sense, but there are consequences, implications. Not sure it makes sense, but I am open to this discussion.
Transparency: It must be possible for any participant in the election to follow their vote cast through the process to the final tally. Human readable vote records, the law, means, methods, all available for anyone to understand.
Oversight: Needs transparency, but also the enabling force of law and norms. Not only can people evaluate the elections, but they should do that! We all live by the outcomes. It only makes sense that we all have an interest in trusting the process, and understanding votes cast are legit.
I arrived at them being deeply bothered by the input trust problem. Prior to voting machines, this never crossed my mind. We take it for granted.
(see my other top level comment)
Trustworthy elections embody those ideas to the maximum extent possible.
Notice secrecy is not in there? The only place where it is really needed is when the voter casts their vote.
All these companies, and to a large degree elections people, and processes, have conflated secrecy with security. Doing that is either a mistake, or nefarious. Which is it?
With paper votes, anyone is able and allowed to check that every step of the voting process is done fairly. You can see the empty voting boxes before the election starts, you can see that everybody puts their one vote in it, you can be present when they're counted. Dutch law made an explicit exception to these rules to make voting machines possible, and that that was even necessary, should have been a massive red flag. And yet this obvious lack of transparency was not enough to get them banned. Only when hackers proved that anonymity could be threatened, were voting machines banned.
Personally I think transparency is more important than anonymity. Compromised transparency enables much larger scale fraud than compromised anonymity, so I'm surprised proving compromised anonymity was even necessary.
Here, we have fraud, but we have a high degree of anonymity. People don't seem to really get it. I am confused as to why.
I am reasonably secure myself and could make the same trade off you did, but an awful lot of my peers would not.
Super interesting observations. Thanks!
(Optical scan, basically)
The basic problem with voting machines casting actual votes is that they do not make a record of voter intent. They make a record of what they think the voter intent was. The actual record of voter intent walks out the building, or is a grease smudge on a screen somewhere. Lost.
They are all votes by proxy.
Because we make votes anonymous, the chain of trust between the voter, their intent, and the record of that intent, is super important. That record needs to be used directly, or we've got issues with how trustworthy the process is. That record is needed should there be issues too.
Electronic machines do not record voter intent, and that is the primary thing of value in an election.
When a voter makes some mark on a physical ballot with a pencil, pen or some other device, they can verify that record is accurate, in line with their intent, and that means that they have confidence that their vote was correctly cast. The vital chain of trust between voter intent and the record of said intent is intact.
Follow that up by using that record directly, (optical scan or simple, human count) and we've got a process, that if necessary can be hauled into a court of law, or manually evaluated by humans, to ensure that we understand the outcome of the election.
Paperless machines are so untrustworthy that nobody can even know who won an election! No potential for review, oversight exists! (Looking at you South Carolina)
When we provide input to a touchscreen, or via button press, or any other simple means, the intent is expressed at that time mechanically, but the electronic record isn't the same as a physical mark.
The truth is, the voter cannot see the vote record directly, because it is electronic. It is just a temporary state. Enabling technology is required for the voter to see the record. In this sense, the record of their vote is isolated from them, and they are forced to trust the input and output of the machine, again a vote by proxy.
This is fundamentally untrustworthy.
An example I have given to kind of simulate this involves a person in a secure room alone. You the voter walk up to a window, and you tell them what your vote is. They go make a record somewhere that you cannot see, and then they verbally confirm your vote, or maybe hold up a card or something. The key thing to understand is the voter has to trust that person, they don't see the little invisible record off to the side, nobody does.
Essentially, there's no chain of trust between the voter and the record of their vote. And no reasonable person would trust such an arrangement to be an accurate election, a trustworthy one. That is basically what voting machines do. They get between voters and the enduring record of their voter intent.
I struggle to communicate this in ways that ordinary people can understand. It's really a problem. People see computers and all the magical things that they do, and they want to trust them. And in a lot of places means and methods they can trust them. Voting is different because it's anonymous, and we don't link people to votes, which means those intent records really matter.
Given the input trust problem that I outlined above, I personally would have a huge ethical problem even making one of these machines. Our current level of technology does not offer a solution for this problem set, where we have to record intent and not make it personally identifiable. Think about it. That doesn't come up very often.
The other odd thing is that election security has been somehow linked with secrecy. It really doesn't need to be that way.
The whole process should be wide open, and it can be that way because we do not associate voter intent with voters. All we really need is a bit of privacy for the voter to vote, and from there, all else can be human readable, under the public eye.
Again, I struggle with that, Because it looks for all the world like a clever sleight-of-hand, almost deliberate intent to deceive. Conflate these things to create problems that are not, and then sell solutions to non problems!
Do these voting machine companies truly not understand what I put in my parent comment? Really?
If they do, why are they making machines? They do not work, and are untrustworthy. And go look like I did. We simply do not have trustworthy methods that do not also require votes be personally identifiable.
Simple paper ballots with marks on them work great.
If they do not understand, then there is a competence issue.
Neither scenario is good. Which is it, or is it really something else, something nefarious?
I do not like any of it. The whole mess is ultra shady.
how about taking a photo of the screen (not a screenshot, but an actual photo with a separate camera) to record what the voter sees?
another idea is to enter the vote twice. like when we create new passwords. do make sure the same result is recorded. do it on two machines side by side. if there is a difference, start over.
In the case of the photo, recording what the voter sees does not impact whether the actual record used reflects their intent. At best, we can verify they saw what they said they saw. The real problem is whether what they see actually reflects the record used for the tally. Edit: Also, if we personally identify votes, this could be used to reconstruct an election. Again, why not just use paper?
Inputting a vote twice can improve mechanical to intent translation accuracy. However, the actual record still requires enabling tech to read, is still a temporary state, and is still not part of the chain of trust between voter intent and the record of the vote used for the tally.