Conclusion:
1. Internet voting is not compatible with democracy
2. No amount of technology can change this
3. Whom you voted for ought to be secret
4. Who voted should not be secret -- it should be known as widely
as possible
5. And who counts the votes, and how, certainly ought not be secretBut that is not synonym with electronic voting. You can go into a polling station and vote on machines; a well designed system shares all vital properties of paper systems while adding many useful features, including ways to reliably count votes algorithmically in an independently verifiable way (any citizen can do the math and check the results, and check that each person only voted once, even though they can't tell who voted for whom).
Don't say it's impossible, with a bit of ingenuity and a bit of mathematics it can be done. :P
EDIT: For more info read https://en.wikipedia.org/wiki/End-to-end_auditable_voting_sy.... It's fascinating.
Chaum, Rivest, Pret a Whatever... Fine. Prove it. With a real world election (simulated). Burden of proof lies with the technophiles, not with the proud defenders of Democracy.
To say Internet voting is impossible is to admit a failure of imagination and ingenuity - I'm confident that some novel application of cryptographic and distributed-computing techniques could come up with a secure system.
But I'm wary of how complicated it would necessarily be to the average voter, or even someone with an advanced knowledge of computing - concepts like hashing and the blockchain are almost impossible to correctly explain to a layperson, let alone do they have the way to be confident of the results themselves - it would immediately cast doubt on the system because you cannot trust something that you cannot understand.
Moratorium on blockchain-based voting system proposals.
If you were to actually build something like this, you would use blind signature. But having this toy example shows that the two concepts aren't exclusif.
Alas. Our actual elections are administered per precinct (say 0 - 1000 voters) and have complicated ballots (dozens of issues, races). Meaning your utterly unique ballot is unlikely to collide with any other cast ballot.
The blockchain-based schemes I briefly studied rely on fuzzing timestamps. A different kind of collision.
Here's the takeaway: Our elections administration is finely tuned to the Australian Ballot. Any new voting system has to specify the style of ballot, which circumstances are applicable, and how the laws, rules, and procedures will have to be modified to adopt it.
You're right. No hash collision (shuffle) in that arm waving proposal. So there's no attempt to anonymize the voter's identity.
Votes cast and recorded in order deanonymize the ballots. If the order is not recorded then the election cannot be audited.
As I hinted (elsewhere), the blockchain proposals fudge the timestamp (when the ballot is cast) in an attempt to anonymize the voter.
This is equivalent to preventing double-spending in the blockchain analogy - a solved problem. There are already companies trying to exploit this to develop electronic-voting systems: https://followmyvote.com/.
I imagine it working like this:
1. Government announces an election. Registered voters are asked to submit the public addresses of their 'vote wallets'. The public addresses of the wallets are derived from the registered voter's public information (e.g. their name). This information need not be secret. The public blockchain would ensure that everyone who is registered received their votes to cast: the equivalent of receiving an empty ballot paper.
2. A Monero-like system (confession: I don't fully understand it yet) acting as a blackbox or Bitcoin-tumbler system then distributes empty-ballots from the public wallets into new private wallets that each registered-voter will have created in an anonymous fashion.
3. Registered voters then cast their votes by 'spending' their ballots (coins) by sending them to the public wallets of electoral candidates. These transactions may be cryptographically signed by a private key derived from the registered-voter's public wallet. This allows a voter to verify that their vote was correctly cast, without allowing others to see whom they actually voted for.
4. There is a problem in that the results of the election cannot be kept secret until the election is over, as the blockchain must be distributed publicly, so observers will see the votes going towards a particular candidate as-they-happen, which will necessarily change voters' votes in cases of tactical-voting; though possible solutions exist, such as having blockchain network nodes agree to only allow final vote transactions during a tiny time window of only a few seconds - or a Proof-of-Work system that will necessarily take more computation time to verify a vote was for a particular candidate the more in-advance of the polls-closing deadline the vote was made.
Because votes cannot be 'mined', and all originate from a single 'government wallet' it means that votes cannot be manufactured - as far as I can tell the only weakness is that false voter-registrations could exist and be used to cast votes, but then we have that situation with postal-voting now any way. If a tumbler approach is used to distribute ballots from public wallets to private wallets it's possible for participants to create more than one private wallet, and claim more ballots for themselves - but this can be detected by:
1. As part of step 1 above, voters create a private secret K.
2. Each voter then creates a hash of K, known as K'.
3. The blockchain would also contain a Bloom-filter, or equivalent implementation of a "Set contains a member" test that does not reveal the actual set contents. Then a strictly sequential process begins, where the Bloom-filter is publicly modified by each registered voter to add their own K' values - this necessarily means that the K' values of the first voters will be publicly known entirely, but as the table becomes more populated it will be impossible to fully know what the entire values of each K' value is. This can be mitigated by having the system create known 'sacrificial' / dummy voters to provide some initial protection for the first actual voters.
4. When the private wallets are created by each voter (in step 2 above) they must be tagged with the K values each voter created as Kp. A private wallet will only be able to participate in the ballot distribution process if the hash of its Kp value (Kp') appears in the Bloom filter populated in the previous step) - thus restricting each voter to only 1 private ballot wallet.
Right to voter privacy (secret ballot) is in my state's constitution.
I'm open to relaxing absolute privacy with something like temporary time boxed privacy. Like until an election is certified.
I wrote about it in another comment, but I'll quickly go over the technology here as well.
When a citizen votes, the vote is encrypted with the government public key. The vote is then put into a container and is signed with the citizens private key.
The vote gets sent to the voting-system servers and is then stored until counting day.
Once the voting period has ended the encrypted votes are pulled out of the signed containers and the signed containers themselves are destroyed, this anonymizes the votes. Only then are the encrypted votes decrypted with the government private key.
We keep the signed containers around the encrypted vote to make sure no-one can be forced to e-vote. Every citizen can vote as many times as they want during the e-voting period and only the last vote counts.
So if someone is standing behind you and telling you to vote somehow, you can do as they say and just vote again afterwards.
As an extra backup, if you do cast your vote on paper as well, then the paper vote is counted and the e-vote is discarded.
The system has a prerequisite though, each citizen must have a private key. We do, as it's stored on our id-cards(equivalent to passports within the EU).
Procedures are in place to protect the Government private key during the voting period and just like with paper-voting observers are involved in every step of the process.
The English documentation is scarce, but the software itself is open source: https://github.com/vvk-ehk/evalimine
https://en.wikipedia.org/wiki/Electronic_voting_in_Estonia#C...
Their whole analysis is based on this claim: "claiming they could be able to breach the system, change votes and vote totals, and erase any evidence of their actions if they could install malware on the election servers."
That is true for all server based software.
Procedures have been set up to mitigate this risk. All software, servers and anything else running in the network is audited before, during and after the election. The whole process is public and anyone can become an observer, just like with paper ballots. There really are many eyeballs making sure the election servers are running the software they are supposed to.
AFAIK everything else they pointed out was purely procedural and the guidelines there have been updated. There really weren't any reproducible technical issues and some of the procedural ones are outright misrepresentations.
The full sources for these two are sadly in Estonian, but the main response to that research can be seen here in English: http://www.vvk.ee/uudised/vabariigi-valimiskomisjoni-vastula...
""" Posted Wi-Fi credentials — The official video of the pre-election process reveals credentials for the election officials’ Wi-Fi network, which are posted on the wall. """
The credentials were actually just for a separate public guest network, that has nothing to do with the "election officials network" or the network where the voting servers are.
or: """ Keystrokes reveal root passwords — Videos posted by officials during the election show operators typing, inadvertently revealing root passwords for election servers. """
Although this was recognized as a possible attack vector and the procedures for filming have been updated, the "operator" typing a password was actually one of the observers typing a password in their own computer.
Obviously any hints to compromised processes are taken with 100% seriousness and if needed processes and guidelines are changed to protect the integrity of the system.
Any technical reports are also (publicly) analyzed and if an issue is indeed found, they're fixed. The thing is, "https://estoniaevoting.org" said everything is bad, and the system should not be used, but never published/provided any re-producible test-cases that could be verified.
OSCE has audited all of our elections, they've given some procedural pointers that have been implemented, but OSCE observers have not deemed our elections compromised.
This is their 2011 audit: http://www.osce.org/odihr/77557?download=true, the pointers given to improve our system have been implemented by now.
And here is their 2015 audit: http://www.osce.org/odihr/elections/estonia/160131?download=..., which has some new pointers, but again has not deemed the system to be insecure.
We're having another parlamentary election in 2019 and again, the system will be improved for that(more verification methods are being implemented), but so far, all of our elections have been deemed acceptable and work is being done to keep the track record.
Our internet voting system has solved this problem. The votes secrecy is guaranteed by allowing citizens to vote multiple times and only the last vote is counted. No-one is obligated to vote on the internet and a paper system works in parallel. If you cast your vote on the internet AND with a paper ballot as well, the paper ballot is counted.
This guarantees no-one can be forced to vote, as they can simply vote again in the 7 day e-voting period.
Each vote is encrypted and signed, but in two different containers. The goverment public key is used to encrypt the vote. That container is then wrapped into another that is signed with the citizens private key. Before counting, the double-casts are filtered out and the votes are anonymized by removing the signed wrapper.
And, it's open source: https://github.com/vvk-ehk/evalimine
As an example faking even 10% of the abstained votes would have swayed the brexit results.
I don't see how this makes online voting insecure, fake paper ballot risk has always been there and has been deemed to be a mitigated risk.
- Bribing voters because you've implemented voting over the network
- Once you figure out you need to switch to voting booths I'll do timing attacks to deanonymize the vote
- As soon as you have a count I'll seed distrust in the count as only a very small percentage of the population actually understands the blockchain and cause a riot or two
Electronic voting is the reverse FizzBuzz. FizzBuzz tests if you know the absolute basics about technology to implement something extremely simple. Electronic voting tests if you are able to reject the use of technology when it doesn't add any advantage and instead creates very hard to solve failure modes. Voting benefits from the use of the extremely simple ballot and box. It benefits from it not being programmable and from the counting method being something that almost all the population is able to carry out let alone understand.
Red Herring. Not counting voter registration fraud, https://www.washingtonpost.com/news/the-fix/wp/2016/12/01/0-... Washington Post only found 4 cases in the 2016 election.
That said, bit-coins and their ilk are obvious tulip bubbles now, their hope is gone under the greed. Yes, the ideas are interesting and may have applications that are unseen. But as communism was great in theory and genocides in practice (mostly), I think the relentless blockchain spam is missing the point as well: Theory ain't shit because humans are assholes. And no, the blockchain is not going to solve that too.
Because of the rhetoric surrounding blockchain spam, it's easy to make the comparison to marxist theory. But I think it's really just a technology, not some grand scheme to alter our economic system. It's more realistic to simply say that it's a tool that could provide some very useful real world applications.
I'm not confounding bitcoins with marxism and that mess of history. I'm just saying that there are things out there that were good in theory, but the applications of the theory are just plain terrible, marxism is a famous example of that.
That's ridiculous. So we can't do anything, whatsoever, to improve our society and our systems, because eh people are assholes or whatever? That's nonsense. Of course we can improve our systems, our laws, our government, and of course mathematics and technology can help. For example, our voting systems are woeful because they are holdovers from a time when nobody thought about these things from a mathematical point of view (and nobody still does: "let's just vote for people and let the person with the most votes win" sounds perfectly simple and correct for 99% of the population). But we can do much better, and it can objectively improve our democracy a whole lot.
Will applying mathematics and technology to democracy fundamentally change and transcend human nature? Duh, of course not, but nobody claimed it will. There's plenty between implementing an utopia and going nothing at all.
Yes, that is correct. Also, me and Mr. Strawman could use some more hay, do you have any?
> ...and nobody still does...
This is what I'm talking about. Trying to force math/tech onto a democracy/republic where the voting population doesn't understand it cannot end well. The voter will be suspicious, and rightfully so, of the process if they do not understand it. This is especially true if there is any corruption in the system at all. Since we know men are not angels, then the vast majority of voters must have belief in the processes as we all know men do become devils. Democracy demands this faith to work. Yes, we can make the system mathematically better, but if that comes at the cost of the faith in democracy, then it's mute. Even with the system we have now, simple as it is, there is already a large lack of faith in it by a lot of the US, and this is already a threat to idea of self-determination.
Paper works fine.
I don't know if it would be inherit to making a secure voting system on a block-chain that you had such (unnecessary for voting) fine-grained timestamps, but just a thought.
This one was alway wierd to me. Other countries always manage to count the votes in one night, and while the USA is very big, you'd think there would be that many more people counting the ballots available to make up for it. I guess it doesn't scale that easily.
Around here, it's recommended to have a polling location for every 800~1000 voter and a booth per ~300 voters.
When the polling location closes, staff (volunteers) start counting very publicly: anyone can watch and the procedure is legally specified such that overseers can trivially keep track and count of everything, the results are certified, displayed at the polling location and (if necessary e.g. for non-local elections) sent up to the next administrative division.
Unless there are improprieties, it doesn't take very long to count a thousand votes.
- Not enough polling stations
- Long lines as the norm in many places
- Polling stations closing early -- I've seen 8pm
- Exit polls released before voting closes
I guess these are generally due to the US being a large country spanning several time zones, with low population density. But they seem fixable, especially if you care enough about the integrity of the vote to commit enough staff and money to the process.
Given the ridiculous amount of money that gets ploughed into campaigning in the US, it seems bizarre that you can't spare a fraction of that to improve the process. But it seems like what little money is available is captured by hucksters selling electronic voting machines.
(Not that the UK system is perfect! Our voting system is pretty bad. Both countries could really use PR.)
Lack of resources can also be an issue, voting is generally administered at the county level and state/federal funding is pretty limited.
An employer cannot legally fire a person for taking time off to vote, or to attend jury duty. That doesn't mean they don't, it just means they can't do so legally.
But if you are someone who depends on that job (such that you can't just go out and find another), losing that job because you voted is something you're going to weigh, and regardless of how you vote (though typically - at least historically, notwithstanding our last election - those on the lower-end of the economic scale vote for Democrats), you are more likely to want to keep your job to pay for rent and food, than to vote.
Furthermore, even if your principles matter more to you than your job, if you do get fired, you aren't likely to then go thru the process of taking your former employer to court for firing you (because that would mean more time and expense away from finding another job).
Heck - the employer may not even need to break any laws, just make implicit threats as to what will happen should an employee take time off without consent of the employer, while neglecting to tell the employee about their voting rights (and nobody reads those employee rights posters in the breakroom either). Even though such threats aren't arguably legal either, nobody is going to fight against them because again - that means loss of job and money to pay rent.
Weekend voting would feel weird to me, but you're probably right that it's better.
http://www.bbc.co.uk/news/av/uk-england-hampshire-39795769/w...
At least the weekday election day somewhat levels the playing field, at least as I see it.
You could have small festivals at every voting booth to encourage people to come out to vote, and you could even move the ballot boxes as part of a big parade so there's absolutely no doubt in the process.
That said, we also have so many voting booths (I have never not been able to walk to my closest - usually only about 1km away) that the worst line I've ever waited in was probably 10 minutes.
Too few voting places, and voting days are not holidays (or on sundays) like in other countries. There are very obvious steps that could be taken to solve both counting time and voter turnout as well as suppression.
But when the party in power counts on that low turnout due to suppression in order to remain in power, do you really think they will do anything about it during their time in power?
The issue is with having a single machine that we have to trust end-to-end.
There are 5 digits voting precincts in the US, give or take.
Let's say the scanner costs $500 and there are 9999 locations. That's 50M of investment per party that can be re-used. It's a lot of money but it's an order of magnitude less than their campaign cost[2]
I think a larger issue is proving that the machine's the parties brought are not tampering with the ballots.
If one party brought machines to a poll place and managed to make trouble it is likely that the other party will show up too. Eventually it would just be part of the political landscape and the results would be better vetting all around.
This is also why the complaints of government being slow to respond to things are hollow: It's meant to act that way.
Which is why Trump and his "Executive Orders" are problematic; that isn't to say they are problematic for previous administrations as well. At any rate, they exist for a certain purpose, but that freedom to use them has been abused pretty much from day one of their existence.
Arguably, they are there mainly for emergency purposes, when something needs to be done fairly quickly, but is something that would take too long to have done thru ordinary channels (Congress, etc). It was probably (wrongly) assumed that Presidents would be wise and restrictive about their use (completely ignoring how human nature works of course).
I only mention Trump because he jumped into using them almost before he was sworn in; but so far (from what I understand) he has yet to break any records for usage of them during a term (except for maybe speed). In fact, the record actually belongs to FDR (who had 4 terms, also WW2, so that might be a part of everything).
In an emergency, I can understand their use, but outside of that, we (everyone, including the President) should trust in the slow plodding of our system. It may not be perfect or work out perfectly every time, but its lack of speed gives us time to reflect and make better decisions, IMHO.
You just need regulation to prevent the abysmal idiocy of manufacturers
Also standardize the interface (ADA compliant, etc)
Hire the Nevada Gaming Commission to regulate it and so be it. Make the machine print a paper backup that is put on a ballot box as well
I am assuming they count electronically, still maintain secrecy, and have a paper trail that can be manually counted if needed.
Assuming that we can achieve exactly equal security to paper voting for a moment, what have we achieved other than adding another few layers of middlemen and cost?
That being said here in France polling stations manage to count the ballots in a couple of hours at most after closing, so I'm not sure if it's really worth it.
and uncertainty. In principle the idea of hitting a button and having the non-networked machine spit out a paper trail that can be electronically counted sounds like a good compromise. But this leaves 1 or potentially 2 sources of uncertainty that are not necessary: (1) that what the machine prints actually corresponds to what you wanted and (2) that the counting machine actually counts what it says on the paper.
(1) is not such a risk if you can verify it visually before submitting. There do exist I believe some ideas in cryptography to allow you carry a record of your vote that allows you to verify it without revealing the choice, these could have interesting application here, but I don't know the pros and cons.
(2) is even a risk in the non-machine case, although it can be mitigated by having multiple independent parties do the count. But it can't easily be done by machine and have the same level of certainty.
Range voting is arguably more democratic than some other voting methods. But much more complicated to put in place with paper ballots :/.
(1) systems which limit rankings to a fixed number (the most extreme case for real ballot methods being two) of ranks (approval and FPTP both are two-ranks methods), and
(2) system which use numerical ranking systems that draw fiber distinctions than mere ordinal ranking (range/score voting being the main example.)
The problem is minimized in ranked-ballots methods, though there is room for debate over whether forced or unforced rankings are better in this regard.
For this reason, I would reject range voting for most public elections independent of practical difficulty (there might be exceptional cases where a consistent meaning can be attached to range ballots, but it's not the case in normal public elections.)
OTOH, ranked ballots Condorcet methods which need to compare pairwise results are probably more tractable with e-voting (or, rather, e-tallying.)
http://scorevoting.net/BayRegsFig.html
There's even a theorem that it tends to elect Condorcet winners under plausible models of voter strategy.
Believe me, its advocates have heard every criticism you can imagine, and the counter-argument is robust. I recommend you check out the book "Gaming the Vote".
Doesn't that just make it more expensive than paper ballots?
This gets you the speed and ADA compliance of the machines, but the reliability of paper.
This commission also helps to insure that there was no funny business at the aggregation center.
There needs to be more than this.
Actually this could work. Defcon this year was at Caesar's palace and despite a voting machine, a couple cars, multiple IoT devices, an electric scooter, and the Beijing Noodle's menu being pwned, nobody managed (dared?) to fuck with any one of the thousands of electronic slot machines.
It would make more sense to check whether your vote was counted post-facto but it's hard to especially if you still want to make it difficult to figure out who voted for whom. There are a few mathematical possibilities but they don't seem particularly practical to me: https://www.youtube.com/watch?v=BYRTvoZ3Rho
So for the time being paper ballots are probably the way to go.
You vote, confirm the receipt says the right thing, and stuff it in a box. Verified hardware is a solved problem. Given that it'd have no idea of the candidates, there's no way to program it ahead of time to favor a candidate without being caught.
We already have election officials watching each other. This needn't be more complicated than a late 70s calculator. It can even be so simple it doesn't know the candidates names or party affiliations.
Well, in theory... In reality, it will be a black box, connected to any open wireless, and made by the company who donated the most. But, in theory, it'd not be too hard to make a fairly dumb device that provides verification. It only has to count and print.
That's fair. I guess you could also force a recount for a random portion of the polling stations to detect anything fishy. I'm not sure if it'd be really worth it though, would it be such a massive improvement over simply counting paper ballots directly?
>Checksums and verified hardware.
Nah, that won't work. Who checks that? Who makes sure nobody tampers with the hardware after it's been verified? How can you even verify a piece of hardware, it could tell you it's running code X when really it runs code Y, no amount of checksumming will help. You could easily backdoor a processor to run a special ROM instead of the official one. Good luck detecting that.
Are we supposed to send electronic engineers at every poling station to check for any tampering with the hardware? How about people making copies of the chips that look exactly the same from the outside but are backdoored? You can't just trust whatever is printed on the package.
So what do you do? Take a random sample of machines during every election and decap all the chips to look for something abnormal?
IMO it's just too much hassle for very little benefit.
In the end a trusted minor miscount would be less bad than a count that is widely untrusted despite being correct: one would mean living through a "what if some people would have cast a different vote" alternative reality, the other would be a full blown legitimacy crisis, with unconstrained cheating in the next election as the least bad of all possible outcome.
Just stick to paper ballots.
We can put people in space. Verified hardware and software is a solved problem. Well, assuming you don't make it absurdly complicated. All it needs to do is eliminate manual counting - while providing a paper printout should someone need a recount.
You don't even need special tokens to make sure the person in the booth is only voting once. You can be as simple as making it flash an LED to be monitored by the election officials. They have volunteers from the various parties already there watching. Watching a light may even be easier.
It doesn't need to be complicated. It need only be a glorified abacus.
Separate from that, what does using millions of one-use physically collected chips buy you that just writing to a paper ballot doesn't?
You buy millions of chips so that we never have to go through this again. This way, the validated code and hardware don't have to be worried about, for a long time. The chips go in the machine, they run it. You'll have spares for a very long time.
What does this do? When the polls close, they don't have to count them. They need only count if there's a demand to do so, like a recount today. That's it. Problem solved.
*(maybe not ones where you actually have to product a documented reason why you can't vote in person)
When the process is simple and low-tech, like pen and paper ballots, it's possible for all ordinary people to contribute to keeping the process fair.
With electronic voting, you just have to trust the manufacturers and the voting officials. Everybody else is out of the loop.
There are normally multiple observers from all parties, plus election officials, that scrutinize the counting process in real-time. Disputes are raised and resolved on the spot where possible, and it not are normally escalated. This is why elections can have provisional results on the same day/night, and the official results are a few days later.
The extent of the competition over seconds and minutes here is bound to lead to some mistkaes, and I would question whether the checks are sufficient (the checks themselves may also be susceptible to mistkaes). The word "childish" comes to mind. We are talking about the future of the country and it's treated like a TV reality/quiz show.
That seems very odd. Can you actually do that? Around here the polling location must first ensure that the number of votes cast matches the number of votes recorded during voting, only after that can they open the urns and start tallying.
PS: Saying you need to change the law when that must happen with any significant change anyway is kind of a minor hurdle.
That's only because the government hasn't bothered to invest in electronic voting security, regulations, and rigor to the extend that they invested in fraud protection and financial regulations. I'm willing to be that if the government cared more about modernizing, securing, and distributing our voting systems, we'd see improvement on the level of electronic banking and finance.
Not that I expect this to happen anytime soon. For various reasons, voting rights and opportunities in the US are steadily declining and don't seem to be taken seriously. You can fill in the blanks yourself there.
Each voting machine tallies votes and prints out a filled-in paper ballot. Then you get quick electronic counting but with an auditable trail.
There are definitely issues with printing paper ballots (low toner, running out of supplies at polling locations, handling cases where the voter discovers an error after printing &c.), but similarly there are problems with optical tallying machines (e.g. hanging chads, incompletely filled in bubbles).
Incompletely filled-in bubbles is a real problem, true, but in my experience I've never seen this.
We trust billions of dollars every day to electronic banking, so why not a vote? Electronic banking comes with many types of federal guarantees to protect against fraud. The government can step in to investigate and prosecute the fraud as well. But there is no such guarantee for the voting to select the government itself!
But it takes so long to aggregate the votes if done with paper ballots. Precisely the point. Electronic voting allows scalable attacks where the number of weak points is dramatically reduced. It is very hard to scale attacks on paper ballots. You would need a coordinated effort in many voting stations to make it work as opposed to hacking a more central electronic system.
That is why I moved from thinking that electronic voting is the logical next step to thinking that we probably need to revert back to paper ballots.