USPS patent for reliable voting system secured with blockchain
patentscope.wipo.int
patentscope.wipo.int
Election systems need to be simple and transparent above all else, and old fashioned paper ballots that can be counted and audited by hand (if necessary) are the best way to achieve that.
Any automated system, whether using blockchain or the existing electronic voting systems, is hopelessly opaque and serves only to decrease confidence in our democracy.
If your goal is full transparency of every component, I don't think a lot of people here could understand that. There's a lot happening and to understand everything at a low level takes a lot of expertise. Understanding at a high level "experts use fancy math to make my vote private and immutable" is easy but also requires a lot of trust. Something we have to do regardless of full transparency, because someone could in fact hire a bunch of people to rig things.
The problem is the breakdown of trust. Fox has been talking about voter fraud for well over a decade and while they do have clear examples, they are over inflated at their significance (sorry, one person counting twice does not constitute a rigged election with a million voters. It constitutes a mistake). The president has a devout fanbase and echos this rhetoric from a very privileged platform. The problem is trust. There's the old saying "trust but verify." The transparency is the ability to verify, but does not require that every person is able to. So don't needlessly tie your hands behind your back.
A lot more trust than the current system, where voters and observers can at least believe what their eyes see. With a digital system nothing they see happening necessarily means anything! They can't verify it, and thus they shouldn't trust it.
I'm all for paper trails, and I think they are necessary, but our current system isn't exactly transparent either. Logistically this is impractical. We've always relied on a subset of people to tell us it is trustworthy. I.e. the volunteers that count. If you can have that and do cryptographic arithmetic so that anyone with a computer can verify the results, that's more transparent, not less.
I mean I think there's this idea that people counting physical ballots in a rushed and hectic environment has high accuracy. That sounds like a naive assumption to me. But what is that error? 0.5%? 1%? For the most party elections are not won by that margin and when they are a recount is issued (and low and behold, numbers change slightly. Proving that the counting is imperfect but good enough).
I think we need a more secure system to expand options for electronic voting and counting but it needs to have a paper backbone for true reliability.
The funniest thing is, those are the same people claiming the election is ironclad now, now that it's them who have won.
I think Turkey can be a great example in this regard. No one trusts the opposing parties with cheating, so every single vote is watched by six parties simultaneously during the count:
1. There are paper ballots thrown in the cases
2. After the election ends, the votes are counted openly. It is a constitutional right to be there and watch while votes are counted.
3. After the count ends, the tally is written down and posted for anyone to see. People take pictures of the tally in order to upload into a central database - be it a party's or an unaffiliated third party's
4. The cases are then shipped to the election committee, which aggregate the results. They also post the results of each case, so that anyone can check against the database and contest of there are mistakes.
The whole process is open to be checked by everyone and since trust is so low these days, everyone keeps track very closely. Even in Turkey's conditions, this gives some kind of peace of mind.
No they aren’t.
The "fraud" alleged by Trump in this case consists mostly of his observers being asked to stand 6 feet back so the election officials could actually work. And the officials themselves are selected from each party.
In some states it's illegal to photograph your own ballot due to concerns of vote selling and coercion.
The distance thing is complicated. I assume it's in reference to Pennsylvania. 6 feet was due to COVID-19 social distancing rules but watchers were actually originally segregated much farther away (at least in Philadelphia). I think the PA Supreme court overturned the 6 feet thing and watchers are farther away again but unsure.
I don't think PA law actually says how far watchers have to be. https://www.legis.state.pa.us/cfdocs/legis/li/uconsCheck.cfm...
Some of the complaints are fair. For example, IANAL but the PA law clearly states that mail-in ballots have to be received by election day.
> For example, IANAL but the PA law clearly states that mail-in ballots have to be received by election day.
If it clearly states X you might as well link to it.
From what I understood they can allowed if the votes were cast in time (which a stamp proves), even if received a few days later (though with a deadline).
I'm a regular counter in Dutch elections, and observers should not interfere with counters (such as photographs, talking, noise, etc). It gives them unnecessary stress which reduces their output. Leave these people alone, let them do their job.
We know paper ballots work perfectly. It's very difficult to commit fraud and as you say, can be audited easily.
Adding layers upon layers of complexity on top of that is not the way.
I agree that we should make the process simple and transparent so that others can verify it, but not because we want the truth-denying anti-science megalomaniac to concede (they won't in the face of hard truth and facts); but because we can have other experts in the field peer-review it.
Blockchains provide a mechanism to audit and eliminate fraud.
For example, by only allowing one vote per person. And allowing the person who voted, to confirm their vote was not changed or tampered with.
Paper ballots guarantee nothing.
Both initial counting, recounting and auditing is quite straightforward. For example, Risk Limiting Audits (https://en.wikipedia.org/wiki/Risk-limiting_audit) are easy to do and provide extremely high degrees of mathematical certainty that an election was not tampered with.
In the US, electoral fraud is not an important issue. Fraud exists to some extent, as in every country, but the scale of it is so small that it’s not a thing worth solving. The application of an opaque and difficult to understand technology to the problem offers no appreciable benefits and comes with its own baggage.
The same applies in the UK, which has piloted systems requiring voter ID. There is no appreciable benefit, and some obvious costs.
VoterID makes perfect sense and is used in some parts of the UK already for 20 years. How anyone can possibly object to it, especially given how many and varied options are acceptable, astounds me.
It's a discussion regarding the UK.
It's a problem when coupled with very active, intentional voter suppression from those who control what forms of ID are acceptable, and how easy it is to obtain them, which is known to vary across demographics.
None of this is controversial: I live in North Carolina and the courts have repeatedly struck down Voter ID implementations from the (supermajority Republican) state legislature because the implementations "target African-Americans with almost surgical precision."
Such unsupported conspiratorial vagueness about disenfranchisement, equally tired and transparent a duplicity, seeks only to assert I believe an underlying hope/fear that there is some systemic problem that might be unwound. For it makes no sense to shore up a system once designed in a world were communities knew each other better to remove any question of doubt from that component of the process.
Are the people of Northern Ireland disenfranchised at the polls? France too?
I think not
And yes, voters in areas with ID laws are disenfranchised. Voter ID laws can routinely be observed to disproportionately affect people who are otherwise legitimately able to vote. You may, in some clear and well-scoped circumstances, be able to make an argument in support of them when they solve some problem. NI might even be an example of that. Since electoral fraud does not appear to be a problem in the rest of the UK, it does not seem possible to avoid a net negative impact.
Without VoterID it's literally impossible to know it's scope
> introduce a system that disenfranchises people?
It doesn't. Do you even know how the process works?
Wouldn't national identity cards with 2fa be a requirement?
Right now the systems are so convoluted that it gives a fertile bed for ignorance to grow, which can directly results in lower public trust in voting and or conspiracies.
The electrification of voting (e.g. Diebold Election Systems) have only worsened that through a combination of questionable bidden processes and outright low competence.
This takes complexity and multiplies it, then expects the public to "get it" when you show them the algorithms, and ask them to do the raw mathematics themselves. It is completely out of touch with the general public.
We need to get back to where average Joe the Bartender doesn't need to delve into the election process to call it "FAKE NEWS MEDIA IS TRYING TO STEAL THE ELECTION". We need to get out of the rut of "If I can't verify it myself, it must be false." There are a lot of things you can't verify yourself - the measuring tape you have at home requires that you trust the manufacturer for its accuracy. No one asks for a NIST certificate and proof that it traces to the original standard of a meter. Or the pharmacist who gave you the pills. Or that the money is safe in your bank.
We trust our doctors, our scientists and our institutions. That's how the society works. Please stop the nonsense of empowering each citizen to make their choice of truth - because they would be pretty horrible at evaluating it.
I personally think that voting should be based on physical ballots and ink/paper. Not electronic. Yes, it is painful but there is a physical evidence to every vote, not some transaction line in the database - although we already trust our banks with the bank balance.
I'm not saying this is a good policy, just a hard technical tradeoff of what would be required for verifiable voting without cryptography.
It's an all or nothing proposition too. A hybrid system (opt in to make your vote public so it can be publicly verified) would create additional vote pools that would differ based on ideological preferences and thereby attract calls to disenfranchise, as we've seen for mail in voting.
It's illegal to take a photo of your ballot to prove who you voted for, think about why. Because if you could then the local crime-ring would force every person they can intimidate to go vote for their boss and take a photo to prove it or the thugs will hurt their kids.
Or maybe think about the pressure some people surrounded by family and friends of the opposite party feel. Sure they will wave the right coloured flag and throw dirt at the TV with their friends at the "wrong" guy, but when safe and sound in the voters box they can do what they like. Imagine if all your friends "opted" into revealing their vote and you didn't.
> I'm not saying this is a good policy, just a hard technical tradeoff of what would be required for verifiable voting without cryptography.
It's a hard technical fact, that's all I was saying. Either trusted third parties (that are currently being maligned to shake our faith in the voting system), open ledger voting, or steep cryptography. I'm personally in favor of the cryptography, but I'm also able to understand it. Unfortunately as it goes against making the system widely understandable, I'd expect the same kind of "fake news" attacks were seeing presently.
FWIW mail in voting also gets rid of the secret ballot. Organizations haven't really adapted to taking advantage of this, but we're on borrowed time. The fundamental problem is that our current system relies on trusted third parties, and that trust is breaking down.
Or doctor who has a single opinion about a practice they want legalised, but doesn't want to admit it to anyone.
A gay teen surrounded by homophobic people who wants to legalise their sexuality but has to fake it and shout at rallies because otherwise people might question what's up.
Those weren't hard to come up with...
Lobbyists and donors vulnerability is better alleviated by fixed salary and functional impartial criminal/justice system.
Added to that, it could lead to 'always on' political discourse, which tends to be divisive and disruptive. This is the reason that companies prohibit political discussion in the workplace, or family get-togethers can become nightmares when strongly opinionated relatives clash.
It also excludes the possibility of agreeing to disagree, which is based on an acknowledgement of human fallibility, that people have intrinsic worth apart from their opinions, and that a diversity of opinions may be overall more beneficial than strict uniformity.
Mexican politicians buy votes with a one time delivery of grocery bags or 20-50 dollar grocery store cards.
[0] A recent example: https://news.yahoo.com/mystery-blue-dots-painted-near-225645...
https://nypost.com/2020/11/06/aoc-facing-heat-for-wanting-to...
It's obviously possible to read what she wrote as a plan to compile a list of enemies for future consequences, but just to give an alternate perspective, I read what was said as a simple call for accountability of public figures (i.e: politicians within the Republican party who have supported Trump, but now are disavowing him), rather than being intended as some sort of mass-surveillance programme targetted at anyone who disagreed with her.
Yes, yes it is.
https://www.washingtonexaminer.com/washington-secrets/first-...
Tapper doesn't appear to be directing employers to do anything here: he's stating how employers judge candidates, and noting how considering one's actions in that frame might cause one to act differently. There's not a threat here, it's an acknowledgement of how society works.
My point is that these aren't reasonable interpretations of what these figures have said. Like I said, it's certainly possible to read into their words as much as you want, but it requires far fewer assumptions about the unknowable mental states of others to simply take their words at face value. Divorced of the spin placed onto what's been said by the press, these statements are entirely innocuous.
[1] https://en.wikipedia.org/wiki/End-to-end_auditable_voting_sy...
There are many challenges in voting systems. Blockchains address none of them and, in fact, without great care they could easily obliterate election security-- enabling coercion and vote buying. A true blockchain would also degrade security in other ways.
Anyone pushing blockchains for voting are extremely confused by the technology, -- e.g. calling any cryptographic audit log a "blockchain"-- at best, or exploiting buzzword hype to defraud you, at worst. None of the possibilities suggest that they have a system you should want to adopt.
Does anyone have a counterargument to paper ballots? I know that there were some murmurings of fraud in the paper ballots this election but from what I’ve seen, paper ballots are head and shoulders above (hackable) electronic systems.
If you are clever about how you make the ballots, you can even get end to end verifiability, and you can publish the ballots so the anyone who wants can verify that the count was done correctly. See [1].
The former could be addressed by having ballot MARKING machines instead of voting machines. Some jurisdictions have already moved in this direction. With a ballot marking machine, you can use a device such as a touchscreen that then prints out a ballot with your selections on it. You verify the correctness of your votes on the ballot. If they are not correct, it is a spoiled ballot that you witness being destroyed before using the ballot marking machine again. If the ballot is correct, you turn it in just as you would a completed human marked ballot. Given the standardized printing of the ballots by the ballot marking machine, they should also be machine readable without bar codes or any other non human readable form. If there is a question about the accuracy of the machine count, a human count can be used to verify correctness. The real gain here is the elimination of ambiguity in voter intent and the elimination of ballots that only exist inside of a computer.
It has the drawback that each person needs to know a bit more about cryptography than a layperson should so it is something that is currently not accesible for general elections.
There are some countries that deprecated their FPTP method to great results.
Sure you can. That's what the cryptographic field of zero-knowledge proofs is about. For example there's well developed protocols where every voter can verify that their vote was counted exactly once, anyone can verify that no non-authorized voters were counted, and perfect secrecy is preserved.
As an aside, having the verifier choose the desired confidence interval is a rather nice property of interactive zero knowledge proofs.
There are a variety of teams and projects working on that problem. I wouldn't say that limitation is really an absolute.
But it may be possible to bypass those issues with a hybrid physical/digital system. Or it may be possible to construct a system that is unsound in theory but works decent in practice. I'm thinking about technologies like DRM and Proctoring.
But deep fakery could become a threat to proctoring in the near future.
It probably is a very unpopular opinion, but what about thinking again about an open vote?
What is wrong with openly standing by your choices?
The fear, that there are too many bullies around who would bully people to change their vote? Well, if bullies do this, they also do other things, so we need to protect the vulnerable of them anyway. The way I understand it, the secret vote comes from a time where henchmen of the local warlord/prince/ were much more common and secret vote is a defence against organized bullying to votechange. But are we( as in people of western states) still living in such times?
People with money will buy votes if you make it possible for them to know how people voted. We have historical examples in US elections pre-1885 and in Congress post-Watergate.
So good democratic leaders make rules to protect the people from themself? Well, I know that one. I simply disagree. I rather think powerful people make rules to let them stay in power. An obscure vote system can help with that.
Imagine suddenly a change to open-vote:
would Trump really have won, if he just would have tried to buy votes? I doubt that.
Would you have voted for him, if offered money?
No? Well, why do you think then, the average joe wants to sell out in meaningful numbers?
Meanwhile, my vote is completely irrelevant. There are plenty of people who didn't vote, and you only need to pay/suppress a couple thousand in AZ, NV, and GA to flip the election.
History says that buying and threatening voters is effective and can move elections. Seeing as there isn't much of an upside to public votes, why would we risk it on the naïve notion that the average joe would think that far ahead when there is cash or a gun on the table?
But as far as I know, many people dare to show open support of one party or the other and the newspapers are not full of people being beaten up because of it?
I can't convince you not to be naïve about the world, but I'm also not going to apologize for supporting laws that reinforce Democratic norms. I have no interest in living in a dictatorship, even if a loud minority believes that is the right thing.
I think I have seen quite a bit about this world and I adress most problems direct or indirect to secrecy and lying.
If you believe demcoracy needs secrecy, than this is not for me to change.
I also did not say we need a open vote now. I said, "thinking" about it. If that alone makes you think I advocate dictatorship, than there is no further point discussing.
Suddenly changing that, would create conflicts. But after the initial conflicts would be solved, it could work?
But this is not a good example, because in east germany there were in theory democratic, secret votes.
They were just meaningless, as the voting system and everything else was in control of a totalitarian party.
So you think, the US is like this? People are mostly afraid to speak of their true opinion? Sad state of the land of the free, then.
The whole idea of having multiple humans standing around watching every ballot seems like it's unfeasible, and presents the opportunity to cast doubt on the process.
With the exception of cases involving perpetual motion, a model is not ordinarily required by the Office to demonstrate the operability of a device. If operability of a device is questioned, the applicant must establish it to the satisfaction of the examiner, but he or she may choose his or her own way of so doing.
Use paper and pencil. Anything more is for failures.
If you wanted to sway an election with electronic voting you would just have to hit a few points versus going through the hassle of voting in person... Or even voting by mail.
Plus, the problem is lazy, stupid humans. You could have the most secure system on the planet, but if the password is password then it's worthless. Social engineering is a HUGE thing. Look at defcon videos where they "hack" via pretending to be someone in the company and make a phone call. It happens so fast(minutes). Also, there's a video on electronic voting , it might be a vice video or jon Oliver... districts left voting machines easily accessible, not under lock and key, no cameras, nothing...
America should do what Germany does with voting make the ballots easy to read/use, none of this stupid chad shit. There are big circles on our ballots to mark the vote, it's simple/clear.
The xkcd comic sums it up perfectly too.
First, it’s a system that people without any degrees can understand and execute without the aid of computers. Trust involves far more than cryptographic trust: understanding how the system works and that it is being carried out faithfully is also important: how do people do that when not only is the system opaque but the means of executing it is also a black box?
Second it’s really difficult to attack. The physical aspect and coordination required makes it difficult to pull off and harder to do quietly without anyone noticing.
People are commenting that you can’t verify or anonymize digital voting, but with blockchain voting, you could mail each voter a “wallet” private key (probably a QR code). Each “wallet” contains one vote, and voters create a transaction that sends their vote to the “wallet” for the candidate they are voting for.
The private keys would be randomly generated and not recorded by anyone, so anonymity would be preserved. Normal blockchain double-spend protections would prevent double voting.
Am I missing anything?
I guess one downside would be that everyone could see a running tally of the votes as they happen. Is that really a big problem?
1. Issue everyone a cryptographic signature card with a private key on it.
2. This card has a pin and allows the user to digitally sign any document with a private key associated with that citizen/resident.
3. This private key never leaves the card. Overall, this is fairly similar to how hardware wallets for blockchains work.
4. For voting, you can either download your ballot online, or go in person to a voting booth.
5. You fill out your digital ballot, then sign it with your private key / card (again, either on your own computer or a ballot box machine does this for you), and your signed ballot is uploaded to a central repository. This central repository is not a blockchain, it is merely a place that has each users public key. When you need to count the votes, you validate everyone's ballots with public keys and tally them in a second.
6. As proof you have voted, the machine you uploaded your ballot to in turn sends you a payload which has been signed with your public key, and therefore can only be decrypted again using your private key card. This is how you know your vote will be counted.
[Pros]
- Replaces the need for mail-in voting and worrying about ballots arriving at some arbitrary election deadline. A citizen with their auth card can vote from anywhere in the world. Elections during crises would be a non-issue (unless the crisis brings down the internet, though I doubt a crisis that brings down the internet would leave a postal system unscathed – esp. once Starlink is live).
- All votes can be counted and re-counted in seconds.
- Fraud seems difficult with this system. You would either need to compromise the auth card issuing authority (difficult and assuredly traceable, though still a SPoF) or steal people's cards and know their PINs. Cards could obviously be canceled / reissued with a high-level of identity verification.
- The auth card becomes your new identity provider for basically every authentication you need (getting a bank account, signing a document, etc) and is much, much superior to the current status quo where the validation is either a digital (handwriting) signature or "what is your address and SSN?".
[Cons]
- You need to give everyone a card. Not a big deal because we already are dishing out SSN cards, DLs, passports, birth certs, etc without too much difficulty. Also you'll need to give a card terminal[1] to anyone who wants to do things from home.
- You need to be able to interact with digital technology. Honestly this shouldn't be much of a problem. If you can use a credit card, you should be able to use this digital auth card.
[1] https://upload.wikimedia.org/wikipedia/commons/thumb/e/ee/Ba...
There are ways to get around this, but votes must not be attributable to individual citizens.
Our election trouble in the US boils down to corruption, and I will be blunt:
The people responsible for trustworthy elections aren't interested. Where they are, simple, time tested, production proven ways and means work just fine and are very robust.
Trust starts with people being involved in the civics. Every layer of complexity comes with both a barrier to trust and an opportunity to exploit the process to control the election outcome.
Electronic voting started out this way.
So far, no scheme has even made economic sense. At any given time and population size, there are always people available for the civics.
Paper has one property no electronic system has presented to date:
The chain of trust between voter intent and the record of their vote is unbroken. When we mark media, we record a direct expression of voter intent and that expression is used directly to determine the final tally, and with that, the election outcome.
Electronic votes are votes by proxy. Voters must trust the tech to interpret and maintain the accuracy of their vote intent. This, by nature, is untrustworthy given we don't actually need to vote by proxy in the vast majority of cases.
This is all exacerbated by voter anonymity and freedom to vote or not. Having both of those in play highlight the dynamic nature of electronic records.
Flat out, this problem is not well aligned to technology.
Getting people involved and insuring elections embody these ideas to the maximum possible is how we run trustworthy elections:
--Freedom to vote or not.
--Anonymity, no vote shall be personally identifiable to the person who cast it.
--Transparency. The law, means, ways, methods, shall be visible to all impacted by the election. It should be simple to understand how an expression of voter intent is recorded and followed through to the final tally.
--Oversight. Not only must the law, process means and methods allow for public involvement, this must be encouraged and actually done and discussed. Legitimacy is made or broken here.
These are all human problems, not technological ones.
Human solutions are in use all over the world.
The US stands fairly unique in both its general denial about the state of its elections and ongoing efforts to fix human problems with ever increasingly complex technology that does not get at the root causes.
People need to be involved. The process needs to be simple, robust, records available, human readable in a court of law and leaders need to lead by example and cultivate trust and people involvement.
Conflating secrecy with security lies at the heart of many election problems in the US today.
Secure vote records mean those vote records, once cast, are kept consistent and not lost, so that confidence in a publicly visible count remains high.
Not identifying votes to voters is the basis for secrecy being conflated into reasons why the public is denied visibility into the process.
Rebuttals welcome and encouraged!
Now for the rough news:
Reform does not begin with the losers.
They lost and that undermines their position and standing to demand change. While their incentive to reform is high, driven by the need to seek power, how just their arguments are is put into question by what the winners can always argue is a general failure to appeal to the people. Losers can take a case to the people however. When said case is seen as an overall gain, it has a solid chance to succeed.
(True whether their intent is just and true or self serving, nefarious.)
Winners will suggest the remedy for losers is to perform better in the next election and that will carry great weight and serve as a basis to undermine reform efforts seen to also undermine the legitimacy.
Winners have a low incentive to reform as said reforms can put their currently successful path to power into jeopardy.
They also understand legitimacy is core to their power and standing as legislators. If they question the process, they question their own legitimacy.
On this point, democracy depends strongly on legitimacy. No one holding power wants the overall process legitimacy undermined. People consent to governance in democratic society. Legitimacy is the core holding things together.
So, what is the path?
Generally speaking, courts are reluctant to legislate process changes. From their POV, they can sort out an election that needs it and will do so under the current process.
The courts can affirm a successful reform, but cannot generally initiate one.
A better process is a political problem and the courts will make it clear new law / process originates from the politics, not the bench.
(I am ignoring activist courts here and those arguments have merit, but none of that is truly germane to election reform.)
This does not leave much!
Reform comes from the winners, people currently holding power.
Their power may be democratic.
Prime example is they won their election.
It may be a matter of process.
Succession is a prime example. They end up in power because someone else fell out of power somehow. (This does put an additional emphasis on cultivating legitimacy into play, just saying.)
It may be the result of conflict too.
A coup, revolt sufficient to topple prior regime, you get the idea.
In all cases, winners have legitimacy as a primary concern and will be extremely reluctant to put that into question. Reform arguments are more potent when existing power and legitimacy are considered and threats marginalized.
Conclusion:
Successful election reform happens when the winners can propose a solution that does not undermine legitimacy and presents a mutual benefit for themselves and those voters they hold power over.
The people, depending on law in their state, can also propose election reform to be generally adopted. This is the primary path for losers reform efforts. It garners legitimacy and can also be used to augment campaigns.
There are some successes!
Maine abandoned first past the post voting.
Oregon adopted vote by mail with good overall results. Other states, Colorado, Washington... have adopted it for similar reasons and have seen similar positive outcomes.
Anyone interested in election reform would be wise to study past reform efforts that improved democracy.
Notably, perception of reduction in democracy, trust, legitimacy are by far the dominant outcome for electronic elections.
Successful arguments include:
Better elections = better politics = better policy = better lives.
Improved public involvement = reduction in fraud and fraud perception = greater trust = more legitimacy.
On that point, reform efforts aimed at greater participation can also mean more successful GOTV. This can mean consent perception is higher, legitimacy higher, difficult policy easier to live with.
Airplanes and elevators have computers in them but the "adversary" is gravity, which doesn't update its "attacks" to adapt to your countermeasures.
Voting systems don't just have to defend their "planes" against high winds and lightning, they also have to defend them against surface to air missiles and rogue pilots.
Seems like the argument against computer voting leaves the door open for bad actors.
Of course, some deep technological revolution might change that eventually. But in that case we still shouldn't rush to shift to a digital-only elections system. Why? Because it doesn't just have to work. Voters have to have confidence in the system. And that system isn't just technical, it's social.
Everybody understands paper. Everybody understand marking paper. Everybody understands manually counting marks on paper and adding up numbers. Further, everybody has intuitions on how to protect and defend that process by making sure the right people are present. People do generally accept machine counting of paper ballots, because they don't see that as much different, and because they know a manual count can be done if there are problems.
However technically clever and perfect some blockchain system is, it just doesn't have the simplicity and robustness in the face of attack or failure that we need. And it especially doesn't provide the social validation of what is essentially a social result.
I'd go even further and say that if voters believe in the system, it doesn't actually have to work. Belief in the system is the mechanism by which democracy provides stability. The election system actually working as advertised is only necessary to the extent that the system actually working is required to make people believe it works.
If this situation sounds completely A-OK to you then I don't know what to say.
Some people are quite cavalier about switching over to computerized voting system despite not understanding the consequences. Resisting this trend is to the benefit of society.
Explain.