Most of the world is able to hold a reasonable election.

Our problem is not how quickly we arrive at a result. That's not even really a problem. Fast results and "breaking news" both have created more problems than they solve.

Our problem is too many people in power aren't interested in trustworthy, reputable elections.

Doing that well does not require machines. In fact, using them often adds weak, or dubious links to the chain of trust between voter intent and a human readable, suitable for a court room, record of their vote.

Further, the winners of past elections are not very interested in changes or improvements that may impact their future in a negative way. Status quo inertia is strong.

All we really need are people, a publically visible process, and leadership committed to a fair election.

We don't have that right now, and we don't have it because corruption is rampant and the incentives to reform run into conflict with large amounts of money. No joke.

I don't know how to fix it. I have seen nothing electronic that comes close.

Think after you read this. Go take a look at elections, and fair and free election standards around the world, and wonder about how we apply those here. More of us need to.

Non electronic voting is super simple to understand for everyone.

I don't know for USA but in France it goes like this : People put their votes inside an envelope which they then put in a transparent box in front of at least two people, who are making sure you don't put more than one envelope and vote more than once.

When the vote is finished, volunteers and officials count the envelopes from the box and then count the votes. Everything is checked many times, at anytime half of the volunteers are for example checking that the other half is doing well. I did that twice and it's very very difficult to cheat while it's also a simple procedure. The first time, I remember having a lot of people watching me ready to notify any mistake.

Then the results and envelopes are collected and transmitted higher up. If you were part of the vote counting process, or a spectator, you know the numbers and you can check that what is reported is correct. It did happen that dead people were caught voting in the past but usually in France it's simple and secure enough.

Now, try to explain the blockchain to almost everyone.

this research paper claims to describe a secret ballot blockchain but i don't have access to it

https://www.sciencedirect.com/science/article/abs/pii/S01677...

notably, if the individual ballot verification process involves each voter knowing their own ballot, this still provides a proof mechanism enabling a voter to demonstrate to another what their actual recorded vote was.

Roughly: Blockchains primary attribute is that they make everything public and attributed and it thus can be validated. Making everything public is the opposite of what we want to do with voting.

Yes, a further answer might be:

It's tricky to combine convincing people that their own votes were counted, with making people's votes anonymous, and with convincing people that only registered voters were able to cast votes (and only one vote each).

There is computer science research that tries to address this, but it's subtle, and sometimes people are concerned that voters would be unpersuaded that the math is right.

Two other general problems with online voting which are as hard, or even harder, with blockchain-based voting:

(1) If people are voting from their own devices, those devices might be hacked in order to make them submit votes that don't match the voter's intent. This might be an even trickier problem if the votes are being anonymized by a complicated cryptographic protocols.

(2) Someone might try to attack the election by interfering with the vote-collection infrastructure. This is probably easier on a public blockchain, where, for example, miners could try to prioritize blocks that don't contain information about votes, or people might try to do fork attacks specifically during the election to cause fluctuations in estimates of whether votes were included on the chain or not.

Some of these things can potentially be addressed with even more cryptography, but I think the biggest insight is that just having a public ledger won't itself convince people that only registered voters were able to cast votes (and only one vote each), especially when the votes are recorded anonymously. In fact, just having a public ledger won't even interact with this challenge!

(Debian has had a clever solution to some of these problems for a while, just using a digitally signed web page rather than a blockchain, but their voters -- Debian developers -- are technologically sophisticated and publicly listed in a database that all other voters can agree on.)

Why do we not want to make it public? Not trolling but really asking this. It could lead us to a better overall system. If people can find out who I voted for and hence I would be in trouble then we have a broader systemic issue. If we can solve that voting can be so much more efficient and more importantly impactful.

Also could we anonymize the voter that way it’s like any other legally private information that can be accessed with a warrant etc.

You might want to look at

https://en.wikipedia.org/wiki/Secret_ballot

https://en.wikipedia.org/wiki/Preference_falsification

https://en.wikipedia.org/wiki/Social-desirability_bias

(I think Brendan Eich's experience with suffering enormous professional consequences for having donated to the Proposition 8 campaign -- a fact which was public under California electoral laws -- is a concrete example. You could imagine the same kind of consequences for everyone who voted for the proposition, in which case many people would not only choose their political donations based on fear of public opinion and social and professional consequences, but would even choose their votes based on those same fears.)

These issues might not convince you, but they represent a sample of other people's answers to your question. Also, in many places, depending a lot on the culture and history, a powerful local figure such as a large employer, union leader, "patron", or religious leader will actively try to tell people how to vote, if given the opportunity.

https://en.wikipedia.org/wiki/Political_machine

https://en.wikipedia.org/wiki/Electoral_fraud#Vote_buying

https://en.wikipedia.org/wiki/Treating

https://en.wikipedia.org/wiki/Clientelism

Verifiable public votes are the one piece missing for a successful "cash for votes" program, which many people (including myself) would consider highly undesirable, and even a perversion of democracy. Over $14B is being spent on the US elections in 2020[1], or more than $40 per US citizen, or nearly $60 per eligible voter. If a campaign or other party could offer citizens $50 in cash in exchange for verifiable proof that the citizen voted in the specified manner (either above the table or... not), that would change the nature of US elections entirely.

[1]: https://www.cnbc.com/2020/10/28/2020-election-spending-to-hi...

Yes, we have two larger systemic issues (arguably one) that prevent public voting.

Voters could be rewarded for voting a certain way.

Voters could be punished for voting a certain way.

You can't get rid of rewards without eliminating capitalism, and you can't get rid of retaliation without eliminating power structures.

Making voter information anonymous except via subpoena doesn't work for the same reason that backdoored encryption doesn't work.