Not the ballot itself. Here in Oregon, anyway, they must sign the outer envelope. After signature verification and scan for identity, the outer envelope is opened and discarded. The inner envelope with the ballot goes on to be counted anonymously.
Like @suranyi said. 
Here there is some concern about relatives insisting on entering the voting booth along with the voter to ‘help’ them. Particularly in the London district of Tower Hamlets.
This is undoubtedly against the rules, though polling station staff have reportedly not pushed back on it.
It wouldn’t necessarily be against the rules in the US. The last time I checked, it was legal to take someone into the voting booth with you to assist you, as long as that person wasn’t your employer or union representative.
Whether you have to justify your need for assistance, I’m not sure, but you are allowed to have assistance.
One of my friends did that for a local election a few years ago - told her husband, “Just fill out my ballot for whoever you voted for, because I don’t really care.”
And when I was on the absentee board last fall, one thing we had to look for while inspecting the ballots themselves was ANYTHING on it that could potentially identify the voter. I don’t recall that anyone signed or otherwise identified a mail-in ballot, at least not that time.
A proxy voting system could be an interesting discussion.
The only solution you can think of is to “go back” to not having mail in ballots? Are you under the impression that the normal state of affairs is to NOT HAVE mail in ballots? Because you’re a couple hundred years late there bud.
Dunno about that, in my neighborhood lots of people had their stimulus checks stolen, and despite getting the fucker on video nothing ever came of it.
Apparently not very well, sometimes spending as little as five seconds per signature.
2020 election: How does voter signature verification work? - Los Angeles Times (latimes.com)
Because, in the US, the husband can go with his wife to the polling place, she can claim she needs his assistance and he can go into the voting booth with her to make sure she “votes right”.
I guess we now know how to make conservatives take domestic violence seriously. Just rebrand it as “voter fraud” and recommend a bunch of suppressive voting laws that do nothing to address domestic abuse.
Five seconds is a long time to study that amount of writing. I’m surprised it’s that long.
They picked up my daughter’s changed signature from when she registered to when she voted recently.
Section 208 of the Voting Rights Act states that, "Any voter who requires assistance to vote by reason of blindness, disability, or inability to read or write may be given assistance by a person of the voter’s choice, other than the voter’s employer or agent of the employer or office or agent of the voter’s union.”
That would technically include a spouse and proof of disability is not required. In practice, though, I haven’t read of flagrant or frequent violations. If a series of spouses were to come through any local polling place making this claim, they would be flagged and the election inspectors all over it.
I think this is another of the technical loopholes in the law that, as in the OP, are mostly theoretical rather than societal problems that affect elections. Of course, I’d be interested in reported cases to the contrary.
I’ve been interested in liquid democracy as a form of dynamic representation/proxying.
Is this a problem? Does this come up in divorce proceedings or when women leave their husbands? Is there any reason to believe it occurs in sufficient numbers to outweigh the benefit of increased voter participation that comes with mail in voting.
My spouse and I vote the same on almost every candidate and issue. Worldview is part of what we like about each other.
Certainly.
If you split married couples into two sets and one set votes the same 90% of the time and the other votes the same 95% of the time, that probably doesn’t mean anything if it’s just a few dozen couples. But if you have two sets of half a million couples, the difference between them should be negligible if it’s just a random selection and there’s nothing specific to either group. You would expect that group A would be 92.345% similar and group B would be 92.487% similar, or something like that.
But if you have two groups of half a million couples and it’s still 90% and and 95%, statistically that just doesn’t happen unless there’s something changing the results between the two groups. The issue isn’t that the numbers are high (both in the 90s), it’s that they should be the same number.
Now we do know that there are differences between mail-in voters and walk-in voters. The former is more urban and more liberal, on average. Knowing that that difference exists, though, you should be able to correct for it mathematically.
I don’t doubt that there is because there’s a lot people in the US, but is there even one case where this happened?
Seems like you need to have one case before you de-anonymize millions of ballots to study.
There is no need to de-anonymize.
Walk me through the steps. How do you match a husband’s ballot to his wife’s? Both mail in and in person.
For mail-in, you have two envelopes. The outer envelope has an address and the inner envelope contains the ballot.
Team 1 converts each address to a hash value, using specific rules on naming (St vs Street, etc.). They print out a QR code of the hash, stick it to the inner envelope, and pass it to team 2. Outer envelopes are destroyed.
Team 2 opens the inner envelope (never having seen the outer), and processes the ballot, inserting the ballot results and the hash into a database. Inner envelopes are destroyed.
Team 3 assigns new IDs in the database, replacing the hashes. They also drop any records with only a single entry.
For in-person, you just follow a similar procedure of keeping the person’s address and ballot separate.
And note that this is v1 of this procedure. I suspect that you could refine it further.
Ah… so a two step de-anonymizing process is somehow re-anonymized if everyone forgets how to do step one once they move on to step two.