[salvar]

And as soon as any of these individuals came in to vote and saw that someone had already voted fraudulently on their behalf, an investigation would start. And you'd get maybe what, three or four votes in exchange for the very real possibility of facing severe punishment?

Why aren't more instances of voters realizing that someone has already voted on their behalf if this is so easy and smart?

[pwg]

> Why aren't more instances of voters realizing that someone has already voted on their behalf if this is so easy and smart?

I can see two reasons (there are likely more):

1) If someone is going to vote as another, and they are smart about it, they will do some pre-research to pick out likely non-voters to become. Voting as someone else who is not going to vote would not likely be caught, because that non-voter will have no opportunity to notice the 'heist' since they do not go vote.

2) At least in the US, with low voter turnout (55% for the 2016 election is quoted by this page [1]) then someone has a somewhat large chance of simply randomly picking a non-voter as their "surrogate", and if they do win that pick, then that non-voter will not notice due to their not going to the polls.

Now, whether either of these strategies would allow an individual to amass sufficient votes to change an outcome is unknown. Even "close" races in the US often have a few thousand votes difference in the final counts, so for someone (or some group) to change an outcome they either have to find enough #1's above to amount to several thousand votes or have to "win" at #2 picks enough without "losing" at a #2 pick enough to get caught out. So it feels like it would be difficult to pull off a few thousand of these, in a single day, without getting caught unnless a fairly large group is involved. And the problem then (with large groups) likely shifts to keeping the entire group quiet about their activity (i.e., leaks from a group member become the downfall point, not other voters noticing double voting).

[1] https://www.cnn.com/2016/11/11/politics/popular-vote-turnout...

[dragonwriter]

> At least in the US, with low voter turnout (55% for the 2016 election is quoted by this page [1]) then someone has a somewhat large chance of simply randomly picking a non-voter as their "surrogate"

Sure, that gives one person doing it one time a fairly reasonable chance of not being detected.

It doesn't give a repeated, signficant pattern of such fraud a decent chance of avoiding detection.

[XorNot]

The number of "likely's" in your statement should tell you something about this proposal. What's likely? How many operatives get caught on "likely"?

Because you are proposing - at the bottom end - trying to get at least a few hundred fraudulent votes in, to swing a very close race. That's coordinated votes too - somehow you want to orchestrate this (and this is exactly what's always proposed as being the problem by voter ID advocates).

Let's be generous and say each operative can achieve 10 votes at different polling stations. This is hugely optimistic. So to get to a hundred votes (probably not enough in even the closest races) you need to

(a) find 100 non-voters who you know won't turn up that polls (hint: you can't - it's all probabilities)

(b) find 10 people to pull this off. And to be clear: this is recruiting 10 people to commit serious state and/or federal crimes. Which has a couple of problems - how do you find them? Why do they agree to do it? And what's the probability of any one of them getting caught? - and it is a probability. Remember in this example the numbers are hugely optimistic - so the probability someone gets caught rises rapidly to 100% as you increase the numbers.

which leads us to (c): someone will get caught. Throughout this whole effort you're literally one poll worker happening to know a person, or getting a bad feeling and calling that person's address afterwards, or a neighbor or member of the community hearing "I'm Julie Smith of this address" and going "wait that's not her" - because you're trying to infiltrate a community here, and the odds on election day the poll workers and voters know the names and faces of the people you propose to pretend to be, is also pretty high. So - someone, one of your operatives, gets caught.

Why do they stay quiet about the operation? They're facing criminal charges and jail time. They can be offered a deal if they roll over on your other operatives or you. How much are you proposing to buy their silence with? Why do they take the risk in the first place if not for monetary incentive?

Ah you say - but maybe it's a whole lot of independent actors doing it. Which okay, let's go with that - how do these people, in sufficient numbers, do enough local research to not get caught in sufficient quantity? - remembering that, when people do the "I just show up and say I'm someone else thing" as "activism" - they get caught.

EDIT: Also worth noting - unless the election rolls are actually destroyed, you also need to keep this all secret forever. If anyone has a crisis of conscience later and leaks it, then at minimum you - the coordinator - definitely go to jail since everyone will happily roll on you to avoid it.

[pwg]

Ultimately, my point boils down to: "given the possible issues, and multiple avenues of potential detection, any coordinated attempt of sufficient size is likely to be detected".

Which is what is likely the reason why not much seems to be happening, too many angles to "get caught" and so few groups attempt a coordinated attempt.

This was always one of the valid arguments against computerized voting machines and in favor of paper ballots. The paper ballots require "feet on the ground" attacks with high risks of detection. The "hack the machines" attack requires no large coordinated "feet on the ground" groups, and given some of the machines were reported to be internet connected, could be accomplished from a remote (and therefore safe) location.

PS - the number of "likelys" is because I have no sources for anything (beyond the 55% turnout figure) so it is all "guesstimates".

[makomk]

I don't know about the US, but here in the UK voters turn up and discover that someone has voted "on their behalf" merely by accident - the poll staff aren't perfect and sometimes simply mark off the wrong person's name as having voted. This is common enough that, as far as I'm aware, it doesn't lead to any kind of serious investigation. (Apparently, there was even one UK council election in Barnes, Richmond-upon-Thames back in 1976 where an ineligible couple voted, were marked off as being a similarly-named couple who later turned up to vote, and correcting this actually changed the result. This correction was only possible because unlike in the US we don't have a fully secret ballot - every ballot paper is serialized and traceable to the person who cast it. I don't think anyone was charged with anything over this. Probably wouldn't even have been investigated if the affected voter didn't see the close result and kick up a fuss.)

[danaris]

At least in my jurisdiction, it's not simply a matter of the poll worker "marking off" your name: they find your name in the book, then they push the book across the table to you so you can make your own signature in the box next to your name.

If someone is attempting to commit in-person voter fraud, they'll need to sign the name that matches the line they put it in. If they're just mistakenly signing on the wrong line, that will, first of all, probably be obvious to both them and the poll worker immediately, and even if not, it will probably be obvious to anyone coming along behind and checking if there's a discrepancy. (Unless the signature is well and truly illegible, and if that's the case, they can check it against the signature from the previous election for that person.)