If there is voter fraud, it might go away if we got reliable detection. Fraud is less likely when being caught is certain, and it is far more likely when being caught is about as likely as getting hit by lightening.

Even if there isn't voter fraud, the mere fact that people are worried about it is undermining trust in our elections. That needs to be fixed, and the only way to fix it is to have a means to prove that there is no voter fraud.

The reason people in the U.S. are worried about voter fraud is that one side of the political spectrum is pushing the narrative — without any real evidence — that voter fraud is rampant. We shouldn't make it more difficult to vote in order to fix a made-up problem.

Let me ask, why do we have identification for joining a company? Checking employment status? Boarding a flight? Starting a business? Opening a bank account? Getting government assistance? Why would that be necessary but not for voting?

Requiring ID allows us to measure when fraud is occuring by individuals attempting to submit votes and data that is not theirs. Otherwise, there is no "truth" to compare to. I guess we can always call every single person and verify their vote, but then is that not the same thing as an ID check?

No, it doesn't; it is neither necessary (as such fraud can be detected by other means) nor sufficient (as even with an ID requirement fraud can be performed without detection) for that purpose.

Why are other situations not an issue for requiring ID but voting is? How can we detect fraud by other means, and why can't these other services also do that? Does the fact that nothing is infallible mean it shouldn't be used? If you didn't vote but someone used your information, how do you catch that without some form of verification?

(1) There is no evidence of an actual existing problem to be addressed by voter ID requirements.

(2) Because voter ID requirements in the US, past and present, have a history of being introduced as part of efforts, often by the admission of their own sponsors (though usually in venues they didn't expect to become public) to suppress voting by legal voters of particular demographic groups disfavored by those pushing the measures, particularly African-Americans.

Why is voting less important than other services? If someone uses your information, how can we tell without verification of the vote? And if we can't tell, how do we know it didn't happen? And if we don't know, how do we know it's not a problem? If every vote is important then every fraudulent vote is equally important.

I'm only talking about ID verification for voting here, not other programs. If those programs are stopping voters then I agree that's bad, but that's a separate topic. And if IDs are required for so many other private and government services outside of voting, then aren't those services discriminating too? Right to bear arms still requires a background check so does that impede on people's rights? How does that reconcile?

Well, maybe. Absence of evidence isn't evidence of absence. In 2016, we were able to identify four cases of voter fraud in retrospect [1]. How does voter ID make this process more accurate?

Not having faith in our measurements for 2016 is also problematic. Let's say we uncover 100 cases of voter fraud after the fact with voter ID in 2020. Is this because we're able to better detect voter fraud now, or is there actually more? How do we know we haven't accidentally introduced a loophole that makes the problem worse?

I'm not really interested in a general discussion about "why does ID make sense for X but not voting". We can be specific: in what ways does voter ID improve our ability to measure — not deter — fraud? Because as far as we can tell now, it's already exceedingly rare.

[1] https://www.washingtonpost.com/news/the-fix/wp/2016/12/01/0-...

Also if we did enact ID checks, and zero fraud happened (perhaps because it preemptively prevented it all), how is that not a good outcome? Considering it went from non-zero to zero, yes that would be an improvement. What would be the loophole that introduces more voter fraud because of IDs? Can you please give me an example of this?

The comparison to other services and rights is critical to show bias and present a rational framework to tackle the problem. If the right to bear arms is not infringed by background checks, then it follows that the right to vote is not infringed by ID checks. Do you disagree with this statement?

ID doesn't provide verification that it didn't happen, it provides a difficulty bar for doing it.

Verification that it didn't happen, insofar as that is possible, is provided statistically through the absence of signficant patterns of attempted double votes, which you'd expect to see if there was any substantial pattern of voter fraud (especially by more than one party; barely plausibly, a single party with a sufficiently good likely voter model could be problematic to detect.)

It's still unclear how voter ID would improve this. Yes, the people at the polling place can check your face against the ID. But if the results are already in and you suspect fraud, what extra information does the ID give you?

This is not a circular discussion to me, because it's clear where to start. The claim is that voter fraud is a problem, so that's where the burden of proof lies. We don't start with an assumption and then ask people to disprove it.

> Also if we did enact ID checks, and zero fraud happened (perhaps because it preemptively prevented it all), how is that not a good outcome? Considering it went from non-zero to zero, yes that would be an improvement.

A reduction from 0.000002% to 0% is basically meaningless. In order for it to be a good outcome, you need to measure it accurately and show a causal relationship, and also not damage the election in some other way such as significantly depressing turnout.

> If the right to bear arms is not infringed by background checks, then it follows that the right to vote is not infringed by ID checks. Do you disagree with this statement?

We curtail people's Constitutional rights all the time when we think the benefits to society outweigh the consequences. Child pornography and obscenity laws infringe our right to free speech, gun control laws infringe our right to keep and bear arms, felony disenfranchisement infringes our right to vote, exigent circumstances infringe our right to be secure against unreasonable searches. I simply disagree that it's in our collective best interest to infringe on voting rights in this way, given the rarity of voter fraud.

As for the claim, I believe the issue is that without the above checks, there's no possible way to know if that fraud is being committed in the first place, especially with at least 40% of the population abstaining. You're saying the measurement shows that it's rare while I'm asking how the measurement is even possible for this type of fraud. Is that the fundamental disagreement we have? If so then I guess this thread has run its course, but I do wish there was a way to agree.

I never said it was less important.

> If someone uses your information, how can we tell without verification of the vote?

Ant significant effort at voter fraud (and, even more certainly, more than one such effort in the same jurisdiction) will show up in a pattern of attempted multiple votes.

The absence of significant pattern as of this is itself evidence that no significant (even in aggregate) voter fraud is occurring.

In fact, even with Voter ID, that's how you'd identify fraud that gets past the ID checks.

What do you mean by multiple attempted votes? Using the same name? That's a different situation though because I'm asking about a name which didn't otherwise vote. 40% didn't show up in 2016 and that was with record turnout, which makes a large supply of names that can be used, and that's assuming that everyone who did vote was completely legitimate.

Considering that, how can you possibly reconcile that everyone who voted is who they claimed to be without verification? How can this be tested post-election?

Denying the vote to someone who has the right to vote is mathematically just as problematic as granting it to someone who does not have the right.

Why is voting, which is critical to a democracy, considered onerous for requiring identification when driving, employment, banking, business and govt. programs all require it? Isn't that a strange balance?

Depress voter turnout, run third-party, identity-appealing, or confusingly-similarly named candidates. Print confusing ballots. Run complex initiatives or propositions with little or no discussion. Leverage majorities in one domain (Senate, statehouses, governor's mansions) to achieve, retain, or negate majorities or minorities elsewhere (vacancies are often filled by gubernatorial or state legislature votes, redistricting is often controlled by statehouse majority, a GoP Senate nullifies much of the power of a Dem House). Have a highly partisan secretary of state certifying statewide elections, including for national office. Have a Supreme Court structure based on few judges, narrow and long-lived partisan majorities, full-bank hearings, and obstructed nominations. Non-trasnsparent, un-auditablre, readily-hacked, and highly partisan technical voting systems. Grossly manipulated media.

Individual voter fraud simply isn't where it's at.

One side is deliberately leveraging it for that purpose, and it's the same side that has consistently blocked legislation to increase election security, set standards for voting machines, and more, so as far as I am concerned they're forfeited the right to have their arguments taken seriously after years of bad faith politicking.

Seems like we should focus on making sure poor/minorities have easy/free access to IDs, rather than keeping the voting process lax.

I had to sign up with my address, and I have to prove my identity when I vote. That all seems perfectly reasonable to me. Even a low-tech ink mark used by some places seems perfectly fine. The law is full of tradeoffs, and I think proving you have voted only once and where you are allowed to vote is perfectly legitimate.

If you are not going to pay them, and not going to view it -- then they lost the income -- you might as well bypass the pay wall..