22

First, the Photo ID Voting Law was put in place because states wished to decrease voting fraud, right? If not, please explain to me why it was implemented. I think I did well enough research but there's a good chance I'm wrong.

Also, if the Photo ID voting law was in place for decreasing voter fraud, is it working? If it is, how well? Statistics please. I can't seem to find the answer I'm looking for.

Thunderforge
  • 7,707
  • 4
  • 29
  • 64

4 Answers4

59

the Photo ID Voting Law was put in place because states wished to decrease voting fraud, right?

That is the stated reason, yes.

However, in-person voter fraud (the only fraud that would be caught by voter id laws) is essentially non-existent. There were 18 confirmed cases between 2002 and 2012 in Texas.

The 5th circuit appeals court found that the Texas law (SB14) specifically discriminates against minorities (see also here).

A federal court found that a similar law in North Carolina "target[s] African Americans with almost surgical precision". It also found:

In particular, the court found that North Carolina lawmakers requested data on racial differences in voting behaviors in the state [...]
"With race data in hand, the legislature amended the bill to exclude many of the alternative photo IDs used by African Americans," the judges wrote. "The bill retained only the kinds of IDs that white North Carolinians were more likely to possess."

Similar accusations were made against the Texas bill:

The Justice Department argued that “a wealth of evidence” suggests that lawmakers consciously discriminated

The reasons given by the Justice Department are among others:

  • the inevitable discriminatory impact of the law (Hispanics are 195%, Black people 305% more likely to not fulfill the strict id requirements than white people; while free ids may be provided, they are difficult to get, especially - and seemingly purposefully - for minorities)
  • statements by lawmakers which said that they knew that the law will have a discriminatory impact
  • the fact that the law was purposefully designed to not reduce that impact
  • the fact that the law will have little or no impact on voter fraud (eg absentee voting - which is more susceptible to fraud and favored by white voters - was not changed)
  • a history of voter discrimination and invalid voter fraud claims

Some lawmakers have also directly stated that the goal of voter id laws is not to fight (practically non-existent) voter fraud, but to discriminate against minorities who are unlikely to vote for them:

A handful of the GOP Senators were giddy about the ramifications and literally singled out the prospects of suppressing minority and college voters.

tim
  • 37,031
  • 15
  • 102
  • 133
  • Recommend referencing this Dallas News article instead of the Washington post, because they've included the Courts decision. http://www.dallasnews.com/news/politics/2016/07/20/appeals-court-strikes-texas-voter-law – Drunk Cynic Mar 09 '17 at 19:31
  • 11
    Citing the number of cases as proving the lack of incidence is begging the question. – chrylis -cautiouslyoptimistic- Mar 09 '17 at 22:43
  • 9
    @chrylis less than two cases a year is pretty negligible considering the population of Texas. – tim Mar 09 '17 at 22:50
  • 33
    @chrylis yes, it begs the question why people still use voting fraud as an excuse as they can't prove it's a problem. –  Mar 09 '17 at 23:25
  • 3
    As an outsider (non-US), I always hear about the alternative photo ids that black people / other minorities use - which are these? – SBoss Mar 10 '17 at 07:46
  • 3
    @SBoss public assistance IDs, state employee ID cards, Veterans Affairs IDs (which were not accepted in North Carolina), private employee ids, student ids, or possibly credit/debit cards, health insurance cards, or other cards issued by private organizations; or they may only have non-photo ids. States may provide "free" ids, but they may be (purposely) difficult to get. – tim Mar 10 '17 at 09:50
  • @tim Great addition, could you add that to the answer? I'm somewhat conflicted - if it were any other country I'd say only allow government issued IDs, but in the US it seems like those don't exist for most people. Allowing private organizations to verify voters seems impossible to do. – SBoss Mar 10 '17 at 11:02
  • 6
    The law was enacted in 2011, and the cited numbers cover the decade previous to it, @chrylis. There's no question begging. – jscs Mar 10 '17 at 11:59
  • 5
    @tim: I still don't get how people are confident that hey caught all cases of voter fraud; from my perspective there could have been thousands of undetected cases before the ID requirement that now did not try it because of it. – PlasmaHH Mar 10 '17 at 12:45
  • 11
    @PlasmaHH There could. There could be millions or billions, who knows. The thing is that one should have at least some evidence when disenfranchising a considerable amount of people, especially when the potential fraud seems unlikely (there is little benefit, one needs at least some identification to vote even without the law, and one needs to hope that the registered person does not show up). And one should try not to specifically target a certain group of people - a group that is unlikely to vote as one would like, and a group which has historically been targeted by similar measures. – tim Mar 10 '17 at 13:16
  • 3
    @tim: sure, but why is "there were only 18 cases known" an argument against ID while "there might have been an unknown amount" is not an argument for ID? The existence or not (unless someone comes up with a magical way to get the numbers) should not be an argument for either side, as it is currently impossible to get a useful number. – PlasmaHH Mar 10 '17 at 13:46
  • 12
    @PlasmaHH It's not an argument against IDs per se, it's an argument against an argument for IDs. Some people claim - without any proof - that there are millions of cases. The only data we have are the known cases though, and it's important to work with known facts, not with assumptions without any underlying evidence. Based on the facts, we know that there is more fraud in absentee voting than in in-person voting (this isn't surprising either). If one genuinely wants to fight fraud, it would make sense to start where there is a larger - although still not that serious - issue. – tim Mar 10 '17 at 13:52
  • 4
    @PlasmaHH There are a number of statistical approaches designed to adjust for concluding the number of incidents that occur despite an inability to sample every incident, statiticians need these since your never see everything. There are numerous well known approaches used specifically for concluding likely crimes uncaught based off of caught crimes, as this is a question that comes up in numerous statistics. Without going into all of them if 18 are caught that suggests no more then 50-100 crimes in total even with very generous assumptions. – dsollen Mar 10 '17 at 14:03
  • 2
    @tim: I only ever see it used like this: "We should use ID to prevent fraud" ... "but there have been only 18 cases of fraud that could have been prevented by ID" and imho that is wrong, since the number is unreliable. The right way to answer to that request is "If you show evidence that ID is able to prevent significant numbers of fraud, then we can talk about doing it". It can't be "an argument against an argument" since itself is unreliable. Its like saying "I refute your made up number claim by using my own made up numbers" – PlasmaHH Mar 10 '17 at 14:03
  • 4
    @dsollen: I am familiar with such approaches, however they have some prerequisites like people actively looking/reporting. Voter fraud is largely something that is discovered more or less accidentally. In countries like germany there have been hundreds and sometimes thousands of cases of likely or confirmed voter fraud (or attempts), despite requiring an ID. I think the number 18 is very unrealistic because nobody is really trying to look, or even has ways to check things. imho the only thing ID would change is the way voter fraud is done, not the overall amount. – PlasmaHH Mar 10 '17 at 14:20
  • 1
    @PlasmaHH cite your source? Though i would prefer a source that wasn't Germany in any case, as there is a very real possibility that the government is encouraging voter fraud for Putin, and I don't think that republicans are claiming that in the US. In any case Voter fraud is hard to get away with, since there person you pretend to be can vote proving the fraud, and frankly individual criminals are idiots that usually give themselves away (ask any police officer!). Combined with the ineffectiveness of in-person fraud & Occam razor and there no reason to presume unproven fraud. – dsollen Mar 10 '17 at 15:11
  • 3
    @PlasmaHH "however they have some prerequisites like people actively looking/reporting. Voter fraud is largely something that is discovered more or less accidentally" Not quite. They actually do regularly check into this type of thing and run audits, studies, etc. – Shane Mar 10 '17 at 15:18
  • 4
    @dsollen: I am not quite sure how the german government would encourage vote fraud for putin as afaik there is no way to vote putin there, sounds like an intresting conspiracy theory.... anyways there are not much resources available in english, so if you have a good translation plugin, googling for "deutschland wahlbetrug" will give a lot of result, you just have to filter the crooks and conspiracy theorists yourself – PlasmaHH Mar 10 '17 at 15:55
  • 8
    @chrylis That is not begging the question. The burden of proof is upon those who put forth the premise. You don't ask me to prove that there isn't a pink dragon in your garage. – RomaH Mar 10 '17 at 19:35
  • @RomaH I do if I'm considering buying a house from you, and I'd have to assume care of whatever vermin/"pets" come with it. If there is evidence that a pink lizard lives in the garage, I'd need to buy a cage, food, etc. But if not, I need not spend money on pet supplies. Likewise, if there's no evidence of significant voter fraud, a state need not spend tax money on prevention measures. – Damian Yerrick Mar 13 '17 at 03:19
  • https://arstechnica.com/tech-policy/2018/08/georgia-defends-voting-system-despite-243-percent-turnout-in-one-precinct/ 12 in 10 years??!?!? – dolphin_of_france Jul 11 '19 at 21:04
  • @dolphin_of_france The question is specifically about in-person voter fraud. The linked article is about electronic voting machines. The difference is important, because Republicans specifically target basically non-existing in-person voter fraud to disenfrancise black and poor people. They do not care about other, more common forms of voting fraud, because trying to prevent those would not prevent black and poor people from voting. – tim Jul 13 '19 at 09:06
23

Also, if the Photo ID voting law was in place for decreasing voter fraud, is it working? If it is, how well? Statistics please. I can't seem to find the answer I'm looking for.

The problem is that without the law, there is no way to detect the fraud unless people vote twice as the same person. With the law, people aren't allowed to vote without ID matching the voting registration, so there is no fraud. Either way, there's very little or no statistical proof that fraud has been reduced.

The typical story after the law is that someone shows up, gets asked for ID. They can either produce it or say that they don't have it with them. If they say they don't have it, they leave and nothing is recorded. If they produce it, the electors look for a matching registration. If there is one, they check if that person has voted. If the person had already voted, they'd send the person away and nothing is recorded. If the person has not voted, they'd let the person vote. If there is no registration, they might give the person a provisional ballot. If the ID doesn't match the current location, they'd send the person to the correct location and nothing would be recorded.

Unless the person fills out a provisional ballot in one location and votes regularly in another location, nothing is recorded that looks like fraud. And provisional ballots clashing with actual ballots normally isn't counted as fraud but as user error.

Under the system without the ID check, if the person can describe the registration properly, the person is allowed to vote if that registration hasn't already voted. There is only a clash if someone has already used the registration to vote. Or if someone later uses the registration to vote. That's the only time that fraud would have been detected. If there is no clash, then there is no evidence of fraud. Successful frauds will not be detected, only unsuccessful ones.

It is much easier to prevent fraud before it happens than detect it after. The only way to detect fraud after the election is to prove that the registration wasn't valid at the time of the election. So if the person on the registration was dead, a felon, or a non-citizen, that registration could be noticed afterward. If the registration is for someone who is eligible to vote but doesn't actually vote, that would never be seen in any statistics if someone else cast that vote fraudulently.

Brythan
  • 89,627
  • 8
  • 218
  • 324
  • 12
    If the reason was really fraud prevention, why is a license to carry a concealed gun valid, but a student id not? Both should be perfectly capable of preventing fraud that may (or may not) exist. – tim Mar 09 '17 at 21:43
  • 4
    Your description of the system without the law also seems wrong. It isn't enough to describe the registration. A person actually needs a registration certificate. If they do not have that, they need any id (not one of the seemingly arbitrary restricted ones in the law), or some other document identifying them (utility bill, bank statement, etc). – tim Mar 09 '17 at 21:48
  • 6
    @tim Not that crap again. That's because a carry license is a state-issued ID while a student ID isn't. – chrylis -cautiouslyoptimistic- Mar 09 '17 at 22:44
  • 3
    @chrylis How is a student id issued by a state school not a state-issued id? And why are other state-issued ids which contain photographs - such as amateur radio licenses or press passes - not covered? – tim Mar 09 '17 at 23:03
  • 20
    @tim State-school IDs are not held to anything resembling formal standards; there's at least one I know of where I could just walk in, and the printer's unattended. They almost never have residency information printed. And then you're just making things up--states don't issue ham licenses, FCC ham licenses don't have photos, and press passes are done in a distributed manner. All general-purpose state ID cards go through the DPS. (In the case of an LTC specifically, it requires another state ID already and uses the same picture and vital information from it.) – chrylis -cautiouslyoptimistic- Mar 09 '17 at 23:09
  • 25
    This is just false. You don't need the law to track the problem--if indeed there was one. Regardless, this ignores the much bigger point that in-person voter fraud is a ridiculous concept. No one is going to sway elections by having people drive one-by-one to different districts to cast one vote under the guise of some person that they hope doesn't show up to cast their own vote. Given that republicans have explicitly stated that they push for these laws to disenfranchise voters, arguing otherwise is just disingenuous. –  Mar 09 '17 at 23:28
  • 1
    @chrylis nor would a state-issued drivers license help with the residency issue given that students can vote where they go to school. Anyways, it's a stilly argument to nitpick all of these aspects. That they have to be nitpicked to justify them kind of points out the problem. –  Mar 09 '17 at 23:30
  • 17
    It's been widely repeated that the rarity of detected fraud implies that actual fraud is rare, but what basis is there for estimating the fraction of fraud that is detected? If 10% of fraud is detected, that would imply that actual fraud is rare. If 0.0001% of fraud is detected, it could be quite common. Establishing that any fraud that was attempted would likely be caught would seem to have value, whether or not any fraud was actually attempted. – supercat Mar 10 '17 at 00:19
  • 20
    The idea that significant numbers of people would commit a felony to get an extra vote or two is a pretty extraordinary claim that requires some decent proof to support. – ceejayoz Mar 10 '17 at 03:33
  • 2
    @supercat not at the expense of disenfranchising so many. –  Mar 10 '17 at 14:36
  • 1
    @supercat statisticians have addressed these issues came up with what the consider good approaches. In this case there is very good reason to expect a high percentage of frauds to be caught. Since you have no way to know if a voter is going to vote every fraud vote has a non-trivial chance of being detected when the voter actually votes. Furthermore individuals doing this on their own (not a larger conspiracy) are likely to be caught due to the 'amateur criminal' tendency, to be frank they usually do idiotic things to draw attention. More then 10% are likely caught – dsollen Mar 10 '17 at 14:55
  • 2
    @supercat also, from a statistical standpoint if fraud happens it likely happens for both parties. If both sides have an equal amount of fraud it cancels each other out. Thus what matters is only what percentage of these theoretical fraud cases happens to lean towards one party over another. If 45% of fraud is for party X and 55% is for party Y then only 10% of fraud causes have any impact on the election. There is no reason to expect one party to have substantially more fraud then the other, so impact of theoretical fraud is even smaller. – dsollen Mar 10 '17 at 14:59
  • 3
    @dsollen: Is there any mechanism by which an increase in fraud favoring one party would lead to a corresponding increase in fraud favoring the other? If not, they might happen to be roughly equal but there's no particular reason to expect them to be. Note also that there's a huge difference between being caught and stopped from voting, versus being caught, recorded, and prosecuted. If the most likely outcome of someone being caught would be that they would leave with no trace of the event being recorded or counted, there's no basis for estimating how often or rarely that occurs. – supercat Mar 10 '17 at 15:29
  • 3
    @supercat That's not the question I begged. The question is rather, there is reason to believe that more would choose to commit fraud for one party then the other, not about increase in fraud. Everything being equal and assuming 50% of individuals that care enough to vote support each party (which roughly they do or the election is already won regardless of fraud) then there is no reason not to believe an equal number of each side would have the inclination to fraud. Presuming fraud not proven, and then that the fraud is predominately for one side over the other, violates Occam's razor – dsollen Mar 10 '17 at 15:39
  • 1
    @dsollen That's only true for simple majority votes with two options. The rest are affected by the full fraud. And in other matters, there is zero reason whatsoever to enact such a ridiculously inefficient and costly method of cheating when there are obscenely better ways of cheating. It is orders of magnitude easier to falsify absentee ballots than in-person voting. Absentee votes can be automated. In-person requires an actual person spending non-trivial amounts of time during a limited voting window for every single vote. – zibadawa timmy Mar 10 '17 at 22:53
9

The other answers seem unnecessarily focused on racial discrimination. It is perhaps more useful to focus on class distinctions and a bias for sensational anecdotes.

People should have to prove who they are when they vote in order to maintain democratic equality, i.e. one person one vote. What could be wrong that? A surprising amount.

Voter fraud, as pointed out in the other answers is rare, but it is also easy to imagine. In other words, it has availability (in a heuristic sense) and any actual incidents (irregardless of probability) serve as confirmation (bias) that voter fraud is a problem that needs solving.

As for the discrimination aspect, a useful analogy is perhaps anti-smoking laws. They are not explicitly classist, they are about public health. But their effect is entirely classist, because the vast majority of smokers in the US are found in a particular socio-economic group. Does that make them discriminatory? Depends on whether you define such by intention or effect.

Likewise, voter-id laws are probably not intentionally discriminatory (although pretending so apparently makes for a good news story) but they place a burden disproportionately on underclass persons, irrespective of those intentions of the authors/voters.

This whole business is another one of those semantic debates. Person A says that discrimination is about intent and thus the laws are not discriminatory. Person B says that discrimination is about effect and thus they are. When you pile that on top of cognitive bias and implicit class prejudice ("of course everyone can get to and wait in line 2 hrs at the DMV in the middle of the day") you get all this mess.

Jared Smith
  • 8,637
  • 4
  • 26
  • 45
  • 7
    The North Carolina law was intentionally discriminatory. Regarding the Texas law, there is no hard proof that lawmakers went out to discriminate (although that too is a fair assumption), but they did know that the law would have a discriminatory effect and they decided not to do anything about it. I agree though that it is not only a race issue, but also a class issue. – tim Mar 10 '17 at 17:18
  • 4
    Once states make it free and easy to get the required ID, then Voter ID laws stop being a form of illegal poll tax. If you have to spend hundreds of dollars and dozens of hours off work to get the required ID, it becomes a challenge. – arp Mar 11 '17 at 17:05
  • @arp yes indeed, free... and relatively easy to acquire. – Jared Smith Mar 11 '17 at 21:49
  • 2
    A good attempt but simply incorrect. These laws have been shown to be explicitly discriminatory. They are intentionally discriminating. Same thing is happening with jerrymandering: http://www.motherjones.com/politics/2017/03/texas-gerrymandering-ruling-unconstitutional –  Mar 12 '17 at 22:41
  • 1
    @blip of course some of the laws are intentionally discriminatory. The part I object to (and I apparently didn't make this clear enough) is that one cannot assume intentionality on the part of the authors: it is entirely possible to create legislation in good faith that is discriminatory in its effect (again, anti-smoking laws). – Jared Smith Mar 13 '17 at 12:55
  • 1
    @jaredsmith in general, that is true. But that's not what happened with these particular laws. –  Mar 13 '17 at 16:21
7

There has practically not been any cases of voter-fraud, so at best this is a solution to a problem that doesn't exist. Yes, supposedly some dead people have voted, a study done by the DMV in South-Carolina is often cited - but that turned out to be people with family-names eg. a son with the same name as his father... and on a few occation, absentee-voters who died between casting their abesntee-vote and the election. Washington Post, NPR, ThinkProgress

There have also be cases where Republican politician have admitted on camera to why they want these laws, because some groups - like poor blacks - are unlikely to have them, and because these groups usually vote for Democrats.

You should be able to prove who you are when you register to vote - not when you're actually voting.

Besides, it's a pretty bu..-sh.. way of influencing the election; standing in line for hours, to cast maybe one or two extra votes - which probably aren't going to influence anything anyway... and risk a huge fine.

Jon Oliver at Last Week Tonight had an episode about this "problem"... you should check it out:

https://www.youtube.com/watch?v=rHFOwlMCdto

Baard Kopperud
  • 1,516
  • 10
  • 9