"Yes, I am D. B. Sweeney, don't you recognize me? Here's my ID card."
November 15, 2013 8:09 AM   Subscribe

About as many people say they’ve been abducted by space aliens as say they’ve committed voter fraud One of the findings of a new working paper by John Ahlquist, Kenneth R. Mayer and Simon Jackman is that “the lower bound on the population reporting voter impersonation is nearly identical with the proportion of the population reporting abduction by extraterrestrials.” Roughly 2.5 percent of the population effectively admit to one or the other.

From the paper:

"Virtually all the major scholarship on voter impersonation fraud – based largely on specific allegations and criminal investigations – has concluded that it is vanishingly rare, and certainly nowhere near the numbers necessary to have an e ffect on any election … To give one idea of the scale: a review of allegations in the 2008 and 2010 elections in Texas found only four complaints of voter impersonation, out of more than 13 million votes cast, and it is not clear whether any of the complaints actually led to a prosecution (Minnite, 2013:101). By contrast, the 2000 presidential election almost certainly was altered by poor ballot design in Palm Beach County, which resulted in at least 2,000 voters who intended to vote for Al Gore and Joe Lieberman casting their ballots for Pat Buchanan by mistake (Wand et al., 2001)."
posted by MisantropicPainforest (40 comments total) 8 users marked this as a favorite
 
If 2.5% "admit" it, does that mean more people have probably done it?

Having trouble understanding what this means.
posted by grobstein at 8:20 AM on November 15, 2013


If 2.5% "admit" it...

Read the article.
posted by 0xFCAF at 8:24 AM on November 15, 2013 [8 favorites]


And in the center of that Venn Diagram lies my main character.
posted by robocop is bleeding at 8:27 AM on November 15, 2013 [10 favorites]


Thanks.
posted by grobstein at 8:27 AM on November 15, 2013


There has to be some percentage of the population to will "admit" to essentially anything just to fuck with the survey.
posted by LastOfHisKind at 8:27 AM on November 15, 2013 [6 favorites]


Really, folks, RTFA:
There’s some internal evidence to suggest that they are just being careless. But to be sure, Ahlquist, Mayer and Jackman carry out a second set of experiments, which is where the extraterrestrial abductions come in. It is probable that Americans are not being kidnapped en masse by space aliens eager to conduct experiments on their sensitive bits. So, Ahlquist, Mayer and Jackman do a second list experiment in which members of the control group, instead of being asked about voter fraud, are asked whether they have been abducted by aliens. If people answer yes to all questions with roughly the same proportions in both the space alien abduction experiment and the voter fraud experiment, then this suggests either that people are rushing through the surveys carelessly, or that alien abduction is a much bigger problem than any of us knew. And this is indeed what the authors find:
The implication here is that if one accepts that 2.5% is a valid lower bound for the prevalence of voter impersonation in the 2012 election then one must also accept that about 2.5% of the adult U.S. population – about 6 million people – believe that they were abducted by extra-terrestrials in the last year. If this were true then voter impersonation would be the least of our worries.
posted by zombieflanders at 8:32 AM on November 15, 2013 [3 favorites]


Yeah, this is a poorly written post.

The authors are not claiming that 2.5% is actually a good lower bound for the incidence of either voter fraud or alien abduction. They are saying that if we take the naive interpretation of their list methodology, and treat 2.5% as a lower bound on voter fraud, then we also would have to take 2.5% to be a lower bound on alien abduction in the last year. Since that would be absurd, we can't conclude that 2.5% actually committed voter fraud.

Implicitly, voter fraud is actually much rarer than 2.5%, although the article doesn't give us the tools to guess how much.

So although 2.5% would be enough people to swing elections, we don't have to worry about this, because the true figure is probably way smaller.
posted by grobstein at 8:34 AM on November 15, 2013 [8 favorites]


Please read the article. Thanks.
posted by MisantropicPainforest at 8:38 AM on November 15, 2013


Yeah, but how many aliens have committed voter fraud? That's the number they DON'T WANT YOU TO KNOW!
posted by blue_beetle at 8:40 AM on November 15, 2013 [12 favorites]


me: Yeah, this is a poorly written post.

The authors are not claiming that 2.5% is actually a good lower bound for the incidence of either voter fraud or alien abduction. They are saying that if we take the naive interpretation of their list methodology, and treat 2.5% as a lower bound on voter fraud, then we also would have to take 2.5% to be a lower bound on alien abduction in the last year. Since that would be absurd, we can't conclude that 2.5% actually committed voter fraud.

Implicitly, voter fraud is actually much rarer than 2.5%, although the article doesn't give us the tools to guess how much.

So although 2.5% would be enough people to swing elections, we don't have to worry about this, because the true figure is probably way smaller.


MisantropicPainforest: Please read the article. Thanks.



My comment is (obviously) based on my reading of the article.

If what you are trying to express is that you think my reading of the article is incorrect, you will have to spell that out.
posted by grobstein at 8:43 AM on November 15, 2013


No no BB it's not the voters who are aliens, it's the elected officials!

Also grobstein and zombieflanders are correct, the point of the article isn't that 2.5% of voters commit fraud it's that a significant percentage of people are careless when filling out surveys, making things like voter fraud incidence more difficult to assess.
posted by Wretch729 at 8:56 AM on November 15, 2013 [1 favorite]


The alien abduction thing is actually a classic example of the problem of addressing very rare events by conventional random-sampling methods. The "errors" all go in one direction (no one who has in fact been abducted by aliens fails to mention the fact) and unless your sample size is enormous you end up wildly overestimating the prevalence of the problem. Every survey-based study on how often people have used guns to ward off Bad Guys suffers from this kind of error.
posted by yoink at 9:02 AM on November 15, 2013


LastOfHisKind: "There has to be some percentage of the population to will "admit" to essentially anything just to fuck with the survey."

Fine! I admit it! I do this!
posted by Big_B at 9:03 AM on November 15, 2013 [3 favorites]


Interesting survey design, I hadn't heard of this technique. The article has a good explanation of the process, so anyone who's confused should read the whole thing first.

MoonOrb, the suggestion is that 2.5% is the false positive rate. There were some people who would mark down that they'd done all the actions on the list, presumably without actually reading them. How do you figure out which people are actually going by the instructions? You put a dummy item, alien abduction, on the list. If we assume that no one has actually been abducted by aliens in the last year (a pretty safe assumption), then the number of people who mark that they have should be the same as the number of people who aren't reading the questions, or are giving joke answers, or some other reason. Since alien abduction and voter fraud have a similar response rate, we can conclude that they are about equally likely.
posted by echo target at 9:04 AM on November 15, 2013 [1 favorite]


Since alien abduction and voter fraud have a similar response rate, we can conclude that they are about equally likely.

Hmm-mm, hmmm-mmm. So if I hear you right, you're saying that illegal aliens are stealing our elections?
posted by yoink at 9:08 AM on November 15, 2013 [1 favorite]


Please read the article.

Then the aliens would win!
posted by Brandon Blatcher at 9:08 AM on November 15, 2013


"Probe early, probe often." - Al Capone Galactic Vice Proctor Epsilon-9XZ/6
posted by Blazecock Pileon at 9:10 AM on November 15, 2013 [1 favorite]


Wretch729: "No no BB it's not the voters who are aliens, it's the elected officials!"

FALSE. The elected officials are Lizard People.
posted by Hairy Lobster at 9:14 AM on November 15, 2013


The problem is, more than 0% of your respondents will sincerely believe, without any need of survey prompting, that they have in fact been abducted.

So, are people more or less likely to falsely believe they have been abducted, as opposed to delusional belief that they have committed voter fraud? If they are more likely to believe in delusional abduction rather than delusional fraud, your baselines are different.
posted by idiopath at 9:15 AM on November 15, 2013


Since alien abduction and voter fraud have a similar response rate, we can conclude that they are about equally likely.

If you mean the psychological experience of alien abduction, sure, yeah, lots of people get that, you can get it out of a bottle of cough syrup. If you mean actual aliens pulling people up in a tractor beam, well, grats on having something clever to say, but...
posted by This, of course, alludes to you at 9:15 AM on November 15, 2013


Pony request: we should have a mechanism for elevating a comment to be an addendum to the main post, both echo target and MisantropicPainforest comments explain the point of the actual paper.

Now is there any actual research about why voter fraud is a trivial non-issue in the USA? In my town you show up and give a street, house number and name and get a ballot. Was joking about seeing how many times we could vote at a recent nearly empty primary, suggesting a drive around, the cynical opinion was to just walk back to the same old guys and give a different name. I sortof think it'd be cool to walk around with an stained index finger all day.
posted by sammyo at 9:17 AM on November 15, 2013 [1 favorite]


Not too sure what you're alluding to, This, but if you've spotted a flaw in the survey design, by all means tell us what we're missing.
posted by echo target at 9:20 AM on November 15, 2013


(perhaps a "this is what I really mean) flag, MisantropicPainforest was much clearer in that comment. (and this is as love from an essayist challenged person)

Because it seems like quite a good article, or an article with an important point, that I would have passed by without seeing the comments.
posted by sammyo at 9:21 AM on November 15, 2013


Now is there any actual research about why voter fraud is a trivial non-issue in the USA?

The Brennan Center's The Truth About Voter Fraud is a good place to start, and they've got a whole section devoted to voting rights. Mother Jones also did a set of pieces on voter fraud vs. suppression a couple years back.
posted by zombieflanders at 9:21 AM on November 15, 2013


I think one problem with this sort of self-reported data is that respondents are likely to have a restricted view of what "voter fraud" means. They likely think it means stuffing ballot boxes or pretending to be someone you are not. Sure, those things count, but that's not all it is.

A small part of my practice is elections law and I get pretty busy in November. The reports I get from precincts is often shocking. Every election, for example, busloads from homes from the mentally disabled show up at the precincts and their caretakers "help them vote". This is a pretty clear violation of the election code. Any person who needs assistance in voting can receive e.g. an illiterate person can have the ballot read to them, but not anyone can be that assistant. Certain people of undue influence, such as employers and union reps, are barred from "assisting". But, we don't want to touch that one because of the stink it would raise, so we let it slide. Worse are people who hang around outside the polling place and offer to "help you vote".

Regarding electors of questionable eligibility, no one gets turned away on election day. Rather, they vote a provisional ballot, although like absentee ballots, they generally don't come into play in tipping an election because they are such a small percentage of total ballots cast. If you vote provisionally, you can find out if your vote was counted within 30 days after the election.
posted by Tanizaki at 9:22 AM on November 15, 2013 [1 favorite]


And a "pony request" for the scientific community, as echo asked if there were problems with the survey design, it seems like there should be some kind of statistical oversight community that can vet studies results. Not exactly a peer review but a meta review that can say that these numbers in this formula are valid. Now that's about as handway as is possible but as much as the article sounds totally credible and match my personal biases, is there real meaning there?
posted by sammyo at 9:31 AM on November 15, 2013


I think one problem with this sort of self-reported data is that respondents are likely to have a restricted view of what "voter fraud" means.

All violations of the election code are not voter fraud. Voter fraud, in the context of many attempts by people to restrict access to voting via voter ID cards, is a pretty clear thing that social scientists are trying to measure. It turns out that voter fraud--the things like John Fund talks about--doesn't exist.
posted by MisantropicPainforest at 9:35 AM on November 15, 2013 [2 favorites]


A small part of my practice is elections law and I get pretty busy in November. The reports I get from precincts is often shocking. Every election, for example, busloads from homes from the mentally disabled show up at the precincts and their caretakers "help them vote". This is a pretty clear violation of the election code.

Not really. There are a number of differing state laws that revolve around "mental incompetence," but those state laws and federal law generally allows people to designate their caretakers (among others such as family) to help them vote.
posted by zombieflanders at 9:42 AM on November 15, 2013 [1 favorite]


And in the center of that Venn Diagram lies my main character.

That's an UFO and not a Venn Diagram.
posted by Pyrogenesis at 9:42 AM on November 15, 2013


grobstein: "The authors are not claiming that 2.5% is actually a good lower bound for the incidence of either voter fraud or alien abduction. They are saying that if we take the naive interpretation of their list methodology, and treat 2.5% as a lower bound on voter fraud, then we also would have to take 2.5% to be a lower bound on alien abduction in the last year. Since that would be absurd, we can't conclude that 2.5% actually committed voter fraud.

Implicitly, voter fraud is actually much rarer than 2.5%, although the article doesn't give us the tools to guess how much.
"

I buy the first part but not the second. I'm not trying to argue that voter fraud is common (all research suggests that it's extremely uncommon indeed in the USA, and that when it does happen it's done systematically and by exactly the same groups who are using voter fraud as an argument for legislation that would disenfranchise poor and minority voters) but just because we've established a false positive rate doesn't mean that we know anything about the false negative rate. They're separate things and require separate study designs to uncover.

This study tells us that at least 2.5% of respondents will likely yield a "yes" answer to any study designed like the one linked, even when the true positive rate is zero. We can see that just by looking at the second part of the study, the one about alien abductions (assuming that we believe the true positive rate for alien abductions is 0%). It actually weakens their conclusion to see that they got a false positive rate of 2.5% both for voter fraud and for alien abduction, because voter fraud has a non-zero (though small) true positive rate while alien abduction's true positive rate is zero. If they're taking all "yes" answers as false positives, then they're making an error there because they don't know what the true positive rate is for voter fraud. Also it's quite possible that people who believe they've been abducted by aliens are much more willing to admit it than are people who've committed voter fraud, such that the two questions aren't really comparable.

I also question the generalizability of this study. Was the survey design that they used (a list of innocuous questions for both groups, with the treatment group's list having a single question thrown in about voter fraud) one that is commonly used to try to assess the rate of illegal activities? The linked article (no time to read the original paper) is a bit ambiguous about that. It says that those sorts of studies have been used before, but not how often or whether they're used to determine true positive rates or only in studies like this one that are more interested in false positive rates. How much is already known about the accuracy of those types of surveys? How much is now about the degree of reticence that people have about admitting different types of illegal acts? These are unanswered questions (in the Post article anyway) that make me question the generalizability of this study to determining true vs. false positive rates for illegal acts in general or voter fraud in particular.

I feel this is one of those studies that appears interesting on the surface but does not in reality reveal very much useful information. It's already well-established that there's a baseline false positive rate for most self-report surveys somewhere in the low single digits, and I'm not sure what this study adds to that.
posted by Scientist at 9:44 AM on November 15, 2013


There were actually some credible reports of voter fraud in and around NYC this year, though not from anyone the Republicans are worried about, and more about local power mongering than R vs. D.
posted by save alive nothing that breatheth at 10:02 AM on November 15, 2013


Plot Twist: Aliens are kidnapping people and brainwashing them into committing voter fraud in order to sabotage the space program. So far the plan has been going just swimmingly.
posted by teh_boy at 10:06 AM on November 15, 2013


The reason there is so little voter fraud of the kind that the Republicans' vote-suppression programs are premised upon (i.e., voting under assumed identities) is because it's a high-risk, low-reward strategy, and the risks increase geometrically if you try to increase the likelihood of a reward. To have it actually tip an election you'd need either for the election to be one of those massively rare events in which one or two individual votes tips the balance, or you have to do it on a large enough scale that the probability of being discovered becomes overwhelmingly large.

Of course, nobody in fact thinks voter fraud of this kind is a real problem that actually needs combating. The only "gain" Republicans see in pushing this agenda is not "cleaner" elections but a more Republican-skewed electorate.
posted by yoink at 10:14 AM on November 15, 2013 [1 favorite]


This is quite damning evidence that aliens are actively meddling in our elections by posing as dead voters. What are the odds of those two numbers converging exactly? Out of an infinite number of possible numbers? (Do I get my own newsletter yet? Or a Fox News special correspondent gig?)
posted by saulgoodman at 10:17 AM on November 15, 2013


If you mean actual aliens pulling people up in a tractor beam, well, grats on having something clever to say, but...

Ludicrous. Tractor beams are used for seizing star ships, not people.
posted by Marisa Stole the Precious Thing at 10:35 AM on November 15, 2013


FALSE. The elected officials are Lizard People.

Who are from ALPHA DRACONIS. It's a whole other star system, ergo aliens.
posted by Hoopo at 10:48 AM on November 15, 2013


Additional evidence that voter fraud in the United States is a non-existant phenomenon is that the Republicans who want to restrict voting access with new ID laws regularly admit that voter fraud doesn't actually happen.

Pennsylvania admits it: no voter fraud problem
posted by obscure simpsons reference at 11:46 AM on November 15, 2013


You know what always bothered me about the movie Independence Day? Poor Randy Quaid's character. He's been claiming to have been abducted by aliens for years and it's ruined his life. Then, a WHOLE SHITLOAD OF GIANT ALIEN SHIPS SHOWS UP and people still shake their heads and roll their eyes at him when he tells his story. Poor guy can't win. No wonder he's an alcoholic.
posted by mr_crash_davis at 2:18 PM on November 15, 2013 [1 favorite]


I'll repeat my personal anecdote about being pressured by my Republican Womens Group mother to volunteer at Nixon's 1972 re-election campaign (even though I was still months away from legal voting age), and because I had a car - and probably looked young and dumb - I was recommended to join a group that was going out to Nursing Homes and "help" the residents who "couldn't do it themselves" fill out Absentee Ballots. I respectfully declined, but I came to the conclusion then that this was why you always heard about Republicans getting more Absentee Votes. Every exposure I've had since then with members of both Major Parties have reinforced the impression that ever since the 1960s Chicago embarrassment, the Democrats have pretty much given up on Voter Fraud while the Republicans have embraced the practice enthusiastically.

The Democratic Party is a corrupt institution; the Republican Party is a criminal enterprise.
posted by oneswellfoop at 2:48 PM on November 15, 2013 [4 favorites]


oneswellfoop — sounds exactly like the situation in Canada, where the Harper Conservatives have been found to have repeatedly broken all sorts of election laws.
posted by five fresh fish at 11:21 PM on November 15, 2013


« Older Hawaii Winter Surf Season Opens With Tragedy   |   The Swants Dance Newer »


This thread has been archived and is closed to new comments