« Modern farming ain't all bad | Main | Weird analogy »
October 16, 2008
Yeah, but who is counting
But for all of the McCain campaign’s manufactured fury about vote theft (and similar claims from the Republican Party over the years) there is virtually no evidence -- anywhere in the country, going back many elections -- of people showing up at the polls and voting when they are not entitled to.The Acorn Story
This reminds me of a project I did as an undergraduate on the Historical Origins and Difficulties in Statistical Modeling of Fingerprinting. Many had claimed that since no two people had ever been found with identical fingerprints, that there wasn't a problem with the fingerprinting system. I said:
"The theoretical grounding for the uniqueness of fingerprints is disappointing, and poorly grounded. The vast majority of arguments for the uniqueness of fingerprints fall into two categories. First, is the theological argument that G-d never creates any two identical things (or the slightly more scientific argument that we can trust the random processes involved to combine in an unique manner or that all objects are unique when examined in sufficient detail.) The other attempts to argue from evidence, since no one has ever seen a match then there must not be any. The problem with this argument is that we cannot extend it to say that no two fingerprints have ever matched. None of the major classification systems would detect if two fingerprints on two sets of prints matched but the rest did not. Therefore, we do not really know if no fingerprints match. Unfortunately, juries have sent criminals to their death by reaching this false conclusion"
It turned out this wasn't quite true, it could be that if they matched on the very first finger of the classification system that you would find a match on that finger. But one could have a later finger in the filing system that doesn't match, and at least in theory your pointer finger of left hand could match the print on the pinkie on my right hand.
I'm reminded of this story because no one is counting fraudulent registration in a way that would indicate if it were a problem. Given the simplicity of voter fraud (just register to vote in multiple districts under different names) one would imagine it is easy to avoid detection. If you tried to vote as another real person and they had already voted (or were to try to vote that day), you made a fuss or dressed distinctly to be detected, or tried to vote multiple times at the same polling place then you'd be noticed. But you can't infer the size of the undetected fraud from the size of the detected fraud. The ones we detect probably are not representative nor do we know the attempt rate. Therefore, we are not measuring the true extent of registration fraud and the measurements of what we do catch don't describe the extent of the problem.
Posted by OneEyedMan at October 16, 2008 10:27 PM
Comments
While your argument is valid from a theoretical standpoint, I believe you've missed the point of the article you've quoted. While it may be possible, even likely, that some degree of voter fraud has taken place, the groups claiming it have failed to prove that it has occurred and, more importantly, that it is a serious problem, yet have used it as justification to implement measures that have demonstrably excluded legitimate voters from voting.
In the case of ACORN, the Republicans (including the McCain campaign) have been trying to make it out to be a huge conspiracy with the potential to bring about the collapse of our democratic system. In reality, it was just a bunch of hired canvassers being greedy/lazy and filling out forms with bogus information rather than doing the jobs they were being paid to do. As I understand it, ACORN itself was marking the forms as being potentially fraudulent before turning them in to the election commissions.
I heard an interesting fact mentioned on NPR a few days ago. The participant in a discussion of the ACORN debacle pointed out that the US is the only "major" democratic country that uses an opt-in system of voter registration. According to this person, most other countries automatically register their voters using available records (presumably, they have some mechanism to correct inaccuracies in those records, but the speaker did not elaborate on the issue). In theory, this dramatically reduces the risks of voter fraud and disenfranchisement, but I have no idea how difficult it would be to implement in this country.
Posted by: REggert ![[TypeKey Profile Page]](http://www.belligerati.net/nav-commenters.gif)
at October 21, 2008 9:27 PM
That is interesting about opt-in registration. I could see higher geographic mobility as a reason to not do this in the USA. Since people move around more here and there are elections determined not just by your state but county and street, this could lead to a lot of people registered to vote in multiple places or the wrong one.
Barriers to voting are not a problem to me. There doesn't seem to be compelling evidence that higher voter turnout makes for better decision making or a happier public, Since the most informed voters are also typically the most partisan and therefore most likely to be registered, I could see lowering barriers to voting as making for bad compromises and cheapening discourse.
By my reading the article is saying that the real fraud is cutting the valid voters out while worrying about bad registrations doesn't seem to be a problem. My point is that they have no idea if bad registrations are a problem because that isn't measured. You can't make a cost benefit analysis if you aren't measuring the costs and the benefits.
As for the huge conspiracy, I doubt there is one. However, if the registration method they employed accidentally registered large numbers of invalid voters that were more likely to vote Republican, I imagine they would have fixed that method a long time ago. They probably knew their method had problems and given the direction those problems was in their favor, they couldn't be bothered to fix it.
Posted by: TheOneEyedMan at October 22, 2008 8:04 AM
Post a comment
Thanks for signing in, . Now you can comment. (sign out)
(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)