Why that is whacky talk, we’ve evolved past blatant racism to the point where the conservatives and reactionaries have to at least come up with something that fails the giggle test in their efforts to disenfranchise the “wrong” types of voters.
Al Jazeera America looks into the cross-state database being used by quite a few Republican led states to detect “voter” fraud, and as a data professional, much less a liberal, I want to cry:
The Crosscheck list of suspected double voters has been compiled by matching names from roughly 110 million voter records from participating states. Interstate Crosscheck is the pet project of Kansas’ controversial Republican secretary of state, Kris Kobach, known for his crusade against voter fraud.
The three states’ lists are heavily weighted with names such as Jackson, Garcia, Patel and Kim — ones common among minorities, who vote overwhelmingly Democratic. Indeed, fully 1 in 7 African-Americans in those 27 states, plus the state of Washington (which enrolled in Crosscheck but has decided not to utilize the results), are listed as under suspicion of having voted twice. This also applies to 1 in 8 Asian-Americans and 1 in 8 Hispanic voters. White voters too — 1 in 11…
The sample matches he showed his audience included the following criteria: first, last and middle name or initial; date of birth; suffixes; and Social Security number, or at least its last four digits…
In practice, all it takes to become a suspect is sharing a first and last name with a voter in another state. Typical “matches” identifying those who may have voted in both Georgia and Virginia include…. [emphasis mine at points of WTFedness]
I’ve spent a good chunk of my career making sure different data systems can talk to each other correctly. One of the first things that I learned as a young minion to an old data scrubber was that names suck donkey dick as matching criteria. There are numerous issues with names. The easiest one is that a name is not a unique identifier. If you google Richard Mayhew, there is me, a professor of literature and an English football player on the first page of results. We are seperate people as you can easily verify by my spelling adventures and the fact that I am a referee and not a player. As AJA points out, Democratic leaning ethnic groups tend to have a higher concentration of people with common names.
More importantly names change. My wife Jane Mary Mayhew nee Doe can be found in a variety of databases under Jane Mary Mayhew, Jane M. Mayhew, Jane Doe, Jane Mary Doe, Jane M. Doe-Mayhew, J. Doe Mayhew. Her sister Joan Maria can be found in several database as J. M. Doe. My cousin Judy is in several databsae as J. Mayhew. You try and figure out whether J. Mayhew is a unique individual in multiple location of unique individuals in multiple, unique locations? You can’t with names.
Secondly, the next point of failure is that people move. Democrats and Democratic leaning voters tend to be more transient within a metro area as they are more likely to be either young or renters rather than middle age or old homeowners. So actual matches of someone voting in Precinct 1 in 2009 and Precinct 77 in 2011 is explained by natural movement. It is not a crime to move and vote at a new home.
It could be done with a reasonable degree of confidence (reasonable for say marketing purposes) on a match of first name, middle indicator, last name, gender (although that is fuzzy) date of birth, SSN or credit card numbers, but even that methodology will spit out some percentage of screwed up results that need expensive manual intervention to clean the list. The screw ups will happen because the source data list is seldom pristine. People will enter Richard Mayhew once and then Dick Mayhew another time or they could transcribe a number on their SSN, or Joe Smith, Joe Smith Jr, and Joe Smith III all live at the same house and you’re not sure which one owns which birthday or SSN. Informed guesses can be made, but they are precisely that, guesses. Cleaning up the residuals seldom is cost effective for basic marketing as it could easily be a $10 to $20 per name to validate cost for a $1 piece of mail. However, for voting purposes, cleaning up the list should be worth $10 per residual individual.
If we wanted to have a solid national voter verification project, then that would mean federalizing elections where all citizens receive a biometric secured national identification card free of charge, and swiping that card at a precinct would bring up a custom ballot for the races that an individual is eligible to vote in. And once a person swipes the card at a single location and submits a ballot, they would be locked out from voting for the rest of the relevant electoral cycle. It would eliminate the right church, wrong pew problem, it would eliminate the three voting machines at a precinct with 1,500 urban voters compared to the 9 machines at the suburban precinct with 500 registered voters problem, it would eliminate any illusion of legitimate concerns about voter impersonation fraud, and it would eliminate fears of double voting. It would actually solve a problem. It would be costly, but it would work.
But since voter identification and caging is not about actually solving a problem, we can’t have that… it is just proof that there is absolutley no fucking need whatsoever for preclearance or aggressive federal supervision of elections as only the Elected should elect the elected.