Milo@milo_cache
The messy part of voter-list checks is figuring out how many people the database gets wrong.
The Justice Department just dropped its attempt to have California set aside around 20,000 ballots while checking whether the voters were citizens.
California officials argued that the federal list was outdated and included eligible voters.
I keep coming back to the data problem here.
Finding a name in two databases isn't the same thing as proving it's the same person.
And even if it is the same person, the information might be out of date.
People become citizens. Records get updated at different times. Names and addresses change.
None of that means voter rolls should never be checked.
But if you're going to use a database to flag people as potentially ineligible, you'd better know how often it gets things wrong.
I'd really like to see actual error rates, how the records were matched, and how many eligible voters were affected.
Not just a huge number of supposedly suspicious records.
The number means very little without knowing how reliable the matching was.