Wisconsin update

This will be fast because I don’t have much time, but I wanted to share this with you. Yesterday, people started talking about the Wisconsin Supreme Court Justice race. Despite passing their Voter ID law, somehow, the same people voted for the justice who is opposed to Voter ID. It seemed strange, almost as if there was some kind of fraud going on.

I responded initially by extracting the very high clone counts I found in Wisconsin. By state and counties, Wisconsin has one of the highest clone to real person ratios of any state I’ve studied. In New York, about 9% of all records are clones. In Wisconsin, it is more like 6% (a little over half of the 874,455 identified “clone” records). For voter registrations, that is a lot. However, I just realized it is much worse.

This is because they also have duplicates (dups). Not ordinary dups, but really weird ones. Normally, dupes match in all fields, as if you copied, or duplicated, the record. In this case, the ID numbers are duplicated, but none of the personal information matches. For instance, ID number 700,153,440 is assigned to two people with totally different, names, addresses, phone numbers, etc. There are 222,076 ID numbers like this (or 444,152 records). This is the opposite of a clone, but can have the same effect as a clone, depending how it’s used.

A clone record can be used to either fraudulently obtain a mail-in ballot, or to reconcile the “voters who voted” count with the number of fraudulent ballots presented at the polls. A duplicated record, providing the addresses differ, can allow two ballots to be presented at the polls (in whatever manner) for counting as a vote.

There is no overlap between these groups, so it is a combined problem affecting about 17% of the database, only a maximum of 8.5% might be legitimate. That’s a hard problem of 8.5% of all records, which is very close to New York’s 9%.

And here’s the kicker: All of the ID numbers in question end in 0. If the entire database is sorted by ID, there are other numbers between these that don’t end in 0 and aren’t duplicated. The only duplicated records have ID numbers that end in 0.

I found this today because I was wondering if the preference for divisible by 10 numbers I’d noticed earlier was a checkdigit. Hint, the answer is no.