Super Costly Wisconsin Dataset Revealed Multiple Paths to Fraud

My analysis of the Wisconsin election database, based on a dataset downloaded on November 10 cutoff, revealed a number of problems with the state's database, including missing information, mismanagement, no effective mechanism to prevent duplicate ballots, and the ability to create voter counts at any time.

Absurd Download Fees - A Patriotic Citizen Bought the Dataset

In the beginning of this article, I would briefly tell the story of this dataset.

When a citizen in Wisconsin heard about fraud in his local election, he was concerned that his ballot had been tampered with and wanted to check the government database. The election dataset is publicly available (https://badgervoters.wi.gov/) and can be purchased by anyone, but he was shocked when he opened the website and saw that the page stated that it cost $12,500 to download the voter list data, $12,500 to download the mail-in ballot data, and $25,000 to buy the entire data set.

He was furious that a database that could be downloaded for free or at most for dozens of dollars in other states was as expensive as buying a new car in Wisconsin! If voters don't have that kind of money, don't they deserve to see the state's database?! This is not Venezuela, China, or Cuba, this is the United States of America!

He turned his anger into action and decided to defend his right to the truth, literally spending $25,000 to download the entire file. When he finally opened the database, he was shocked again - the extreme confusion of information, the huge gaps in numbers, how to sort through it all? So, he put the database up on the Internet for free download to the public.

I downloaded this hard-to-get database.

As I wrote this, it occurred to me that I am pitching a story to a news media outlet, so it has to be newsworthy! The changing face of the era were news, but how could the thoughts and actions of the common people who shaped the era not be news? I am hoping the editor to allow this description remain here. Hopefully, this citizen will know we were using the database he bought and that his dedication was not in vain. Cheers to this citizen's dedication!

Basic Database Information

The analysis of this dataset snapshot of the November 10 Wisconsin election follows.

First, the basic information: the database provides two tables, Voter Voting History and Registered Voter List.

The voter's history table contains 36 columns, one of which is the Election Name, which, when labeled as the "2020 General Election," is the Record of "mail-in" ballots sent out for this year's Nov 3rd election. If we count them one by one, we got 2,152,453, indicating that 2,152,453 ballots were mailed.

Excluding misaddressed, damaged, unreturned, and multiple ballots sent to the same voters, only 1,949,121 ballots were returned to unique voters. In other words, 1,949,121 people voted by postal ballot.

The voter registration table contains 120 attributes and contains approximately 6,994,368 voters (the database was not properly maintained, so it can only be an approximation).

The voter list table includes a column called "November 2020", which records the two methods of voting in the election: Absentee, which totaled 745,405, and At Polls, which numbered 485,135.

Adding together the votes cast by the two voting methods, we get 1,230,540 people voted in the election.

Huge Gap Between Database Numbers

The Voting History table shows 1,949,121 people voted by mail, while the registered voter list table shows 745,405 people voted by mail. Even if you add in the 485,135 people who voted at the polls, only 1,230,540 people on the voter registration table voted in this election! Between the two tables, which voted number is true?

In addition, the registration table clearly show 1,230,540 people have voted, yet CNBC News reported on November 4 that more than 3.1 million people had voted in Wisconsin and that Biden had won!

What are the valid ballot data released to the news media by the State of Wisconsin? Where did CNBC News get the figure of more than 3.1 million votes cast in Wisconsin?

I recommend the State of Wisconsin and CNBC News explain to the public, openly and honestly, which they are legally obligated to do so.

In my analysis of the fraud in the Pennsylvania database, I have explained in detail that if the number of voters in the database is significantly lower than the number reported in the news at the same time, the "update last voted day" feature is not working and it is impossible to tell whether a person has voted or not.

Again, it should be emphasized again that election databases that do not have the "update last voted day" feature are intentionally designed to allow duplicate ballot counting.

Wisconsin is much worse compared to Pennsylvania in terms of the ratio of the numbers in the database to the numbers reported in the news.

More Voters in the Database than Total Population of Wisconsin.

How can there be 6,994,368 voter records in the database when the total population of Wisconsin is only 5,860,000, and excluding the population under 18 (20%), which can only be 4,680,000 adults?

The difference is 2,314,368, which means that a third of the registered voters in the database should not exist. If such a database were used for elections, no election in Wisconsin would ever be fair again.

As an example: From the 1,230,540 voters counted on the voter registration table for this election, I saw the same person with two voter numbers voted (note: the voter information and addresses in the following comparison are identical, so it is confirmed to be the same person, but the address is not listed here for privacy reasons).

Database Management is Confusing and Not Trustworthy

From the 1,230,540 voted in this election, I also saw 558 Inactive voters voted. The State of Wisconsin requires that a voter becomes an "Inactive Voter" after not voting in four years. However, an Inactive voter can become an Active voter by showing an ID. So the 558 inactive voters who have already voted should actually be counted as active voters. Whether this was an arbitrary addition or an administrative mess is unknown.

When I populated the two datasets into database tables, thousands of voters were each screened out by the parsing program: there were blank voter numbers, garbled fields, broken lines, blank application dates that could not be added, etc. At the end, it took a lot of time to successfully enter most of these kicked-out voter information into the tables, but some of them could not be fixed. This is the absurb nature of this expensive database.

Will deceased people come to vote in such a ridiculous database? Yes, and here it comes. Look at this line of record in the database, this woman voted using absentee.

However, the database listed her being Deceased, and as you can search online, this woman did pass away in January 2018.

Following this thread, I found more than 550,000 voters marked as "deceased" in the database for "Voter Status Reason," a total of 559,201 voters to be exact, which accounted for more than 7.995% of the total number of voters in the database. Or if you divided by adult population (4.68 million) in Wisconsin, it's 12%.

This database makes me completely lose faith in Wisconsin's ability to have a fair election! I don't see an effective mechanism to prevent duplicate ballots, duplicate voter IDs for one physical voter both voted, the number of vote in the election very different between two tables and in turn different from the number reported by the state...and in this state you can vote without ID if you declared "indefinitely confined."

Wisconsin's voter database is one of the most baffling that I have seen so far in state voter databases. It can take a long time to fix, even longer than it would take to rebuild one.

by Ting Mei 2020.11.28