Since our current president is trying to get this claim of election fraud to stick, I thought I’d share my experience with counting absentee ballots on the Ingham County Consolidated Absent Voter Counting Board. We were told the rules we were being required to follow were dictated by Michigan law, and so its reasonable to expect that similar processes were being followed everywhere else in Michigan. In fact, we got threatened with having to appear before a judge and potentially get a felony multiple times. The county clerk staff took ballot counting very seriously! We were not allowed to bring any technology that connected to the internet into the building, including smart watches or kindles. And we were not allowed to leave the building until they let us, which you’ll read about later.
The deputy county clerk said that until the day of the election, the ballots were stored in a vault in the county building that could only be accessed by ID by a couple people, and he said they would run a report of access to that vault the morning of the election to ensure there wasn’t any unauthorized access. The ballots, which were in sealed bags, were then transported from there to the place we were doing the counting by a Democrat (D) and Republican (R). This is a recurring theme: any tasks that involved any potential for mistake or fraud or decision-making had to be done in D-R pairs. If one person needed to get a drink of water or use the restroom, the other had to stop working and wait until their partner returned.
We were allowed to open the bags and start processing the ballots at 7am. There was a whole assembly line of people opening the envelopes, removing the secrecy sleeve, tearing off the stub, taking the ballot out of the secrecy sleeve, flattening, counting, and preparing the ballots to be tabulated. I wasn’t involved in that processing, so my friends could answer questions about how that worked. What I do know is that the process was designed to protect the confidentiality of voters, so that even if you knew the identity of a voter, you wouldn’t know how they voted because the ballot doesn’t get removed from the secrecy sleeve until a couple steps down the line after any identifying information has been removed.
I was on a tabulation team, which meant I and my R colleague were grabbing stacks of ballots from the processing team, feeding them into the scanner, ensuring the scanner read the correct number of ballots, sorting out any errors, sending the scans to the adjudication team, labeling the stacks of ballots with a precinct and batch number, and placing the ballots back in the bags so they can be saved in case there is a recount. The errors I mentioned were infrequent. Sometimes the scanner would miss a ballot, so we would have to reject the batch and rescan it until the number scanned was the same as the number in the batch (which, in our case, was almost always 25).
Every once in a while a ballot was from the wrong precinct, which usually meant a voter somehow voted a ballot from a precinct different than the one in which they lived. Sometimes a ballot could not be scanned because there was a smudge on the “timing marks” along the borders of the ballot. The solution for either of these situations was to send this ballot to the duplication team. The duplication team would work in D-R pairs with a well-defined process to transfer the votes from the original ballot onto a fresh ballot from the correct precinct. Eventually we’d get the duplicated ballot back and tabulate it. Like I said, these were infrequent, maybe 5 ballots out of 1,000 at the most, and both the originals and the duplicates were saved.
Once my colleague and I were finished tabulating a precinct, the deputy clerk would come over and compare the number of ballots scanned with the number of ballots that precinct reported they collected. This was usually a matter of excitement and fanfare. This is important: if the number of ballots scanned differed from the number that precinct reported collecting, even by just one ballot, that meant we had to find the cause of the discrepancy and potentially retabulate the entire precinct.
While we were working, the adjudication team was reviewing ambiguous ballots. They worked in D-R pairs to review any issues the tabulation system identified, such as over-votes, write-ins, or ambiguous markings. Over-votes happen when someone votes for more candidates then there are available seats in a particular contest. My understanding is that write-ins eventually get sent to the “board of canvassers” to consolidate votes for candidates (e.g. Bob Smith and Robert Smith are votes for the same person, which was reportedly not allowed before write-in candidates had to register). Ambiguous markings are resolved by determining what a voter’s most consistent markings are. E.g., if the voter filled in most of the ovals completely and then one oval just had a slash through it which resulted in an over-vote, the adjudicators could rule that the slash was accidental (the person sneezed) and invalidate it, allowing the voter’s vote to count in that contest. Again: none of these decisions were made unilaterally. The D-R pair had to agree on every decision. If they could not agree, they were to fetch the co-chairs (also a bipartisan pair) and the four of them would make a decision together.
I have not yet mentioned another source of accountability: the election challengers. There were representatives of the Democratic and Republican parties there as challengers, watching the process to ensure rules were being followed properly. They spent most of their time watching the screen of the adjudicators, but they would occasionally walk around and observe the other teams working, writing in their notebooks as they went. They were not allowed to talk to the election workers. They could only speak to the county clerk staff, who were neutral, not associated with a party. My understanding was that they could only challenge something if they felt a rule was not being followed. There was only one time the entire day I saw them speaking with the clerk staff in a way that made me think they might be challenging something, but I’m not sure if any challenge was actually made.
Of the 19 precincts we tabulated that day, 16 of them balanced on the first try, meaning we tabulated the exact number of ballots the precinct reported. For the other three, three large teams had to be assembled to hand count every single ballot envelope from that precinct to see if perhaps the precinct miscounted. 1.5 hours later, we learned that turned out to be the case for two precincts, so those precincts balanced. The third, largest precinct (almost 2,000 ballots) ended up with one extra ballot remaining, meaning we had tabulated one more ballot than the precinct reported they collected. At this point it was 8pm, we had all been there since 6am, and the deputy clerk said it’s possible that two of the envelopes stuck together when doing the hand count. I was thankful he didn’t order another 1.5 hour (actually, probably 2-2.5 hours in the case of that precinct, since I think they started hand counting earlier than the others) hand count. He must have deemed this acceptable, and I assume if any race came down to a difference of a single vote, a runoff or recount could be ordered.
This seems worth highlighting: out of 18,795 ballots, at the end of the night we only had one ballot not accounted for.
After all the processing, tabulating, adjudicating, and double-checking was finished, a whole bunch of steps were left that apparently could only be done by the co-chairs while the rest of us either stood around or helped put away chairs, tables, carry sealed ballot bags out to vans, etc. This took 2.5 hours, I kid you not. It seemed every single page in a gigantic 3-ring binder had to be signed by the chairs, multiple reports had to be printed and then signed by every single person in the room, with more threats of tracking us down and putting us in front of a judge if we weren’t compliant. Ballots had to be sealed and signed by co-chairs and then loaded into a van to be driven back to the vault from whence they came by two Dems and two Republicans. And all the results and reports had to be loaded onto a flash drive. Oh, I forgot to mention: we were told that the three computers we were using for tabulation and adjudication and the flash drive used to transfer the results had never touched the Internet. Once all of this was completed and we had all signed our time sheets, we were finally allowed to leave.
I can’t speak for what happened before or after ballots entered that room, but I hope this helps readers understand how incredibly difficult it would be to commit voting fraud in the places where absentee ballots are being processed and tabulated in a strictly-enforced bipartisan process.
Kevin Daum 