On Tuesday, December 6th, I was a Clinton recount observer for Macomb County, MI. I volunteered because, as a precinct delegate, I had already received some data from the Macomb Democratic Party, and I wanted to see for myself, where possible, how the data compared to what I could observe from the actual ballots. The recount was held at the Sports and Expo Center of the south campus, a large fieldhouse where thirty recount stations were set up. Each station had two recount clerks and up to two observers from the Clinton, Stein, and Trump campaigns were allowed. What follows are some of my observations and impressions from my day at the recount.
1. The Trump campaign was highly-organized and out in force. With thirty stations to observe at one time, the Trump campaign had no shortage of “volunteers.” In addition, there were a minimum of four (and possibly more; it wasn’t always clear who was whom) lawyers present to represent the Trump campaign. The lawyers acted as a roving band, who swarmed any table where they thought a challenge might be occurring. I found them as a group to be arrogant, but never intimidating, and in fact, they did provide some comical relief in an otherwise tedious process. The recount clerks all had “assistance needed” signs to be held high if there were ever any questions, issues, or problems. When staff from the county clerk’s office saw these signs, they would proceed to the table to help. It soon became clear that these signs were the signal for the Trump lawyers to descend upon the table, oftentimes making it difficult for personnel from the clerk’s office to reach the table. I admit to laughing out loud at them when they swarmed the table I was observing; the assistance our clerks were requesting was for someone to bring wire-cutters when we were ready to open our ballot box. Six testosterone-soaked, pompous, supercilious lawyers slinked away in unison when they learned the nature of the help we needed.
Two days after the recount was ended by the courts, I was in contact with the Clinton recount organizer. He had been told by a Stein lawyer that the presence of so many Trump observers and lawyers was no fluke. Two superPACs, the Stop Hillary PAC and Great America PAC, paid a stipend to Trump volunteers and paid for the lawyers. He was also told that a number of the lawyers were from out-of-state. Despite that, the Clinton campaign (real) volunteers were able to have one observer at each of the thirty tables while I was there; oddly, despite their head start in recruiting volunteers and from a greater pool than just the Greens, the Stein campaign had far fewer volunteers and very few tables had Stein observers. Since Clinton volunteers were specifically instructed to NOT challenge ballots, this could have been a serious problem if discrepancies had been found.
2. Despite greater numbers, Trump observers were poorly trained. The training made available to Clinton observers was required phone training; my session lasted about one hour, but I know of some people that were on the phone for 1 ½ hours (the difference probably being due to the number of questions from participants). I was nervous going in that the training might be insufficient, but it had been thorough and covered in detail even the most unlikely scenarios. On the other hand, the Trump observers I sat with or talked with really had no clue beyond “don’t touch the ballots.” At my first station/precinct, the Trump observer kept announcing the vote on the ballot before either of the recount clerks did. Despite being told two times to stop, she continued to announce the vote; the third time she was warned, the recount clerk also told her it was her last warning, and she would be removed as an observer if she continued. At my second station/precinct, the observer was better, but he had been a recount observer for a local election in his hometown and knew the rules. Even he admitted that he was dismayed by some of his fellow Trump observers behavior.
3. Actual numbers from one precinct: There are 337 precincts in Macomb County, so if we had had roughly 30 more votes/precinct for Clinton, Michigan would have been won (statewide, if we had swung 4 votes/precinct, we would have won Michigan). At the first precinct I observed, I was testing the waters a bit to see how much I could actually see on each ballot (beyond the topline votes, the reason I was there), as well as noting, in general terms, the incidents of third-party voting and undervotes. It became clear that I could do my job as an observer and note additional items of interest, so by the time I moved to another table with another precinct, I was prepared.
These numbers are from one of the largest cities in Michigan (in the top five), which is 85%+ white, and is generally considered one of the safest cities in Michigan. The precinct recorded 1224 ballots: Trump=683; Clinton=474; Johnson=30; Stein=11; Castle=3 (U.S. Taxpayers Party in MI/Constitution Party elsewhere); Soltysik=1 (Natural Law Party in MI/Socialist Party USA candidate elsewhere); and 22 undervotes. During the course of the recount, Clinton gained one vote due to a challenge by the Stein observer (we actually had one for this precinct) which the Trump observer actually agreed with. The voter, rather than filling in the bubble, had traced the outline of the bubble. The recount clerks were undecided. When the Stein observer called it for Hillary because the intention was clear and the pen was clearly touching the bubble, the Trump observer agreed even though he could have fought the challenge.
4. Details about undervotes: Of the 22 undervotes, three were for Evan McMullin. For the purposes of the recount, these were considered undervotes, but for the purposes of the election, they counted because McMullin was a registered write-in candidate. On the other hand, there were seven undervotes for Bernie Sanders. These did not count in any way, since Sanders was not a registered as a write-in candidate. There was also a vote for Ben Carson; another vote that was not counted in any way. The balance of the undervotes were totally blank: no topline candidate was selected, no registered write-in candidate was written in, and no straight-ticket party option was chosen. (In the other precinct I observed, there were undervotes for John Kasich and Mike Pence, as well as multiple votes for Bernie Sanders, but the total of undervotes was far fewer).
5. Random observations, in no particular order: For me, the biggest surprise was the downticket voting of Johnson voters. Because of the way the ballot was laid out, it was impossible to see every local race in the time I had to view the ballot. Instead, I decided to focus on the State house race for third party voters. More than half of the Johnson voters voted for the Dem candidate for the State house; my notes are a bit of a scrawl, but I show 22 hash marks for Johnson voters with Dem downticket (House) votes.
I was also surprised by the number of voters who indicated they were voting straight ticket, but then proceeded to fill in the bubbles for individual candidates. In most cases, this was consistent, but in other cases, it changed the vote. For example, one ballot had the straight ticket bubble filled in for the Green Party, but then the voter filled in the Clinton bubble. In accordance with Michigan election laws, this became a vote for Clinton. There was also a voter who voted straight-ticket Dem but wrote in Bernie Sanders. Because Sanders was not registered as a write-in, this remained a vote for Clinton.
There were also a number of voters who voted for the presidential election only. It wasn’t a huge percentage of voters; I didn’t track this specifically but my notes say “about 4%.” The good news, from a partisan perspective, is that the clear majority of these topline-only voters were Trump voters, but in a state that is totally controlled by Republicans, it does highlight an area in which Democratic outreach and education could be improved.
Finally, only because it bothers my perfectionist self, what the hell is so difficult about coloring in bubbles? Any mark within the border of the bubble counts as a vote in Michigan, but why take any chances? The vast majority of voters did it per the directions (and this is even illustrated on the SoS website), but the number of people who put tiny, tiny check marks or Xs was surprising to me. My personal favorite, however, was the voter who indicated a straight-ticket Republican vote, as well as individual votes down the entire ballot...by circling well outside the bubble. The entire ballot became an undercount/no vote. In this case, it was NOT OKIYAR.