transparency pilot project
Scanner Count Off
Uncooperative Board of Elections
698 More Ballots Than Voters
Uncooperative Board of Elections
Uncooperative Board of Elections
Uncooperative Board of Elections
JUNE 2021 BROOKLYN NY PRIMARY
Preliminary Report
June 2021
New York City Primary
SMART Elections is currently conducting The Brooklyn Transparency Pilot Project. It is part of our ongoing #CountTheVote initiative to monitor, protect & verify election results. The goal of #CountTheVote is to improve our elections and increase public confidence in them.
The Brooklyn Transparency Pilot Project seeks to compare the number of voters who vote, with the number of ballots cast in one City Council District in the 2021 Brooklyn Primary. We chose the 45th City Council District, partly because there is an ongoing civil court case here alleging corruption and poll workers "fraudulently feeding ballots into the scanner." We are not making those allegations, or taking a position on them. We are simply aware that allegations are moving forward in the courts, and they indicate that there is some lack of confidence in the process. Nationally, there is also a lack of confidence in our elections. In a 2019 C-SPAN poll, only 50% of Americans said they believed the election would be “conducted openly and fairly.”
We are issuing this preliminary report because, in our opinion, we are experiencing hostility and obstruction from the New York City and the Brooklyn Board of Elections, as well as some coordinators at Early Voting poll sites.
We have discovered discrepancies that raise concerns and deserve further exploration. These were reported to the coordinators at each poll site and directly to the Board of Elections. The Board of Elections responded by removing our project leader as a poll watcher and changing the information our poll watchers were allowed to view. Now we cannot collect the data necessary for the successful completion of the project.
We deeply respect the voters, candidates and poll workers who have committed so much time and effort to the election. They deserve to have confidence in it.
At this point, we cannot assure the public that the results in all these locations are accurate, because the process is not transparent and we have not been allowed to check.
This lack of cooperation and transparency is unfortunate and unnecessary. We hope that the Board of Elections will examine these issues fully and whole-heartedly commit to a respectful, transparent partnership with those engaged in public oversight.
FDR HS - 698 MORE BALLOTS THAN VOTERS
Comparing the number of voters to ballots is a common reconciliation
The number of voters who check in must always be the same or close to the number of ballots scanned. This has nothing to do with affidavit or void ballots. Those are accounted for separately. Each voter who checks in successfully is given one ballot and is supposed to scan it themselves, unless there is a problem with the scanners. So the number of voters who check in is supposed to be close to the number of ballots scanned.
Sometimes voters leave without scanning their ballots - called "fleeing voters". Because of this, the number of ballots scanned can be lower than the number of voters who check in. But the number of ballots scanned is not ever supposed to be more than the number of voters who check in.
On June 16th, 2021 we found 698 more scanned ballots than voters who checked in
This occurred at the close of polls at FDR High School at 5800 20th Ave. in Brooklyn.
At closing, there were 480 voters checked in, according to the numbers provided us by the coordinator, and one electronic poll book that we were allowed to view personally.
We examined the number of ballots scanned personally and it was 1178, meaning there were 698 more ballots scanned than voters checked in at that point in the Early Voting.
Board of Elections Response
We told the coordinator about the discrepancy. In our opinion, she showed no concern. It was subsequently reported to the Board of Elections, in person, and also via an email thread that included members of the Campaign Finance Board and the Chief Democracy Officer from the Mayor's office.
The Board of Elections responded with an email that our "concerns may stem from the public count on scanners vs. pieces of paper. Public counter counts the first sheet of a ballot as it is representative of the number of voters, not the number of ballots."
We are aware that with a 2-page ballot, the public counter counts one ballot, but the protective counter counts two. If the discrepancy were due to that, then there would be 480 voters and 960 ballots scanned. But that is not the number that we found. Furthermore, at other polling locations, the number of voters checked in and the number of ballots scanned was very close. So from our experience, this does not explain the problem.
FDR HIGH SCHOOL FOLLOW UP
On June 17th we returned to FDR High School
We had a different poll watcher review the data. At that point the number of total voters checking in more closely matched the number of total scanned ballots. There were 1505 voters checked in and 1491 ballots scanned. So there were 14 less ballots scanned than voters.
But the number of new voters since the day before was 1025. According to the public counter on the scanners 313 voters cast their ballots since the day before. So this creates a 712 vote discrepancy between the number of new voters checking in and the number of new ballots scanned. Again, basic reconciliation of voters to ballots is not possible.
Input from experts
We consulted with an experienced high level election official from another state who has managed and reconciled polling locations for more than a decade. HIs response was, “Something is off because these numbers don't make sense. That's all I can say ... And it's weird they caught up."
Public Counter and Protective Counter on the Scanners
The public counter shows how many voters have scanned their ballots during this session. The protective counter is like an odometer on a car and is supposed to show every page scanned for the life of the scanner.
In this election, the protective counter counts 2 pages because it's a 2-page ballot, but the public counter counts 1 for each voter. It is possible to reconcile these numbers by multiplying the number of ballots scanned (public counter) times two, then adding that to the previous protective counter number.
At this location the public counter and protective counter are not reconciling. There was a 39-page difference between the public counter and the protective counter, across all scanners on 6/17 at the close of polls.
It is hard to say why. One explanation is that voters are not feeding in both pages of their ballots.
Election technology experts we spoke with said that there are programming or software errors that can cause the protective counter to advance when the public counter does not. In some instances the vote might not be counted. They said it can also be from the machine being subjected to a heavy blow or drop earlier, which can cause an invisible internal alignment that will not show up in diagnostics.
We found this pattern in at least one other location, but the discrepancies at this location were the largest.