Why does the Post Election System Check matter to all New Mexicans?
Can we disregard State Law and expect public confidence in our elections?
As a software developer for several decades, I know that I can write code to implement just about any algorithm you can think of. I also know that I can make mistakes which can lead to erroneous results. So, I have always had doubts in my mind about our voting systems.
Several years ago, when I cast my vote at an early vote center, I asked the people at the tabulator how can I know that my vote was recorded as I intended? One of the nice people responded with “did you hear the ding?” I responded “yes I heard the ding” which would indicate that my ballot was registered. However, my question is how can I know that my votes were cast as I intended them to be cast? A second person at the tabulator responded that I can’t know. This was an honest answer.
However, now that I have had the opportunity to investigate how the tabulators are used in Bernalillo County, I have a much better answer. We may not be able to know directly how our vote was cast but there are measures in place to detect faulty tabulators during the counting process so we can have confidence that our votes are counted as intended.
One of the most convincing measures is enshrined in state law. That is the Post Election Voting System Check. In this process all paper ballots (remember that piece of paper you filled out and inserted into the tabulator?) for randomly selected precincts are counted by hand and the hand counts are compared to the original machine counts. A report written by an independent auditor is then made available to the public.
The law, New Mexico Statute 1-14-13.2, reads, in part, as follows (emphasis mine):
B. …The random sample shall be chosen in a process that will ensure, with at least ninety percent probability for the selected offices, that faulty tabulators would be detected if they would change the outcome of the election for a selected office.
So, state law indicates that, when done properly, we should have a 90% probability of detecting faulty (for whatever reason) tabulators if they would change the outcome of the election. This is a catch all test of the tabulators. This check is essential to assure us that nothing happened during the election that could lead to incorrect results.
The implementing New Mexico Administrative Code (NMAC 1.10.23.9 Voting System Check Procedures) reads, in part, as follows (emphasis mine):
(c) Precincts will be randomly selected using a process that is visually observable, such as rolling dice or selecting pieces of paper from a box, with the probability of selection being proportional to the number of persons registered to vote in the last election in each precinct.
(d) The random sampling process shall be open to public observation. At least seven days prior to the random sampling conducted pursuant to this subsection, the secretary of state shall post notice on its web site of the time, date, and location of the random sampling.
Unfortunately, the way it has been deployed since I started observing, has not been completely open to public observation because the dice rolls are entered into a tool that is not visible to the public which announces the precinct associated with the dice roll. Last year I requested access to that tool so I could understand how the dice roll is converted to a selected precinct. I was not granted access and still do not have access to date.
To make matters worse, when the audit results were published for the 2022 system check, it became clear that the precincts were in fact NOT randomly selected. The report itself, along with a history of reports, is available on the State's Audit Reports page. A presentation I put together on this issue can be found here.
This invalid implementation of state law and NMACs calls into question the integrity of our election process to the extent that we cannot be confident in the results. We must resolve this before the next election or there is no guarantee that the votes of New Mexicans will be tallied correctly.
What are your thoughts on the issue?
Please see my post below entitled: April 28, 2024 - UPDATE on Random Precinct Selection Process. Access to the precinct selection tool was granted and it became apparent that the non random selections were most-likely human error. Thank you!