Before I get started, I would like to say what this article is NOT ABOUT. I will be presenting data on the Ramone-Burns recount that occurred in the 2022 general election. This article IS NOT about contesting the results. Quite the contrary. Since that race did have a hand count audit, I have more confidence in that result than others. So this is not an article that puts the results of that race in question, but let’s look at the recount in a little more detail, and see what it does put in question.
The original results of the race were shown on the DOE Website and they were still there at the writing of this article.
And you will also see that there that there was a re-count result. I will only provide the summary portion.
So that is where the story begins. The conventional “press” thought that the highlight of this data was that the NET difference, changed by 6 votes and that didn’t change the outcome, but nobody thought the NEON FLASHING LIGHT BURIED HEADLINE was important. Have you figured it out yet? Let me make it a little more obvious.
Look at the recount. The number of votes for both candidates went down. Department of Elections processes reported more votes cast and counted than The Board of Canvass was able to find to back-up that count. The total error was .48% or 42 ballots.
In June 2022, well before the election referenced above, I tried to report to the Board of Elections that there were significant vulnerabilities in our election systems. They had no interest. I further documented those issues, well before the 2022 election in an open letter to the board here: 9/22 Letter to BOE
Among other things I highlighted a huge vulnerability. Specifically, the concern that the DOE does not perform publicly viewed post-election audits on our Absentee and Early vote. I pointed this out BEFORE the race above.
The result highlighted above was and is a FLASHING WARNING SIGN regarding our elections processes. Can we get any further insights regarding the problem(s)?
Cast Vote Records:
Cast Vote Records (CVRs) are files reported by the Election Management System. Computer based voting systems are required to have them. A snapshot component of what they look like, showing the columns for only one race is presented below.
The CVR’s provide a chronological order of the vote tally for a given election by machine. These records provide vote results for each ballot, but do not provide any voter identity. I selected a sample that includes an undervote, just so you could see an undervote. Note that the CVR’s above are not sequential by increments of 1 because this list has been filtered to only include ballots associated with RD-21. The sample above also only includes results for 1 race, but on any CVR row, there would also be the results of all the other races in additional columns. One final note regarding CVR’s. Although they are in chronological order, they do not provide an actual time stamp. Given the CVR’s and a time stamped photo of a person entering a specific voting booth, you could still not tie a voting result to an individual. This is why photos of people voting during the day are NOT sensitive.
By looking at the information provided in the “batch” column, you can segregate votes by type. Results are shown below.
Again, with this exercise we are reminded that ballots are not votes. Sometimes people don’t vote on a specific race and that results in an “undervote”. Above we see how many pieces of paper (ballots) there should have been in each category. In the articles referenced above I indicated particular concern in the Absentee and Early Vote categories. Can we learn more?
The File of Absentee Voters:
The File of Absentee Voters (FOAV) is a file (list) that is required to be produced and distributed by the DOE. Among other things this lists all the people that are on the permanent absentee list, and all the other people who aren’t on the permanent list, but that did request an absentee ballot. It shows when the ballots were requested by the registrant, when they were sent out by the DOE, when they came back to the DOE and whether the ballot was accepted or rejected and why. I obtained such a list for the 2022 election.
The last ballot receipt on the update I had was recorded as of 11/7/22 3:29 PM. The list would not have included any Absentee Ballots delivered after that time and through 11/8/22 which was the “Day of Election”. Still, let’s see what we can learn from a review of the FOAV and then also pick up a few more gems from the CVR’s file.
Working from top to bottom and referencing the line numbers in the table below 1) We start with the total ballots received according to the FOAV and subtract 2.) the number of those that were rejected and we get 3.) 527 pieces of paper. But this doesn’t account for any of the ballots that may have come in after 15:29 on the 7th. The CVR’s showed that 4.) there were two lots of Absentee Ballots processed at the very end, and after all the Day of In Person voting results. Those two batches comprised 11 ballots. I have assumed they were the ballots arriving after 3:29 pm on Monday. After you add those back 5) you have 538 pieces of paper. Finally, to get “votes” for the respective candidates, you subtract 6.) six undervotes reported for this race and 7.) one write-in vote. You get 8) a balance of 531 Absentee Votes that should have been cast and counted for the candidates. But the DOE reported 9.) 544 Absentee Votes cast and counted, resulting in a difference of 13 shown on line 10. So with the DOE’s own data, the numbers don’t appear to reconcile in the Absentee category.
I showed over two years ago, before the 2022 election that the vulnerability existed and that you couldn’t catch problems without audits. How about that, we had a recount and it revealed a problem(s). The non-reconciliation of the Absentee vote confirmed it was at least in one of the two vulnerable areas I had highlighted. If the 13 ballot analysis holds, the error in number of ballots was 2.4% of the Absentee result. If we apply that same percentage to the “Early Vote” in this race it would amount to 22 additional ballots for a total of 35. That is NOT confirmed at this point but hypothetically that would be consistent with accounting for the lion’s share of the 42 ballot discrepancy. Another note. In this race, about 17% of the vote was Early or Absentee. In some races that combined component of a race was over 40%. What would the error level have been for those races?
The result could be that of simple non-nefarious systemic process error. That said the exploitation of the vulnerability could be much worse than this example. Electronic adding, or algorithmic flipping of votes could occur and you would never know it. In 2022 these two components represented over 20% of the vote. In 2020 with Covid in play it was above 30%. You could flip 5% of the ballots for a 10% net difference in a race and without an audit you wouldn’t catch it.
The primary job of the DOE is to provide an accurate count. The inability to balance the vote checkbook to within .48% caused alarm right? Clearly they would understand the ramifications referenced above right? Clearly they would initiate investigations and studies and get to the bottom of it along with corrective actions right? If that is the case, why did I take additional time and effort to pinpoint the Absentee Ballot concern? Well, this is a concern only associated only with the DOE systems, and one that doesn’t involve the Board of Canvass at all. You may see why that is important later. In general it’s simple. I didn’t see those expected responses from the Board, or the DOE. Did they have a concern, or curiosity, or accept any responsibility, or was their focus something else?
Below is an audio interaction between myself and Board Member Ms. Sandra Battaglia. We were talking about the recount at the BOE public meeting in February 2023. This was discussed within a broader context of many process issues raised regarding the 2022 election, however, for this article I have selected a clip that references the recount presented above. At the time of this discussion, I had not confirmed the suspected error in the Absentee portion of the vote, just the total difference between the two counts.
Audio Clip: Public Board Meeting: Dan Lesher/ Sandra Battaglia 2/2022 meeting.
It appears that Ms. Battaglia seemed to think I was questioning the decision of the Board of Canvass regarding the winner. I was not. I wasn’t even trying to show that vulnerabilities exist. I had already done that in prior correspondence. I was trying to show that the result, suggest that actual problems exist and that audit revealed that they exist, and that proves the value of audits. I was trying to communicate that to the group responsible for those systems and apparently failed.
I am involved in the discussion so I may not be able to be objective so I ask the reader a few questions. Did you think my concern was the result of the race, or that I was saying there was a nefarious conspiracy? Did Ms. Battaglia seem interested in any details or like she wanted to get to the bottom of things or did it seem to you more interested in taking the position that it was not their problem? Did she display the kind of attitude and the curiosity level you want in the board that is supposed to be the watchdog over our election department and its’ systems?
I would be the first person in line to say we should not judge a group or a person on a short edited clip. But what about when you send the entire board and the Commissioner a letter in detail that suggests that they should audit Absentee and Early results not only because it is the right thing to do, but because it may also be required by code as I did here? https://delawarecounts.com/p/post-election-audits-is-delaware If they would ever address the questions or concerns with a counter analysis I may agree with them but they don’t.
Now, I understand that the DOE only has a multi-million dollar budget, with an Information Technology department and more. And I understand the 11 BOE members probably don’t get a huge stipend for phoning in to each meeting so maybe they just don’t have the resources to address security and quality control concerns associated with our elections. But wait, on the other hand, I have zero budget, I am one person with no stipend. I do know my way around an excel spreadsheet and have a friend or two that also don’t get paid but took some initiative to get some CVR’s. So it is not resources. Why have they not even acknowledged these issues that could impact the quality control of over 20% of the vote?
Maybe that’s it. I am just one person. There are very few of us who go to BOE meetings. Maybe it’s as simple as that. Maybe they think if nobody else cares about what I / we are saying why should they. I think and hope that many people are concerned about election integrity but maybe they don’t have anything specific they can point to for a tangible action item. If you agree with the concerns of this article now you do.
Myself and others have fought for over 9 months to get the Board of Elections to put an email address on their website. They finally did. That email address is StateBoardOfElections@delaware.gov. If you believe that the post-election audit of the Absentee and Early vote is important and you want the BOE and DOE to do that, please drop them a note and let them know. Feel free to copy me on that note at delawarecounts@protonmail.com. Better yet, attend the next board meeting in person or by phone and register your concerns in the public comment section. Thank you for your consideration and help in advance.