Read on Twitter
Understanding VA-7 Turnout Part 2 : From the last thread on VA-7, we explored two key anomalies :
1. Drop in R turnout in Henrico and Chesterfield Counties
2. Uniform >40% D turnout
This thread will explore a comparison in turnout with VA-4. Why VA-4? Henrico and Chesterfield.
1. Drop in R turnout in Henrico and Chesterfield Counties
2. Uniform >40% D turnout
This thread will explore a comparison in turnout with VA-4. Why VA-4? Henrico and Chesterfield.
2. For some background, there are 14 counties in VA that cover more than 1 district. See image. Here we see the counties on each row and the districts they cover in the numbered columns. Henrico and Chesterfield cover both district 4 and 7.
3. One more note on Chesterfield and Henrico. These two counties create two halves of a donut that encompass Richmond City county. This provides an interesting option of comparing turnout differences within a county across different districts. Let's see if they are.
4. VA-7 on top, VA-4 on bottom. The yellow rows represent Chesterfield and Henrico. Both of these counties show a significant decrease in R turnout in VA-7 compared to VA-4 (2016 vs 2020 difference). Conversely, the D turnout shows a significant increase in D turnout. Why?
5. Said differently : Republicans in the district 7 part of these counties disappeared, while the Republicans in district 4 they didn't. How is it that R turnout in these populous counties are so different depending on district?
6. Likewise, why are Ds much more reluctant to vote in VA-4 compared to their peers in VA-7? VA-4's D turnout was a "paltry" increase of 20% compared to VA-7's 44% increase. Even within the same counties!
7. These results look more like the outcome of concerted effort to warp the election rather than organic voter trends. Given the high number "tabulation errors" in Henrico and Chesterfield, it suggests malfeasance and election interference. Let's revisit the Change Log.
8. I performed an analysis to calculate the number of votes affected across the change log entries. Since the log doubles as an audit trail, we can see what the total votes were, and what they became after the change. The hope was to understand why changes occurred.
9. Here's a look at VA-7 (Freitas) change entries. The highlighted record shows his votes were decreased from 42,244 to 39,821 with the "Change reason" being "Tabulation error in the precinct". From digging into these reasons, they don't appear to used appropriately/ consistently
10. Many entries are classified as "Data entry error in central office". Let's take a peak at a very unique county where this occurs. Prince William County is covered by Districts 1, 10, and 11. My analysis from last month showed votes being shifted from District 1 to 10 & 11.
11. When we look at the change log to understand those shifts, this is what we see. The top rows show a uniform decrease of 59% of the votes that were reported on election night. But only for state wide races (Prez / Senate). What makes those two races special?
12. Those are the only two races that will appear on every ballot in in the state regardless of precinct, county, or district. Therefore, the only votes that could be shifted to different areas. What follows the initial 59% decrease are many changes that are multiples of 100. Hmm
13. Anyone familiar with financial crimes will be familiar with perfectly round amounts / adjustments being a red flag for suspicious activity. Why do we the same patterns in our voting results? Were all of these accidental typos in central office in the days after the election?
14. The 59% decrease in the Prez / Senate races in Prince William County on 11-7 @ 1:25PM is hard to track. From my analysis last month, it seems some of those votes don't get shifted to other counties / precincts, but some do. I was told by someone with local knowledge that
15. election officials has explained to campaigns that votes were mistakenly associated with the wrong districts. However, this raises a plethora of questions about how the tabulation process works.
16. Key questions : if two ballots from different districts are fed into a tabulator, how does it count the votes? Are tabulators configured to a specific ballot type? If so, would only state wide races (overlap) between the diff ballots be counted?
17. Could this explain "tabulation errors" between VA-4 and VA-7 within Henrico and Chesterfield? I'm not a VA local, and don't have many details on how officials organized the absentee ballots. But for those reading this thread, here are some questions if you have such knowledge
18. What was the return address for absentee ballots? Was it different by district or county? Understanding where these ballots went could provide clues for how they were counted. Do VA-4 Henrico voters mail their ballots to a diff address than VA-7 Henrico voters?
19. For the tabulators, can they count votes properly across multiple ballot types (different districts)? Are ballots segregated by type before being fed into the machines? Is feeding different ballot types into a machine "normal"?
20. The more I dig into the data, the more questions emerge as to what happened in VA. I think we all would like to have more confidence in our election processes, and if you have any information to share, please let me know.
@threadreaderapp
unroll
unroll
