What We Know About the TN Special Election

Was Tennessee District 7 a fair election?

Photo from CBS News

My Cell Phone Is Like 911 for People With Election Doubts

At 9:22 on Tuesday I got a text from a SMART Elections Team Member:

This is my life. Almost every day, people send me their doubts about our elections. These worries are detailed in articles, emails, texts, recordings and hundreds of pages of lawsuits that cover pretty much everything that is going wrong - and has gone wrong - with our elections. I have contacts on the left and the right, so I get the full experience of shock, agitation, fury, panic, cynicism and despair that is happening across the political spectrum—in regard to U.S. elections.

Did Something Happen in Tennessee?

On Tuesday, at 10:13pm, after receiving the text from our team member, I reached out to our technical advisor, Bennie Smith, who is a State Election Commissioner in Tennessee. He had some data from Davidson County, a strongly Democratic County.

You can see Bennie’s data here. His company is called TurnOut Gurus. His data is dense, but there are simple things you can understand from a quick look. If you click the Primary_Identity button in the upper right corner and toggle between Always Democrat and Always Republican you will see that Davidson County is a predominantly Democratic County.

• Always Democrat F= 8,093 | M = 5,060

• Always Republican F= 836 | M = 995

  

Below are the total preliminary results. You can see in the 2nd chart that Davidson County is the second largest county. So a 10% reduction in Democratic votes on Election Day would have a big impact on the outcome.

But even 10% of all the votes in Davidson County would not be enough to make up a 15,000+ margin of victory. There would have to be more than that.

On Wednesday, Nathan from Election Truth Alliance texted to say they had data from Montgomery County that looked suspect and wanted to know if we had other counties.

From Nathan Taylor:

Election Truth Alliance (ETA) Analysis Indicates That Candidate Vote Totals in Montgomery County Strongly Correlate to Turnout in the Tennessee Special Election

If you don’t know anything about statistical analysis of elections, this won’t mean much to you. But if you think about it in common sense terms, you’ll understand that there’s very little reason why as more people vote, they would change who they’re voting for.

More Democratic voters will generally result in more votes for a Democratic candidate and the same for Republicans. But we don’t expect either candidate to get more votes, simply because more total voters are at a location, compared to, for example, a location with the same breakdown of Democrats and Republicans that has a smaller number of total voters.

Statistically, we don’t expect to see a candidate’s vote totals change from a location with low turnout to a location with high turnout, unless there’s a demographic explanation.

Generally data scientists expect the results of surveys or voting to settle into a standard percentage that pulls toward the mean, (or average support) for each candidate. Meaning, in a graph, they’re expecting people’s choices to settle into roughly straight lines, like this. especially as you get into large data sets:

Above is an example of a graph from an election in Columbia County New York in the 2016 primary, that received a hand-count audit. We’re confident the results are accurate, and the graph has a statistically expected pattern. This graph is from research that I led with a statistical team in 2016.

The precincts (or districts as they’re called in New York) are arranged from the smallest to the largest number of votes. After some early fluctuation, there is no meaningful difference between a candidate’s support in small or large precincts.

Data scientists expect the same to be true of turnout. Unless there is some overriding demographic issue, they expect candidate’s support to settle into a relatively straight line as the turnout increases and the dataset becomes larger.

This expected statistical pattern is not what’s happening in Montgomery County, TN, according to the analysis from ETA. Here is their graph.

As the turnout gets larger, the Republican Matt Epps gets a larger and larger percentage of the vote. This is not an expected statistical pattern.

ETA says, “Montgomery County’s precinct-level results exhibit statistical signatures consistent with known forms of ballot-inflation or digital vote manipulation. These patterns warrant additional scrutiny.”

Montgomery County is the largest county in the election, with 42,542 votes reported. It’s more evenly divided between Democrat and Republican vote, than Davidson County. In the preliminary published results Republican Matt Van Epps won the county.

ETA bases its claim that election results that strongly align with turn out are suspect, partly on research that was published at Arxiv, a curated research-sharing platform run by Cornell University.

An Academic Response

Thursday, I met with one of SMART Elections’ academic advisors and showed him some of the data that’s been generated so far, particularly the ETA analysis.

Academics are generally cautious and I was not surprised that he had some questions and wanted to look into the data further himself. He doesn’t have the bandwidth to do that until December 15th, by which point any window for a challenge will likely have passed. But he might be someone that a candidate could turn to at a later date for input.

If a court case is brought by Behn’s team, they will need academic data experts to testify.

Next Steps

SMART Elections’ #CountTheVote Team has researched the statutes and timelines regarding recounts, audits and legal challenges to elections in Tennessee.

• There is no mandatory audit for this election. The statute calls for elections held in August and November to be audited, not December.

  • Tennessee Code Title 2. Elections § 2-20-103

    • (b)(1) Following each August and November election, a county election commission utilizing a precinct-based optical scanner and not selected for audit shall, in accordance with this section, conduct automatic mandatory audits of the voter-verified paper ballots cast for the top race on the ballot, as described in subdivisions (b)(2) and (3).

• Any challenge to the election results can only be brought by a candidate or an incumbent, and it must be filed within five days of the election being certified.

  • 2024 Tennessee Code Title 2. Elections Chapter 17 - Contested Elections

    • § 2-17-101

      • (b) The incumbent office holder and any candidate for the office may contest the outcome of an election for the office.

  • 2024 Tennessee Code Title 2.

    • § 2-17-105

      • The complaint contesting an election under § 2-17-101 shall be filed within five (5) days after certification of the election.

• Davidson county certified their election results on Friday 12/5

• We believe that all counties have certified now, as the TN elections website says “View official results”

• That means any challenge to the election, would have to be filed by Aftyn Behn by Wed. 12/10, at the latest, or possibly earlier, if other counties certified prior to Davidson County.

• We are open to speaking to Ms. Behn, if anyone would like to put us in touch. You can contact us at SMARTelectionsAct@gmail.com.

• We’re non-partisan and share data with any candidate and the public.

There’s more in the works.
Stay tuned, share with your networks & support.

Please Donate & Subscribe

We are gaining a better understanding what happened in previous election cycles, as we put protections in place for the 2026 midterms. If you find our work valuable, please subscribe and make a donation as well! All donations are tax-deductible.

We’re trying to raise $200,000

We are currently in a fundraising drive to raise $200,000 by January 2026 to protect the midterms, support our current work such as this investigation, and file more lawsuits. You can help.

Thanks for all that you do!

Think for yourself, or others will think for you without thinking of you.

~ Henry David Thoreau

---