After the results of Tim Canova’s August 2016 primary against Rep. Debbie Wasserman Schultz, a number of election-integrity experts found the certified results problematic. Lulu Friesdat, a journalist and documentary filmmaker who has investigated election issues for over a decade, worked with a team of data analysts, and they found statistical red flags in the results.
Canova wanted to inspect the ballots, initially in about a dozen precincts, under Florida’s public records law. By inspecting the ballots, Canova, a Nova Southeastern University law professor, sought to verify the vote, which he hoped would put people’s concerns and suspicions to rest.
In November of 2016, Friesdat filed two public records requests with the Broward County Supervisor of Elections Office. The office did not comply fully with the public records requests, and falsely stated that they “do not capture Ballot Image Files.” In March of 2017, Canova and Friesdat filed a third public records request with the office. In June of 2017, after the Broward Supervisor continued to stonewall the requests, Canova filed a lawsuit to gain access to the ballots.
On September 1, 2017, while the lawsuit was pending, and prior to allowing any inspection, Brenda Snipes, the Broward Supervisor of Elections, personally signed off on the illegal destruction of all of the ballots and accompanying documentation in the race. Even after Snipes destroyed the ballots, she concealed the fact, until forced to disclose the illegal destruction during the discovery phase of our lawsuit. As the court suggested in its May 11, 2018 order granting Canova summary judgment, even after Snipes admitted to the destruction of ballots, she continued to act in bad faith by litigating for another eight months, thereby harming Broward County taxpayers by needlessly driving up the litigation costs for both sides.
Without the original paper ballots, and using only digital scanned images, elections experts that were consulted say it is impossible to verify the results of our primary against Wasserman Schultz. There are an unfortunately large number of ways that the process could produce an incorrect ballot count. Ballots could have been lost or replaced before the scanning; ballot on demand machines could have produced extra ballots; some digital images could have been either accidentally or deliberately repeated numerous times. Digital images themselves can be altered, and there is no convincing chain-of-custody evidence for these digital images. The process of creating them involved using third-party proprietary software, as well as assistance from a third-party vendor, Clear Ballot. Even the chain-of-custody documents for the original paper ballots were not filled out fully.
When Ms. Friesdat’s data team was able to review some of the information from the election, they found large and unexplained discrepancies between the number of voters who voted and the number of cast ballots. In all, there were more than 1,000 discrepancies, and out of 211 precincts only 19 had the same number of voters and ballots. These irregularities were highly concerning to election experts.
Duncan Buell, a professor of computer science at the University of South Carolina, said, “I see what I would call a high likelihood of massive incompetence. Either that or there is fraud. I don’t think you should see numbers this big in this many precincts.” Buell has examined election records extensively in South Carolina.
Douglas Jones, a computer science professor at the University of Iowa stated that the county ought to be reconciling the number of voters with ballots and if they’re not doing it, “they’re grossly negligent.” Jones served on the Election Assistance Commission’s Technical Guidelines Development Committee for four years, but said “I’ve never seen a county that looks like this.”
The large discrepancies between the number of voters and the cast ballots, plus the inability or refusal of the Supervisor of Elections’ office to produce the original ballots, all raise questions about what the true totals for the race may have been. The experts we spoke with concurred that the certified results must be considered suspect. “They destroyed the evidence,” said Karen McKim, a member of the Wisconsin Election Integrity Action Team and a veteran of hand-counts in that state. “They can’t defend their results.”
According to U.S. intelligence agencies, there are serious risks to the integrity of our elections from hacking or other electronic manipulation. Those risks could be from either foreign or domestic sources. The ease of hacking into electronic voting machines – even those not ordinarily connected to the internet – was recently reported in a New York Times Magazine article by Kim Zetter entitled “The Myth of the Hacker-Proof Voting Machine” (Feb. 21, 2018).
It is incumbent on us all to protect our democracy, and to make sure that we fully verify the accuracy of our elections.