Verified Voting Blog: Letter to Georgia Secretary of State regarding Verified Voting’s position and involvement with risk-limiting audit pilots
The following letter was sent to Georgia Secretary of State Brad Raffensperger on December 16, 2019. The letter addresses Verified Voting’s concerns following the November 2019 election in Georgia and provides clarity on Verified Voting’s position and involvement with risk-limiting audit pilots in the state.
Dear Secretary Raffensperger,
I am writing to address a few issues that have concerned us since the November election and so that you and your staff have clarity on Verified Voting’s position.
As an initial matter, Verified Voting did not recommend that Georgia purchase all ballot marking devices for all in-person voters. We made our position clear in a letter to the co-chairs of the SAFE Commission dated January 4, 2019 attached for your reference. Verified Voting stands by its position and notes that this continues to be our recommendation for jurisdictions who are deciding what system to purchase among commercially-available voting systems. The fact that Georgia did not follow our recommendation and purchased Dominion BMDs for all in- person voters does not change our position.
Since the summer of 2019, Verified Voting has been working with the staff of the Secretary of State to implement post-election risk-limiting audits. Mark Lindeman, Director of Science & Tech Policy at Verified Voting has been the primary contact for your staff and is a subject-matter expert on RLAs. Our work with you on the implementation phase in no way endorses Georgia’s decision to move forward with BMDs instead of our prior recommendation of both hand-marked paper ballots and ballot marking devices in the polling place.
A risk-limiting audit is a tabulation audit: it uses statistical methods to provide confidence that the paper ballots were correctly tabulated. It checks only the tabulation, namely whether a full hand-count of the cast paper ballots would reveal something different than the reported outcome. It does not check — among other things — that voters actually verified their paper ballots, or that the paper ballots being tabulated are exactly those paper ballots that should be tabulated. Nor does it check whether strong chain of custody procedures, proper ballot accounting or other processes necessary to create a trustworthy record were observed. To express or imply that doing an RLA pilot demonstrates the security of the system is simply not true.