December 5, 2016
North Carolina State Board of Elections
A.Grant Whitney Jr.
Rhonda K. Amoroso
Joshua D. Malcolm
Honorable Board Members,
Verified Voting is national, non-partisan, non-profit advocacy and research organization committed to promoting and supporting secure and trustworthy election technology and practices. We have long supported the adoption of post-election audits as a means to provide transparency and confidence in the voting process. We write to you today in support of the request by Ms. Marilyn Marks of Charlotte, North Carolina for the Board to perform an investigation into the conduct of hand-to-eye audits (“sample audits”) in all counties using the ES&S iVotronic DRE voting machines with Real Time Audit Log (RTAL) in the November 8, 2016 elections.
Before 2004, North Carolina used electronic DRE voting machines without an independent paper records in a number of counties. In the 2004 General Election, the machines in Carteret County lost over 4,500 votes. 1 The vote loss was greater than the margin of error in a state-wide contest, and North Carolina moved to make positive changes to restore voter confidence in its systems and procedures. The General Assembly required all voting systems provide voter-verified paper records/ballots. In addition, the General Assembly passed a requirement that a hand-to-eye comparison be done of the voter verified paper ballots to ensure the vote counts are correct, an essential safeguard against errors or malfunctions and against tampering.
Conducting such post-election audits is a critical part of ensuring the integrity and correctness of the vote count. The 2007 report by Brennan Center for Justice at New York University School of Law and the Samuelson Law, Technology & Public Policy Clinic at the University of California, Berkeley School of Law “Post-Election Audits: Restoring Trust in Elections,” the authors emphatically stated the importance and necessity of using the paper ballot to audit the election results.
“Paper records will not prevent programming errors, software bugs or the introduction of malicious software into voting systems. If paper is to have any real security value, it must be used to check, or “audit,” the voting system’s electronic records.” [Emphasis added.]
According to State statute § 163-182.1(b)(1) the Board is required to provide for the hand-to-eye sample count of the paper ballots or paper records of a state-wide ballot item in every county which includes the “real time audit log” (RTAL) produced by the iVotronic voting machines. We have been advised that Mecklenberg County conducted its hand-to-eye sample count using electronic images of the ballots taken from the iVotronics, rather than the voter-verified paper record provided on the RTAL. We have independently confirmed that this was the practice in at least one other county following the November 8, 2016 elections.
Using the electronic image to conduct a hand-to-eye sample count does not conform to state requirements nor to best practices. It cannot confirm the correctness of the vote count; the electronic image was not visible to the voter at the time it was recorded and printed, and the law clearly intends for the voter-verified paper record –not a record no voter has seen—to be used for the sample count. In addition, an electronic image could be altered undetectably by a programming error, bug, glitch malfunction or malicious attacker after it is recorded, therefore it is not a reliable or meaningful record of voter intent. Conducting the hand-to-eye count by using the electronic ballot image undermines the usefulness and effectiveness of the sample count and subverts the intention of the statute.
We further note that the task of reviewing the paper record verified by the voter is a feasible one; we have not only helped develop protocols for such audits using identical equipment in other states, we have observed the conduct of those audits, and note that they were successful. In one audit we observed in a county of comparable size, only a very small number of the RTAL printouts were unreadable. In such instances, best practices would require both RTAL records of voter intent and non-voter-verified printouts should be used, not the latter instead of the former.
Verified Voting urges the Board to conduct a full investigation into the hand-to-eye counts performed by all counties that use the iVotronics for voting in either early voting or Election Day voting (or both). To protect the intent of the State statute and to ensure the integrity of the vote count in North Carolina, the hand-to-eye sample count should be conducted using the voter-verified paper record. We urge the Board to determine if counties opted to use the electronic record instead of the paper record when the paper record was available, and to provide clear guidance to counties to conduct all hand-to-eye sample counts with each paper record unless it is absolutely impossible. In such cases we strongly recommend guidance as suggested above, to examine every available RTAL and supplement with a review of the non-voter-verified printouts only where unreadable records exist.
We thank you very much for your service and consideration.
- “More than 4,500 North Carolina votes lost because of mistake in voting machine capacity,” USAToday, Nov. 4, 2004