The Voting Equipment Audit Report

This report prepared by Pro V&V, dated March 12, 2015, was alarming and provocative but inconclusive.

The auditors physically examined 27 AccuVote TSX machines in Virginia Beach and found all had issues of eroded capability due to age. This is valuable data but there is no actionable rating of EOL expectations that localities can use.

In Henrico County, where a “high level of anomalies” had been cited in the November 2014 election, the auditors were unable to examine any AVS WINvote machines because of an impending special election. Conclusions were offered that were not specific to Henrico’s machines and this data is not valuable. In my opinion, there was no audit here.

In Spotsylvania County, where “a high level of anomalies” occurred in Precinct 302 in November, 2014, the auditors tested “four spare machines” rather than the actual machines which had been in use in Precinct 302. In my opinion this is not a valuable comparison and not an audit.  The auditors also declared that they were able to “connect” to the WINvote wireless network with their cell phones but there is no allegation they were able to transmit data in either direction, nor to crack the encrypted code on the WINvote DREs. This is a technical area beyond my expertise but I see no evidence from the auditors that they were able to do anything more serious than determine from their cell phones that a wireless network was active in their immediate vicinity. If there is more to it, we need details.

ELECT’s Interim Report

This report offers details of the Spotsylvania County issues in Precinct 302 including the allegation of a smartphone streaming music from the public library wireless network being the primary suspect for WINvote DREs “crashing” in succession. These actual voting machines have now been turned over to the State Police for examination in a formal inquiry. It will be very instructive to learn whether this anomaly can be replicated and whether the encrypted data on these machines was at any time vulnerable to alteration or erasure.

The results of this formal investigation should provide actionable data for the State Board of Elections. Without empirical proof of actual failure or open vulnerability to result manipulation, any action to de-certify WINvote DREs would be reckless, and the damage to the integrity of the elections community would be incalculable. The cost to affected localities would be severe.

The recommendations for substantial improvement in the ability of the Department of Elections to conduct comprehensive audits, regular certification review, and thorough post-election analysis is well founded and long overdue.

I applaud the Department for its investigation and for the transparency with which it is being conducted. Please continue to move with caution and deliberation on this most challenging time.