UPDATE: On September 23, EFF learned that the IEEE's flawed e-voting standard (P1583) was rejected by its own working group in a 13-6 vote. The standard is being sent back to the drawing board, where EFF and others will continue to fight for security and accountability in the election process. Thanks to the hundreds of people who joined us to make this happen!
I write to express my deep concern about the IEEE Voting Equipment Standards Project 1583 and SCC 38 and to urge IEEE to immediately stop the balloting on both. This standard is of utmost importance, since it involves the most fundamental elements of democracy -- the right to vote and to have one's vote accurately counted. IEEE's duty to maintain an open, transparent, clear process, which always exists, is even more important here, as is its duty to ensure that the substance of the standard represents the best scientific thinking about securing elections. Please note that I have great respect for the IEEE and believe that, if handled correctly, the IEEE standard for electronic voting machines could provide much needed assurance to people around the world. I understand that the Project was created with this goal and believe that it has been unfortunately diverted from that goal. I am concerned that IEEE's failure to correct the path of this standard will tarnish the IEEE's reputation. First, I am deeply concerned that the draft standard that has been put to ballot does not reflect the current scientific consensus about the security and other performance requirements for electronic voting equipment. For example, one of the most important goals of voting equipment standards is to ensure that the equipment has accurately recorded a voter's intention, something that only the voter herself can confirm. A large number of scientists and security experts have raised this basic issue, and have embraced a requirement that voting machines be voter verified or verifiable. One such method that is already available from multiple election machine companies is a provision that the machines produce a paper ballot for each voter that allows a voter to see a summary of her votes to confirm them. The paper ballot is then separately stored and is available for audits, in the case of dispute or for a recount. I am shocked to learn that the IEEE standard does not require at least this level of security. Additionally, I am alarmed at reports that the P-1583 and SCC 38 processes have been seriously flawed, with the result that those who are critical of the draft have faced unreasonable and shifting procedural roadblocks to their participation. I am deeply troubled by reports that new members of the various committees are being denied the right to participate until current members vote on their participation, that additional barriers (with obscure deadlines) are being placed before those who wish to vote on the ballot. Additionally I am concerned about reports that there are ongoing failures to timely circulate the current standard document and the comments received, or to give sufficient notice of meetings and to publish minutes of meetings. All of these make it difficult for interested parties who are not affiliated with large companies to participate, the exact thing that I believe is necessary for this standard to be perceived as secure and fair. Finally, I note that the IEEE Code of Ethics indicates that IEEE members should avoid "real or perceived conflicts of interest." I am concerned about reports that the leaders of these committees, who have reportedly been rushing the process, are affiliated with the vendors of electronic voting machines. Again, given the paramount role that voting machines play in democracy, the IEEE should take extra care to avoid even the appearance of conflicts of interests. Accordingly, I call upon IEEE to take the following steps: 1. Immediately stop the balloting currently on P-1583 and SCC 38. 2. Ensure that P-1583 and SCC 38 follow IEEE's best practices, including, at a minimum: a. Membership in the committees should not require a vote by current members and members who wish to vote must be given at least 30 days written notice of any requirement that they register to join a ballot group. b. Substantive meeting agendas, the current version of the standard and all pending comments by members should be posted or circulated 30 days in advance of each meeting; minutes should be posted no more the 7 days after each meeting. Agendas, drafts of the standard, comments and minutes shall remain posted. Members should receive at least 30 days notice for all meetings, whether in person or via telephone. c. Elections of officers of the working group should be made by majority vote of members of the working group, with recall by a vote of three-quarters of the members. d. Clear disclosure of all corporate and other affiliations of all participants should be required. Membership in the Committee especially, should be carefully balanced to avoid the actual, or the appearance of, control by those affiliated with voting machine vendors. 3. Require that the IEEE standard reflect the current views of the necessary security for voting machines and, at a minimum, that the standard include a requirement that voting be voter verifiable, with at least the same security level as current voter verifiable paper ballots. Thank you for your time. Sincerely, |
