The States ' Inability to Comply with the Commission Voting System Standards Poses Grave Risks to the Nation
The Commission 's standards, as defined in their standards, with its more than 1,000 pages of software, hardware, and procedural specifications, contains certain standards that are quite likely impossible to meet. But the technoelection bill HR811effectively makes them the law of the land. What does this mean to our nation?
- There is no voting equipment in use or on the market today that complies with the 2007 standards never mind the next generation, due out in 2010, which is a complete software architecture change.
- The complex Commission standards have not been scoped by the industry and nobody knows what would be the ultimate price tag for a system that might meet the standards. With current systems, which rely on sloppy, rudimentary and primitive code base, priced at $6000/each, one can only imagine with the multifold increase in The Commission's requirements that a compliant system could easily run as high as $20-30,000. Public funding simply can not support this endeavor.
- The increased complexity of the requirements places a solid opaque wall between the voter and his vote. Democratic elections must be observable to any citizen. This violates that principle.
- HR811 mandates are broad and deep, and all are untested new requirements. Rolling out such deep and broad changes on a massive scale without preliminary pilot programs is a risky and reckless endeavor, with the nation 's stability at stake.
- Every change in election administration bears a heavy management cost. Personnel must be retrained, voters must be re-educated, storage, configuration, maintenance of equipment must be secured, etc. etc.
- The 2007 standards will be in place only 2-3 years before the next version becomes effective, thereby rendering moot all overhauls based on the first set of standards, and requiring another wholesale and expensive overhaul to the system.
- E-voting vendors will have free reign to build and charge whatever they like in accordance with The Commission standards.
- Given the impossibility of even trying to implement The Commission 's standards, states are faced with either citizen lawsuits or destabilization of their election systems through wholesale overhaul. Neither choice is healthy for the country.
- State election administrators lose the ability to administrate; they will just be implementing court orders, which, as we know, can be fairly random and hardly reflective of the will of the people.
New Added Step in the Voting Process
With the new mandated read-back technology ( "convert ballot selections into accessible form "), HR811 integrates one of the higher cost requirements from The Commission 's voting system standards: ballot text read back. With this new step in the voting process, election administrators will need to do a lot more between the primary and the general elections, adding to their workload and subtracting from their ability to focus on the basic requirements of managing secure, accurate, accessible and observable elections.
The ever increasing complexity of the software-driven mandates will result in an increased number of ballot configurations, etc. This will drive up costs both in configuration and election management (training, implementation, etc.)
By adding a whole brand new extra step in the voting process (the read back and text conversion), the technoelection bill is adding one extra step which accordingly spikes overall configuration and election management costs.
Untested, Uncertified, New Technology
The text read back device mandated in HR811 is, as yet, completely new, untested, and uncertified. In sworn testimony in Congressional hearings about HR811 , ES&S Senior Vice President John Groh, representing the Election Technology Council (e-voting industry lobbyist organization) made the following statement about the read back requirements:
It must be noted, however, that legislative language which requires the disabled community to verify a VVPAT record independently is currently problematic: there are many issues of feasibility and usability that require more thought and supporting research to identify solutions to these concerns.
It may be easy to state that an automated reader can be used to convert the text on the paper to an audio output, but that will only help for those languages that are supported by common text to speech readers.
There are concerns that graphical languages, such as Cantonese cannot be so easily accommodated. For verifying the VVPAT with audio, there are usability concerns. If the text to speech reader is to be a separate independent system, then the voter must transfer his or her headset to that independent system.
Considering the general capabilities of human memory retention, verifying anything but a short ballot would involve transferring the headset back and forth between the voting device and the independent verification system.
This process would greatly decrease the usability of the voting system for the voter. This requirement is a good idea, and one day there may be a solution that is practical, usable and cost effective, but there needs to be more meaningful dialogue and supporting research in this area before it is made a statutory a requirement.
This requirement for an accessible VVPAT can be met through alternative systems. This will require jurisdictions to provide extra voting systems at each polling location where the use of electronic voting is the chosen method of voting.
Interestingly, ES&S, once the sole distributor of the Automark, is now claiming that the Automark can actually meet HR811 read back requirements.
Due diligence warrants a secondary inquiry to the vendors as to why they have within a matter of months issued completely conflicting statements on this matter.
