Back on 31.Oct the EAC‘s Technical Guidelines Development Committee (TGDC), released its recommended voluntary voting system guidelines (VVSG) for public comment during a 120 day period ending around 05.March.2008.

The 600 page guideline (which is why the online version is so valuable) includes significant changes that would eliminate the use of existing
direct recording electronic (DRE) voting systems in the majority of states if
the EAC adopts the recommendations.

During a recent EAC public hearing in Austin, TX, ITIF Senior Analyst Daniel
discussed these recommended guidelines and the importance of innovation
for improving our voting technology.

Specifically, Castro’s statement discussed the
need for functional standards, as opposed to design standards, to promote
innovation in voting systems. Interestingly, Castro also disagreed with the recommendation that
all voting systems be "software independent" and described how the real goal of
any voting system should be to be "human independent." ITIF is actively involved (as increasingly will be the OSDV Foundation) in the VVSG conversation and held an interesting workshop back on 06.Dec with audio available on that page.

The relevance of the VVSG to the OSDV Foundation cuts across all of our work. It is our goal to have all OSDV Foundation design specifications certified by the Elections Assistance Commission, and importantly for our work to complement the VVSG.

To better explain the relevance of Castro’s remarks, I’m going to borrow liberally from Castro’s statement in commenting about the relevance to the work of the OSDV Foundation.

Functional standards define the minimum operational requirements to which a system must conform. Since functional standards do not define any specific technology or process, they are flexible enough to allow researchers to develop new approaches to solve existing problems.

Design standards define specific technical requirements for voting systems. Some design standards may encourage competition and innovation. For example, the EAC can set common data formats to facilitate transparency and interoperability of voting system technologies and data.

The ITIF (and Castro) believe the EAC should refrain from proscribing standards that limit voting to specific technology. For example, the earlier version of the VVSG (in 2005) restricted, but did not prohibit the use of wireless technologies. These restrictions were intended to balance both short-term goals of improving voting system security with long-term goals of allowing innovation.

In contrast, the new recommended guidelines ban all radio frequency wireless communication in voting systems which presupposes that no future use of wireless communication can ever be conducted securely. Castro points out that enacting such design standards that prohibit certain technology creates a barrier to innovation. And we agree.

Castro points out further, that this is a chicken and egg problem. For example, while a future version of the VVSG could be modified to allow wireless communication, the EAC would likely only modify the VVSG if a vendor could demonstrate a working prototype that securely uses
wireless communication. However, a commercial voting system vendor would be less likely to invest in the R&D required to create such a system since there would be significant risk that the vendor would not be compliant with the VVSG.

This is where the OSDV Foundation plays in. Without requirements for commercial advantage or mandate, the OSDV Foundation can, indeed, produce the R&D necessary to design and prototype systems that leverage VVSG requirements. We strongly encourage the VVSG to embrace functional, not design standards. And the OSDV Foundation can handle the "heavy lifting" of applied R&D to advance innovation.

While I’m thinking about it, have a look at the survey of voting machine use in 2006.

Incidentally, something kewl: the audio relied on TalkShoe, which allows users to listen to a live audio stream online at or
use a standard telephone, Skype or ShoePhone to call in directly to the
conference .