vn9

VotingNews Brief #14:

Ciber Official Says His Company Will Test Only the "Voting Part" of SERVE, not the "Internet Part"

Los Angeles, California
November 25, 2003

By Marc Strassman
Voting Technology Reporter
The Latest VotingNews

Before the SERVE remote Internet voting system can be used in seven states in the 2004 elections, it needs to be certified by election officials in each of those states. Before that can happen, it needs to be tested by an Independent Testing Authority (ITA) and found "qualified" by the National Association of State Election Directors, acting through The Election Center.

The tests run by the ITA are to see if the system conforms to the standards set by the Federal Election Commission (FEC), as set out by it in its Federal Voting Systems Standards (FVSS).

NASED's "General Overview for Getting a Voting System Qualified"

In these NASED Guidelines it says:

"The FVSS contains all the requirements for punchcard, lever, optical scan or direct recording equipment systems. Provisions for Internet enabled voting systems will be included in future revisions.”

So there are not currently any standards at all for remote Internet voting systems, such as SERVE, which therefore needs to be qualified under standards that don't exist.

Not only that, but responsibility for updating the FVSS to include Internet voting (something I asked them to do in 1997, 1998, and 1999, to no avail) has been transferred (by the Help America Vote Act [HAVA] of 2002) from the FEC to the Election Assistance Commission (EAC), and, unfortunately, the Republicans in the Senate haven't yet signed off on allowing President Bush's four nominees (two Democrats and two Republicans) to be members of this commission to proceed to confirmation by "UC" (unanimous consent) during the waning days of this session of Congress. Without their confirmation, there is no EAC.

Therefore, SERVE needs to pass muster under standards that don't exist and that can only come into existence under the authority of a commission that itself doesn’t exist. As an exercise in logic (or illogic) these are exciting developments, but as a practical matter they present some real problems for the successful launch of SERVE.

So I spoke with Shawn Southworth, ITA Practice Director at ciber, one of the very few ITAs and the company that has been contracted to conduct the software tests on SERVE. He told me that the contract to test SERVE was between ciber and the Defense Department but that it included a non-disclosure component that prohibited him from saying anything more about it, including how much money was involved.

Asked about the ability of his company to test the SERVE remote Internet voting system against a standard for remote Internet voting systems that didn't exist, Mr. Southworth replied that "nobody has a standard for an Internet voting system," confirming the assertion that the current FVSS don't have such criteria.

He repeatedly said that ciber would be testing the "voting part" of SERVE, but not the "Internet part". When told that SERVE was "essentially an Internet voting system," he said that ciber would be testing the "voting" part of SERVE, not the "Internet" part.

Asked to account for the "Internet part," he said that the Internet was merely "the medium it’s transferred on," which he said was no different than a floppy disk, hard drive, telephone, or pencil currently used as part of the voting process and, as hardware, was therefore outside the purview of ciber's testing contract.

In this way, the ciber spokesman was able to brush away the seeming difficulty of testing something to a non-existent standard by saying that the non-existent standard (for remote Internet voting) didn't enter into their part of the process at all, although one might think that it was central to it.

Mr. Southworth finished up by referring me to Brian Hancock at the FEC (and maybe the EAC) for further information.