Response to Jim Dickson's Recent Statements to Authorities

Three recent hearings included testimony by Jim Dickson, the Vice President for Governmental Affairs of the American Association of People with Disabilities (AAPD). In all three hearings, he gave very similar testimony regarding the 4400 votes lost by UniLect Patriot voting machines in Carteret County, North Carolina. His testimony shows a lack of knowledge about the facts and a lack of understanding about the operation of computerized voting systems.

In Mr. Dickson's testimony to the Baker-Carter Commission in April, 2005, he said:

"In the famous case in North Carolina where 4,000 people voted and the machine did not record their vote which was horrible – that was an old machine. Had that county purchased a machine that is accessible, it would have been impossible for those 4,000 people to cast a ballot. The machine would have shut off when the memory was full."

Mr. Dickson also testified at the election forum held on June 15, 2005, convened by Connecticut's Secretary of the State, Susan Bysiewicz. During his testimony, he asserted:

"Professor Fischer from Yale mentioned the problems with DREs in Carteret County, North Carolina. There have been a lot of stories in the news about problems with DREs. The problems stem from either old, obsolete, and inaccessible DREs, or they came from the challenges that our elections officials and poll workers face with any new system. In Carteret County, the voting machine that we're talking about was 20-year-old technology. Had Carteret County purchased one of the new, accessible touch screens that meet federal standards, the machine would have shut off when the database was full. There would not have been 4,000 unregistered ballots." (See http://www.ctn.state.ct.us/ondemand.asp. Click on "Forum to Explore Ideas for Connecticut's New Voting Machines (Date Recorded: 6/15/2005) " to see the video. Mr. Dickson's testimony starts at about an hour into the video at 1:15:53.)

In the Senate Rules and Administration Committee hearing on June 21, 2005, he repeated the same statements about the UniLect Patriot machines used in Carteret County:

"Nobody is buying those machines. Had they purchased accessible voting machines, it would have been impossible for somebody to vote and not have it recorded. That machine is 20 years old. The accessible machines now would shut off when the memory is full. The voter would know and the poll worker would know."

These statements, which Mr. Dickson repeatedly makes, are incorrect. Here are the facts:

	Malfunctions do not just occur on old equipment and new equipment unfamiliar to election administrators. Hundreds of malfunctions of new, federally qualified machines have been documented in news articles and election reports, and the majority of these occurred on systems that had been in use in the counties for at least several elections. Mr. Dickson appears to be unaware of, for example: - The hundreds of votes recently lost in Miami-Dade County on new, federally qualified equipment they have been using since 2002. - The efforts of ES&S to cover up the fact that certified software they had installed in several Indiana counties did not tabulate the votes correctly. - The recurring problems with phantom votes (more votes than voters) that the clerk of Bernalillo County, New Mexico has encountered with the new software from Sequoia. - The vote-switching that frustrated voters in at least 10 counties during the 2004 General Election, from Florida to Washington State.
	Mr. Dickson's assertion that the Carteret County equipment was 20 years old is wrong. Patsy Hardesty, Elections Director in Carteret County since 1991, told VotersUnite.Org that the Unilect Patriots used in Carteret County were purchased by the County in 1995 and first used in 1996, so they were about 9 years old, not 20, when they were used in the 2004 election. Furthermore, they are touch screen machines, similar to the "new" technology of Diebold, ES&S, and Sequoia.
	Mr. Sanderson, maintenance contractor for Carteret County, said that the software was upgraded last year before the 2004 election, so while the hardware is 9 years old, the software which provides the vote-recording and counting function is the latest technology from UniLect. Note that this is consistent with the claim on the UniLect website, which says: "the PATRIOT System is almost 100% software and UniLect will upgrade your county's software on an annual basis. Thus, in 5 years, your county will have a current year system, not a 5 year old system."
	In Michael Shamos' Power Point presentation regarding the UniLect Patriot, he points out that there are no federal standards covering capacity. VotersUnite.Org checked the proposed 2005 standards as well and found that they do not address capacity either. Therefore, there is no justification for Mr. Dickson's claim that newer equipment would have shut off when it reached capacity.
	In fact, the opposite is true. The new, federally qualified ES&S tabulation software does not shut off when it reaches capacity. It starts subtracting votes, as Ken Carbullido, Vice President of ES&S Product Development, admitted in a Nov. 8, 2004 letter to Guilford County. This malfunction occurred in three counties during the 2004 election: Broward and Orange Counties in Florida, and Guilford County in North Carolina.
	There is no justification for Mr. Dickson's linking of accessibility to functionality. They are independent features of the systems. Accessibility features affect the process of casting the vote, functionality affects the processes of recording and counting votes. Mr. Dickson's claim that accessible systems shut off when they reach capacity is not only false but also shows a lack of understanding of computer systems.

Note that the "federally certified accessible machines" Mr. Dickson advocates can only be references to the AccuPoll AVS 1000, the Avante Optical Vote Trakker (an optical scan system), and the AutoMark ballot-marking device. These are the only three systems that meet the FEC 2002 voting system standards (currently the official guidelines adopted by HAVA) and also provide standard features by which the blind and those with visual or mobility impairments can cast their votes independently. While two other Avante Vote Trakker systems — both electronic voting systems with a voter-verifiable paper audit trail — are fully accessible, they have not yet passed the testing required to ensure that they meet federal standards.

Other manufacturers' systems, such as the Diebold TS and TSx, Sequoia Edge and Advantage, ES&S iVotronic and iVotronic LS, and Hart Intercivic eSlate, either fail to meet the FED 2002 voting system standards or fail to provide important accessibility features needed by the disabled community, or both. In fact, the limited accessibility features each of these systems offers are options and might not be available on every system in a polling place.

In fact, voters with disabilities who have used the optional features of these partially accessible systems have serious complaints about their operation. (See for example: https://www.votersunite.org/info/RunyanOnSequoia.htm).

Regardless of Mr. Dickson's many inaccurate statements, he is correct that voters with disabilities must be provided an opportunity to vote independently. Therefore, we encourage decision-makers to exercise the same caution in purchasing voting systems as they use when making other large purchases. Rather than simply trusting the claims of the salespeople, decision-makers should investigate those claims, assess which products truly allow voters with disabilities to vote independently, and then take the actions necessary to comply with the spirit of the HAVA accessibility requirements.

It's easier to make it up
than look it up.
~ Lynn
ASU graduate student,
whose last name I can't remember