From: Ellen Theisen at VotersUnite!
To: Logan, Dean
Cc: 'Pam Eakes'; 'Kelly Lippman'; 'Kathy Barker'; email@example.com; Constantine, Dow; 'Holly Jacobson'; John Gideon; Huennekens, Bill
Sent: 8/11/04 1:22 PM
I spoke with Bill today, telling him my concerns about installing new, unexamined software in King County (and the other five WA counties) so close to an election, especially after so many of the candidates were common knowledge. He pointed out to me that the changes would have been required regardless of whether the consolidated ballot was used, and I would like more technical information about that. Is there someone I could talk to who could tell me the details?
In addition, may I have the version numbers of the new firmware for the precinct scanners and central count scanners? Bill told me the new GEMS version is 1.18.23. The most recent version I've heard of before is 1.18.19, and I'd like to know:
- What were the changes for the versions in between? Are those changes included in 1.18.23?
- What jurisdictions are using the versions in between? - Has 1.18.23 been developed specifically for WA or is it installed somewhere else?
As Bill probably mentioned to you, I am concerned about the higher potential for errors in rushed, uninspected software -- and the much higher potential for vendor fraud when new software is installed after candidates are known, even higher than installing new software a month before the election. And considering the legions of attorneys that both parties are gathering for the November election, this rushed, last-minute change is ripe for legal challenges that could hold up certification and cost the state much money that could be better used elsewhere, especially since the new software has not been through the ITA process and no one has looked at the source code but those who programmed it.
Although you are running hundreds of ballots through the scanners during the certification and acceptance testing, it is crucial to also conduct random, manual audits of the machines after the election. Diana Housden is planning to manually count one large precinct and two smaller ones in order to ensure that the election in Klickitat is accurate. This is about 16% of her precincts. We were thinking of 30% in larger counties just to make sure that the sampling was large enough to catch errors and any other types of anomalies. At least that percentage of votes for federal offices should be audited.
On another note: I was telling a friend in Palm Beach County about the excellent security procedures you have in place in King County. He and a friend are meeting with Theresa LaPore to discuss security procedures there. Both were surprised to learn that observers in King County are allowed to watch the tabulation of the votes on the central computer. She won't even tell them where the central tabulator is. They were also surprised to learn that the precinct totals are posted at the polling places in King County, and were hoping to get some documentation from you about both these procedures. The idea would be to show to her that an election director in a populous county is using such procedures, hoping to get her to adopt them as well.
Do you have a written guide outlining the procedures you have told us about -- including the fact that one person is never allowed to use the central tabulation computer alone? It would be of great help to them.
Thank you so much,