2004 Ballot Recount: Observer Report
- December 21, 2004:
Report by Green Party Observer
"The Board hand-counted a precinct with 609 votes cast ... I don't think that the precinct chosen in Guernsey was chosen randomly ..."
"There was also a man there from Triad, the company that manages the machines and computers that make the tally. He had come earlier and tested the equipment. I asked about this, wary of someone fiddling with the machinery. Malinak said that she and the deputy director had had him test 'prelogical and accuracy.'"
"... ballots in Guernsey had been put in ... metal boxes whose lids had been sealed with wire loops and wax ... The boxes were stacked up in the room with none of the seals broken ... I think that Guernsey has a good system for keeping the ballots secure between counts."
"Guernsey is a Republican county ... [Another observer and I] went through the signature book page by page, and we found no funny stuff regarding provisionals and absentees. Specifically, in this precinct there were no provisionals and all absentees were accounted for."
"... when we asked for the signature book they brought it out immediately"
Green Party Observer
December 21, 2004:
Report by Green Party Observer
The recount in Guernsey went smoothly. The Board hand-counted a precinct with 609 votes cast ( = 3.3% of the 18,182 votes countywide).
Two of us counted the signatures in the poll book for the precinct in question; the poll book showed no discrepancy (609 voters were recorded).
Guernsey uses punch cards, and the machine tally of these 609 votes exactly matched the hand tally. The Board then recounted the rest of the county by machine. I left before the machine recount was finished. I thought that the Guernsey board did a professional and accurate job. A more detailed account follows.
The Board began the recount around 5:30 p.m. on Wednesday, December 15, 2004. On hand were all four members of the Board (Donald Brown, Richard Carter, Mary Ann Hendershot, and Peggy Neuhart), the director Jacqueline Malinak, and the deputy director, Sandra Tilton. Guernsey is a Republican county, so the Republican Brown is chair; they have, however, made a Democrat, Malinak, the director.
In addition to the board and the staff, there seemed to be two of three clerical workers present. At the start there were five witnesses: myself, Witness N and Witness M for the Green Party, a man for the Kerry campaign, and one other gentleman who left after about a half an hour. No one ever asked us to identify ourselves or to show the letters we were carrying signed by David Cobb.
There was also a man there from Triad, the company that manages the machines and computers that make the tally. He had come earlier and tested the equipment. I asked about this, wary of someone fiddling with the machinery. Malinak said that she and the deputy director had had him test "prelogical and accuracy." I'm not sure what that means.
The punch card ballots in Guernsey had been put in "transfer cases" after the election. These are metal boxes whose lids had been sealed with wire loops and wax (?). The boxes were stacked up in the room with none of the seals broken. I took a photograph of this, should anyone care. I think that Guernsey has a good system for keeping the ballots secure between counts.
The recount began by hand-counting and then machine-counting a "test deck." This was at best an enigmatic procedure. A deck of about 50 test cards was hand counted. The majority of them had no vote for president (i.e., they were "undervotes"); I have no idea why. One of them was an intentional "overvote" with all four presidential chads punched out. After the hand count, the deck was run through the machine counter, and the tally declared to be the equal of the hand count.
It took Witness N and myself about 5 minutes to decide that this was the case, for it was hard to figure out how they had summed the hand count, and how that matched with the machine printout.
There was one short spat as we tried to figure all this out. Witness N. kept saying that he didn't understand the system and at one point the Democrat on the board, Richard Carter, said "we're not here to educate you on the computer technology."
In the end, I think that the test deck showed that the machine was counting accurately. There was one "team" doing the hand count. The question of "teams" was a bit confusing. We were told before we arrived that there would be two teams; when we got there they said there would be four. In fact, the hand count was done entirely by the director and deputy director a single team, and though there is supposed to be only one witness per team, there was no complaint about all of us leaning over and watching.
I can only surmise that the "four" mentioned teams referred to the staff on hand ready to work on the countywide recount but that recount was done by machine, and there was not much for a team or a witness to do.
After the test deck was run, they broke the seal on one precinct, "Monroe/Wheeling," and began to work through it card by card. Malinak & Tilton were the team. Malinak would hold each card up and announce "two" or "eight" or some other number, depending on which chad had been punched out. Tilton would keep a tally. Only when there was no vote or an over- or under-vote would both of them look at the ballot.
Provisional ballots and absentee ballots had been added to the ballots cast at the polling place.
It was Witness N. and I who went through the signature book page by page, and we found no funny stuff regarding provisionals and absentees. Specifically, in this precinct there were no provisionals and all absentees were accounted for.
I should say, too, that when we asked for the signature book they brought it out immediately and left us alone with it. As we went through it we had some questions, and the staff was helpful.The result of the hand count was:
342 for Bush
6 for others
= 609 total
When the cards were run through the machine the results were the same.
While the hand count was going on the staff had been opening the other transfer cases and preparing the ballots for a machine recount. This amounted to fanning all the cards to make sure they would feed through the machine and had no loose paper in them.
Once we all decided that the precinct count was accurate, and the signature book was accurate, they began to run all the other precincts through the counting machine. I left halfway through this, and don't know the final result.
There were one or two anomalies to be noted. The "volunteer manual" we received says "the board must randomly select whole precincts whose total equals at least 3% of the total vote." I don't think that the precinct chosen in Guernsey was chosen randomly; I think they said "3% of our vote is about 550 votes; what precinct is larger than that, but not too large?"
According to the manual, "the test deck...must not be the same test deck used for the official count." And: "At the conclusion..., the program must be retested using the pre-audited test deck."
I suspect the test deck was the same they had used before (none of us asked); and I didn't wait, but I would not be surprised to hear that there was no retest.
On the whole, however, I was impressed by the professionalism of the Guernsey Board of Elections.