Now for all you activists jumping with glee over the opportunity to sue the pants off your corrupt state and local election officials, bear one single thought in mind: Florida 2000.
This single litigative challenge to an election outcome from a single state threw the nation into chaos and resulted in the Judiciary Branch (about as nonrepresentational a branch as we have) selecting the President. Is this the future you want?
Hold that thought and consider four things now:
1) The testimony of Jill Lavine regarding the time required to count DRE toilet paper trails by hand
2) The results of a Cuyahoga County study regarding error rates in counting toilet paper trails
3) An analysis of the conflicting and impossible mandates combined with the nonsynchronous effective dates of the Holt Bill provisions
4) Another choice different from the false choices of "Holt Bill now or something worse or nothing at all for 2008"
My name is Jill LaVine, I'm from Sacramento, California.I'm the Registrar and I have been working in elections for a little over 20 years.... A total of 1,612 valid ballots were counted at the early voting locations. And at this time I believe we were the first jurisdiction in the nation to go out with the voter verified paper audit trail. This experiment was very closely watched and it was controlled under some very controlled conditions. We had experienced people at each staff. We had -- from the vendor at each of the polling places.
The equipment had to meet of course all of the Secretary of State's requirements and our requirements and expectations. At the end of this project, knowing the California code requires that during the canvas of any vote that 1 percent of the precincts chosen at random will be manually recounted to verify the equipment, as part of the canvas, we chose one of the units or one of the polling sites and recounted these ballots.
The precinct we selected had 114 ballots. Because it was possible for a voter from any one of the 246 ballot types in the county to vote at the early sites, it made this recount very difficult and our tally sheet consisted of not just one page but several pages to accommodate all the choices.
We had four teams of two sit at tables with tally sheets to handle all the contests. The paper ballots were held together with large binder clips. Because they varied in length from 11 inches to over 20 inches, they rolled and it was very difficult to handle. I was watching several of the teams. They would put a brick on paper and paper weighted each end of these little curled ballots and start counting, and as soon as it moved or something bumped it, it rolled back up again and they'd be starting all over. The vendor also used a heat sensitive thermal paper that left kind of an icky residue on our recounters' hands, and so they said, can we have some rubber gloves? So those were provided, too.
We allowed provisional voting for this early -- for this ballot project and processing the provisional ballots was a very quick and easy process. We also allowed for write-in votes in this process and that was very quick and easy also, because in one case presentation of the reports made it very easy to count those write-in votes. So knowing that this project was under scrutiny, we verified the number of voters on the machine with the report. We verified the report with the paper record. Then we verified the machine totals with the paper records, so we did several cross checks to make sure we got it all together. And when the counting was all completed we were off by one ballot.
So what we learned is after printing out a report, that a fleeing voter who actually voted who didn't, you know, push the cast button, cast ballot button didn'tproduce a paper record for privacy reasons. So therefore, going back to the report, we found the fleeing voter and then we actually confirmed the number of that voter and we took that activity report and everything came out right.
But it took 127 and a half hours to recount 814 ballots, or approximately an hour and 15 minutes for each ballot.
The number from the machine count did match the paper of votes for the paper ballots exactly. Now, we are very thankful that this project was a November election, because had it been a primary election in California with our eight parties and our three non-partisan crossover opportunities that California allows, I think we'd still be counting. Also the paper audit trail did not print in Spanish, so we recounted in English only. This is before we had any true rules about what a paper audit trail should be and we were kind of stabbing in the dark here.
We were grateful that there were no challenged contests and it was not necessary to count any more than the 114 ballots. Otherwise, there would be significant delays in those election results.
I want you to know that we canceled that RFP and we've learned now with the third time after that.2) Study on toilet paper trails:
In April 2006, prior to the May 2 primary, the Cuyahoga County Commission contracted with the Election Science Institute (ESI) to conduct a comprehensive review of how their new voting system actually worked on an election day. ESI's report, including the performance of the Diebold Accuvote TSX voting system, was released by the Cuyahoga County Commissioners:
"…members of the manual count team found that 10 percent of the paper ballots were physically compromised in some way."