Wisconsin held its 2006 federal primary on September 12, 2006. As usual on election night election officials stated that everything went smoothly. But, even with these election night assurances the reality is there are 3 counties in Wisconsin with reported troubles. The counties are Winnebago County (which contains the City of Oshkosh), Waukesha County, and Milwaukee County.
In Winnebago County the blended system of Diebold AccuVote OS optical scanners and Diebold AccuVote TSx DRE's did not integrate as advertised. The main technical problem is the 2 sets of components to the blended system do not actually blend. Fidlar Election Company is the Diebold representative in Winnebago County. Fidlar stated a product called the "accumulator" would allow the two technologies to "blend" and produce precinct-level reporting of candidate totals.
There are two problems with the "accumulator" product sold in Winnebago County. The first problem is technical. The accumulator does exist or at least is not available for sale. The second problem is a legal one. Wisconsin has not certified the use any Diebold system using the "accumulator" so even if the accumulator were delivered by November 7, 2006, state law would prohibits its use. Using uncertified voting systems is violation of Wisconsin statute WI 5.40(2). The matter of the possible fraud of selling this system has been turned over to the County Counsel for Winnebago County. County Supervisor Jef Hall is also following up by proposing a resolution to hold back a portion of the payment to Diebold equal to the overtime required to hand count the votes the “accumulator” vaporware could not accumulate.
In Waukesha County the blended system of Sequoia Edge DRE’s and BRC/ES&S/Sequoia Optech Eagle IIIP scanners did not blend as expected. A Milwaukee Journal Sentinel article focused on how inexpedient the system is without interoperable equipment. Buried in the article are the following gems:
• [County Clerk] Nickolaus and her staff resorted to correcting the city's results manually - a process that continued until 1 a.m
• Waukesha City Clerk Thomas Neill said city officials had been unable to test the county's program for reporting results because they were busy readying their touch-screen voting machines. Those machines arrived just a week ago, Neill said, adding that the new equipment turned out to be inoperable Tuesday. [Such testing is required by Wisconsin state statute WI 5.84]
• Computer glitches, inoperable equipment and other problems troubled Tuesday's primary balloting in Waukesha County, resulting in one candidate mistakenly being posted as winner of a race only later to be declared the loser.
To date Ms. Nicklaus has refused to explain the nature of the interoperability defects, exactly how the results were updated by hand or how the candidate went from winning to losing after the corrections. Many in the area are asking the obvious question what happened between 8:00 pm Tuesday and after 1:00 am Wednesday?
In the City of Milwaukee and Milwaukee County have three reported problems to date which seem unrelated. The system use in Milwaukee is the blended system of AutoMark and BRC/ES&S/Sequoia Optech Eagle IIIP scanners.
The first reported problem is the two system components failed to blend as advertised. In this case the ballot for which the Automark were a different size than the ballots used for the Optech scanners. This interoperability problem was know to the City Election commission. The “solution” was to place every AutoMark’ed ballot into the segregation box for ballots with write in candidates. At the end of the night the AutoMark’ed ballots would be recreated on an Optech ballot using the procedures described in statute for an absentee ballot which a scanner will not read. The second Optech ballot would then be fed into the scanner to contribute to the ballot totals. The problem reported early in the day was the AutoMark ballots were mistakenly given to general voters who then marked a ballot which the Optech scanner would not accept.
The second problem is the number of ballots cast for a polling location hosting more than one ward printed the total of ballots cast for all wards in the location as the number of ballots cast in each ward. Thus a location with 3 wards of 10, 20 and 30 ballots cast for wards 1, 2, and 3, would print an end of day report stating 60 ballots were cast for ward 1, 60 ballots for ward 2, 60 ballots for ward 3. That the simplest statistic on the report (number of ballots scanned) was in error, was seen no reason to doubt the more complicated calculations on the poll tapes (e.g. number of votes in the primary race for County Sheriff). The error was then propagated to the central municipal Unity software when the memory packs were uploaded. This prompted a hand recounting of the total number of ballots for each ward which as of today is not yet complete.
The third problem is a video taped documentation (link to be added) that the Optech Eagle in the combined location (wards 258, 259, and 265) displayed 586 total ballots cast for the three wards, printed 576 total ballots cast for the three wards, and the poll books show 588 ballots distributed to electors (absentee or in-person). Clearly, the Optech Eagle has at least two counters (one for the LED display) and one for the printed tape report. It is fascinating these two machine counters were not consistent.
Comment on This Article
You must login to leave comments...
Other Visitors Comments
You must login to see comments...