|
Why isn't the North Carolina State Board of Elections having the Unilect machine re-examined?
Why is the NC BOE allowing Burke and Carteret Counties to use these machines after all of the votes lost on them, and after these same machines failed examination and were decertified in Pennsylvania? Why are citizens being forced to vote on equiptment known to lose votes! Why isn't the State Board of Elections doing more to protect our votes?
Note to Carteret Citizens - Take Action:
Contact your county commissioners and your county Board of Elections, tell them you want the machines replaced!
1. County commissioners email their individual email, phone and fax numbers here:
2. County Board of Elections office and tell them the same. County BOE members here :
3. Email Senator Daniel Clodfelter and ask him to let the Public Confidence in Elections Act, bills, S223 and H238 out of committee for a vote. Senator Clodfelter's contact inf here _________________________________________________________
Burke and Carteret County problems with Unilect Patriot Voting System
Media coverage, and new information about undervotes caused by Unilect in both counties. ___________________________________________________________
Burke County, North Carolina. 34,604 people went to the polls, 3,842 did not cast vote for president.
http://www.islandpacket.com/24hour/politics/story/1942288p-9916474c.html http://newsobserver.com/news/story/2029356p-8413353c.html
Burke has the Unilect Patriot DRE, purchased in Feb 2000, firmware 2.52 installed May 2000, with Intellect VS 2.45 tabulating software, installed in May 2000. Has modem capability. This is the same as was decertified by Pennsylvania. __________________________________________________________
Carteret County problems with Unilect Patriot
Carteret County, North Carolina. The UniLect Patriot System.
4,438 votes lost by faulty control unit. One statewide election could not be settled,
the democratic contender for Commissioner of Agriculture finally withdrew from the
contest as there was no other solution except for a new statewide election.
"the problem in Carteret County was in a control unit. That is, the Patriot is
a daisy-chained voting system where a bunch of touchscreens are all
plugged into a central control device (from which voting machines are
activated, etc.). The central unit in these counties didn't have
enough memory (essentially) to hold more than a small number of
electronic ballots."
__________________________________________________________
What the Media Missed:
For Burke County, the media only paid attention to the Presidential contest, and failed to question the ballot design theory proposed by the NC BOE.
For Carteret County, the media only latched onto the obvious problem of the 4,436 votes lost on the control unit. They completely missed the undervote story.
Burke County's undervotes are much more shocking when you look at the precinct data.
The ballot design along with our "straight ticket voting law" was blamed for the problem,
Yet this would not account for the high undervote for Governor:
with the lowest undervote rate at the Morganton 4 precinct at 3.86
and the highest undervote by precinct for Governor at 14% at the Smoky Creek Precinct.
Overall, Burke County's total average "undervote" for the Presidential contest was 11%, and the total for Governor was 7.117%, indicating that more was involved in voting problems than the straight ticket voting and ballot design.
25 of the 29 regular voting precincts had a 6% or higher undervote for Governor, indicating that the problem with casting or counting the votes was likely to be a voting system problem.
How did Burke County compare to other counties in their undervote for President and for Governor?
The undervote rate for all DRE counties for president averaged 2.05, with Burke at 11%.
The undervote rate for all DRE counties for governor averaged 3.03, with Burke at 7.12%.
(A detailed analysis of DRE by precinct showed that some had extremely large overvotes for president, therefore causing the overall average undervote to seem lower.)
OpScan counties fared much better overall.
The undervote rate for all OpScan Counties for pres averaged 2.19.
The undervote rate for all OpScan Counties for gov averaged 2.00
Cartert County undervote problems by precinct examined -
It was not highly publicized, but Carteret County had problems at various precincts with a large number of undervotes. This is aside from the votes lost during one stop voting due to the 3,005 vote limit on the software. These undervotes are from ballots cast, but no vote registered for a particular contest.
Citizens should not be compelled to vote on systems that can and do throw their votes away in such astonishing numbers.
IF banks threw away 5 to 10% of the money deposited, the government would shut them down. However, when this happens to our votes, the "currency" of democracy, nothing is said.
Some Carteret Precincts with astonishing undervote rates:
At the Peletier Precinct, at the COMMUNITY BUILDING, 252 Peletier Loop Rd, of 455 voters who went to the polls,
7.91 % did not cast vote for President,
9.45% did not cast a vote for governor, and
8.35% did not cast vote for US Senate.
At the Morehead Nr 1 precinct, at the MOREHEAD PARKS & REC. BUILDING, 1600 Fisher St.
Of the 807 voters who went to the polls,
7.063 did not cast vote for president,
9.045 did not cast vote for Governor, and
7.56% did not cast vote for US Senate.
At the Morehead 4 precinct, OPEN DOOR BAPTIST CHURCH, 227 Highway 24, out of 1,039 voters,
6.256% did not effectively cast a vote for president,
8.18% did not cast a vote for Governor, and
5.87 did not cast a vote for US Senate.
At the Stella precinct, of 457 voters,
5.69 % of all ballots did not register a vote for president,
8.75 did not count for governor, and
6.35% did not count for US Senate.
At Cape Carteret, Ceder Point, at WESTERN PARK COMM. CENTER, 275 Old Highway 58, of the 1,271 voters who turned out,
5.43% did not cast a vote for president,
7.24 did not cast a vote for governor, and
5.43 did not count for US Senate.
Here are North Carolina's statutes regarding electronic voting:
"§ 163-165.4. Standards for official ballots.
The State Board of Elections shall seek to ensure that official ballots throughout the State have all the following characteristics:
(1) Are readily understandable by voters.
(2) Present all candidates and questions in a fair and nondiscriminatory manner.
(3) Allow every voter to cast a vote in every ballot item without difficulty.
(4) Facilitate an accurate vote count.
(5) Are uniform in content and format, subject to varied presentations required or made desirable by different voting systems
"§ 163-166.7. Voting procedures.
(C) The State Board of Elections shall promulgate rules for the process of voting. Those rules shall emphasize the appearance as well as the reality of dignity, good order, impartiality, and the convenience and privacy of the voter. Those rules, at a minimum, shall include procedures to ensure that all the following occur:
(1) The voting system remains secure throughout the period voting is being conducted.
"§ 163-165.7. Voting systems: powers and duties of State Board of Elections.
The State Board of Elections shall have authority to approve types, makes, and models of voting systems for use in elections and referenda held in this State. Only voting systems that have been approved by the State Board shall be used to conduct elections under this Chapter, and the approved systems shall be valid in any election or referendum held in any county or municipality. The State Board may, upon request of a local board of elections, authorize the use of a voting system not approved for general use. The State Board may also, upon notice and hearing, disapprove types, makes, and models of voting systems. Upon disapproving a type, make, or model of voting system, the State Board shall determine the process by which the disapproved system is discontinued in any county. If a county makes a showing that discontinuance would impose a financial hardship upon it, the county shall be given up to four years from the time of State Board disapproval to replace the system. A county may appeal a decision by the State Board concerning discontinuance of a voting system to the superior court in that county or to the Superior Court of Wake County. The county has 30 days from the time of the State Board's decision on discontinuance to make that appeal.
Subject to the provisions of this Chapter, the State Board of Elections shall prescribe rules for the adoption, handling, operation, and honest use of voting systems, including, but not limited to, the following:
(1) Types, makes, and models of voting systems approved for use in this State.
(2) Form of official ballot labels to be used on voting systems.
(3) Operation and manner of voting on voting systems.
(4) Instruction of precinct officials in the use of voting systems.
(5) Instruction of voters in the use of voting systems.
(6) Assistance to voters using voting systems.
(7) Duties of custodians of voting systems.
(8) Examination of voting systems before use in an election.
Election Rewrite/Ballots And Voting http://www.ncleg.net/gascripts/BillLookUp/BillLookUp.pl?Session=2001&BillID=S17 http://www.ncleg.net/Sessions/2001/Bills/Senate/HTML/S17v8.html |
New: See the 2nd video exam that got the Unilect Patriot decertified in PA:
The videos show that:
- the votes are lost because of the defects in the voting machines.
- The fault is not the election director's
- The election officials/precinct workers cannot prevent most or all of the undervotes
The video shows that the huge undervote in Burke countyNC can be attributed to the voting machines, not the election director. Carteret also has an undervote problem.
2% undervote is considered a normal phenomenon, but in NC we seem to have a much higher one.
The North Carolina State Board of Elections has taken no action to decertify or even examine the Unilect used
in Burke and Carteret Counties.
Unilect Decertified in Pennsylvania - Fails Test AGAIN!
Patriot fails second examination, new concerns rasied about the security and accuracy of the machine:
The state of Pennsylvania just released it's official report on the second Unilect re-examination. The exam was conducted on April 22, and Unilect's performance raised more concerns than even the first examination in February. The report only became available
via the SOS website May 8th, 2005.
05/07/2005 Report gives Patriot Voting System failing grade after UniLect Corp.'s Patriot touch-screen voting system is unreliable, and company officials have given unsatisfactory answers when questioned about problems in two re-examinations, Pennsylvania Secretary of State Pedro Cortes said in a report released Friday.
Cortes' report was based on the findings of Carnegie Mellon University computer professor Michael Shamos, who conducted the most recent re-examination of the Patriot system in Harrisburg on April 22. Article
Some highlights from the PA Secretary of State's report:
Page 4 -"During the reexamination, the Department and its expert again observed, similiar to the February 15th reexamination,
that the System failed to reliably sense touches.
This behavior continues to raise concerns that voters may inadvertently cast undervotes for various offices."
The vendor has been unable to explain the significant undervote problems experienced in all three of the Pennsylvania counties that utilize the system.
See page 4 where it states:
Therefore, the System is not "safely and efficiently useable in the conduct of eletions"
nor is it "suitably designed and equipped to be capable of absolute accuracy," as required
by section 1107-A(11) of the Election Code, 25 P.S. 3031.7 (11). The failure to
recognize voter choices also violates section 1107-A(13) of the Election Code, 25 P.S. 3031.7(13),
which requires accurate tabulation of votes cast.
See page 5 for issue of security deficiencies:
"However, the System continues to allow manual adjustment of vote totals
without a sufficient audit trail. The system merely records the fact that an adjustment was made and the time at which it was made. These observations continue to raise questions about the System's ability to meet the requirements of section 1107-A(12) of the Election Code, 25 P.S. 3031.7(12), regarding ballot security procedures to prevent ballot tampering.
IN other words, an operator can alter vote totals manually without leaving a real audit trail.
There is nothing to prevent tampering with the ballots.
First Examination Details:
State decertifies voting machines used in 3 counties
Tribune-Review Thu, 07 Apr 2005 9:10 PM PDT
Pennsylvania officials on Thursday barred three counties from continuing to use a touch-screen voting system that apparently contributed to a larger-than-usual undercount in the November election. |
Mercer County Pennsylvania: The UniLect Patriot system
7 or 8 out of every 10 votes thrown away in 41 precincts.
Two reports outline the defects in the way the Unilect touchscreens operate, including failure to count votes that are "cast":
Report by Professor Shamos, examiner of the equipment
Two most compelling statements from Shamos' report:
"At the Central Station, a capability for
"manual edit" of vote totals is provided. This
means nothing less than the ability change vote
totals to any desired numbers." ..
"Among the complaints received about the
Patriot system is that the touchscreen does
not function reliably. That is, when a voter
touches the screen, the touch is not
necessarily sensed, which results in the voter
incorrectly believing that she has cast a
vote."
More highlights from the Shamos report:
Shamos never recommended that PA certify the machines..
he had examined them twice, and advised that corrections to the equipment must be made before he would recommend certification. The state of PA certified them anyway without a 3rd examination by Shamos. In other words, the state of PA is at fault for certifying the machines. They were warned!
"In fact it is still the case today, as it was in 1993, that the touch panel is difficult to use, and some problems detailed in my prior reports have still not been eliminated." (page 14)
"It is simply not responsible to compel Pennsylvanians to vote on a system that results in an undervote for one out of every 14 voters."
Note - Professor Shamos said "The code provided did not correspond exactly to the system submitted for reexamination.
The source code for the PCU was labeled version 2.54, but the device presented was version 2.52."
"Worse, the scenario observed makes it possible for an insider to alter vote totals not by manipulating the proprietary files on the Central Station (which is difficult), but by changing the election setup on the Central Station, which is easy since software is provided to do it.
The mode of attack would be to code an election normally and make InfoPacks. Precincts would test the election coding and find everything to be in order. An insider would then change the election coding at the Central Station at his leisure anytime after the InfoPacks had been created up until counting on election night.
When the InfoPacks were returned, the votes on
them would no longer correspond to the original voting positions on the InfoPacks, and the vote totals would be "correct" but would be credited to the wrong candidates. This attack is plausible because we observed it in an inadvertent way and on a small scale during the examination."
...
"At the Central Station, a capability for "manual edit" of vote totals is provided. This means nothing less than the ability change vote totals to any desired numbers."
...
"Among the complaints received about the Patriot system is that the touchscreen does not function reliably. That is, when a voter touches the screen, the touch is not necessarily sensed, which results in the voter incorrectly believing that she has cast a vote."
"A different but possibly related problem is that the system occasionally enters a mode in which no touch at all can be recorded anywhere on the screen. This behavior can be observed at timing mark 2:30 on the videotape. The screen froze up and would not respond to any input."
"Between 2:30 and 2:52 on the videotape I can be observed operating the system properly, yet it did not record all of the votes I cast while being observed by Mercer County Commissioner Michele Brooks. It
was at times necessary to touch the screen three times in order to get a vote to register.".....
Pennsylvania has developed a 12-ballot test to determine whether a
system tabulates ballots in accordance with Pennsylvania law, and in
particular tests the Pennsylvania Method of tabulating overrides to straight party votes. This test is not intended to stress the system or exercise all possible logical paths in its software, but merely to check whether the system correctly implements the particular requirements of the Commonwealth.
This test was performed and all 12 ballots were tabulated correctly.
It was sometimes necessary to press the screen several times during the test to cause choices to register properly."
Pennsylvania uses the same equipment as Burke County,North Carolina:
County-level Windows personal computer for configuring ballots and
accumulating votes, software version 2.45.
Precinct control unit for activating voter terminals and accumulating
machine totals, firmware version 2.52.
Patriot touchscreen voting unit, firmware version 2.52
Burke Co., North Carolina has the Unilect Patriot DRE, purchased in Feb 2000, firmware 2.52 installed May 2000, with Intellect VS 2.45 tabulating software, installed in May 2000. Has modem capability.
Carteret Co. North Carolina has 45 Unilect Patriot purchased May 96, firmware 2.54 installed May 96, with Intellect 2.49 tabulating software installed in May 96.
Download all of the precinct data for Burke and Carteret County for the November 2004 election:
Burke & Carteret Undervotes 
|
