Content uploaded by Ron Baiman
Author content
All content in this area was uploaded by Ron Baiman on Nov 12, 2016
Content may be subject to copyright.
Page 1 of 17
Official States Electronic Voting System Added Votes Never Cast
In 2004 Presidential Election; Audit Log Missing
By Peter Peckarsky, Ron Baiman, and Robert Fitrakis
Copyright 2006 by Peter Peckarsky
An electronic vote counting system included votes never
cast in the total vote count reported for the 2004 Presidential
election according to an official directly in charge of
conducting the election, the Republican Director of the Miami
County (Ohio) Board of Elections. The audit log for the system
is missing all information for the 2004 Presidential election.
The statement of this official is supported by the records
of the election obtained pursuant to law from the Miami County
(Ohio) Board of Elections and from the office of the Ohio
Secretary of State, Kenneth J. Blackwell, Chair of the Bush-
Cheney 2004 campaign in Ohio.
The votes involved were reported as having been cast on
optically scanned ballots. Under state law, the records
(including the audit log) were to be available to the public at
8:00 a.m. Eastern Standard Time the morning after the election.
This exclusive article is the first published report
disclosing the manner in which, according to the directly
responsible election official, an electronic vote counting
system whose audit log is missing relevant information caused
votes never cast to be included in the officially certified
result of the 2004 Presidential election.
When asked during a scheduled in-person interview in the
Board’s offices about the matter of the votes never cast which
were apparently added to the reported total by the electronic
vote counting system, the Democratic Chair of the Miami County
Board of Elections, attorney Robert Luring, for his complete on
the record response, said: “We thank you very much for the
information and the Board of Elections will consider it.”
An audit log provides an internal but not tamper-proof
check on the operation of an electronic vote counting system.
The portion of the audit log which should have been generated by
the electronic voting system for the November, 2004, election is
missing. The audit log contains no information from July 24,
2004, through and including April 13, 2005. This means the
Page 2 of 17
information was either deleted or was never generated.
Director of the Miami County (Ohio) Board of Elections
Republican Steve Quillen stated the audit log information for
the November 2, 2004 election should have been generated by the
system.
Former Deputy Director of the Miami County (Ohio) Board of
Elections Democrat Diane Miley referred to the absence of the
audit log information as “mysterious” and indicated the audit
log for the 2004 general election should have been generated by
the system.
Neither Mr. Quillen, the current Director of the Miami
County (Ohio) Board of Elections, nor Ms. Miley, the former
Deputy Director of the Board, had any explanation for the fact
that the audit log information for the November 2, 2004 election
was completely missing. Mr. Quillen has been Director of the
Board since February, 2002. Ms. Miley was Deputy Director of the
Board for over three and one-half years (from February, 2002, to
November, 2005).
When asked during a telephone interview about the missing
portion of the audit log and for any information he had which
might explain how and why the audit log information was deleted
(or never generated), Board Chair Roger Luring said he had
nothing to add beyond what Director Quillen said about the
missing audit log.
In the Ohio system of public election administration, a
bipartisan balance of major party supporters is theoretically
maintained at each level of the system. If the Chair of the four
member county Board of Elections is a Democrat, the other
members of the Board will be a Democrat and two Republicans. If
the Chair of a county Board of Elections is a Democrat, the
Director is typically a Republican and the Deputy Director is
typically a Democrat.
Miami County, Ohio, used a central count optical scan
voting system for the November, 2004, election.
The scanning hardware portion of the electronic vote
counting system (two Model 550 Central Ballot Scanners) was
manufactured by American Information Systems, Inc., of Omaha,
Nebraska (a corporate predecessor of Election Systems and
Software, Inc. (“ES&S”), of Omaha, Nebraska); the application
Page 3 of 17
computer programming for the election was performed by ES&S.
According to records reviewed in the Miami County Board of
Elections office, the application software and certain hardware
for the November 2, 2004, election was not shipped to the BOE
until Friday, October 29, 2004.
According to the minutes of the Special Meeting of the
Miami County BOE which was called to order at 2:00 p.m. on
Monday, November 1, 2004, the Media Test (sometimes referred to
as a logic and accuracy test) of the recently received hardware
and software commenced some time after 2:00 p.m. the day before
the election and was concluded sometime before 4:00 p.m. the
same day.
The electronic vote counting system used in Miami County
included a desktop computer used to tabulate the votes and
prepare reports. The system also included two vote scanning
machines (Model 550 Central Ballot Scanners) whose output on
Election Night was a series of twelve floppy disks containing
information from the scanned ballots.
The data from the floppy disks was read by an internal
floppy drive on the tabulating desktop computer running the
Windows 98 operating system manufactured by the Microsoft
Corporation of Redmond, Washington. The vote counting
application software being run on the tabulating desktop
computer was Election Reporting System version 1.6.1
manufactured and then programmed for the election by ES&S.
While the Board of Elections designated (using paper signs)
the scanning machines as Machine 1 and Machine 2, the ES&S
software recognized the scanner designated as “Machine 1" as
“Scanner 2.” The software recognized the scanner designated as
“Machine 2" as “Scanner 1.”
In the November 2, 2004 election in Miami County, Ohio,
there were three types of ballots which could lawfully be
counted.
The first type of lawful ballot was a regular ballot cast
by a voter at the polls on election day.
The second type of lawful ballot was a provisional ballot
cast by a voter either at the polls on election day or at the
office of the Board of Elections before the polls closed on
Election Day. After Election Day and before the official vote
Page 4 of 17
total was certified on December 1, 2004, the staff members of
the BOE decided which provisional ballots to count. The
decisions of the staff members were approved by the BOE when the
BOE approved the certified results.
The third type of lawful ballot was an absentee ballot.
To be counted, an absentee ballot from a person registered
to vote in Miami County and with a mailing address in the United
States was due at the BOE office by the time the polls closed.
To be counted, an absentee ballot from a person registered
to vote in Miami County but with a mailing address outside the
United States had to be postmarked on or before Election Day and
had to be received at the BOE office by November 12, 2004.
According to Director Quillen, all absentee ballots had to
be returned to the office of the Board of Elections; absentee
ballots were not accepted at the precinct polling locations.
The regular ballots were marked at a precinct voting
location and deposited in a ballot box by the voter. After the
polls closed, the ballot boxes were transported to the office of
the Board of Elections by the Presiding Judge, the chief of the
four election officials at each voting location. The votes were
not counted at the precinct voting locations. There was no means
to scan the ballots at the precinct voting locations.
Thus, the only non-provisional votes available to be
counted at the office of the Board of Elections as soon as the
polls closed were the absentee ballots received at the BOE
office by the time the polls closed. These absentee ballots were
the first ballots counted by the electronic vote counting system
on Election Night.
The absentee ballots had two numbered stubs attached which
indicated the ballots were to be used as absentee ballots. The
regular and provisional ballots had two numbered stubs attached
which did not indicate whether the ballots were to be used as
regular or provisional ballots. Regular ballots were used for
voters casting a provisional ballot. Once the stubs were removed
from the ballots, the absentee, provisional, and regular ballots
were indistinguishable.
According to Director Quillen, each scanning machine
produced six disks on election night. When the twelve disks
Page 5 of 17
produced by the scanners on Election Night were provided by
Director Quillen for inspection by a reporter each disk was
marked with a machine number (“1" or “2") and a disk number
(“1," “2," “3,” “4,” “5,” or “6").
According to Director Quillen and former Deputy Director
Miley, the absentee ballots were scanned in the first scanning
run on each of the two machines. This means the absentee ballots
from all 82 precincts were scanned in the first run on each
machine. In total, there were six different sets of ballots
scanned on the 6 scanning runs on Scanner 1 (Machine 2) and six
different sets of ballots scanned on the six scanning runs on
Scanner 2 (Machine 1). Each set of ballots was comprised of
ballots from a different set of precincts.
On Election Night, the ballots from somewhere between 6 and
11 precincts were scanned on each scanning run after the first
run on each machine.
(Director Quillen recalled that there was an informal
handwritten paper log made indicating which precinct’s ballots
were scanned on each of the last 5 scanning runs on each of the
two scanning machines. The Board of Elections was unable to
produce any copy or original of the log.)
According to Director Quillen, those ballots which could
not be read by the scanners and were thus rejected by the
scanners were immediately referred to Resolution Boards (each
Board consisted of an equal number of Republicans and
Democrats). A Resolution Board met immediately in the office of
the Board of Elections and marked a new replacement ballot
containing the same votes as those marked on any ballot which
could not be scanned. According to Director Quillen, these new
ballots were scanned through the appropriate scanner before the
ballots from any other precinct were scanned through the same
scanner.
According to Director Quillen, on all scanning runs after
the first scanning run on each scanner (in which first scanning
run absentee ballots from all precincts were scanned according
to the reports generated by the electronic voting system) all
ballots from one precinct were scanned as a group by a given
scanner.
According to former Deputy Director Miley, the replacement
ballots (sometimes referred to as “remakes” of ballots which
Page 6 of 17
were “remade”) were not always scanned through the appropriate
scanner before the ballots from any other precinct were scanned
through the same scanner and were not always scanned on the same
scanning run as the ballots from the same precinct which did not
need to be replaced. Deputy Director Miley said she did not
recall what actually happened but that it was possible that all
of the replacement ballots were run through the scanner on the
last scanning run.
According to Director Quillen and former Deputy Director
Miley, if ballots from 9 precincts were scanned on a given run,
the system should show a large block of votes from each of the 9
precincts and no votes from any other precinct. In the case of
the computer generated reports from some of the scanning runs,
this is what the records show.
However, on several runs, including the second run on
Scanner 2 and the last scanning run on Scanner 2, the records
show that there were not just a large block of votes from the
several precincts whose ballots were scanned but also relatively
small numbers of votes from many precincts.
The fifth scanning run on Scanner 1 contained large blocks
of votes from 7 precincts and 4 extra votes from 2 other
precincts.
The second scanning run on Scanner 2 contained large blocks
of votes from 9 precincts and 87 extra votes from 47 other
precincts.
The fourth scanning run on Scanner 2 contained large blocks
of votes from 11 precincts and 4 extra votes from 3 other
precincts.
The last set of ballots scanned through Scanner 2 (the
sixth scanning run on Scanner 2) and read by the tabulating
computer represented the last opportunity for the software to
add votes not actually cast to the election night results. In
fact, on the last set of ballots scanned through Scanner 2 there
were votes reported from every one of the 82 precincts in the
county.
The sixth and last scanning run on Scanner 2 contained
large blocks of votes from 9 precincts and small blocks of votes
from every one of the other 73 precincts. The number of extra
votes in small blocks on the sixth scanning run on Scanner 2 was
Page 7 of 17
522.
The four scanning runs referred to immediately above (the
fifth run on Scanner 1 and the second, fourth, and sixth runs on
Scanner 2) produced a total of 617 extra votes. In the view of
BOE Director Steve Quillen, these votes were added by the ES&S
electronic vote counting system.
Of these 617 ballots, 357 contained a vote for the George
W. Bush/Richard Cheney ticket and 225 contained a vote for the
John F. Kerry/John Edwards ticket. Thus, these ballots provided
an additional margin of 132 votes for George W. Bush and Richard
W. Cheney. Thirty-five of these 617 ballots contained a vote for
a third party candidate, an undervote (i.e., no vote), or an
overvote (attempted votes for two or more tickets or candidates)
for the Presidential/Vice Presidential election of electors.
Because of the manner in which the votes now deemed
improper by Republican Director Quillen were detected, it is
possible that additional, as yet undetected, unlawful blocks of
votes were added when a large block of votes was scanned from a
precinct.
If most or all of the replacement ballots were scanned on
the last scanning run as former Deputy Director Miley suggested
was possible, this would explain the presence of ballots from
each precinct in the last scanning run on Scanner 2.
About 9:30 a.m. Central Standard Time on November 1, 2006,
a telephone call was placed to Mr. Aldo Tesi, President of
Election Systems and Software, Inc., in Omaha, Nebraska, to ask
whether ES&S hardware and software added votes never cast by a
voter to the total vote count reported for the November 2004
election. The telephone call was referred to Mr. Rob Palmer,
Marketing/Communications Director for ES&S.
Mr. Palmer asked that questions be sent to him by e-mail
but was unwilling to promise a response by any specific time or
date. The e-mail was sent to Mr. Palmer on the morning of
November 1, 2006, and a prompt response was sought. In a later
telephone call on the morning of November 1, 2006, Mr. Palmer
confirmed that he had received the e-mail and said he would
attend to the questions in the e-mail as soon as he completed a
few telephone calls. At mid-afternoon on November 1, 2006, Mr.
Palmer said in a telephone conversation that he had not yet
turned his attention to the questions. Mr. Palmer said he might
Page 8 of 17
possibly be able to give some attention to the questions on the
evening of November 1, 2000, but offered no assurances of a
response on November 1, 2006, or on any other specific date. A
reporter requested that Mr. Palmer respond before the close of
business on November 1, 2006. As of 9:30 p.m. Central Standard
Time on November 1, 2006, no response to the questions in the e-
mail to Mr. Palmer had been received.
There were supposed to be several types of documents
available against which the results of the counting process
could be checked. These types of additional documents included
for each precinct a signature book, a pollbook, and a Report
Accounting for Ballots certified by all four precinct election
judges. In addition a countywide absentee database, countywide
report of provisional ballots cast and counted, and a list of
the precincts within the county indicating how many ballots of
each type (absentee or regular) were ordered for each precinct.
For various as yet unexplained reasons some of these
records are incomplete or unavailable from the Board. In
addition, a leak in the building where the ballots and other
records were stored required that these records be moved. The
records are now in a smaller space than before the leak and
somewhat disorganized.
The essential information on the Report Accounting For
Ballots (“RAB”) for each precinct is required by Ohio Revised
Code section 3505.26 and was supposed to be certified by the
four election judges at each polling place immediately following
the closing of the polls.
The RAB lists: 1) the number of blank ballots received, 2)
the number of ballots counted in the ballot boxes, 3) the number
of “walk-in” (i.e., provisional) ballots cast, 4) the number of
ballots spoiled, 5) the number of ballots not used, 6) the total
number of ballots used and unused, and 7) the total number of
voter signatures in the signature book.
If properly completed, an RAB provides a means to check the
results written on the floppy disks produced by the scanning
machines which are part of the ES&S electronic voting system.
However, in 5 precincts (in which the BOE certified 3,655
votes were cast) all entries on the RAB certificates were not
completed. In 72 other precincts, the number of votes reported
by the ES&S electronic voting system on Election Night does not
Page 9 of 17
match the number of voters listed on the certified RAB as having
cast regular ballots on Election Day.
Each election judge certified by the judge’s signature,
under penalty of perjury, that the judge would “discharge to the
best of my ability the duties of Judge of Election” and that the
judge would “endeavor to prevent fraud.” Thus, there is reason
to believe the numbers entered on each completed Report
Accounting For Ballots are correct.
Audit Logs
The ES&S system produces an audit log which, if it were
present, might allow an observer to determine whether there were
anything reflected in the log indicating an irregularity in the
operation of the vote counting hardware and software.
The audit log for the November 2, 2004 election is
completely missing. This means the information was either
deleted or was never generated. Director Quillen and former
Deputy Director Miley have stated that as far as they know the
audit log should have been generated. Neither was able to offer
any explanation as to why the audit log information for the
November 2, 2004, election is missing. The audit log contains
entries beginning on July 8, 2004 and ending on July 23, 2004,
and then beginning again (one line after the last July 23, 2004
entry) on April 14, 2005, and ending in 2006.
According to Director Quillen and former Deputy Director
Miley, the only elections in Miami County, Ohio in 2004 were a
primary in March, 2004 and a general election in November, 2004.
Former Deputy Director Diane Miley said the absence of the
audit log information for the November 2, 2004, election was
“mysterious.”
When asked about the missing audit log information, Miami
County Board of Elections Chair Roger Luring said he did not
have anything to add to the comment of Director Quillen who had
no idea why the information is missing.
About 9:30 a.m. Central Standard Time on November 1, 2006,
a telephone call was placed to Mr. Aldo Tesi, President of
Election Systems and Software, Inc., in Omaha, Nebraska, to ask
why all information concerning the November, 2004, election is
missing from the audit log. The telephone call was referred to
Page 10 of 17
Mr. Rob Palmer, Marketing/Communications Director for ES&S.
Mr. Palmer asked that questions be sent to him by e-mail
but was unwilling to promise a response by any specific time or
date. The e-mail was sent to Mr. Palmer on the morning of
November 1, 2006, and a prompt response was sought. In a later
telephone call on the morning of November 1, 2006, Mr. Palmer
confirmed that he had received the e-mail and said he would
attend to the questions in the e-mail as soon as he completed a
few telephone calls. At mid-afternoon on November 1, 2006, Mr.
Palmer said in a telephone conversation that he had not yet
turned his attention to the questions. Mr. Palmer said he might
possibly be able to give some attention to the questions on the
evening of November 1, 2000, but offered no assurances of a
response on November 1, 2006, or on any other specific date. A
reporter requested that Mr. Palmer respond before the close of
business on November 1, 2006. As of 9:30 p.m. Central Standard
Time on November 1, 2006, no response to the questions in the e-
mail to Mr. Palmer had been received.
Recount Day
The ballots were counted three times: on Election Night
(November 2, 2004), to prepare an official certification of the
results (signed on December 1, 2004), and at a recount (on
December 16, 2004).
An understanding of the disarray in the official records
(and the consequent difficulty of determining exactly how many
extra votes were added to the vote totals) may be gained from
considering the events on Recount Day - December 16, 2004.
The three recounted precincts were Precincts 004 (“Piqua 2-
A”), 030 (“Troy 3-E”), and 063 (“Bradford”).
At the recount, the ballots from each of the 3 recounted
precincts were counted by hand. Then the ballots from each of
the recounted precincts were run through the scanning equipment
and counted again. This was the first machine count on recount
day.
In two of the precincts, the hand count matched the machine
count. In the third precinct, the hand count was one less than
the machine count. The BOE decided the difference could be
disregarded.
Page 11 of 17
Then, all of the ballots from the entire county were run
through the scanning equipment as a group and counted yet again
in the second machine count on recount day.
The vote total for each of the three recounted precincts
obtained during the second machine count on recount day differed
from the total for each of the three precincts obtained during
the first machine count on recount day.
In Precinct 004, the difference between the hand recount
total of 558 and the second machine recount total of 566 (which
matched the officially certified total) was 8 votes.
In Precinct 030, the difference between the hand recount
total of 661 (the first machine count total was 662) and the
second machine recount total of 675 (which matched the
officially certified total) was 14 votes.
In Precinct 063, the difference between the hand recount
total of 440 and the second machine recount total of 443 (which
matched the officially certified total) was 3 votes.
The total difference between the hand recount totals and
the certified results was 25 ballots or an average of just over
8 ballots per precinct. A shift of just less than 6 votes per
precinct statewide from the officially certified total for the
Bush/Cheney ticket to the Kerry/Edwards ticket would have been
enough to change the outcome of the election in Ohio and the
nation.
The first machine count was compared to the hand count
number rather than to the certified vote total for each
precinct. This action amounts to a test of the vote tabulator
rather than a comparison of the recount totals with the
certified totals in the recounted precincts.
After verifying for himself in 2006 the difference between
the hand recount vote totals (and first machine count totals)
for the 3 recounted precincts and the certified vote totals for
the same 3 recounted precincts, Director Quillen remarked that
the vote counting machines must have added votes to reach the
certified totals.
When asked why the BOE at the recount compared the hand
counted number for each precinct to the first machine count
conducted that day rather than to the certified vote total for
Page 12 of 17
each recounted precinct, Miami County BOE Chair Roger Luring’s
comment was: “We thank you very much for the information and the
Board of Elections will consider it.”
On Election Day 2004, the members of the Miami County BOE
were Democrats Roger Luring and Miami County Democratic Party
Chair Kelly Gillis and Republicans Dr. Richard Adams and
Attorney Robert Huffman, Sr. Attorney Huffman died a few weeks
after the election of atypical pneumonia according to his son,
Robert Huffman, Jr., who succeeded his father as a member of the
Miami County BOE in 2005.
In a telephone interview, Robert Huffman, Jr., indicated
that, as far as he knew, his deceased father and former Member
of the Miami County Board of Elections Robert Huffman, Sr.,
thought the November, 2004, election was conducted properly and
the votes were counted accurately.
In an in-person interview, Board member Kelly Gillis,
indicated he did not know why extra ballots appeared to have
been added to the vote total by the electronic vote counting
system. In this regard, Mr. Gillis mentioned that he was a
member of a Resolution Board and that he did not know what
happened to the new ballots made by the Resolution Board to
replace ballots which could not be scanned by the scanning
machines.
After Mr. Gillis mentioned that he did not know what
happened to the replacement ballots made by the Resolutions
Board, Director Quillen was asked about the matter. According to
Director Quillen, the new ballots (or replacement ballots) made
by a Resolution Board were scanned through the appropriate
scanner before the ballots from any other precinct were scanned
through the same scanner.
According to former Deputy Director Miley, the replacement
ballots (sometimes referred to as “remakes” of ballots which
were “remade”) were not always scanned through the appropriate
scanner before the ballots from any other precinct were scanned
through the same scanner and were not always scanned on the same
scanning run as the ballots from the same precinct which did not
need to be replaced. Deputy Director Miley said she did not
recall what actually happened but that it was possible that all
of the replacement ballots were run through the scanner on the
last scanning run.
Page 13 of 17
Former Deputy Director Miley’s observations about the
replacement ballots were not obtained until mid-October, 2006.
When Director Quillen was contacted to ask whether replacement
ballots accounted for the extra votes he said were added by the
ES&S system and to seek access to the replacement ballots, Mr.
Quillen said he was too busy with the 2006 election to discuss
the 2004 election or to provide access then to records from 2004
but would be pleased to discuss the matter and provide access to
records after the 2006 election. In compliance with Ohio law and
when not faced with an impending election, Director Quillen and
the entire BOE staff typically provided prompt access to public
records requested.
With respect to the recount, Mr. Gillis said he did not
know why the Board failed to compare the hand counted vote total
for each of the three recounted precincts to the certified vote
total for each of the three recounted precincts.
Due to scheduling difficulties, it was not possible to
arrange an in-person interview with Dr. Adams. In a telephone
interview, Dr. Adams said BOE Chair Roger Luring was the
spokesperson for the Board and that Dr. Adams would let Chair
Luring’s comments be the comments of the entire Board.
ABSENTEE BALLOTS
There were discrepancies in the handling of absentee
ballots. For example, about 59 voters registered to vote in
precincts other than Precinct 030 were issued (and returned)
absentee ballots from Precinct 030 which was one of the
recounted precincts.
In mid-September, 2006, two boxes of absentee ballots were
counted in the Miami County BOE storage facility. The first box
had 2,338 ballots and the second box (which was found in
September, 2006, through the extraordinarily diligent efforts of
Miami County Board of Elections Deputy Director Pam Calendine)
had 2,268 ballots for a total of 4,606 ballots.
The Board certified to the Secretary of State that it
supplied 5,630 absentee ballots to voters asking for an
absentee ballot. The Board’s absentee ballot database indicates
that the Board received requests for 5,687 absentee ballots.
The Board of Elections certified that the number of
absentee ballots cast as of the time the polls closed on
Page 14 of 17
Election Night was 5,191. Thus the total number of absentee
ballots counted on Election Night should have been 5,191.
In fact, according to the reports generated by the
electronic voting system, the total number of absentee ballots
counted on Election Night was 4,667. In the three recounted
precincts, 134 absentee ballots were counted on the scanners on
Election Night. In order to facilitate the recount, the absentee
ballots cast in the three recounted precincts were removed from
the absentee ballots in storage and placed with the other
ballots from the appropriate precinct according to Director
Quillen.
Thus, if the Election Night count of 4,667 were correct,
there should be 4,533 absentee ballots in storage instead of
4,606. Director Quillen had no explanation for this difference.
If the BOE certification that 5,191 absentee ballots were
cast by the time the polls closed on Election Night were
correct, there should be 5,057 absentee ballots in storage, not
4,606. Director Quillen had no information as to where the
missing 451 absentee ballots were.
Because of the theoretical possibility that the extra 617
(or more) votes may have been due to misplaced absentee ballots,
a detailed examination and count of 4,606 absentee ballots
contained in two boxes was conducted beginning in September,
2006, by a reporter and various assistants.
A detailed examination, count, and analysis of the
available absentee ballots on a precinct-by-precinct basis
indicates the extra votes were not the result of misplaced
absentee ballots.
The 2006 Outlook
In 2004, only about 15 Ohio counties used a central count
optical scan electronic voting system.
In 2006, about 29 Ohio counties will use a precinct count
optical scan electronic voting system. The remainder of the 88
counties will use Direct Recording Electronic (DRE) voting
systems.
One advantage of some precinct count optical scan systems
is that they can alert the voter to an undervote or an overvote
Page 15 of 17
before a vote is cast irrevocably. This allows the voter to
change the ballot to avoid an undervote or an overvote.
However, essentially the same counting process is used by a
precinct count optical scan system as by a central count optical
scan system. Instead of the central count method of scanning
ballots at a central location and putting scanning data in
electronic form at the central location for input into a
tabulating computer at the central location, the precinct count
optical scan systems involve scanning the ballots at the
precinct and putting the scanning data in electronic form at the
precinct. Then the electronic form of the data is carried
(instead of across a room, across a county) to the central
location for input into the tabulating computer at the central
location.
When using either system, the ballots are typically
transported from the precinct voting location to the central
counting location. When the ballots become available for
inspection or a recount, the vote totals generated by the
optical scan electronic voting system can be compared to the
vote totals obtained by hand counting the optical scan ballots.
In Ohio, the ballots did not become available for a recount
until after the results were officially certified on December 6,
2004. United States Senators John Kerry and John Edwards, the
2004 Democratic nominees for President and Vice President,
respectively, ended their campaign and any official inquiry on
their part into the accuracy of the 2004 vote count in Ohio over
a month earlier about 1:00 p.m. Eastern Standard Time on
November 3, 2004.
Another advantage of precinct count optical scan systems
over central count optical scan systems is that precinct count
optical scan voting systems can produce a vote total which can
be posted at the precinct voting location after the polls close.
The central count optical scan systems do not have his
capability because they have no means to scan the ballots at the
precinct voting location.
The posting of the results at each precinct can provide
protection against error or fraud in the vote counting process.
Under one reading of Ohio Revised Code section 3505.30, the
results are supposed to be posted at each precinct on election
night before the election judges leave the polling place.
Page 16 of 17
Inquiries at various Ohio Boards of Election indicate that some
counties will post the results at each voting location and
others will not.
Once the results are posted and copied by the public at a
given precinct, the vote for a given precinct should not be able
to be changed on Election Night at the central counting point
without explaining how extra votes were cast (or how votes
counted at a given precinct were removed from the total of votes
cast) between the time the votes were counted and the results
were posted at the precinct (after the polls closed) and the
later time when the results were tabulated and announced at the
BOE’s offices.
In 2006, in Miami County, Ohio, according to Director
Quillen, voters will use a Diebold TSX Model 100 DRE system with
a voter verified paper audit trail at the precinct voting
locations on Election Day. Absentee voters who wish to cast
their absentee ballots in the BOE office may do so either on the
DRE system or on a paper absentee ballot. Absentee ballots cast
on paper will be counted by a Diebold Absentee Voter Optical
Scan (AVOS) Accuvote system according to Director Quillen.
So-called independent testing authorities are paid by the
electronic voting machine vendors to inspect and certify the
software sold by the electronic voting machine vendors. Such
software certification is required by most state and local
authorities before they will authorize the use of electronic
voting equipment.
When asked whether he or the Miami County Board of
Elections had any way to verify physically or electronically
that the software on the Diebold DRE voting system and the
Diebold optical scan voting system being used in Miami County in
November, 2006, was the same as the software certified by so-
called independent testing authorities, Director Quillen
indicated neither he nor the BOE had any way to make such a
verification.
This article was reported by Peter Peckarsky, Ron Baiman,
and Robert Fitrakis. The article was written by Mr. Peckarsky,
who was lead trial counsel for the contestants (plaintiffs) in
Bill Moss, et al., v. George W. Bush, et al., the only judicial
contest of the 2004 Presidential election in Ohio. Dr. Ron
Page 17 of 17
Baiman was an expert witness for the contestants in Moss v.
Bush. Prof. Robert Fitrakis was co-counsel for the contestants
in Moss v. Bush and is executive editor of The Free Press. Prof.
Fitrakis is also an independent candidate in 2006 for Governor
of Ohio. Mr. Peckarsky does not support Prof. Fitrakis’
candidacy and has not had any involvement in the Fitrakis
campaign.