Many people do not even realize that their ballots in elections in the United States are generally being counted by
near secret computer programs under the control of the companies which design and sell electronic voting systems
to election jurisdictions throughout the nation. However, many who have learned of this fact, have chosen to stop voting because of the belief their vote is not accurately counted.
Some of the problems that have people convinced their vote is being stolen are
the percentages of votes cast for losing candidates ending up being the same; or the same ending digit for the number of votes for various candidates or from precinct to precinct; voting tabulation equipment breaking down while a certain candidate is ahead and then the expected winner surfacing ahead after the repairs; projected
winners often having nearly exactly the number of votes announced (coupled
with bloopers such as ABCNEWS.COM announcing election results calling every major race in the United States, twelve hours before the polls opened for the 1998 general election).
In most Missouri election jurisdictions and throughout
the United States, other than the programmer, there is no one, including the
election officials who ever certify the accuracy of vote totals. (Of course, the equipment and software are
supposed to be certified by the state, but no one in the local elections office nor secretary of state's office generally knows what is going on inside the tabulation machines nor software while the ballots are fed through or as pre set and/or adjusted by the programmer.)
Although the National Association of State Elections Directors and the Federal Elections Commission have established some
VOLUNTARY national standards and some independent testing of vote counting systems; lack of public confidence
in the integrity of electronic voting; coupled with the negative implications of marking a ballot card being all too much
a reminder of school testing, or appear too difficult for the undereducated, elderly, nervous, or disabled to feel they
can vote correctly, has driven many from the polls.
Before the 2000 Missouri Presidential Primary the secretary of state's
office needs to assure the people that the elections systems used by Mo. election authorities, at least meet the minimal voluntary standards set forth by
the Federal Elections Commission. If necessary, the Missouri legislature needs to appropriate emergency funds for
National Association of State Elections Directors (NASED)
qualified systems and equipment to be provided to the communities. As an alternative, smaller communities,
with systems, such as ETNET, that are
most likely unable to comply with the standards, should be encouraged to vote by paper ballot and hand counting even if
only for the Presidential Primary.
In November, 1999, I obtained a copy of the Voting Systems
Standards from the Federal Elections Commission (a large notebook of fairly detailed materials) and began to study the
voluntary plan to improve the integrity of electronic vote tabulation. I also found on the internet a list of
the voting systems that were qualified by NASED as of June 22, 1999. (See votenet.org site linked below.)
The
voluntary national standards are obviously outdated, and are not due for updating until the 2004 election cycle. The
original standards were released in 1990 and it had been recommended that all computerized voting systems
in use within the states comply with the standards by January, 1996. Unfortunately provisions were also made for grandfathering
equipment already in use, when the standards were introduced. However, Congress did not fully fund the plan
and the states failed to adopt the standards. Missouri Congressman Roy Blunt, as secretary of state, served on the Federal
Elections Commission Clearinghouse's Advisory Panel for the standards, but has never taken the initiative to
help fund the project after he became a member of Congress. . (The FEC needed only $100,000 more for both FY99 and FY00 to have the standards revised
and adopted by some of the states for the November, 2002 election.)
As of June, 1998, Missouri was
shown by the FEC, without formally adopted standards, without statutes, and without regulations on the standards, but
requiring an independent testing and NASED certification. However, most of the systems used in Missouri
did not seem to appear on the list of approved systems provided by the Election Center.
As Eva Waskell, Director of Elections Project for Computer Professionals for Social
Responsibility, Washington, DC Chapter stated in an article in 1993, "Overview
of Computers and Elections" (as linked below), "Most of what we can know about the accuracy of
vote tabulation is what the company representatives tell us." She continued,
"... since it is impossible to verify that vote tabulating programs are doing what they are supposed to do and nothing more, it is impossible to resolve any dispute about whether or not manipulation of votes is taking place. The public
debate about alleged vote tampering will continue forever." Waskell also noted
that all of the pre- and post-election testing of the tabulating programs proves nothing about the accuracy of the vote totals. And she said that
recounts of elections only demonstrates a program is tabulating consistently,
NOT accurately. She stated, it is impossible to ever detect what lies inside
without being privy to the detailed instructions for the program.
"As long as vote tabulating logic remains part of a set of rules known only to the programmer who wrote the software, the votes of every American citizen will continue to be tabulated secretly. Secret vote counting has no place in a democracy." Eva Waskell, 1993.
The Federal Elections Commission reports
that in 1996, 37.3 per cent of the registered voters in the United States had punchcard voting; 24.6 per cent used mark sense (optical scan) voting (the method on the increase); 20.7 per cent still used mechanical lever machines (which are no longer made and are being phased out); 7.7 per cent had access to direct recording electronic voting machines; and only 1.7 of the registered voters had the opportunity to vote paper ballots.
By June, 1998, only half of the states had adopted the FEC national standards
for electronic voting systems or the testing of the system against the standards by independant testing authorized by the National Association of State
Elections Directors. Those states were--Alaska, Arizona, California, Colorado, Connecticut, Delaware, Florida, Georgia, Hawaii, Indiana, Iowa, Kansas, Kentucky, Massachusetts, Michigan, Minnesota, Missouri, Nevada, New Mexico, New York, Oklahoma, Rhode Island, South Carolina, South Dakota, Tennessee, Texas, Virginia, Washington, and Wyoming. But, the Federal
Elections Commission admits the standards will continue to need updated and the whole system is voluntary.
The first electronic tabulation system was approved in Missouri, April 25, 1972. By 1974, thirty-one states, including Missouri,
had passed permissive punch card voting legislation.
From the beginning only a few companies were manufacturing
electronic elections equipment and supplies. Over the years other companies surfaced and the records
appeared to me to reflect that Missouri, especially under Congressman Roy Blunt's term as secretary of state, some companies
appearing more secure than some
that were approved were turned down. Companies also went out of business, merged, changed names, or addresses, but a
few people remained in place. When Congressman Blunt became Missouri Secretary of State, there appear to have been nine
electronic election systems approved for use in the state. Although approximately 15 such systems had been approved by
the end of his term in 1992, only seven or eight companies appeared to be in actual use.
By April, 1992, 43 Missouri election authorities were using the punch card system, 21 were using the scanner system, 2 were using lever machines, and 50 were still using paper ballots. In 1992, 20 election authorities were using the Thornber
system and Business Records Corporation, was marketing an optical scan system to at least 14 Missouri counties.
In 1998, eleven counties in Missouri
used paper ballots. Those counties were: Carter, Chariton, Clark, Dade, Mercer,
Reynolds, Schuyler, Scotland, Shannon and Worth. 43 election authorities used punch card systems in 1998. The remaining 62 election authorities used optical
scan systems. Jackson County, Kansas City, and St. Louis City each used VOTEC
punch card systems. St. Louis County also voted by punch card, but used software manufactured by Business Records Corporation. Obviously the majority of Missouri voting was done by punch card and tabulated on VOTEC software, which does not appear to be NASED, FEC qualified as having been
tested by the voluntary national standards.
In 1974, the General Accounting Office, requested the National Bureau of Standards to undertake a systems
analysis of computerized voting. The report was released in 1975 as NBSIR 75-687 or NBS SP 500-30. Although
I have not seen that specific report yet, I have reviewed a 2/12/93 report by Roy G. Saltman of the National Institute of
Standards and Technology, (available at the below link)
noting the views expressed are his own and not those of NIST.
Saltman states the report made recommendations to protect the security and assure the accuracy of all aspects of
computerized vote-tallying. And it recommended research on election equipment and systems, independent of
return on investment; leading to national minimum standards, assuring accuracy and security.
In 1980 Congress passed legislation for a preliminary study for the development of national standards for election
equipment and systems. Although no more research was then done, the Federal Elections Commission issued
performance standards in 1990. The National Institute of Standards and Technology made comments during
the 1989 comment period and pointed out the draft standards "contained no quantitative accuracy requirement whatsoever
for sensors reading voters' choices." Although that was then added, the FEC standards focus on testing of equipment, not on
the whole system of voting. It would appear that state certification and the current national standards and independent testing still emphasizes the testing of equipment versus the whole system. Saltman points out the FEC standards were characterized as seriously flawed by Election Watch,
a California based citizens group. Election Watch complained of the approval of the continued use of pre-scored punch
cards, and the failure to require vendors to make the source codes of the vote-tallying programs public. Saltman contended the national standards do not require an audit
trail or an independent audit of the software. (However, the FEC, elections administration, contends the current voluntary
standards DO require this.)
In 1984 the State of Illinois began a study of the electronic voting systems used in the state. They found 28 per cent of those tested had tabulating errors, BUT FEW SEEMED TO CARE.
In 1986, the John and Mary R. Markle Foundation, a privately funded foundation based in New York City approached
the National Institute of Standards and Technology and asked for a new study of computerized voting. NIST then
published Special Publication 500-158, August, 1988. I also obtained a copy of
a report called An Election Administrator's Guide to Computerized Voting
Systems , published in 1988 by ECRI through a grant from The Markle Foundation.
In the NIST report and ECRI reports numerous recommendations were made, including
state certification of the software used for vote-tallying and recertification for any changes to software, and retention of copies of
the software by the state.
NIST also recommended that all software associated with the vote-tallying process be isolated from influences over which
the election administration loses control. They said the vote-tallying operations should be run only on dedicated
computers, under the control of the election authority and there should not be any outside access to the software and the
system.
Saltman added the 1988 NIST report recommended the concept of "one vote" be treated as if it were currency
and that election authorities try to have internal-control specialists on staff.
The 1988
NIST report recommended the elimination of pre-scored ballot cards because of inaccurate vote counts, incorrect punching,
chad that falls off or fails to fall and gets back in a hole (even from ballots sold as chad free), etc. In spite of the recommendation the use of
pre scored cards actually increased.
NIST also expressed concern of the mark-sense ballot cards, because of the long-
term inaccuracy of the mark-sensor in the reader affecting its capability to distinguish mark from no-mark and smudges. In
Saltman's 1993 writing he knew of no reported examinations of the accuracy of the sensing of voters' choices achieved in
real elections by the mark-sense equipment.
Saltman comments that the non-ballot systems, such as
direct entry and on-line voting, systems, also cause accuracy to be an issue, AND raise the question as to whether the
hardware and software are actually designed to LIE. He believes there is no real assurance of a fair election without
a neutral third-party being involved.
During the time the 1988 study was underway by NIST, Congress passed the Computer Security Act, but amazingly left
out vote-tallying systems. Saltman noted that still in 1993 there was no technical staff within the Federal Government
to advise Congress on technology and the election process.
The ECRI report noted that the passage of time was seeing election administrators increasingly learning how to do their own election setup and vote tabulation. The use of outside computer programmers was supposedly on the decrease. Missouri never seemed to meet that trend and certainly it has appeared most
election authorities in the least contract for vote tabulation to be programmed
by someone other than those actually part of the election authority's office in most jurisdictions.
Saltman also raised an important question as to
whether tallying software should be privately owned at all. PERHAPS THAT IS THE LONG TERM SOLUTION FOR
CONSISTENT, RELIABLE, SECURE VOTE TABULATION ACROSS THE STATE OF MISSOURI AND ACTUALLY THE ENTIRE
ELECTRONIC VOTE
TABULATION SYSTEM DEVELOPED AND UNDER THE DIRECT CONTROL OF THE SECRETARY OF STATE?
Local control of elections is needed, however, that has long been
compromised through turning over the tabulation to vendors. Uniformity
in how the votes are tabulated and real security so the public could be reassured of the integrity of elections
would be refreshing even to those of us
highly supportive of as much local control as possible, most likely increase voter turnout in the years to come, and perhaps also be more economical. State versus outside vendor support would seem to solve a lot of problems and
should be more acceptable to the public if properly presented.
The ECRI report also noted as early as 1988 that the states would do well to require audit trails to be recorded for every stage of vote-casting; and that audit trails also be recorded for the vote-tabulation process; that the original system source code be deposited with the state so it could be compared with that used in problem elections, to detect possible discrepencies; that remote, non-election terminals be disconnected during ballot tabulation (that appeared to be a problem in Greene County where use of on line computers gave an appearance of manipulation of the returns); that manual recoutns be performed on a prescribed percentage of ballots (which would avoid some Mo. election authorities always choosing the precinct with the least ballots cast and eliminate any waiver of a manual recount).
Several Missouri election authorities are doing vote tabulation with systems that violates many of the don'ts set forth in the ECRI report. ECRI recommended that components of a system should not be purchased from several different vendors. New software or any new equipment should not be purchased two or three months before an election because there is not time to put it through acceptance testing and the staff does not have time to learn how to operate it. Most election authorities and those paying for the election systems also are not requiring the vendors to meet performance standards or suffer non performance penalties. (Legislation is needed to protect the public in this
area and to establish penalities.)
Voters perception of the lack of integrity in the counting of votes has been repeatedly influenced by articles such as one which ran in New Yorker 1988 by Ronnie Duggar, reporting Shouptronic Company was accusing Business Records Corporation (earlier known as Cronus) of having a monopoly on the counting of votes in the USA. Collier's book VOTESCAM
also refers to Cronus as being a major problem for vote fraud. Interestingly
Shouptronic applied for approval in Missouri in October, 1982, and May 19, 1983,
they were sent an outline of procedures for certification. In the records I
reviewed at the secretary of state's office, 1997, their record then appears to
stop. On the other hand Business Records Corporation incorporated in Mo. November 4, 1988, was dissolved
for failure to pay franchise taxes, August 12, 1994. BRC continued to be used for vote tabulation for the 1994 election and was then reinstated to do business, October 10, 1995. In 1998 Affiliated Computer Services, Inc. of Dallas became the majority share holder of BRC Holdings, Inc.
and in 1999 the sole share holder. In 1997 ACS merged with CDSI, which was granted the technology contract for all the U.S. Courts in 1996, in addition to
contracts with the U.S. General Service Administration, U.S. Department of Transportation, Federal Communications Commission, and Argentina Lottery. November 27, 1997, American Information Systems, Inc. acquired BRC Holdings, Inc. of Dallas and Sequoia Pacific Systems to form Election Systems and Software, Inc. of Omaha, Nebraska. ACS thus is actually the primary shareholder of ESS, Inc. In 1998 ACS was awarded contracts with the U.S. Postal Service, U.S. Senate, U.S.
President, U.S. Vice President, FDIC, U.S. Air Force, and U.S. Naval Reserve
THE HENRY M. ADKINS AND SON, INC. AND THE CHICAGO AND
LITTLE ROCK INFLUENCE
December 28, 1983 Henry M. Adkins and Son, Inc. of
Clinton, Mo. was approved as a corporation. August 18, 1984 they were administratively dissolved by Secretary of State Kirkpatrick for failure to pay
franchise taxes. Fourteen years later they were still proclaiming they were
a valid Missouri corporation until Secretary of State Cook notified the
election authorities they were not properly registered to do business as a
corporation. In a December 18, 1984 letter Secretary Kirkpatrick noted
Henry M. Adkins and Son, Inc. was distributing the electronic tabulation
system for 29 Missouri counties and to the cities of St. Louis and Kansas City.
He noted they were using the Thornber Voting System from Chicago, Illinois. Thornber Voting System was a division of Business Records Corporation.
During the 1996 election the St. Louis County Board of Election Commissioners told me someone they called only "Hank" had come from Chicago for many, many years to monitor all the vote tabulation. For the 1996 general elections however, "Hank"was not present for the St. Louis County election, although
his replacement, also from Chicago, refused to be identified.
By the 1998 elections Henry M. Adkins and Son, Inc. were boasting they
provided election materials to 57 Missouri election authorities and had at
least 19 people in the field on election day to "monitor" the tabulation.
Henry M. Adkins and Son, Inc. also distributed the ETNET system, from Election Resources Corporation of Little Rock, Arkansas. For the 1996 primary
election some Missouri election authorities relied on tabulation software from ETNET, which had not been certified for use in Missouri nor
by the national standards.
March 5, 1993, Floyd Gilzow, then Missouri Deputy Secretary of State, reviewed the
ETNET system and made some
noteworthy observations. Gilzow had some alarming concerns that do not appear to have ever changed. He said that the
counting system for ETNET was not physically secure, although it had password access. And he noted there
was nothing to prevent vote totals from a previously run stack of cards to be substituted for the official election ballots.
The log file could be easily altered or eliminated. The system could be used to show how individuals voted their ballot. He
found it was possible to edit the file before it was printed and show precincts were counted when they actually were not.
Gilzow's observations were agreed to by ETNET's representatives.
Voters tend to be skeptical of international interests
possibly influencing elections in the United States.
The FEC contracted with ManTech Advanced
Systems International, September, 1997, to undertake a "Requirements Analysis for the FEC voting system standards".
Mantech has provided a variety of support to the U.S. Department of State since 1980. And ManTech has provided
support for the Army, Air Force, Navy, and numerous other classified agency organizations. The telecommunications
section of the ManTech website states the company's support to the Joint Interoperability Test Center in
Ft. Huachuca, Arizona, since 1983, supporting Defense Information Systems Agency, "tasking to certify the
interoperability of systems and subsystems of all agencies of the U.S. Government, military services,
and commercial industry users." This sounds as if election standards revision recommendations have
been placed in the hands of a company with near unlimited power already. Should a nation where
the people are supposed to govern place trust of vote tabulation integrity to such a powerful company?
NASED and the
FEC are using Nichols Research. and Wyles Labs for qualifying elections systems. Wyle Labs does the independent
testing for the hardware and Nichols Research does the independent testing of the software. Although greater
research needs done on these companies and how they were chosen for this most important task, a red flag is
the fact both companies are in Huntsville, Alabama. Nichols Research, Inc. has a web site and. is actually headquartered in
Sunnyvale, California and has collected market researsh data in the San Francisco Bay Area since 1978. It has not been
determined if this is the same Nichols Research in Huntsville, Alabama.
The Election Center, Houston, Texas,
maintains among other things, the records of the NASED, FEC qualified election systems. The Election Center, notes
on its website that it is "An international service association of election and voter registration officials". A 501(c)(3) the
Election Center states its purpose is "to promote, preserve, and improve democracy". The international relationships
are not apparent from the website.
Election Systems and Software, Inc. became the largest company in the world
in the automated elections industry, when they merged with Election Products, Inc. Unfortunately the web site for ES & S, Inc.
links to international associations, which link to the United Nations. ES & S, Inc. had two voting systems NASED qualified as of
June 22, 1999. They and Sequoia Pacific and MicroVote were actually among the first systems qualified in 1995. Interestingly
MicroVote ran into the most trouble of virtually any company when trying to obtain certification in Missouri under Secretary of
State Blunt.
As noted in the section on laws and code of state regulations, Missouri needs to make some
revisions in elections laws and regulations. However, in order to fully understand
electronic vote
tabulation one must look to some definitions as established in the current Mo. statutes. For instance, "automatic tabulating equipment"
includes the apparatus necessary to examine and automatically count votes, and the data processing machines which are used for counting and tabulating results." (115.013 (1) RsMo.) "Ballot" includes the ballot card or paper ballot on which each voter may cast all votes to which he is entitled at an election. (115.013(2) RsMo.) A "ballot card" is a ballot which is voted by making a punch or sensor mark which can be tabulated by automatic tabulating equipment. (115.013 (3) RsMo.) A "ballot label" is the card, paper, booklet, page or other material containing the names of all offices and candidates and statements of all questions to be voted on. Earlier known as the "vote recorder" (115.013 (4) RsMo.) An "electronic voting system" is a system of casting votes by use of marking devises and counting votes by use of automatic tabulating or data processing equipment. (115.013 (9) RsMo. ) And a "marking device" is either an apparatus in which ballots are inserted and voted by use of
a punch apparatus, or any approved device for marking paper ballots with ink
or other substance which will enable the votes to be counted by automatic tabulating equipment. (115.013 (14) RsMo.
Missourians need to be made more familiar with the terms and equipment used as part of the elections system. People tend
to trust more that which they understand. The secretary of state's office needs to encourage all election authorities to
make sure the counting of ballots is done so voters can view the process. (The judicial system could
actually help with this by encouraging use of courtrooms for vote tabulation, which would also help some voters
feel more confident in the integrity of the process.) And the election authorities need to be encouraged
to better display posters showing the proper way to mark ballots. Sample ballots should be identical to those used at the polls
except for the weight of the paper and color perhaps. Sample ballots should be available throughout the community and
encouraging people to practice marking a ballot.
All counting of ballots should be done by local people without
the election authority hiring vendors, non residents of the election district, to come in and monitor the tabulation.
Since Missouri law requires an oath for elections
workers it appears the law already requires this, but many election authorities do not seem to accept it. Perhaps that law needs
to be more specific. Even as early as 1969, when electronic voting was first being discussed vendors and officials
were suggesting anyone who touched any ballot should be under oath to uphold all the laws. Missouri
needs to require anyone who helps with any vote tabulation to be under oath.
So called bi-partisan teams assisting with vote counting are essentially a joke in many jurisdictions. Including all the political
parties on the ballot in this would help some, but would not entirely
preserve the integrity of the process. Often those assigned by the major political parties to observe or assist with the counting are elderly or undereducated or convinced the system both of government and electronic vote
tabulation always works perfectly.
One of the important duties bi partisan teams are used for in the counting process is for what are termed makeover ballots. The practice of makeover ballots probably needs simply ended and th ose ballots either totally counted by hand or added to the spoiled ballot stack. What is actually happening is the bi partisan teams sometimes
leave one person to claim to copy the voter's intended votes onto another ballot card. Or the team members sit and talk and let stacks of ballots that need to be copied to be tossed into the storage box and then have been observed asking each other what they had looked at and what should be put on the ballot to be counted. How many simple human errors are made by this method of claiming to make a copy of a ballot someone voted? How many members of the teams have deliberately slipped in some ballots marked as they would want versus the intent of the voter who initially voted the ballot?
One horror story from Lawrence County, Missouri in 1998 was, of course, the elderly lady on the bi partisan team opening the absentee ballots and doing a manual recount, who could not see to perform a manual hand count of punch card ballots early evening, but suddenly claimed to count perfectly near the midnight hour. Such happenings certainly send a bad message to the public.
Election authorities and political parties need to be encouraged to take the bi partisan (which should include all parties on the ballot) teams' duties much more seriously and to select people who can stay alert, who are willing to become knowledgable of the
entire process, and to weigh each part of the process as if each ballot were
valued at one million dollars.
The board of
state canvassers should become something more than a rubber stamp for pre printed results. Even if the election authorities
need to submit TWO abstracts the board of state canvassers should actually open and canvass a set of the abstracts. Both in
1996 and 1998 I personally canvassed several of the abstracts and found numerous errors had been made by the election
authorities. The law needs changed to permit the secretary of state and/or the board of state canvassers to call blatant,
serious errors to the attention of the Attorney General as an elections law violation, subject to the seriousness of the violation,
to sanctions and/or forfeiture of office by the election authority.
CONCLUSION
Missouri vote tabulation
has been overly influenced by certain non trustworthy people, such as Henry M. Adkins and Congressman Blunt and his
former deputy secretary of state, Paul Bloch (who went on to profit from affiliation with what was known as American
Information Systems). As a result of the conflicts of interest by non trustworthy persons or entities, much of the
equipment currently in use by Missouri election authorities should not be relied on as either secure or accurate for
vote tabulation.
Missouri needs to require NASED qualified hardware and software by all election
authorities and needs to develop even more stringent standards for certification of the election systems used in our
state. The Missouri Secretary of State needs to take a leadership role in reassuring the people that vote tabulation
will be done with the utmost integrity and each ballot treated as if it were worth a million dollars; the importance
of one person, one vote must be emphasized.
Links to other sites on the Web
Waskell Report, 1993(click- search-Waskell)
Saltman Report 1993(click-search-Saltman)
Approved Voting Systems
ManTech International
Mo. Voter Registration and Casting Ballots
Part Three-Absentee and Write-in Voting
Copyright Marvalene Pankey 1999, marvalenpankey@yahoo.com, 417-876-4626, e fax 810-314-0929.
Get your own Free Home Page