Voting Devices Tested
by Chuck Campbell
A mock election testing
the Vote-PAD Voting-on-Paper Assistive Device was held December
15, 2005 at the Minnesota State Office Building. Vote-PADs brochure
touts its ability to facilitate “Independent
Voting for People with Disabilities,” describing the system
as: “(A)n inexpensive, non-electronic, voter assist alternative
that helps most people with visual or dexterity impairments to vote
independently.” The brochure’s background section states: “Some
people with visual or dexterity impairments cannot mark a paper ballot
without assistance. The Federal Help America Vote Act of 2002 (HAVA)
requires that every polling place must offer a method by which people
with disabilities can vote independently.” Vote-PAD’s
owner, Ellen Theisen, said she “invented the system with input
and cooperation from people with disabilities and people interested
in transparent elections.”
Rick Cardenas, a person with quadriplegia who has limited use of
his hands, was one of the first to try the Vote-PAD system. As co-director
of Advocating Change Together (ACT), which facilitates self-advocacy
with others who have disabilities, Cardenas has both a personal and
professional interest in accessible voting systems.
“Make sure the holes are open —some of them weren’t
open,” said Cardenas, referring to the Vote-PADs transparent
ballot sleeve, which is designed to protect the ballot from stray
marks and has holes where a voter can mark choices. Other than the
closed holes, Cardenas said the system worked well. “The desk
level is a good height; a lot of times election judges push you over
to the accessible voting booth, which is too low. The guide makes
it much easier than free handing.”
“It’s a really good idea to test run the Vote-PAD system,” Cardenas
said. “I’ve been able to vote independently by marking
the circles, but the Vote-PAD is quicker and easier. We’ll
see if it works and elects the people I want to elect,” Cardenas
added with a smile.
“This [Vote-PAD] machine can be helpful to someone who can’t
read big words,” said Melvin Haagenson, who has cerebral palsy. “I
think it’s good, and more people with disabilities will vote,
but it might be a good idea to take the Vote-PAD to group homes,
assisted living facilities, and day activity centers for more testing,” said
Haagenson.
Those with visual impairments
had less positive assessments. “It’s okay to call me
blind; it’s shorter,” said
Judy Sanders, adding: “I get to vote for president again; will
it change anything?” Sanders main criticism of Vote-PAD was
that the system requires visually-impaired voters to count down the
side of the ballot both to find the appropriate hole to mark for
the desired candidate and to verify she’d marked her ballot
as intended. “How do I know I did it right? If I’m in
the hole I thought I was in?” she asked.
Though Vote-PAD includes
a verification wand that vibrates when passed over a marked hole,
the only way for Sanders to find that hole was to count down to
it after listening to the audio instructions. Sanders said the
only way to be certain she’d marked her Vote-PAD
ballot as intended would be to ask a sighted person for verification. “The
holes are too close together and too small,” said Sanders. “I
can verify that I marked a hole, but not if it was the hole I intended.”
Sanders voiced her preference
for one of Vote-PAD’s competitors: “AutoMARK
is very well designed. It tells you if you put your ballot in upside-down,
and the keys are easy to read. AutoMARK allows audio voting. It gives
you audio prompts for each race, and audio verification of the candidate
for which you voted. AutoMARK seems to have thought of anything that
can happen and has a solution.”
AutoMARK solutions include a sip/puff tube for voters without use
of their hands. Vote-PAD accommodates such voters by mouth stick.
“The Vote-PAD is very labor intensive. People have to teach
you how to use it,” said Thomas Heinal, who gets around with
the help of his seeing-eye dog. “It’s too slow and cumbersome.
I can’t see that at a polling place you’re going to have
an hour’s worth of instructions or that you would remember
them. I don’t think they should use it – period,” said
Heinal, adding that he wants a voting machine that leaves a paper
trail.
State Representative
Bill Hilty from District 8A amplified Heinal’s
concern. “Without paper records there is no way to verify that
votes have been accurately counted, and you trust a (voting machine)
company, none of which have a good record on transparency. They claim
all their software is proprietary and cannot be made available for
third parties to examine. We’re being asked to trust private
companies with private interests to protect the security of our voting
system with no way to verify, and that’s a leap of faith we
should not be willing to make,” said Rep. Hilty.
The federal Help America Vote Act of 2002 (HAVA) mandates the following:
HAVA Section 301 3A: “The
voting system shall be accessible for individuals with disabilities,
including non-visual accessibility for the blind and visually impaired,
in a manner that provides the same opportunity for access and participation
(including privacy and independence) as for other voters.”
HAVA Section 301 2A
i: “The voting system shall produce a
permanent paper record with manual audit capability …”
The need for equal accessibility
is obvious, and Rep. Hilty, a member of the Minnesota House Civil
Law and Elections Committee, provided a wealth of information on
the necessity for a voter-verified paper trail that can be manually
audited. Many current touch-screen voting systems are entirely
paperless, and tally votes on an electronic memory card. Quoting
documentation Rep. Hilty provided, he pointed out the memory card’s
vulnerability:
“Finnish security expert Harri Hursti proved that Diebold
lied to Secretaries of State across the nation when Diebold claimed
votes could not be changed on the memory card. A test election was
run … with a total of 8 ballots – 6 ballots voted ‘no’ on
a ballot question as to whether Diebold voting machines can be hacked ….
Two ballots voted ‘yes.’
“At the beginning of the test election the memory card programmed
by Harri Hursti was inserted into an Optical Scan Diebold voting
machine. A ‘zero report’ was run indicating zero votes
on the memory card. In fact, however, Hursti had pre-loaded the memory
card with plus and minus votes.”
“Correct results should have been Yes: 2, No: 6. However … the
results tape read: Yes: 7, No: 1.”
“This exploit, accomplished without … any
password and with the same level of access given thousands of poll
workers across the USA, showed that the votes themselves were changed
in a one-step process. The hack would not be detected in any normal
canvassing procedure, and it required only a single credit-card sized
memory card.”
Rep. Hilty provided
further documentation stating that Diebold is the defendant in
a class action lawsuit alleging, among other things, “Concealment
of Known Flaws in Voting Machines,” and loaned Access Press
a DVD titled Invisible Ballots, highlighting vulnerabilities of paperless
voting systems, including:
• Professor Rebecca Mercuri, Ph.D. of Harvard’s Kennedy
School of Government, referring to Florida’s 2002 primary: “… a
quarter of a million votes vanished …,” and, “… in
Dallas, Texas … you had machines where when you pressed for
the Democratic candidate it would only light up for the Republican
candidate.”
• Computer Technician David Allen, describing what Bev Harris,
author of Black Box Voting, discovered on a publicly accessible Internet
site: “Pretty much the blueprints to the vault – everything
you would need to know if you were going … to hack one of
these (Diebold) machines.”
• Professor David Dill of Stanford University’s Department
of Computer Science: “the 2002 Governor’s race in Alabama
being apparently decided by machine error.”
Invisible Ballots also
raises security concerns regarding ownership of paperless voting
systems by: major government contractors whose future depends on
federal policy, foreign companies, and a U.S. senator who was owner
and CEO of the company whose machines counted 80% of his state’s
vote.
“We need random post-election audits to make sure electronic
results correspond with paper ballots,” said Rep. Hilty. Both
Vote-PAD and AutoMARK comply with HAVA’s mandate for a paper
trail enabling such audits. Furthermore, both systems use a paper
ballot that is manually marked by the voter, and can be verified
before being optically scanned into an electronic vote tabulator.
Rep. Hilty sees voter-verifiability as crucial:
“A permanent paper record is not necessarily a voter-verified
paper record. Many touch-screen machines merely have an internal
tape similar to a cash register receipt which is usually referred
to as an ‘activity log.’ It presumably records every
transaction which has taken place on the machine, including … a
record of every vote recorded. Keep in mind that some of the major
discrepancies between exit poll results and machine results may be
explainable in terms of what the voter thought he or she was selecting
being different from what the machine recorded.
“In some devices
the voter is able to see this paper tape through a glass cover;
in others it is entirely internal. The ones I have seen are exceedingly
difficult to read, but even if you can clearly see that what you
intended is what is printed on the tape, since the actual tabulation
is electronic, there is no way to know whether what the machine
printed is what it tabulated, or that when the data for this particular
device was transferred to a central tabulator it was done accurately,
or without being altered, intentionally or otherwise. Of course,
if you are blind, you totally have to trust what the machine tells
you. I do not see any way around this, no matter what system is
employed.
“So the important thing is that the voter … be able
to actually see the paper record to … verify that it truly
and accurately represents his or her choices. If it cannot be voter-verified,
a ‘permanent paper record’ will only (possibly) tell
us what the machine recorded, not necessarily what the voter intended.
But even if we do have a voter-verified paper trail, the question
still remains … whether what was marked or recorded on the
paper ballot or paper tape is actually what was electronically tabulated.
Thus the importance of doing post-election hand-count audits of a
statistically significant number of precinct records, randomly selected.
If this is not done there is no way that one can ever be confident
that the electronic results were recorded as intended. So no, a permanent
paper record in and of itself does not ensure transparency.”
Both Vote-PAD and Auto-MARK,
and some other devices in the process of being certified, could
meet Rep. Hilty’s transparency requirements,
which are more specific and more stringent than HAVA’s, provided
a random, statistically significant, post-election hand-count is
conducted. In regard to such systems, Rep. Hilty said, “… I
think that the only criteria for choosing between them would be cost,
ease of operation, maintenance and storage requirements, etc.”
“It would cost about $5,000 per polling place to deploy AutoMARK,” said
Ramsey County election official Joe Mansky. “There are 104
voting precincts in the city of St. Paul,” said Dorothy McClung,
Director of Ramsey County’s Department of Property Records
and Revenue. Thus, the initial expense of equipping St. Paul’s
precincts with AutoMARK totals $520,000 minimum.
“I’d like to test the AutoMARK by pushing it off a table
and seeing if it still works. We transport voting machines to polling
places by private vendors,” said Mansky, comparing trucking
companies’ treatment of voting machines to that of luggage
by airport baggage handlers. “Given primary and general elections,
any voting system would be trucked to or from polling places 4 times
per year. We want the machines to last at least 10 years – that’s
40 trips,” said Mansky. Mansky compared Vote-PAD to a notebook,
saying it would be much cheaper and more durable than AutoMARK.
“Anoka, Ramsey, Washington, and Dakota counties currently
use Diebold machines,” said McClung, adding that Diebold-compatible
optical scanners could enable Vote-PAD without replacing existing
machines. “I think Vote-PAD could be further improved. We will
continue working with Ellen Theisen to make it better,” said
McClung.
“We think there is a device out there that will create an
audio message from text. This could be used in conjunction with Vote-PAD.
This would also potentially benefit ESL voters, such as our Hmong,
Spanish, and Somali speaking populations,” said Mansky.
“Whether we ultimately purchase Vote-PAD or not, we have an
interest in making as many good products available as possible. We
want the disability community involved in picking the equipment that
will serve their needs best,” said McClung, thanking those
who came out to test Vote-PAD despite the weather.
For further information see:
www.Vote-PAD.US
www.fec.gov/hava/law_ext.txt
www.vogueelection.com/products_automark.html
www.invisibleballots.com
