Transcript Myth Breakers for Election Officials

Myth Breakers
for Election Officials
A Brief Summary
 HAVA Facts
 E-Voting Problems in Recent Elections
 Election Complexities with E-Voting
 HAVA–Compliant Alternatives to DREs
 Costs Considerations
 Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
What Myths?
 Many local election officials don't have the information they need to
make wise decisions about voting equipment. They hear many
conflicting stories, and it's hard to tell truth from myth.
 This presentation gives facts that dispel many of the e-voting myths.
 For details download:
www.votersunite.org/takeaction/mythbreakers.pdf
Preface
Copyright ©2004 www.VotersUnite.org
[email protected]
For Example: did you know?
 A repeat of a failed election was held in 2004
• Hinds County, Mississippi had to hold its November 2003 election all over
again because so many of the DREs broke down that they couldn't determine
the will of the voters.
 Dead batteries had to be replaced before an election
• Neglecting to keep the DRE batteries charged between elections cost
Arapahoe County, Colorado over $100,000 in battery replacements just
before a recent election.
 Logic and Accuracy testing is labor-intensive
• If it takes an hour to do the Logic and Accuracy testing on one DRE, San
Diego county would have to spend 1275 person-days testing before every
election in order to comply with California law.
Preface
Copyright ©2004 www.VotersUnite.org
[email protected]
Facts about The Help America Vote Act
(HAVA)
 HAVA Does Not Require the Use of DREs
• States are required to allow the disabled to vote unassisted, and they have
until the first general election of 2006 to comply.
•
Three non-DRE systems currently satisfy this requirement:
• Ballot marking devices, such as the AutoMark.
• Tactile ballots like they use in Rhode Island.
• Open Voting Consortium system, which is free software that runs on
standard computers.
 HAVA Does Not Prohibit Punch Card and Lever Systems
•
A jurisdiction may continue to use its punch card or lever system if it adds a
training program for voters to prevent over-votes and accidental under-votes.
HAVA Facts
Copyright ©2004 www.VotersUnite.org
[email protected]
More Facts about HAVA
 HAVA Allows Partial Replacement of Old Systems
• A state can take HAVA funds to replace punch cards or levers in some
counties and not in others.
 HAVA "Audit" Requirement is Not a Meaningful Recount
• Vendors and some election officials say that an end-of-day printout
satisfies the HAVA audit requirements. However, if a DRE has made any
errors in recording or storing votes, its end-of-day printouts will be
incorrect, and no meaningful audit can be done.
•
When a machine produces results a second time, it's merely a reprint, not a
recount.
HAVA Facts
Copyright ©2004 www.VotersUnite.org
[email protected]
And More Facts about HAVA
 HAVA Preserves the Right to Use Paper Ballots
• HAVA explicitly preserves jurisdictions' rights to use paper ballots.
Section 301(c)(2) specifically says that the term "verify" may not be
construed to forbid the use of paper ballots.
 EAC Guidelines and Standards Are Strictly Voluntary
• HAVA charges the Election Assistance Commission with developing
guidelines and voting system standards, but compliance with these
standards is not mandated for the states, nor is compliance required in
order to receive HAVA funds for voting equipment upgrades or purchases.
•
This means that states retain control over whether or not they upgrade
voting equipment to the FEC 2002 standards, which are the current
standards.
HAVA Facts
Copyright ©2004 www.VotersUnite.org
[email protected]
Map of E-Voting Problems
12) Indiana, 2004:
1) California, 2003:
ES&S installs uncertified
13)
Hinds
17) Montgomery
County, 2003:
County, 2004:
Diebold2)
installs
uncertified
California, 2004:
software on iVotronics; admits
18)DRE
Sarasota
County,
2004:
DREs overheat,
Diebold
break
presents
down;
incomplete
software
without
notifying
Senate
leader
certified version won't tabulate
3)majority
California,
2004: 14)
Floyd,
Coweta
Counties,
2002:
ES&S
to
count
election
ballot
invalidated,
whenDREs
font
isfail
magnified
authorities
19)
Wake
County,
2002:
introduces
urgent
bill
to
votes
4) Alameda County, 2004:
Sec’y
of
State
decertifies
8) Arapahoe County, 2004:
Diebold
DREs
lock
up;
access
cards
189
votes.
re-held
later
15)and
Bryan,
Terrell
Counties,
2002:
Flawed
ES&S
iVotronic
ban computerized
voting
Diebold
control
modules
Diebold
for
November
5) Orange
County,
2004:
20)
Georgia,
2004:
24) Napa
County,
2004:
of battery
charge
9) Failure
Bernalillo
County,
2002:in malfunction;
Diebold
DREwrong
ballots
omitballots
races;
software
losescandidate
436
fail
to
start
up
16)
Muscogee
County,
2003:
Hart
DREs
trip
circuit
breaker
Diebold
ballot-encoding
Optical
Scan
calibration
error
fails
DREs
costs
over
$100,000
to
marked
onwrong
screenraces
6) Orange County,
2004:DREs fail to count
Sequoia
and
present
21)
Miami-Dade
County 2002:
10)
Dallas
County,
2002:
25)
Lubbock
County,
2004:
DREs register
and shut
down
when County,
batteries
mix-ups"yes"
prevent voters from
to
thousands
of votes
replace
batteries
Hart
access-code
confusion
7)tally
San
Diego
2004:
12,000
out
of 48,000
votes –
iVotronics fail to count
iVotronics
mark
Optical ScanES&S
programming
error
when
votersES&S
vote
"no"
run causes
out;Diebold
voters
voting
in primary
7,000turned
votersaway
to insufficient
receive
DREs
lose
votes;
memory
11) Harris County, 2003:
8.2%
ofCounty,
the votes2002:
22)
Broward
incorrect
the screen
prevents
allchoices
votes on2004:
from
the
polls
26)
Scurry
County,
the
wrong
ballotsfail
control
modules
Hart DREs won’t start; voters
ES&S iVotronic error misses
from being
counted
Scan paper
to start up properly Defective
write chip
votesinonOptical
make-shift
counting
22% of
the votes
23) Broward
County,
2004:
gives landslide victory to
ES&S
lose 2004:
134 votes;
the wrong candidate
27)iVotronics
Bay County,
winning
is 12 voteserror
Optical
Scanmargin
ballot alignment
hands thousands of votes to
opponent
These are just a few examples
of the E-voting problems in U.S elections in the last
two years
E-Voting Problems in Recent
Elections
Copyright ©2004 www.VotersUnite.org
[email protected]
Some of the Recent DRE Problems
Many people advocate the use of voter-verified paper ballots (VVPB) on
DREs, but when machines break down or voters receive the wrong ballots,
a VVPB printer would be no help.
California, 2003: Diebold installs uncertified
software without notifying authorities
San Diego County, 2004: Diebold DREs lose
votes; control modules fail to start up
Arapahoe County, 2004: Failure of battery
charge in DREs costs over $100,000 to
California, 2004: Senate committee passes urgent
replace batteries
bill to ban computerized voting in 2004
California, 2004: Secretary of State decertifies all Bernalillo County, 2002: Insufficient memory
causes failure to count 12,000 out of
Diebold DREs for the November election
48,000 votes
Alameda County, 2004: Diebold control modules
Dallas County, 2002: ES&S iVotronics mark
fail to start up
incorrect choices on the screen
Orange County, 2004: Hart DREs trip a circuit
Harris County, 2003: Hart DREs won’t start;
breaker and shut down when batteries run
voters write votes on make-shift paper
out; voters turned away from the polls.
Access-code confusion causes 7,000 voters Indiana, 2004: ES&S installs uncertified
to receive wrong ballots
software on iVotronics; admits certified
version won't tabulate votes
E-Voting Problems in Recent
Elections
Copyright ©2004 www.VotersUnite.org
[email protected]
More of the Recent DRE Problems
Hinds County, 2003: DREs overheat, break
down; election invalidated, and re-held later
Sarasota County, 2004: ES&S DREs fail to
count 189 votes
Floyd, Coweta Counties, 2002: Diebold
DREs lock up; access cards malfunction;
wrong candidate marked on screen
Wake County, 2002: Flawed ES&S iVotronic
software loses 436 ballots
Bryan, Terrell Counties, 2002: Diebold
DRE ballots omit races; present wrong races
Muscogee County, 2003: DREs register
"yes“ when voters vote "no"
Montgomery County, 2004: Diebold DRE
presents incomplete ballot when font is
magnified
Georgia, 2004: Diebold ballot-encoding mixups prevent voters from voting in primary
Miami-Dade County, 2002: ES&S iVotronics
fail to count 8.2% of the votes
Broward County, 2002: ES&S iVotronics lose
22% of the votes
Broward County, 2004: ES&S iVotronics lose
134 votes; winning margin is 12 votes
So, some people question the wisdom of using the current crop of DREs,
even if they have a printer attached. To make sure voters aren’t
disenfranchised, each precinct would have to have paper on hand anyway
in case the machines malfunction.
E-Voting Problems in Recent
Elections
Copyright ©2004 www.VotersUnite.org
[email protected]
Optical Scan Machines Miscount, Too
Problems with optical scan machines have been severe but recoverable.
These examples illustrate the importance of auditing optical scan machines
by performing random recounts of the paper ballots.
Napa County, 2004: Optical Scan calibration error fails
to tally thousands of votes
Lubbock County, 2004: Optical Scan programming
error prevents all votes from being counted
Scurry County, 2004: Defective chip in Optical Scan
gives landslide victory to the wrong candidate
Bay County, 2004: Optical Scan ballot alignment error
hands thousands of votes to opponent
E-Voting Problems in Recent
Elections
Copyright ©2004 www.VotersUnite.org
[email protected]
Illegal Use of Uncertified Software
 Most states require electronic voting equipment to be certified
before it can be used in an election.
•
If the state requires federal certification, only equipment with a NASED
qualification number may be used. Some states also require state
certification for all voting systems.
 In violation of state laws and without the knowledge of election
officials, uncertified software has been installed by vendors and
used in elections in at least these states:
•
•
•
•
California – Diebold
Indiana – ES&S
Maryland – Diebold
Arizona - Diebold
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
The NASED Qualification Process
 Independent Testing Authorities (ITA) are selected by NASED
• The ITA testing process is a secret from election officials and the public
• All contracts and contacts are between the vendors and the ITA
• ITAs test a machine’s design, not individual voting machines
• Currently qualified machines meet 1990 standards, not 2002 standards
 NASED-qualified, state-certified machines have performed like this:
• New Mexico – Sequoia voting systems lost 12,000 votes
• North Carolina – ES&S iVotronic lost 436 votes
• California – Diebold TSx lost 10 votes
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
Chain of Custody is More Complex
 Election equipment is particularly vulnerable to tampering after it
has been installed tested but before it is used.
•
Every single DRE must be secured to ensure that the software is not
illegally altered between elections.
•
Each DRE, after being tested and zeroed out, must remain secure until
election day.
•
Electronic machines break down during an election and are removed for
repair. Procedures must be developed for retesting the machines before
they are placed back in service.
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
Ballot Boxes are the Size of a Credit Card
 Using electronic voting equipment does not eliminate the need to
track and preserve the physical records of votes.
 The physical records are now in the form of ballot memory cards
rather than paper ballots.
•
Ballot cards are not a permanent form of storage since the data can be
erased or overwritten
•
•
Ballot cards are the size of a stack of about five credit cards
Ballot cards, which can be easily lost or slipped into a pocket, must be
kept as secure as ballot boxes
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
Logic & Accuracy Testing – the Reality
 Logic & Accuracy testing on electronic voting machines is
essential.
•
In November 2001, the failure of Registrar of Voters managers to do L&A
testing on ballot counting machines caused votes for some candidates to
go to other candidates. All 82 elections were subjected to a hand recount,
with the results in 13 local water and school board races overturned.
 Testing DREs is a major undertaking
• To comply with California law, the Registrar of San Diego County must
test 10,200 DREs before every election. If it takes an hour to do the Logic
and Accuracy testing on one DRE, San Diego county would have to spend
1275 person-days testing before every election.
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
Logic & Accuracy Testing – What if it Fails?
 Ideally, every L&A test would show that the machines are
operating correctly. But if a machine fails, then:
•
It would be necessary to take the machine out of service or have it
repaired.
•
If the software were flawed, the flaw would be present in all DREs
using that same software.
•
It would be too late to have a software patch developed, tested,
certified, and installed in time for the election.
•
It might be too late to print absentee ballots for the entire county to
use for the election.
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
Increase in Human Error
 While many look to e-voting machines as a way of reducing human
error, the fact is that human error may be on the increase because
of the new and complex problems they present:
•
•
Houston, Texas: Poll workers assigned the wrong ballots to voters.
•
San Diego, California: Poll workers gave voters the wrong
provisional ballots.
•
Walker County, GA: after six elections were held on the same
equipment, even vendor technicians couldn’t operate it
successfully.
Orange County, California: Poll workers gave thousands of voters
the wrong ballots.
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
Potential Problems Inherent in Electronic
Devices
 Touch Screen Misalignment (Florida, 2003)
 Miscellaneous Breakdowns on Election Day (California, 2004)
 Power Surges or Static Electricity Discharges
 Electrical Outages and Inadequate Battery Charges (Colorado, 2004)
 Maintenance Challenges for Poll Workers (San Diego, 2004)
 Rapid Obsolescence and Toxic Waste Disposal
• When they are obsolete, HAVA funds won’t replace them
• Rechargeable batteries in DREs wear out and are toxic waste
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
Election Official’s Challenges
from the San Diego Report, 2004
 Recruiting more poll workers, and technically savvy workers
 Additional training for poll workers
 Testing thousands of machines before each election
 Providing troubleshooting and hotline support during an election
 Providing field support
 Educating the public about using the machines
 Handling technical problems on election day
 Creating a back up plan for emergency problems
Election Complexities with EVoting
Copyright ©2004 www.VotersUnite.org
[email protected]
HAVA-Compliant Alternatives
HAVA does not require DREs. Here are some alternatives:
 Precinct optical-scan with ballot-marking devices.
• Two major studies of voting systems determined that precinct-count
optical scan systems outperformed DRE voting machines in terms of
residual voting errors and cost per voter.
•
Ballot-marking devices provide all the support for disabled individuals
that DREs provide.
 Tactile ballots for the vision-impaired and hearing-impaired
• Tactile ballots have been used in Rhode Island very successfully.
 Open Voting Consortium System
• Free voting software to install on standard computers provides all the
features of other computerized systems.
HAVA–Compliant Alternatives to DREs
Copyright ©2004 www.VotersUnite.org
[email protected]
The Possibility of Paper
 Several touch screen systems provide an integrated printer for
printing voter-verified paper ballots:
•
•
•
Avante
AccuPoll
TruVote
 Some citizen groups advocate the exclusive use of paper ballots in
systems where no vote is recorded electronically:
•
Ballot Integrity Project
HAVA–Compliant Alternatives to DREs
Copyright ©2004 www.VotersUnite.org
[email protected]
Comparison of Approximate
Acquisition Costs
System Type
5-Booth Precinct
Paperless DRE System
$19,000
DREs with integrated VVPB Printer
$20,000
Optical Scan + Ballot-Marking Device
$10,000
Optical Scan + Tactile Ballots
$6,000
Open Voting Consortium System
Cost Considerations
Free software + $6,000
Copyright ©2004 www.VotersUnite.org
[email protected]
Hidden Costs of DREs
 Increased warehouse costs for secure and environmentally-controlled
storage for the machines when they are not in use.
 Increased energy costs for keeping the backup batteries charged
between elections.
 Increased labor costs for security when these machines are stored
overnight at the polling place before an election.
 Increased costs for hardware maintenance and software upgrades for
each of the thousands of such machines for a typical large county.
 Increased costs for expendable parts, including the backup batteries
and smart cards used by these machines.
 Increased labor costs for verifying the software and firmware version
on each machine before every election.
Cost Considerations
Copyright ©2004 www.VotersUnite.org
[email protected]
More Hidden Costs of DREs
 Increased personnel costs for performing logic and accuracy tests on
every one of the machines prior to the start of every election.
 Increased labor costs for hiring additional poll workers (San Diego
required twice as many when it switched to DREs).
 Increased training costs for longer training sessions and larger
number of poll workers to train on using a more complicated system.
 Massive costs for replacing these machines in 10 to 15 years when the
technology that they use is no longer maintained or supported by the
vendor, and HAVA funding is no longer available.
Cost Considerations
Copyright ©2004 www.VotersUnite.org
[email protected]
What Computer Experts Say
 Johns Hopkins/Rice Report on Diebold DRE software
• “Our analysis shows that this voting system is far below even the most
minimal security standards applicable in other contexts.”
 SAIC Report on Diebold Software
• “The system, as implemented in policy, procedure, and technology, is at
high risk of compromise.”
 Compuware Report on Diebold, ES&S, Sequoia, and Hart
• The study showed that all four of the voting machines had serious security
problems. These problems are described in great detail in the report,
which is over 200 pages long.
 RABA Technologies Report on Diebold
• "I was really surprised with the totality of the problems we found. Just
about everywhere we looked we found them."
Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
How the Vendors Respond
 RABA expert
• William Arbaugh, a University of Maryland assistant professor of
computer science who participated in the test, graded the system an "F,"
"with the possibility of raising it to a 'C' with extra credit -- that is, if they
follow the recommendations we gave them."
 Diebold President
• Bob Urosevich said in the release that the RABA Technologies report
confirmed "the accuracy and security of Maryland's voting procedures and
our voting systems as they exist today."
Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
Georgia Secretary of State Cathy Cox
 From a Walker County, GA news report:
•
“The voting machines have been used for six elections, three of which
were for the same State House District 1 race. Problems have cropped up
at every election.”
 Claims made, on Atlanta television, by Cathy Cox days after
the problems in Walker County.
•
"Though Georgia Secretary of State Cathy Cox said the state’s 26,000
elections voting machines performed without any problems on Super
Tuesday earlier this week, some lawmakers Thursday said the machines
may nonetheless be vulnerable to fraud and wanted printed receipts to
serve as proof of the computer tabulation."
Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
Florida Secretary of State Glenda Hood
 U.S. Representative Robert Wexler (D-FL)
•
One of the many recent severe election problems in Florida has led to
the federal lawsuit filed by Representative Robert Wexler (D-FL). It is
currently scheduled for a hearing in August of 2004.
 Florida Secretary of State Glenda Hood
•
On Lou Dobbs tonight, March 8, 2004, Secretary of State Glenda Hood
said this:
“Well, I have a high confidence level. And it's based on the fact that,
since 2002, when we put new equipment in place in the state of Florida,
that we have had no problem whatsoever, according to our 67
supervisors of elections.”
Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
Clearing up Misconceptions
 League of Women Voters' opposition to Voter-verified paper
ballot (VVPB) is not a member-endorsed position.
•
Many rank and file members strenuously object to the policy and
have launched a website in opposition to the Executive Board's
action.
 No systems require voters to verify their ballots
•
If machines are required to provide a method by which voters
could verify paper ballots, voters would NOT be required to verify
them.
Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
Clearing up More Misconceptions
 VVPB does not violate the Americans with Disabilities Act
•
Some people are concerned that providing different verification methods
to sighted and blind individuals would be a violation of the law. The
United States Department of Justice disagreed in an official opinion issued
through its Office of Legal Counsel.
 Optical scan machines are a reliable way to count paper
ballots
•
Jim Dickson, vice president for governmental affairs at the American
Association of People with Disabilities, says: "As a matter of fact, not
theory, whenever paper ballots are counted by an automatic tabulator you
never get the same results twice."
However, a Caltech/MIT report says, “"Optical scanning has the best track
record of all equipment types currently in use."
Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
Clearing up More Misconceptions
 Election Center Received Large Donations from Vendors
•
In March of 2004, it was discovered that, for years, the Election Center
has been receiving large donations from the three major manufacturers of
paperless electronic voting. Optical scan machines are a reliable way to
count paper ballots
•
Executive Director R. Doug Lewis, a major defender and proponent of
DREs, confirmed that the center had taken donations from makers of
electronic voting machines – Sequoia, Electronic Systems & Software,
and Diebold.
•
Lewis said he did not think accepting donations from the
manufacturers presented any conflict of interest or breach of ethics.
Distinguishing Truth from Myth
Copyright ©2004 www.VotersUnite.org
[email protected]
Election Transparency
 Election transparency minimizes fraud and miscounts.
 With transparency, ballots are collected and counted in public view.
 With electronic voting, ballots are collected and counted by software
processes, which are:
•
•
•
•
developed by anonymous software engineers, who are hired by vendors.
federally qualified by anonymous testers, who are hired by vendors.
installed and maintained by technicians, who are hired by vendors.
trade secrets of the vendors and therefore not open to public scrutiny.
“An election that uses electronic ballots is not
transparent.”
Copyright ©2004 www.VotersUnite.org
[email protected]
Myth Breakers
for Election Officials
A Brief Summary
For more details, download the entire “Myth Breakers” file:
http://www.votersunite.org/takeaction/mythbreakers.pdf
Copyright ©2004 www.VotersUnite.org
[email protected]