The Real Story:
Seventeen Months of Fumbling and Bumbling by the SF Department of
Elections and Elections Commission
Timlines, actions, inaction
and poor decisions related to IRV implementation
Background: The Department
of Elections, under its Director John Arntz, has made a series of
poor decisions throughout the 17 months since instant runoff voting
was passed on March 5, 2002. The Elections Commission, which was
created little more than a month before the passage of instant
runoff voting, has sat by ineffectually while their director has
fumbled the ball.
Director John Arntz, who has been the director
since the firing on April 22, 2002 of previous director Tammy
Haygood, first began working at the Department of Elections in 1999
as a desk clerk, behind the counter answering phones and giving
forms to the public. Previously he had ZERO election experience, and
ZERO administrative or supervisorial experience. His previous employ
had been writing product descriptions for a seafood company. The
seven members of the Elections Commission also had very little
election experience, prior to their current tenure. While one
certainly can appreciate the way Director Arntz has climbed the
internal ladder at the Department of Elections, his election
experience is not deep, and the Commissionĺ─˘s is even less.
matters worse, neither the Director nor the Commission has sought
expertise or counsel of known experts in the field about how to
implement instant runoff voting, just as it has failed to involve
the public in many key decisions about elections, despite the clear
lessons from the 2000 presidential elections about the value of
greater transparency in election administration. Consequently, San
Francisco has been treated to 17 months of fumbling and bumbling by
the governmental agencies charged with fulfilling the will of the
voters and implementing the law. This view was confirmed by Judge
James Warren, who presided over the lawsuit filed against the City,
and who characterized the Departmentĺ─˘s efforts as ĺ─˙fumblingĺ─¨ and ĺ─˙at
best, bumpy and haphazard.ĺ─¨
Below is a 17-month timeline of poor
decisions and failures related to implementation of instant runoff
voting, compiled from various letters and meeting notes. What
follows is a remarkable record of the failure to heed the will of
Before going through the detailed chronology, here are
two particularly poor decisions that directly threatened the success
of IRV implementation:
Elections Director John Arntz goes
against advice from his vendor, members of the Elections Commission,
and staff from the Center for Voting and Democracy, and makes a
decision to have ES&S upgrade the Optech Eagles (precinct-based
scanner) for counting IRV ballots, rather than using the Optech IV-C
(centralized scanner) or touch screens. Arntz rejects central scan
solution because it requires transporting and handling ballots,
rejects touch screen because he claims there is a lack of time and
money. Yet in March he will create a ĺ─˙partial hand countĺ─¨ method
that requires extensive transport and handling of ballots. And other
counties implement their touch screen systems in a matter of a few
months. This decision made the conversion of voting equipment far
more difficult than it needed to be.
Dept of Elections
submits its own application for its own designed procedure for
counting IRV ballots, called a ĺ─˙partial hand count,ĺ─¨ to Secretary of
State. This is meant as a backup plan, but it is untried and
untested. CVD staff Steven Hill contacts John Arntz and tells him
his method has problems. Hill tells Arntz that if heĺ─˘s going to
design a ĺ─˙backup planĺ─¨ that utilizes a manual tally, better to model
it after the kind that has been used successfully for decades in
Australia, Britain, and Ireland, utilizing paper ballots, a pure
hand count, and no voting equipment. Arntz ignores this advice, he
plunges forward with the flawed ĺ─˙Arntz method.ĺ─¨ Says Hill, ĺ─˙John
Arntz reinvented a wheel that did not need to be reinvented, and he
created a square wheel besides.ĺ─¨ The Arntz method later will be
decisively rejected by the Secretary of Stateĺ─˘s office.
Proposition A passes with 55% of the vote, repealing December
runoff elections and adopting instant runoff voting.
Meetings occur with Elections Director Tammy Haygood and members of
the new Elections Commission (EC). Haygood states that she does not
have the staff to implement IRV. It quickly becomes apparent that
Haygood is not going to move quickly to implement IRV by Nov. 2002,
and that Elections Commissioners do not know much about elections,
much less IRV. Hence, they defer to their Director.
ES&S representative Joe Taggard (the cityĺ─˘s voting equipment
vendor) writes a letter to the Elections Commission, asking for
guidelines regarding certain issues related to IRV implementation.
At the request of EC President Tom Schulz, Center for
Voting and Democracy staff draft a letter responding to ES&S/Joe
Taggard. But the EC does not mail out the letter until May 10.
Elections Commission fires its Elections Director Tammy
Haygood. John Arntz, with little elections experience, is plucked to
be acting director.
New EC President Michael Mendelson
finally mails out April 3 letter responding to Taggard/ES&S.
Another letter also mailed to SoS Elections Chief John Mott-Smith.
Letter from ES&Sĺ─˘ Joe Taggard to the Elections
Commission recommends one of two possibilities for implementing IRV
in San Francisco: 1) utilizing existing equipment by counting the
IRV ballots in a central location (not in the precincts) on the
Optech IV-C high-speed scanners (not on the Optech Eagles), or 2)
count IRV ballots in the precincts on ES&S touch screen
equipment. Both of these technologies are more readily adaptable to
IRV than precinct-based Optech Eagles.
Meeting of ad hoc
task force for IRV implementation composed of Joe Taggard
(ES&S), Michael Mendelson (Elections Commission), Brenda Stowers
(Elections Commission), Steven Hill and Caleb Kleppner (Center for
Voting and Democracy), Deputy City Attorney Julie Moll, and a few
others to establish overall policy and guidelines.
with Michael Mendelson and representatives from Diebold (another
voting equipment vendor). Diebold expresses interest and ability to
implement IRV (they are the vendor that does the ranked ballot
election in Cambridge, MA).
ES&S representative Joe
Taggard e-mails a memo to Deputy City Attorney Julie Moll with a
comparison chart that clearly demonstrates that touch screen voting
equipment is the easiest and a relatively inexpensive method for
running an IRV election due to availability of Prop 41 money for
upgrading voting systems.
Meeting of the ad hoc task force
for IRV implementation.
Acting Director John Arntz writes to
Mayor Willie Brown asserting that he will exercise the
charter-permitted option of not running an IRV election in November
2002. He states ĺ─˙Although the Department of Elections will introduce
IRV in 2003, the Department will not delay its responsibility to
plan for IRV.ĺ─¨
ES&S presents its proposal for using
Touch screens to implement IRV at meeting of ad hoc task force for
Second meeting with vendor Diebold
representatives in which they present their IRV solution. Diebold
again expresses interest and ability to implement IRV.
ES&S presentation on ballot layout, IRV algorithm, and other
Letter from ES&Sĺ─˘ Joe Taggard once again
outlines the difficulties of upgrading the Optech Eagles for the
purposes of counting IRV ballots, and urges Director Arntz to go
with other available options such as using touch screens.
Michael Mendelson requests that Center for Voting and Democracy
staff draw up implementation policies. Over the next month, CVDĺ─˘s
Caleb Kleppner drafts memos related to ballot design, interpreting
ballots, overall policy details (how to interpret skipped rankings,
release of results, etc.), dealing with write-in votes, Logic and
Accuracy testing, and more. Those memos mostly end up being ignored.
Meeting of Brenda Stowers, Julie Moll, John Arntz, Caleb
Kleppner, and Steven Hill. Elections Director John Arntz makes
decision to have ES&S upgrade the Eagles, rather than upgrade
the Optech IV-C or use touchscreens. Arntz instructs Julie Moll to
tell vendor to upgrade the Eagles and not to use a central scan or
touch screen solution. Arntz rejects central scan solution because
it requires transporting and handling of ballots, rejects touch
screens because of lack of time and money (yet in March he will
create a ĺ─˙partial hand countĺ─¨ method that requires extensive
transport and handling of ballots. And other counties implement
their touch screen systems in a matter of a few months, particularly
with millions of dollars available from Prop 41 to pay for this).
This ends up being a crucial decision that endangers the success of
the IRV project.
Deputy City Attorney Julie Moll says a
letter to ES&S has been mailed, asking ES&S if they can and
will implement IRV on the Eagles. Otherwise, the City will have to
seek another vendor.
Meeting with President of EC,
ES&Sĺ─˘ Joe Taggard responds to letter
of October 15, saying ES&S is willing to upgrade the
precinct-based scanners known as Optech Eagles, but reiterates once
again that touch screens or centralized scanners (Optech IV-C) is a
better way to count IRV ballots. Nevertheless, Director Arntz opts
for the difficult course of upgrading the precinct-based scanners,
the Optech Eagles, which ES&S says will cost $1.5 million.
Arntzĺ─˘s decision made the conversion of voting equipment far more
difficult than it needed to be.
The original date
for signing a contract for upgrading the Eagles comes and goes.
ES&S delivers its proposal for upgrading the Eagles
with $1.5 million price tag
EC President Michael
Mendelson tells Caleb Kleppner that contract between the City and
ES&S will be signed in 10-14 days
EC President Michael Mendelson says
contract will be signed in 10 to 14 days
Michael Mendelson says again that the contract will be signed in 10
to 14 days
EC President Michael Mendelson says once
again that contract will be signed in 10 to 14 days. When Steve Hill
reminds Mendelson that he said this nearly three weeks ago,
Mendelson slams down the phone on Hill and refuses to speak to Hill
again until April 2 at the Elections Commission meeting.
The Department of Elections begins thinking about coming up with a
ĺ─˙backup planĺ─¨ in case for some reason the contract negotiations with
ES&S donĺ─˘t work out or take too long.
Attorney Julie Moll issues memo to John Arntz, stating that whether
or not the department may ĺ─˙adopt its own rules for evaluation,
testing and approval of a ranked-choice voting systemĺ─¨ without
approval from the Secretary of State is a legal gray area. Her memo
states, ĺ─˙The California Constitution provides that charter cities
such as San Francisco may govern their municipal affairs, including
the manner and method of electing local officeholders. Cal. Const.,
Art. XI, Č▀ 5. When a charter city adopts a law governing municipal
elections, and that law poses a genuine conflict with State law, the
local law prevails unless the State law advances a matter of
statewide concernĺ─ÂIt is also possible that a court would conclude
that all matters relating to the manner and operation of a charter
cityĺ─˘s voting system are municipal affairs.ĺ─¨
from ES&S says the contract will be signed in two weeks
At the Elections Commission meeting, President Michael Mendelson and
Director John Arntz say the contract will be signed within two
Director John Arntz submits his own application for
his designed procedure for counting IRV ballots, called a ĺ─˙partial
hand count,ĺ─¨ to Secretary of State. This is meant as a backup plan,
but it is untried and untested.
CVD staff Steven Hill
tells John Arntz at Elections Commission meeting that his untried
and untested ĺ─˙Arntz methodĺ─¨ has problems. Hill tells Arntz that if
heĺ─˘s going to design a ĺ─˙backup planĺ─¨ that utilizes a manual tally,
better to model it after the kind that has been used successfully
for decades in Australia, Britain, and Ireland, utilizing paper
ballots, a pure hand count, and no voting equipment. Hill tells
Arntz this several times over the next two months, but Arntz ignores
this advice. He plunges forward with the flawed ĺ─˙Arntz method.ĺ─¨ This
ends up being the second crucial decision that endangers the success
of the IRV project.
The Elections Commission says the
contract will be signed next week.
Joe Taggard from
ES&S says the contract will be signed in early April, with
application for certification being made in June or July.
Secretary of State responds to March 14 application by the
Department of Elections and requests additional info and money
City Attorney Dennis Herrera says signing the
contract next week ĺ─˙is realistic.ĺ─¨
Joe Taggard (from
ES&S) says itĺ─˘s likely the contract will be signed April 10.
Elections Department representative Jennifer Novak says ĺ─˙we
will probably sign next week;ĺ─¨ Taggard says by the end of the month.
John Arntz sends Public Education Plan for IRV to Supervisor
Sandoval. Arntz does not release it publicly or seek community
DOE responds to March 27 letter from Secretary of
State and submits full application
Taggard, Arntz, Herrera
all indicate the contract will be signed ĺ─˙next week.ĺ─¨
First the DOE cancels SoS hand count demo, then SoS cancels next
Memo from John Mott-Smith requesting answers from
Department of Elections to specific questions about election
Arntz finally releases the Public Education
Plan to the public. The Plan has a price tag of $2.9 million, a huge
sum that everyone knows is ĺ─˙dead on arrivalĺ─¨ at the Board of
Supervisors, particularly given the budget deficit. There is much
speculation that perhaps that is Arntzĺ─˘s intention.
responds to Mott-Smithĺ─˘s May 16 memo
application to Secretary of State; Secretary of State says
application in ĺ─˙incomplete.ĺ─¨
As expected, and with the
support of IRV advocates, the Finance Committee of the Board of
Supervisors slashes the Public Education Plan from $2.9 million to
Secretary of State Kevin Shelley sends a
sternly-worded letter to the Department of Elections, saying
ĺ─˙obviously this process would be much further along had you not
waited over one year to submit your application for certification.
Your delay is particularly baffling given the fact that my office
began advising you of the need to certify your new system just days
after the March 2002 election, and has continued to remind you of
this requirement over the past year.ĺ─¨
demonstrates the ĺ─˙Arntz methodĺ─¨ for Secretary of State.
Board of Supervisors approves $1.6 million for ES&S contract to
upgrade voting equipment, and $750,000 for Public Education Plan.
Arntz tells Elections Commission they are preparing
December runoff as backup plan in case nothing gets certified by the
Secretary of State.
Short response by John Arntz to
Secretary of Stateĺ─˘s June 5 letter. Arntz declines opportunity to
respond to direct queries made by Secretary of State Shelley
regarding ĺ─˙constitutional, statutory, and procedural defects in the
manual count IRV system proposed in your application.ĺ─¨
CONTRACT BETWEEN ES&S AND CITY FINALLY IS SIGNED -- nearly six
months after the contract was supposed to have been signed on
January 1, 2003.
City attorney Dennis Herrera issues a
public opinion about IRV which partly contradicts the confidential
memo from his office dated February 19 (see above).
ĺ─˙Arntz methodĺ─¨ for a manual IRV tally is denied certification by the
Secretary of Stateĺ─˘s Voting Systems Panel. Mr. Arntz has difficulty
answering questions from the panel about his own method, including
ones as simple as ĺ─˙how do you resolve ties?ĺ─¨
Arntz, without authorization from the Elections Commission,
announces to the Finance Committee and then San Francisco media that
IRV is dead for this year.
The Elections Commission
declines to stop its director from killing IRV, or from planning a
December runoff, despite a December runoff being a violation of the
charter as well as state law.
A lawsuit is filed against
the Department of Elections, Elections Commission, and Director John
Arntz for failing to implement the law and comply with the will of
the voters to use instant runoff voting.
Summary of bad decisions
The Department took no concrete steps to determine
definitively a method of implementation from the passage of Prop A
on March 5, 2002 until October 7, 2002. But even then, the first
decision about implementation included several mistakes.
Director Arntz decided to rule out the use of touch screens, even
though touch screens are easiest for the voter and make
implementation of instant runoff voting a simpler task. Instead, he
decided to upgrade the Optech Eagles. He also ruled out the use of
central scanning on the Optech IV-C (which would have been the least
expensive solution, and applicable even after the city moves to
touch screens, since the IV-C is used to count absentee and
provisional ballots). Arntz rejected the ĺ─˙central scanĺ─¨ solution
because he did not want to have to transport ballots on Election
Night from the precincts to a central counting location.
Director Arntz began preparing a backup solution in February ĺ─ý his
ĺ─˙partial hand countĺ─¨ ĺ─ý this approach involved not only moving
ballots on Election Night but handling them repeatedly. Even in
drafting a backup plan, Director Arntz failed to consider more
workable solutions than his ĺ─˙Arntz method.ĺ─¨ More workable solutions
would have included a reconsideration of the ĺ─˙central scan,ĺ─¨ which
is not only logistically easier than a partial hand count, but costs
must less money and is much much faster. He also failed to consider
a pure hand count on paper ballots, as has been used for decades in
places like Cambridge, MA, Australia and Britain, which also is much
less expensive and faster. One vendor has estimated that the cost
and time for doing a hand count of paper ballots would be two days
to count all of San Franciscoĺ─˘s ballots for the November election,
for a cost of approximately $225,000, which is only a tenth of the
cost of the ĺ─˙Arntz methodĺ─¨ and a fraction of the cost of a second
citywide election in December.
Lack of public involvement and
assistance from experts
Director John Arntz also consistently
excluded the public from any meaningful discussion about
implementation of instant runoff voting. This should have included
input about ballot design, voter education, and the method of
implementation of IRV.
Director Arntz and the Elections Commission
also failed to solicit input from recognized experts who could
assist them with implementation, including experts from the Center
for Voting and Democracy, experts in Cambridge, MA, Britain,
Australia, and elsewhere.
Neither Director John Arntz nor the
Elections Commission has significant election administration
experience. Unfortunately, the implementation of instant runoff
voting in San Francisco suffered as a result. Director Arntz made a
series of poor decisions throughout the 17 months since instant
runoff voting was passed on March 5, 2002, and the Elections
Commission sat by ineffectually while their director fumbled the
ball. The implementation of IRV was plagued not only by poor
decision-making, but also by a lack of responsiveness to the
community, a lack of leadership, and a lack of can-do attitude.
Consequently, San Francisco has been treated to 17 months of
fumbling and bumbling by the governmental agencies charged with
fulfilling the will of the voters and implementing the law.
have claimed that, compared to his predecessor, Director Arntz has
done a satisfactory job for the past couple elections and helped
turn the Department around. There may be some truth to this
assessment, which is a big feat considering the state of the
Department when he took the reins. Nevertheless, since March 5, 2002
and particularly since the end of April when he took over the
Department of Elections from his predecessor, a significant part of
Director Arntzĺ─˘s job entailed developing and implementing a workable
voting system to be used for instant runoff voting. On this score he
has performed extremely poorly, resulting in litigation against the
City of San Francisco.