Re: Experience at Tuesday's Presidential Primary Election at a Precinct in Santa Clara County

From: Ron Olson <ron_at_caseohio_dot_org>
Date: Thu Feb 07 2008 - 18:35:31 CST

Dylan,

 

I saw this on another email list

 

The courage campaign is asking concerned citizens to sign a petition to
ask that the 94,000 Decline to State ballots cast in Los Angeles County
in the primary be hand counted. A failure by voters to fill in a
redundant democratic bubble on a democratic ballot caused these ballots
to be rejected by the optical scanners.

http://www.couragecampaign.org/page/invite/counteveryvote?stg_signup_id=2339
58.

Ron

 

From: ovc-discuss-bounces+ron=caseohio.org@listman.sonic.net
[mailto:ovc-discuss-bounces+ron=caseohio.org@listman.sonic.net] On Behalf Of
Dylan Hirsch-Shell
Sent: Thursday, February 07, 2008 2:40 PM
To: Open Voting Consortium discussion list
Subject: Re: [OVC-discuss] Experience at Tuesday's Presidential Primary
Election at a Precinct in Santa Clara County

 

I voted in Los Angeles County. My polling place was using the Inka-Vote
system. For those of you who don't know, InkaVote consists of paper ballots
that slide into a plastic ballot holder with holes through which an ink pen
with a flattened circle shaped tip can be pressed to mark specific circles
on the ballot (scantron style, but with a pen that fills the empty circles
on the ballot with a perfect circle of the same size). Each plastic ballot
holder has laminated pages on either side indicating which hole represents a
vote for which candidate/preference. Only one column of holes, consisting
of choices for a particular office or state propositions or city
propositions, is accessible at a time, such that when you first put your
ballot in and turn to the first laminated page you can vote for one
candidate for president, then you turn the page and you're presented with
the various state propositions that you vote "yes" or "no" on, then you turn
the page and you're presented with the city propositions that you can vote
on. Apparently, because of the situation with different parties allowing
different groups of voters to participate in their primaries, there were
different polling stations, each with a different ballot holder presenting
different sets of accessible holes for marking, set up for voters in each
party and for the Non-Declared voters.

As a Decline-to-State voter, I had heard before the election that I would
have to request a Democratic ballot if I wanted to vote for a Democratic
presidential candidate. When I requested a Democratic ballot at the polling
place, the poll worker still gave me a Non-Declared ballot (it clearly said
Non-Declared across the top), but said that I should go to the Democratic
booth that had been set up and mark the circle indicating that I wanted to
vote for a Democratic presidential candidate, and then mark the circle
corresponding to the candidate of my choice.

I hope this doesn't mean that my vote in the Democratic primary wasn't
counted.....

On Feb 7, 2008 5:10 AM, Arthur Keller <voting@kellers.org> wrote:

I was Precinct Inspector (chief poll worker for a polling place) at
Tuesday's California Presidential Primary at Precinct 2113 in Palo
Alto in Santa Clara County, California. I have posted messages
describing my experiences at previous postings (see
http://gnosis.python-hosting.com/voting-project/November.2006/0052.html
and
http://gnosis.python-hosting.com/voting-project/June.2006/0008.html ).

Santa Clara County previously used Sequoia AVC Edge voting machines
with VeriVote printers. However, these were decertified by Secretary
of State Debra Bowen. So instead we had one voting machine per
precinct together with paper ballots.

Santa Clara County provides ballots in 5 languages. All of the
ballots are bilingual with English and another language. If someone
does not ask for a ballot in a specific language, the English/Spanish
ballot is given to the voter. So there are 4 different language
ballots.

Because this was a primary election, there were ballots for each of 6
parties (Democratic, Republican, American Independent, Libertarian,
Green, and Peace and Freedom), as well as Non-Partisan. The
Non-Partisan category is for the Decline-to-State voter. I suspect
that a significant number of American Independent Party registrants
intended to register as Decline-to-State voters. I noticed some of
the ballots from other parties had Democratic Presidential candidates
written in on their ballots. I doubt if those preferences will be
included in the tally, but I assume that their preferences for the 7
propositions will be included.

At 4 different language ballots times 7 (6 parties plus
Non-Partisan), there were potentially 28 ballot styles in Santa Clara
County. There were about a dozen actual ballot styles based on our
exact combination of language preferences and party registrations.
That we did not need all the ballot style combinations is largely
because the Registrar of Voters has made a concerted effort over the
last few years to get all voters to state their language preference.
Part of the intent is to provide the election materials to the voter
in the right language and avoid unnecessary paper. Previously the
materials were in all languages. Now they are in English/Spanish,
Chinese-only, Tagalog-only, and Vietnamese-only. Printed in smaller
batches, the Chinese, Tagalog and Vietnamese materials are printed in
a sturdier paper, why the English/Spanish materials inside the
booklet are printed on newsprint. All ballots, however, include
English and one other language, so that English readers can hand
count the ballots if necessary.

Those who registered as Decline-to-State were given a Non-Partisan
ballot unless they specifically requested a Democratic ballot or an
American Independent Party ballot. Those are the only two parties
that allowed Decline-to-State voters to participate in their primary.
Specifically, the Republican party chose not to allow
Decline-to-State voters to participate in their primary, and those
who requested a Republican ballot were to be offered a Voter
Registration Card to reregister. Decline-to-State voters who
requested a Democratic ballot were not to be offered a Voter
Registration Card to reregister unless they specifically requested
it, as per Poll Worker training. I did, however, offer several Green
party registered voters a Voter Registration Card when they requested
a Democratic ballot, because I could not give them a Democratic
ballot.

Giving Decline-to-State voters a (paper!) Non-Partisan ballot by
default was in general easily changed for Precinct Voters who wanted
a party ballot, but it presented a problem for Vote by Mail voters.
They were sent Non-Partisan ballots by mail, and had to specifically
request a Democratic primary ballot if they wanted one. Many of the
33 Provisional voters in my Precinct were Decline-to-State voters who
voted a Non-Partisan ballot absentee and came to the polls to request
a Democratic ballot to vote in the Presidential primary. I have no
idea how the Registrar of Voters will handle those duplicate ballots,
but I hope that they are not prosecuted for voting twice. This issue
was not covered in training. Perhaps there should have been ballots
that included only the Presidential candidates (and not the
propositions) for the two parties that allowed Decline-to-State
voters to participate. Doing so would have allowed those voters who
had cast a Vote-by-Mail ballot to have provisionally voted for
President. Since we in the polling place new that they had received
a ballot, but did not know whether they received a non-partisan or
partisan ballot, the ballot would necessarily have to have been
provisional.

Many Decline-to-State voters in the Precinct specifically requested a
partisan ballot for the Democratic party. Some of those that
received a Non-Partisan ballot at the polls complained when they did
see the Presidential candidates. We swapped the Non-Partisan ballot
for a Democratic ballot (after asking which party, Democratic or
American Independent) after marking the preference as required in the
voter rolls.

We were issued 200 English/Spanish Democratic ballots as part of the
total of 430 ballots in about a dozen ballot styles. Late morning, I
discovered that we had gone through more than half of the
English/Spanish Democratic ballots (and we had 10 Chinese/English
Democratic ballots, but no other languages). With about two-thirds
of the time left until poll closing, I was worried about running out
of paper ballots. So I called my Field Inspector and then the
Registrar of Voters hotline requesting additional ballots. I was
told to call back when I had 30 or 40 ballots left. Based on
Secretary of State directive, we had been discouraged from offering
the use of the electronic voting machine to anyone unless they
specifically requested it. I had fully expected not to use it at
all. Several hours later, someone from the Registrar of Voters in
San Jose showed up with 100 paper ballots, or rather photocopies of
paper ballot proofs (they had all the trim marks on them).

At that point, I decided to offer Democratic voters the choice of
"paper or plastic," that is, paper ballot or electronic voting
machine. One Democratic voter chose the electronic voting machine.
The second voter who chose the electronic voting machine was a
Decline-to-State voter who did not request a party ballot. After the
Decline-to-State voter cast a ballot on the voting machine, the voter
returned to me and complained that the ballot did not offer a choice
for Presidential candidate. I told the voter that the voter had
specifically to request a ballot for Democratic or American
Independent Party. The voter said that she was not told of this
requirement and that how the option to vote in the partisan primary
worked. The fact that the voter did not appear to be a native
speaker of English indicates that there may have been some language
barrier in the instructions. At that point, I decided to offer use
of the electronic voting machine only to those who were voting a
Democratic ballot. No one else specifically requested it (if they
had, I certainly would have allowed them to do so). Had we not been
running out of paper ballots, I doubt anyone would have used the
electronic voting machine, since it was used only by request. As the
number of Democratic ballots dwindled, I switched to requesting that
everyone who cast a Democratic ballot (including Decline-to-State
voters) do so on the electronic voting machine. In the end 78 voters
used the electronic voting machine (for 77 Democratic ballots and one
non-partisan ballot), and I had seven Democratic ballots in
English/Spanish left. So we had 270 Democratic ballots cast, when we
were originally supplied 200. While I was given 100 spare
photocopied paper ballots, first 50 were taken away for use by other
precincts, and then 25 more were taken away. I had 25 unused spare
photocopied paper ballots. None were used at my precinct.

We had a steady stream of people coming all day, with the line to
sign in never longer than 7 voters, and the line for the electronic
voting machine never longer than 5 voters. As far as I could tell,
most voters had no problem with the electronic voting machine. But
some were confused about how to review their ballots and get the
paper trail printed. When presented with a rounded rectangle saying
something like, "Press here to review a paper copy of your ballot,"
they instead pressed the "Review" button. To get back, they had to
press the "Return" button or something similar but non-intuitive. I
encountered this particular problem previously at other elections.
Clearly the human factors engineering for the Sequoia AVC Edge
electronic voting machine related to the paper trail is faulty. But
such problems often arise when a process or software is modified with
a "bolt-on" step instead of being rethought and re-engineered.

In the November 2007 local election, precincts were given paper
ballots (I think they were only in English/Spanish). These paper
ballots were only to be given to those who specifically requested it
(I think non-English speakers were to use of the five electronic
voting machines provided). Few voters specifically requested paper
ballots, while no voters specifically requested use of the electronic
voting machines. Several voters, when handed a paper ballot, said
something like, "Back to the Future." Although some were
disappointed that so much money was spent on voting machines that
were then not to be used, no one handed a paper ballot requested to
vote on the electronic voting machine. Only those who the poll
worker requested to vote electronically did so. In the November 2006
election, when I specifically gave voters a choice of paper or
plastic, as the paper trails on the electronic voting machines ran
out, most voters chose paper ballots since they could get to vote
faster (as the line grew for the one electronic voting machine left).
This unscientific study indicates that voter preference is dominated
by poll worker instruction.

One person (a computer expert very knowledgeable about security) at
my precinct specifically said he has surprised that I was requesting
he vote on the electronic voting machine, as he knew that I knew
about their security risks. I told him that Secretary of State Debra
Bowen's directive to hand verify the paper trail against the
electronic voting machine totals for 100% of the ballots for 100% of
the machines assured me that the electronic voting machine ballots
were more likely to be correctly tallied than ballots marked on
photocopied paper that would have to be hand-transcribed for machine
tallying. Knowing my knowledge in the area, he was reassured and
proceeded to vote electronically, as did his wife. One voter asked
whether the ballot was encoded on the activation card returned to the
poll worker. So I demonstrated the status key that showed the card
contained only that the card was used to vote on a specific voting
machine (by serial number) at a specific time.

The card activator had a security seal over the whole were it is
designed for a PC card to be inserted.

The card activator has a menu provides access to a bunch of
undocumented features, including the ability to change the date and
time on the card activator. (I did not check whether that feature
was still available, but I noticed previously that it was a feature.)
One security risk is that a card activator may have its date and time
changed, a supply of cards activated with the time stamps of during
the election, and then the date and time restored. Those cards could
be used to vote in one or more busy precincts. However, the number
of votes cast would exceed the number of voters who signed in to vote.

When opening the polls in the morning, we had one small glitch.
There are four long strips of paper seal for sealing both sides of
two cardboard ballot boxes. The strips had backing, but these were
also in strips several inches wide crosswise. It was hard to pull
off all of the backing strips. One of the poll workers accidentally
caused the half-exposed sticky part to stick to itself while trying
to remove the remainder of the backing. I had to call the Field
Inspector for a spare. The Field Inspector quickly brought a spare
prior to opening the polls. Later, as the ballot box for
vote-by-mail ballots filled up, one side sagged, with the paper seal
failing to stick very well. I surmised that the seal that failed to
stick was the spare one, and perhaps it was from a batch of a
previous election. I note that the backing of the strips for sealing
the box of cast ballots and the box of unused ballots were cut the
long way, which was easier to fully remove with less risk of the
strip sticking to itself. I suggest that long backing strips also be
used on the strips for sealing the ballot boxes used during Election
Day.

The white cardboard ballot box was for regular paper ballots, and the
brown cardboard ballot box was for vote-by-mail and provisional
ballots, as well as voter registration forms, preferred language
surveys and permanent vote-by-mail requests. Separation of those two
types of ballots is mainly a capacity issue. The large paper ballots
stack neatly by themselves in the white ballot box. Had vote-by-mail
ballots been added to the mix, the combination of short and fat
envelopes along with large and flat ballots would have required a lot
more space. I understand that some election protection volunteers
were concerned that some poll workers in Mountain View were putting
Democratic ballots from Decline-to-State voters in the brown
cardboard box. While this was a training issue and potentially a
voter privacy issue, it should not have mattered since all the voted
paper ballots were returned to the Registrar of Voters in a single
box after polls close and the ballot boxes are opened and the ballots
counted (but not tallied).

When all (or most) ballots were cast on the electronic voting
machine, provisionally ballots were identified by activating a card
as provisional. The card activator displayed a number, which was
recorded on a pink paper form. This number was also recorded on the
card along with the activation and presumably recorded with the
ballot on the electronic voting machine so that the Registrar of
Voters can have the electronic ballot included in the canvass if the
provisional ballot is to be counted. The paper form includes the
voter identification and information needed to ascertain whether the
provisional ballot is to be counted. There are two problems with
this approach. First, the ballot is permanently identified with the
voter by provisional ballot ID. Second, it may be possible to
determine how the voter voted and use this information in determining
whether a provisional ballot should be counted. Because the process
of determining whether a provisional ballot is to be counted has
little or no public oversight, the potential exists to systematically
disenfranchise voters who vote a particular way. Of course, even if
the ballot is not available, the voter's party affiliation can be
used to disenfranchise. Furthermore, certain ethnic groups who use
multiple last names or who write their family name first and given
name last can be systematically disenfranchised by failing to check
for these variations when trying to match a voter to the voter rolls.
This problem also arises with voter registration databases that are
to be checked against social security records or drivers license
records. While the problem of voter registration database
verification-caused disenfranchisement has been dealt with by the
courts, I am not familiar with court adjudication of this issue in
provisional vote counting. Because the result of whether provisional
ballots are counted is communicated only with voters who inquire and
only in jurisdictions providing this service, it is not clear whether
the information is available to enable a court case to be filed and
adjudicated. That such ethnic groups are less likely to be native
speakers of English can compound the problem.

For the November 2007 election, the usual pink paper form was used.
If the voter cast a paper ballot, the paper ballot was to be placed
in an envelope along with the pink paper form. A code using the
voters birthday was to be used for the voter to be able to find out
whether the ballot was counted. In a private meeting with Assistant
Registrar of Voters Elaine Larson several months ago, I suggested
that envelopes be printed with the pink form and serial numbers
pre-printed on both the envelope and a tear-off stub. The voter
would be handed either the provisional ballot envelope or a ballot
secrecy sleeve but not both. A ballot returned in the ballot secrecy
sleeve was cast in the white ballot box. A ballot returned in the
provisional ballot envelope was sealed by the voter in the envelope,
the stub returned to the voter, and the envelope cast in the brown
ballot box. If a provisional voter was to vote electronically, the
activator card provisional ID number was written on the envelope,
with the stub handed to the voter with the activation card and the
provisional envelope placed in the brown ballot box without a paper
ballot. I made sure all provisional voters used paper ballots, so
their envelopes were not opened until it was determined that their
ballots were to be counted. Provisional ballots are separated from
their envelopes and counted in batches by high speed scanners, just
like vote-by-mail ballots. I talked to poll workers and election
trainers, all of whom were happy about the system with the pink
provisional envelopes. One quibble is that the portion of the flap
that stayed with the envelope was not long enough to seal well, and
voters did not like licking the envelopes. Postal workers at the
post office service counter have a solution to this
problem---moistening sponges---that should be considered to address
this problem.

There were 340 ballots cast in my precinct, of which 78 were cast
electronically. There were also well over 100 vote-by-mail ballots
dropped off at the polls. The bag in which vote-by-mail envelopes
were to be placed did not seal well. And this bag was so full, it
would not fit into inspectors bag. I recommend that better sealing
vote-by-mail ballot bags be used in the future and that multiple such
vote-by-mail ballot bags be provided in case the number of ballots
did not fit in the bag. (I had that problem previously.)

With only one electronic voting machine in each precinct, the process
now follows California Election Code Section 19370. After the vote
tallies are printed on the sealed paper trail, the paper trail box is
removed and another paper trail box is installed. Two more copies of
the vote tally are printed and the paper trail box is removed. Some
of the paper tape is removed from the box, and the two copies are
separated. Both copies are signed by all the Precinct Board members
(i.e., poll workers), with one copy posted outside the precinct and
the other copy placed in the roster index.

Voters were encouraged to mail their ballots in early to ensure that
they were properly received and counted, and also that their ballots
were included in the totals announced as the polls closed rather than
days later. However, with the compressed primary cycle, and one
candidate withdrawing a week before the election, I suspect many
voters will wait until the last minute in the future. Not only will
this mean more work for Registrar of Voters staff after election day
dealing with the large amounts of paper ballots submitted at the
polls, but it also will mean more ballots dropped off at the polls
(hence the need for more bags).

Having enough qualified poll workers is an issue. My two best poll
workers were one who has worked with me several times before and one
high school student. The one who worked with me before was
experienced, knowledgeable, worked fast, and followed my directions.
The high school student was a quick study, worked fast, and also
followed my instructions. I had one more poll worker who worked all
day and a pair of poll workers who split the day. Those three poll
workers were slow and did not always follow instructions. For
example, there were 325 voters on the tally sheet (in part because
one poll worker was confused whether provisional ballots are to be
marked on the tally sheet---they are, but vote-by-mail ballots are
not!), 340 ballots cast, and 343 voters signed in on the roster
index. I suspect the problem with the extra 3 voters on the roster
index is that 3 provisional voters were told to sign the roster index
where they appeared as well as on the provisional ballot pages.
These were voters who had been sent vote-by-mail envelopes but could
not surrender them at the polls. I recommend that the instructions
with the vote-by-mail envelopes should indicate that they may be
mailed in, dropped off at the Registrar of Voters, or dropped off at
the polls, and that they may be surrendered at the polls on Election
Day if the vote wishes to vote in the precinct (for example, if the
voter has spoiled the ballot). Doing so would increase the number of
surrendered ballots with a corresponding decrease in the number of
provisional ballots. Much of this problem had to do with
Decline-to-State voters or with voters who had chosen a candidate no
longer actively running for office.

One of the poll workers parsimoniously kept removing pens from the
Pollstar voting booths. I figured it was easier to leave them there,
since we had more than enough. Few pens if any were taken from the
polling place. Although I requested poll workers leave a few pens at
the polling place, I noticed there supply of pens on the table would
periodically increase, with all the Pollstars bereft of pens. This
particular poll worker hastily carried materials out, when I wanted
to take them to my car systematically. He dropped some of the bags
because he was unable to carry them all at once, and the bulging
vote-by-mail bag seal failed and spilled out a few ballots. I
resealed the bag and warned election officials to the problem at the
drop off point for election materials at a local middle school.

Another poll worker (the one who kept on getting confused whether
provisional ballots are marked on the tally sheet that counts the
number of voters) was rather slow. For the rush of voters in the
last two hours, I moved him off the roster index and replaced him
with the "good" experienced poll worker. I moved the other poll
worker to marking the tally sheet (the street indexes aren't marked
after 6pm).

One other poll worker arrived at around 9am (as we agreed) and left
not long afterwards. This poll worker is a Spanish-speaking Stanford
student who specifically volunteered to work as a bilingual poll
worker. She left because there were few if any voters who needed her
services, and the high school student also knew Spanish. I
encouraged her to volunteer earlier next time and to specifically
request being placed at a precinct with a Spanish speaking population
that is located convenient to public transit. The transportation
issue is why the voter arrived at around 9am.

In some earlier elections, we were told to record the tally number in
the roster index or street index. However, starting with the
November 2007 election, this practice that risked voter privacy was
stopped. There are 3 street indexes, one from 7am to 11am, one from
11am to 3pm, and one from 3pm to 6pm. Since voters are crossed off
when they vote only for the street index when the vote, one can tell
the biock of time when a voter voted. Since this does not include
sequence information, it does not present the same risk to voter
privacy.

I recommend that the Non-Partisan sample ballot include the option
whereby the voter can request a partisan ballot, as part of the
regular form for requesting a mail ballot. Along with the
Non-Partisan vote-by-mail ballot should be instructions for how to
exchange the ballot for a partisan ballot either by mail or at the
precinct. Similarly instructions should be provided for how voters
should deal with spoiled ballots, such as when a voter's desired
candidate withdraws. Considering that vote-by-mail turnout in Santa
Clara County, California exceeded precinct voting by nearly 50%, the
problems of Decline-to-State choosing partisan ballots and candidates
dropping out are likely to recur in the future. In contrast, three
times as many non-partisan vote-by-mail ballots were cast than
precinct ballots. Given that turnout was about 60% vote-by-mail and
40% precinct. It is likely that as many non-partisan voters chose a
partisan ballot in the precinct than voted non-partisan. My
guesstimate is that at least 6500 Decline-to-State voters chose to
cast a Democratic ballot at the polls. This number contrasts with
nearly 353,000 total ballots cast and just over 220,000 Democratic
ballots cast (including Decline-to-State). I do not know how many
Decline-to-State voters received and cast vote-by-mail Democratic
ballots.

It may be helpful give poll workers instructions on how to count the
number of paper ballots and the number of envelopes. While this may
seem as a trivial task, I find it best to have poll workers create
group of 10 at a time. This process can be done by multiple poll
workers in parallel. The paper ballots are placed crosswise, with
each group of 10 perpendicular to the next. All of the piles of
paper ballots are combined into one pile, with an extra short group
of 0 to 9 ballots kept separate. Then the groups of 10 are counted
with the count of the short group added to the overall count.
Counting them without making groups is a more error prone process
that can't easily be restarted if someone gets confused or
interrupted in the process. This process is well known to those who
hand-tally ballots, but out of practice when in this era of
electronic voting machines.

This election had the first widespread use of optical scan paper
ballots in the polling place in Santa Clara County. (Fewer voters
used them in the November 2007 election when they were given only to
those who requested them.) Because these ballots were virtually the
same as vote-by-mail ballots (although it is possible to tell them
apart other than by the creases), there was some familiarity with
them, and the Registrar of Voters already had the high speed scanners
for counting them. However, some voters were confused. To make a
selection, the voter was instructed to complete the line (connect the
two line segments) with the arrow pointing to the voter's selection.
One voter showed his ballot to a poll worker to ask if he had done it
right. He had circled the selections instead, so we took back his
ballot and marked it spoiled and give him another one after
instructing him how to use it. At that point, I had the poll worker
who handed the voter the ballot demonstrate how to mark the ballot
(fortunately, we were given a pad of sample lines to connect). We
added that to the instructions to mark both sides of the a ballot,
remove the stub, and return the ballot to us in the secrecy sleeve or
provisional ballot envelope. I think this step should be added to
poll worker training.

Clearly this problem was voter error. I do not think the problem was
ballot design. But it does confirm the quip that the problem of
making systems idiot-proof is that they keep on making more clever
idiots. This experience confirms the need for ballots to be checked
at the precinct when the voter is still able to correct them. A
precinct-tally optical scan system provides feedback for overvotes.
However, unless the precinct-tally optical scanner provides feedback
to the voter for all the selections (either visually or by auditory
feedback), the voter can't be sure that scanner has interpreted the
ballot correctly. I have previously written about the benefits of
such a system to providing true voter-verification using a consistent
process for all precinct voters (including the reading impaired), and
this case confirms the benefits of such a system.

The electronic voting machines in Santa Clara County as Sequoia AVC
Edge Direct Recording Electronic (DRE) voting machines. If these
were converted to print paper ballots (that is, to be Electronic
Ballot Printers), then the Decline-to-State voter who was given a
card activator for a Non-Partisan primary ballot could have
complained that the printed paper ballot was wrong, had it marked
spoiled, and then given a card activator for the party ballot of the
voter's preference. This case demonstrates one benefit of the
Electronic Ballot Printer architecture over the Direct Recording
Electronic voting system architecture. I believe that none of the
paper trail systems are accessible based on the directive of
California Secretary of State Kevin Shelley as later enacted in state
voting law. (See California Election Code Sections 19250 and 19251.)
The auditory report of the ballot is supposed to come from either the
paper trail itself or a by-product of printing the paper trail.
("'Accessible' means that the information provided on the paper
record copy from the voter verified paper audit trail mechanism is
provided or conveyed to voters via both a visual and a nonvisual
method, such as through an audio component." -- California Election
Code Section 19251(a).) With an Electronic Ballot Printer coupled
with a Precinct-Tally Optical Scanner with voter feedback, true
accessible voter verified paper ballots are possible. However, this
approach does require that the voter handle the paper ballot, which
some accessibility advocates criticize. I believe that a consistent
manner in casting ballots and providing voter verification to all
voters provides a greater benefit.

Another benefit of Electronic Ballot Printers is that they are
supplied with "blank" ballot stock that is customized to the specific
ballot style by the voter activation card. The "blank" ballot stock
has numbered stubs for inventory control as well as identification of
the date of the election and county. However, the ballot style
details, such as lists of candidates, are printed along with the
selections made to the voter, such that a ballot printed by an
Electronic Ballot Printer is functionally identical to a hand-marked
paper ballot. One key benefit of an Electronic Ballot Printer is
that they could be used to create ballots of specific styles, but
without selections marked, for marking by hand by voters. Electronic
Ballot Printers could be used to print ballots for particular parties
or languages when those run out as well. The Registrar of Voters
could safely reduce the number of pre-printed ballots and supply more
"blank" ballot stock for the Electronic Ballot Printer. It would be
far easier for the Registrar of Voters to have a supply of "blank"
ballot stock for distribution to precincts if turnout was higher than
expected. Electronic Ballot Printers are also useful for early
voting, when a single polling location has multiple voting
jurisdictions and many potential ballot styles.

For ease in dropping off vote-by-mail ballots, ballot collection
boxes should be provided in the City Hall for each city or town in
the county. These could be serviced daily during the week before the
election, and perhaps less often previously.

I recommend that vote-by-mail ballot also list the "home" precinct
location, which is useful when the voter wishes to surrender the
ballot to cast a precinct ballot.

The process of administering elections in quite complex. It is
important not only to get the voting machinery to be reliable,
accurate and safe, but also to get the processes and procedures
right. Systematic study of these processes and procedures is
warranted with the goal of sharing best practices. This is not to
say that one-size-fits-all, that is, that every jurisdiction should
follow the same practices. If nothing else, whether a jurisdiction
is urban or rural or in between, and its consequences on precinct
size, will affect what the best practices should be. Large
jurisdictions will be able to afford larger dedicated elections
staff, while smaller jurisdictions may need to outsource more to
vendors or use staff usually dedicated to other government functions.

Our system of democracy is plagued by comparatively low voter
participation. I believe this is also reflected in the shortage of
qualified poll workers and the need to continue to use poll workers
who are less effective. I can only imagine how our government would
be different if voter turnout were in the 90 percent range.
Considering that the high school student who worked in my precinct
was nearly as effective as my best experienced poll worker, we would
do well to encourage high school seniors to work in the polls. It
would do well also to encourage young people to vote. I daresay that
the student learned valuable lessons, not only about the nuts and
bolts of democracy, but also about how people function together as a
team and how to deal with problems and systematically improve to
resolve them. While that learning experience will differ from
precinct to precinct and from Precinct Inspector to Precinct
Inspector, I do think that the learning does justify taking a day off
from school. As Mark Twain said (paraphrasing), "Never let your
schooling interfere with your education." Unfortunately, California
state education code discourages high schools from having their
students work in the polls by penalizing them in the ADA (Average
Daily Attendance) money for the day. A bill to remove this penalty
was vetoed by the governor. Perhaps the culture in those
jurisdictions that hand-count paper ballots at the close of polls
after Election Day is different than the culture in jurisdictions
with mechanized counting. With the shortage of qualified poll
workers, I'm skeptical about the finding enough people to tally
ballots.

It is an honor to serve as a Precinct Inspector, as well as a
learning experience. I recommend that all voting system researchers
and election integrity activists do it at least once. Jan L. A. van
de Snepscheut reportedly said, "In theory, there is no difference
between theory and practice. But, in practice, there is." This
concept clearly applies to elections.

I certainly don't work 15 hours or more on Election Day for the
money, or for the lack of sleep the night before or after.
Increasing funding for poll workers increases their supply, but not
their quality. The Declaration of Independence states, "That to
secure these rights, Governments are instituted among Men, deriving
their just powers from the consent of the governed." Thomas
Jefferson said, "The price of freedom is eternal vigilance." It is
time to retire that tired statement: "Don't vote -- It only
encourages them [politicians]." Actually not voting encourages them
even more. [Adapted from an article by Lee Strong in the WSFA
(Washington Science Fiction Association) Journal; hopefully increased
voter turnout isn't just science fiction.]

Please attribute or cite if you quote from this article. This
article was publicly posted on the web at
http://gnosis.python-hosting.com/voting-project/February.2008/0064.html

Arthur Keller
Precinct Inspector, Santa Clara County Precinct 2113 on February 5. 2008
Researcher, Technology Information Management program,
Baskin School of Engineering, University of California, Santa Cruz

--
----------------------------------------------------------------------------
---
Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA  94303-4507
tel +1(650)424-0202, fax +1(650)424-0424
_______________________________________________
OVC-discuss mailing list
OVC-discuss@listman.sonic.net
http://lists.sonic.net/mailman/listinfo/ovc-discuss
By sending email to the OVC-discuss  list, you thereby agree to release the
content of your posts to the Public Domain--with the exception of
copyrighted material quoted according to fair use, including publicly
archiving at  http://gnosis.python-hosting.com/voting-project/
 

_______________________________________________
OVC-discuss mailing list
OVC-discuss@listman.sonic.net
http://lists.sonic.net/mailman/listinfo/ovc-discuss
By sending email to the OVC-discuss list, you thereby agree to release the content of your posts to the Public Domain--with the exception of copyrighted material quoted according to fair use, including publicly archiving at http://gnosis.python-hosting.com/voting-project/

==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Fri Feb 29 23:17:05 2008

This archive was generated by hypermail 2.1.8 : Fri Feb 29 2008 - 23:17:08 CST