[EL] Palm Beach County (Bush v Gore issues in remedy far worse than ballot design problem)

[Paul Lehto]

Here's what's interesting:* In response to a very small problem everyone
seems to agree will not confuse voters, the choice by election officials of
a response or remedy ("to duplicate ballots") raises numerous much more
serious issues including Bush v. Gore issues,* a few of which I detail
below in bullet points. These problems appear to me to be unavoidable and
fatal wherever ballot duplication is chosen as the remedy, as it is here.

Because election officials know which ballots are affected by this error
(and they intend to "duplicate" those by hand, whatever that means), and
because voting machines are programmed to scan/look in certain ballot
sectors and interpret markings in those sectors are Yes or No votes, it is
clearly not "impossible" as election officials claim in the article below,
to program a machine to count correctly.  But they say it is, and thus
propose their remedy of duplicating these 27,000 ballots.

It is, however, probably impossible or at least very complicated to count
ALL ballots using a *single* ballot interpretation scheme, since *votes on
judicial retention and races down ballot from judicial retention will
appear in different sectors* on the 27,000 affected ballots versus the rest
of the ballots. If a scanner "looks" for a vote in the wrong sector of the
ballot, or otherwise fails to "see" the voter's mark in the correct sector
for a variety of technical reasons, no vote is registered even though the
human eye would clearly see and count a vote in many of these cases.

What's very unclear, and fascinating to me, is why the election officials
have chosen "duplication" by hand as the best method.  Clearly it doesn't
mean having affected voters re-vote and thereby duplicate their votes.
Instead, it definitely appears to mean that *election officials will take
the 27,000 affected ballots and transfer (and thereby re-vote) the ballots
onto properly printed ballots*, and then run those 27,000 votes through the
machines along with all the other ballots and have their votes counted or
handled by a single, straightforward ballot program scanning function. But
this has some interesting implications:

1.  It is significantly more work to duplicate and revote the ballots of
27,000 voters than it is to simply hand count such a relatively small
number of ballots.  This is so especially since only those down-ballot
races past the tax collector race are affected because the yes and no votes
would be in differing sectors for the two different classes of printed
ballots.   It would be like a hand recount, limited to 27,000 ballots, but
only as to the down ballot races affected by the design shift, and
relatively easy to do.  One might guess that officials are animated by
legal advice and/or an understanding of *Bush v. Gore* and concluding that
such a sensible and efficient solution would violate the 'equal dignity'
ideas and anti-partial recount ideas of Bush v. Gore and possibly HAVA,
requiring all ballots to be treated the same, at least for counting
purposes. *(Irony: 27,000 ballots are definitely being treated very much
not the same, for non-counting purposes described herein)*

2.  Given it appears discussion on this list that nobody or hardly anybody
would be confused by the lack of the header being printed, I sure hope no
printer or election official decided to inject this problem into the
election by spotting the omission and then changing the ballot after 27,000
were already printed and/or mailed, but something like this may have
happened. Changing ballot design after finalization and after printing
starts raises a lot of serious issues as we can see and should never be
done lightly or with less then thorough control of defective ballots.

3.  Palm Beach election officials will be the clearest example ever of
repeat voting - they are voting 27,000 times.  Now, they say that when they
duplicate and thus revote these 27,000 ballots they will be 100% error free
in their human frailty, unlike those election officials in Florida 2000.
No, wait, *Palm Beach is not claiming they are perfect interpreters of
voter intent on defective ballots, and has not so much as promised to be
faithful to the voters' intent when they read the defective ballots and
transfer the votes by hand onto proper ballots, and if Bush v. Gore is
taken fully seriously, then serious issues are raised of the interpretation
of voter intent by these election officials as they decide what the voter
meant and then copy that intent by hand onto a new ballot*.  Perhaps all
the extreme fidelity we assume and expect to voters' intent, together with
the agency to re-vote somebody else's ballot are all inherent in the
election officials' use of the word "duplicate?" :)

4.  Given the amount of time it will take to vote 27,000 times, surely this
process will have to start days or even weeks prior to Election day.  This
will give those persons involved a very good feel for the election trends
and results long before election day.  I'm sure some politicians and some
media might be willing to at least buy lunch if not some more valuable
consideration for election officials in order to get the intelligence on
how those ballots are trending, especially for a state senate race where
this number of ballots would be quite material.

5.  If, after a few drinks at lunch, it became clear to an election worker
how valuable it would be if human frailty in duplicating ballots were to
result in a few hundred vote "typographical error" in the ballot transfer
process, it is doubtful such "typos" would be detected by anyone.  Because
they are transferring the ENTIRE ballot to a new ballot, this opportunity
exists for every race on the ballot this November.   Most likely, a worker
wouldn't have a lunch or a bribe, but would simply have a personal
preference in a race or two, perhaps the state senator race.  In addition,
in the incredibly hectic rush sure to exist on and before Election day as
ballots are duplicated, it would simply save a lot of time and avoid
possible supervisor anger for slow progress for workers to more quickly
transfer votes from ballot to ballot by making assumptions or not really
being careful about what the original voter voted and intended.
Administrative efficiency concerns can thereby disfranchise voters.

6. Will officials faithfully duplicate all overvotes and undervotes?  Or
will they correct some "errors"?  How does one possibly mimic the light
pencil or pen marks of a voter when such light marks and their location can
and will effect whether that vote even counts at all? (machines under some
circumstances will not register or pick up light marks or incomplete
arrows.)

Here, *even if we agree that in the vast majority of instances election
officials will properly interpret voter intent, and even if we stipulate
100% honesty and fidelity in transferring that human-eye assessment of
intent to a proper ballot, the human-eye assessment of intent is NOT THE
PROPER STANDARD of duplication, because what really matters is not what a
human reader thinks the vote is, but what an optical scanning machine
thinks the vote is.  Thus,all kinds of things that are hard to impossible
for election officials to truly "duplicate" must be faithfully duplicated:
The thickness of a line, color of ink or pencil, darkness, specific
location, presence of extraneous marks or ballot damage, and many other
things, ALL OF WHICH affect whether the scanner will count something as a
vote or not.*

Even an "objective" photocopier (instead of human interpretation) used on
the defective portion of the ballot and  followed by pasting the races into
a proper location in order to create a scanner-readable ballot countable
along with all other ballots falls far short of avoiding these kinds of
issues - the issues just shift a little: For example, copiers will lighten
or darken, the position of the pasted portion on the ballot is absolutely
critical to proper reading, copies will copy blue ink in black or grey
tones, and blue ink is not "seeable" by many optical scanners that use blue
light to scan ballots, rendering blue-ink undervotes into valid votes by
photocopying (which does "see" blue inks), the paste of the photocopy to
create a proper ballot overall would increase the thickness of the ballot
and perhaps jam the scanner or gum it up, and so on.

*The remedy of duplication of ballots here is very much worse than the
problem. * It's such a forest of issue-spotting opportunities that perhaps
many will just close their eyes and hope to skate over this thin to
nonexistent ice without falling through.   Some will be happy to look the
other way on things like this because it apparently is worth almost
anything in order to avoid what is claimed as the horror of horrors:
"Florida 2000 all over again."

Under this mistaken view, we get a computerized voting system more
defective and laughable than the punchcard/chad system of 2000, yet we
still pretend like we're so much better off now with HAVA and all.  Such
"positive thinking" helps avoid all the angst, debate, litigation, and also
all the accountability for mistakes too.  All we need to do is act like
Pangloss or like Pollyanna.  And if that doesn't work, maybe the judicial
doctrines of standing or laches can save us from being concerned or upset
about any of these things! ;)

Paul Lehto, J.D.

On Sun, Oct 21, 2012 at 5:29 PM, Steve Kolbert <steve.kolbert at gmail.com>wrote:

> For those interested, I've attached a scanned image of an unmarked
> absentee ballot from Palm Beach County that contains the printing error.
> (It comes in two files because this scanner could not scan the entire 9.75"
> x 18" ballot page all at once.)
>
> As you can see, the ballot doesn't really look misprinted at all -- it's
> very hard to tell there's an error even when you know to look for it, which
> may be how Palm Beach County officials never noticed the misprint until it
> was too late. (If you still can't find it, look in the second column for
> the race for Tax Collector. Under that race, there are judicial retention
> races -- but there is no heading that says "JUDICIAL RETENTION /MONTENER AL
> JUEZ," as there are headings for every other race.)
>
> This is an unfortunate error by a vendor, but it's important to note that
> this error probably won't cause any voter confusion, even if Palm Beach
> County's plan to tabulate the misprinted ballots will cause headaches for
> the election administration staff. In other words, this is not another
> butterfly ballot.
>
> Steve Kolbert
> (202) 422-2588
> steve.kolbert at gmail.com
> @Pronounce_the_T
>
> On Oct 21, 2012 12:36 PM, "Rick Hasen" <rhasen at law.uci.edu> wrote:
> >
>
> > Holy Cow! They’ve Already Screwed Up the Ballots in Palm Beach County
> > Posted on October 19, 2012 9:14 pm by Rick Hasen
> >
> > A major theme of The Voting Wars is that we learned the wrong lessons
> after the Florida 2000 debacle.
> >
> > Here’s some good evidence at ground zero of the 2000 battles: Palm Beach
> County:
> >>
> >> The problem started when an Arizona company that printed the ballots
> failed to include a heading over the merit retention elections for judges
> on the Florida Supreme Court and 4th District Court of Appeal. The mistake
> was discovered and corrected after the first 60,000 ballots were printed
> and 50,000 mailed.
> >>
> >> It was soon discovered that the error would affect all races on the
> flawed ballot. When the header was inserted, the races on about half of the
> ballots shifted, Bucher said. The shift will make it impossible for
> tabulation machines to count the votes on an estimated 27,000 of the bad
> ballots. Saying it would be impossible to program machines to read the
> defective ballots, Bucher said the only alternative is to duplicate them by
> hand.
> >>
> >> Late Friday, she sent Detzner a report, outlining how the process will
> unfold Monday when workers begin opening, sorting and copying the roughly
> 15,000 absentee ballots that have been returned. Of the total, she said
> about 8,600 are flawed.
> >
> >
> >
> > Posted in election administration, The Voting Wars | Comments Off
>
> _______________________________________________
> Law-election mailing list
> Law-election at department-lists.uci.edu
> http://department-lists.uci.edu/mailman/listinfo/law-election
>



-- 
Paul R Lehto, J.D.
P.O. Box 1
Ishpeming, MI  49849
lehto.paul at gmail.com
906-204-4965 (cell)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20121022/861ecb44/attachment.html>