Re: I need some help

From: Teresa Hommel <tahommel_at_earthlink_dot_net>
Date: Tue Apr 26 2005 - 01:04:55 CDT

3. I need someone who is well versed in OVC's system to read section 2.4
of Shamos's paper
(http://euro.ecom.cmu.edu/people/faculty/mshamos/paper.htm#_edn1 ) and
to write a technically solid rebuttal. Reward: naming as a co-author.

Ron, I implore you not to make the same mistake that Shamos makes. He
says in section 1.1 that he sees no "engineering difference" between
airplanes and voting machines. That's why he ends up making such
outlandish statements and arguments -- he views elections as an
engineering problem. But elections are not an engineering problem, they
are a democracy problem. If you ignore the needs of a legitimate
democratic election, of course you may write something technically solid
but it is likely to be irrelevant to the needs of our democracy. I
actually gave a whole speech on this topic.
http://www.wheresthepaper.org/RadcliffeFeb12_2005_TeresaHommel.htm

Below are my responses to section 2.4 of Shamos' paper.

Teresa Hommel

2.4. The "Voter-Verified" Paper Trail

            It is alleged that adding a so-called "voter-verified paper
trail" to a DRE machine will either permit tampering to be detected or
at the very least will provide a reliable record of how each voter voted
that can be used for a recount, even if the recount must be conducted by
hand. This is incorrect. A paper trail accomplishes one thing, and one
thing only - it provides assurance to the voter that her vote was
initially captured correctly by the machine.

FALSE. IT RECORDS THE VOTER'S CHOICES ON A "PERMANENT" NON-ELECTRONIC
MATERIAL WHICH CAN BE USED TO DETERMINE WHETHER OR NOT THE COMPUTER
RECORD OF THE BALLOT WAS CORRECT. SHAMOS IS FALSELY NAIVE HERE, BECAUSE
AS A COMPUTER SCIENTIST HE KNOWS THAT THE SCREEN CAN SHOW ONE THING, THE
INTERNAL RECORD CAN SHOW ANOTHER, AND THE PAPER PRINTOUT CAN BE THE SAME
OR DIFFERENT FROM THE SCREEN OR INTERNAL RECORD.

 This is no small accomplishment, but it can be achieved in numerous
other ways, as explained below. That is the only voter-verified part.
The paper trail provides no assurance at all that her vote will ever be
counted or will be counted correctly. The reason simply is that the
paper trail itself becomes insecure at the moment of its creation.

TRUE, AND ALSO THE INTERNAL ELECTRONIC RECORD BECOMES INSECURE AT THE
MOMENT OF ITS CREATION, REGARDLESS WHETHER IT WAS ACCURATE OR INACURATE
TO BEGIN WITH. HOWEVER, THE PAPER RECORD CAN BE WATCHED BY NONTECHNICAL
ORDINARY CITIZEN ELECTION OBSERVERS, AND THE ELECTRONIC RECORD OF THE
BALLOT CAN NOT BE OBSERVED.

            First, if the machine cannot be trusted, which is the
working hypothesis of paper trail proponents, then it cannot be trusted
to deal with the paper trail safely. After the voter leaves the voting
booth, it can mark her ballot as void and print a different one. The
voter will have left the booth believing not only that her vote was cast
and counted properly, but that it will also be counted properly in any
recount. None of these beliefs is correct.

TRUE, AND THIS IS ONE REASON WHY A POPULEX-TYPE DESIGN IS DESIRABLE,
WHERE THE VOTER TAKES THE PAPER BALLOT IN HAND AND DELIVERS IT TO THE
COUNTING STATION (OR OPTICAL SCANNER).

            One might argue that inspection and testing of the machine
would reveal such abjectly bad behavior, but the claim of DRE opponents
is that no amount of inspection and testing is ever sufficient. If
testing is adequate to reveal paper trail flaws, then it is adequate to
uncover other faults in the machines.

TRUE, INSPECTION AND TESTING OF AN EVOTE COMPUTER MEAN NOTHING BECAUSE
THE COMPUTER IS MANAGED BY NONTECHNICAL ELECTION WORKERS WHO WOULDN'T
NOTICE IF IT WAS TAMPERED WITH, AND ARE CERTAINLY INCAPABLE OF
PERFORMING SOPHISTOCATED ROUTINES TO VERIFY WHAT SOFTWARE IS IN THE
COMPUTER.

            Here is a further, but only partial, catalog of problems
with paper trails.

 1. The paper trail cannot be on a continuous roll of paper, since that
would permit reconstruction of each voter's ballot based on the order in
which votes were cast. Therefore, the paper trail must consist of
separate pieces of paper. However, once the pieces of paper are
separated, the integrity of the trail is lost. Looking at a piece of
paper, we will not be able to tell for certain where it came from.
Stuffing and all other paper ballot tampering methods then become
possible. The addition of cryptographic indicia, which has been
proposed as a method to prevent insertion of unauthorized ballots,
cannot work since the voter will never know whether her real ballot
contained the proper indicia when it was created. If it didn't, the
ballot will not be tabulated during a recount.

PARTLY TRUE, NO MACHINE CAN PREVENT FRAUD BY INSIDERS AND COLUSION OF
INSIDERS AND VOTERS. ONLY MULTIPARTISAN OBSERVERS WHO ARE NOT PART OF
THE COLLUSION CAN DO THAT BY OBSERVING THE BALLOT BOX AND THE BALLOTS.

2. Adding a paper printing device to a DRE machine naturally adds
another component that can fail, run out of ink, jam or run out of
paper. If DREs are alleged already to be prone to failure, adding a
paper trail cannot improve that record. In Brazil in 2003, where a
small number of precincts had installed paper trails, failure of the
printers delayed voters by as much as 12 hours, a figure that would be
catastrophic in the U.S.[32]

SHAMOS IMPLIES THAT AMERICANS CAN SEND A SPACE SHIP TO THE MOON, BUT WE
CANNOT BUILD A PRINTER THAT CAN PRINT 200 SHEETS OF PAPER IN 12 HOURS.
SO BUY A JAPANESE PRINTER.

SERIOUSLY, THIS IS WHY MANY PEOPLE ADVOCATE HAND-MARKED PAPER BALLOTS TO
BE COUNTED BY A PRECINCT-BASED OPTICAL SCANNER. IF THE OPTICAL SCANNER
BREAKS, VOTERS CAN STILL MARK THEIR BALLOTS. MY PENCIL MIGHT BREAK, BUT
I COULD BORROW ANOTHER ONE AND IT WON'T COST $3000 TO $8000 LIKE A DRE
WOULD.

3. There is no voter-verified paper trail machine that has been tested
on any large scale.

TRUE, AND THERE IS NO DRE THAT HAS EVER BEEN AUDITED AFTER AN ELECTION.
AMERICAN ELECTIONS ARE A SORRY BUSINESS.

4. States that propose to implement the paper trail have promulgated
regulations stating that the paper shall govern over the electronic
record in the event of discrepancy[33]. This has the effect of making
the insecure paper record paramount over the secure electronic one, a
return to the early days of the Australian ballot.

REPETION IS NOT AN ARGUMENT. NO ELECTRONIC RECORD OF ANY ELECTION HAS
EVER BEEN AUDITED, SO WE DON'T KNOW IF ANY ELECTRONIC ELECTION HAS BEEN
CONDUCTED HONESTLY AND WITHOUT ERRORS. HOWEVER WITH MULTIPARTISAN
OBSERVATION OF PAPER BALLOTS, WE CAN BE SURE THAT EITHER ALL OBSERVERS
COLLUDED WITH FRAUD, OR NO FRAUD OCCURRED.

5. With complex ballots, voters are prone to forget exactly whom they
have voted for. When confronted with a paper record, they may
erroneously claim that the machine made a mistake. This will call the
machine's reliability into question, prompt calls for a recount and cast
doubt even on machines that are functioning properly.

THIS IS ONE REASON WHY MANY PEOPLE ADVOCATE HAND-MARKED PAPER BALLOTS TO
BE COUNTED BY A PRECINCT-BASED OPTICAL SCANNER.

6. Paper trails do not address the problem of DRE failures. If the
complaint is that a machine cannot be initialized for use on the morning
of election day, then having a paper trail mechanism is of no help. In
fact, the presence of the mechanism increases the load on the machine's
power supply and processor and itself increases the probability of failure.

PAPER TRAILS ARE NOT A PANACEA, THEY PROVIDE AN INDEPENDENT MEANS FOR
AUDITING THE COMPUTER. ALL VOTING SYSTEMS SHOULD BE ACCOMPANIED BY
SUFFICIENT PAPER BALLOTS SO THAT VOTERS CAN VOTE IF THE EQUIPMENT
BREAKS. FOR EXAMPLE, IN NEW YORK EVERY ELECTION DISTRICT HAS 300 PAPER
BALLOTS PACKACED IN THE BACK POCKET OF ONE OF ITS MECHANICAL LEVER MACHINES.

7. The paper trail requires a re-examination of meaning of the terms
"ballot" and "official ballot." This is not a mere semantic exercise,
but a question of great legal and, in some states, constitutional
significance. Can a piece of paper be a ballot if it is neither marked
nor touched by the voter? If so, significant statutory changes will be
required. If the paper is the ballot, then what conceivable meaning can
be ascribed to the computer count, which is not derived by counting the
"ballots," but by processing the voters' original inputs that were
separately used to generate the ballots? If the paper ballots are
official, then we are put in the untenable position of having to certify
an election without ever actually counting the ballots, unless an
allegation of irregularity compels a "recount."

ELECTRONIC VOTING HAS BEEN THE CAUSE AND RESULT OF MANY CHANGES IN OUR
LAWS. CHANGING TECHNOLOGY OFTEN CHANGES LAWS. FOR EXAMPLE, BEFORE WE HAD
CARS, WE DIDN'T HAVE SPEED LIMITS ON OUR HIGHWAYS.

CAN AN ELECTRONIC RECORD THAT IS NEITHER MARKED NOR TOUCHED BY THE VOTER
BE A BALLOT? IT CAN IF THE LAW SAYS IT CAN, AND THE CHANGES TO LAWS OR
STATE CONSTITUTIONS ARE UNLIKELY TO BE "GREAT" OR "SIGNIFICANT."

HOWEVER THE EFFECT OF INVISIBLE ELECTRONIC BALLOTS THAT CANNOT BE
OBSERVED BY ORDINARY MULTIPARTISAN CITIZEN OBSERVERS IS GREAT AND
SIGNIFICANT -- THE EFFECT IS TO UNDERMINE THE LEGITIMACY OF OUR
ELECTIONS AND OUR GOVERNMENT. THIS IS BECAUSE UNOBSERVED VOTE-RECORDING
AND VOTE-COUNTING, AND REQUIRING VOTERS AND THE PEOPLE TO "TRUST"
UNOBSERVED THESE PROCESSES IS ANTITHETICAL TO ELECTION LEGITIMACY.AND TO
DEMOCRACY ITSELF.

8. Each losing candidate will claim that the election was stolen from
him by the machine and will insist that the only true indication of the
voters' preferences reside on the paper, even if there is no evidence of
irregularity or tampering. Thus paper recount will become the default
method of vote counting, mitigated only by the high cost of such
recounts. If this is to be the case, why use voting machines in the
first place?

THERE IS NO GOOD REASON TO USE ELECTRONIC VOTING.

9. Paper trails cannot readily be viewed by disabled voters, requiring
them yet again to reveal their votes to strangers in order to have them
verified. It is no answer to say that there are other mechanisms to
review their votes. If paper trail proponents truly believe the paper
trail is necessary for fair elections, then elections will not be fair
for the disabled.

THE PURPOSE OF AN ELECTION IS TO DETERMINE THE WILL OF THE PEOPLE IN
SELECTING OUR PUBLIC SERVANTS. AN ELECTION IS FAIR IF ALL PERSONS
LEGALLY ENTITLED TO VOTE ARE IN FACT ABLE TO CAST A BALLOT WITHOUT UNDUE
BARRIERS, AND THEIR VOTES ARE RECORDED AND COUNTED ACCURATELY.

HAVING A DISABILITY IS NOT FAIR. PERSONS WITH DISABILITIES THAT FORCE
THEM TO DEPEND ON OTHERS SUFFER NOT ONLY FROM THEIR DISABILITY BUT FROM
OUR CULTURE'S DISRESPECT FOR THOSE WHO ARE NOT "INDEPENDENT," AND THIS
IS NOT FAIR.

HOWEVER, WHEN A PERSON REQUIRES ASSISTANCE FROM OTHERS IN COMPARABLE
AREAS OF THEIR LIFE, AND ALSO REQUIRES ASSISTANCE TO VOTE, THE
UNFAIRNESS RESULTS FROM THE DISABILITY AND NOT FROM THE ELECTION. WOULD
YOU SAY THAT IF A PERSON NEEDS ASSISTANCE TO EAT BREAKFAST, THAT THE
BREAKFAST IS UNFAIR?

FOR MORE DISCUSSION OF VOTERS WITH DISABILTIES, SEE
http://www.wheresthepaper.org/Evoting_BadForDemocracy.htm#VotersWithDisabilities
http://www.wheresthepaper.org/links.html#dis

10. A report of the Caltech-MIT Voting project concluded that the
presence of paper trails actually decreases public confidence in the
voting system[34]. This can be understood as follows: would requiring
airplane passengers to inspect the plane's engines before boarding
enhance their belief in the safety of the aircraft?

WOULD REQUIRING AIRPLANE PASSENGERS TO WEAR SEAT BELTS ENHANCE THEIR
BELIEF IN THE SAFETY OF THE AIRCRAFT?

OUR FIRST REQUIREMENT HAS TO BE THAT ELECTIONS ARE LEGITIMATE, AND FOR
ELECTRONIC VOTING SYSTEMS THIS MEANS CREATING VOTER-VERIFIED PAPER
BALLOTS AND USING THEM TO AUDIT THE ELECTION.

            My position on paper trails, despite their problems, is not
an extreme one. If a manufacturer produced a reliable paper trail
device and the remainder of his system were acceptable, I would see no
problem in certifying such a machine. I am firmly opposed to any audit
trail requirement, however, and even where audit trails are used, the
paper record should never govern over the electronic one because it is
vastly less secure. The proper use of audit trails is as evidence. If
the paper trail totals differ from the electronic ones, that is the
starting point for investigation, not the end of the issue..

ALL TRANSACTION-CAPTURING AND TRANSACTION-PROCESSING SYSTEMS MUST BE
AUDITED TO ENSURE ACCURACY. THIS IS A PROFESSIONAL IT STANDARD.

ALL ELECTION PROCEDURES MUST BE APPROPRIATELY OBSERVED. THIS IS A
REQUIREMENT FOR ELECTION LEGITIMACY IN A DEMOCRACY. THIS CONCEPT MAY BE
DIFFICULT TO UNDERSTAND FOR SOMEONE WHO THINKS THAT "the republic will
survive if a president is elected who was not entitled to the office."
(SECTION 1.1)

 

Ron Crane wrote:

> At Alan's request, I am writing a paper rebutting Prof. Shamos's
> recent paper that promotes DREs. I've got most of it in hand, but I
> need some help on three issues. Two of them should be easy for someone
> who knows, and one will require some work. I need all these by
> Thursday if possible. Here they are:
>
> 1. A few reliable cites for the proposition that flight-control
> software is subject to rigorous specification, implementation, review,
> and testing procedures.
>
> 2. A few reliable cites for the proposition that software used for
> major financial transactions (e.g. a bank's check-clearing system) is
> subject to tight standards (though probably not as tight as (1)).
>
> 3. I need someone who is well versed in OVC's system to read section
> 2.4 of Shamos's paper
> (http://euro.ecom.cmu.edu/people/faculty/mshamos/paper.htm#_edn1 ) and
> to write a technically solid rebuttal. Reward: naming as a co-author.
>
> Thanks!
>
> -Ron
>
> _______________________________________________
> OVC discuss mailing lists
> Send requests to subscribe or unsubscribe to
> arthur@openvotingconsortium.org
>

_______________________________________________
OVC discuss mailing lists
Send requests to subscribe or unsubscribe to arthur@openvotingconsortium.org
==================================================================
= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
==================================================================
Received on Sat Apr 30 23:17:14 2005

This archive was generated by hypermail 2.1.8 : Sat Apr 30 2005 - 23:17:22 CDT