Re: A generic best practice document for NewMexicolegislators

From: Edward Cherlin <cherlin_at_pacbell_dot_net>
Date: Sun Jan 02 2005 - 23:08:27 CST

On Tuesday 28 December 2004 12:36 pm, charlie strauss wrote:
> Paper or plastic?
>
> Doug jones has reccomended that neither paper nor electroninc
> voteshave primacy in an dispute. Iinstead he reccomends that
> someone decides which is most likely to reconstruct the
> correct record as best as possible. I agree though it does not
> simplify the process.
>
> I'm now getting flack

Flak is what they shoot at you. You may be accused of *being* a
flack if they think you are just spouting the official line.
English is a weird language.

> for promoting this idea. Many people
> want the paper to be the sole deciding factor in a dispute.
> In which case why count the electroinc ones I ask. Well they
> still want them comapred, but the paper wins in a dispute.
> While this of course would give comfort in nearly all
> elections, since the paper and electroinc reocrds will closely
> match, it doesn't give much comfort in a dispute, and seems
> like it would leads to the electronic records being ignored.
>
> So i'd like to hear form both sides on the issue about why you
> think your position is so obvious.

The question is like asking whether original bank records or
auditor's reports should have precedence. Each is meaningless
without the other.

Paper cannot have simple precedence because paper votes can be
rigged by destroying ballots or adding in spurious ones
(consider "Landslide" Lyndon Johnson). The OVC system has
protections against both, alone or together. The protections
must have precedence over a simple physical count.

Electronic records cannot have simple precedence because they
even easier than paper to fake. Thus the OVC system has
voter-verified paper ballots, with security codes to match up
the paper and electronic records and prevent tampering.

What we can count reliably in an OVC system is the combination of
a paper ballot with a matching electronic record, within a
dataset having correct and identical checksums. In case of
malfeasance, either a less reliable method must be used, or a
new election called. An example of a less reliable method would
be a court judgment that one party was fiddling the vote,
presumably in its favor, and awarding the election to the other
candidate (in a two-candidate election, anyway).

> For example I woul dbet
> that Teresa Hommel will argue that paper is so much more
> transparent than either electronic ballots or some muky
> resoulution proces that this has to take precedence. Is this
> the reason? Or maybe the reason is that perhaps there are
> disputes that no amount of investigation would resolve?

Fiddling both paper and electronic records can create an
unauditable election, which would presumably need to be done
over under heavy outside monitoring while a criminal
investigation was carried on.

Election officials routinely bring audit data into court with
current DREs, optical scan machines, and other voting hardware.
This is not the voting data, but the machine error and status
logs. Judges have had no trouble understanding a report from a
log that a machine jammed at a certain time, was opened,
unjammed, and closed, and then normal processing resumed, each
at a specific time, that jammed ballots were put aside for
manual processing, with results recorded in due form and date
and time recorded, and so on. Particularly if the jammed ballots
and the log tape are produced in court for examination.

> Now there is another complicaiton if you dont pre-designate
> paper as the winner. And that is who gets to resolve any
> dispute?. Surely it cant be an election official since they
> are partisan, have their own agenda regarding election
> efficiency, and there is no adversarial process or
> transparency to their decision?

Courts, as has been customary in such cases for centuries.

> It seems like it should be a sitting judge, not up for
> election. That all political parties could file amicus
> positions. and the judge could consult experts including
> election officials. But in the US, traditionally judges are
> not investigator so should there be some sort of prosecutor to
> conduct the inquiry.

This is of course part of the current problem, since the election
officials and governors in certain key states are of the same
party as the disputed winners and the Administration in
Washington, and control their respective state and Federal
investigative apparatus.

No, political parties and citizens groups have standing to bring
suit under current law, and are doing so in some cases in Ohio
and elsewhere. That will continue to be the case. Then the court
can order further investigation that will not be controlled by
elected officials.

> Should their be a board and how should this board be formed.
> what about timing and delays.

No. Having one elected or appointed board watch another under
guidance from the same administration accomplishes little. We
need to bring in an independent branch of government that cannot
be ordered off the case for political expediency. (Collusion
between courts and executive remains possible, but when
everybody is in on the fix, no legal measures can prevent it.
The people still have to pay attention if they want to get any
results.)

> finally how big a discrepancy should trigger the inquiry?

Anything statistically significant, or anything that could change
the outcome. That's two separate standards, both of them
necessary. One of the fundamental principles of statistical
process control is that you investigate the cause of any
detectable discrepancies, so that they do not have a chance to
grow big enough to become production faults or emergencies. It
is not necessary to have the government do the analysis,
although it is certainly permissible. What is essential here is
that the raw data be made public for any interested organization
to examine as it thinks fit.

> Doug, do you have the answer?
> _______________________________________________
> OVC discuss mailing lists
> Send requests to subscribe or unsubscribe to
> arthur@openvotingconsortium.org

-- 
Edward Cherlin
Generalist & activist--Linux, languages, literacy and more
"A knot! Oh, do let me help to undo it!"
--Alice in Wonderland
http://cherlin.blogspot.com
_______________________________________________
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 Jan 7 22:28:55 2006

This archive was generated by hypermail 2.1.8 : Sat Jan 07 2006 - 22:28:59 CST