Re: CA "Accessible Voter Verified Paper Audit Trail Systems" standards posted

From: Joseph Lorenzo Hall <joehall_at_gmail_dot_com>
Date: Tue Jun 15 2004 - 21:41:53 CDT

I don't have time to list the notes I've written on the document right
now... but will tomorrow morning. There are other weird parts to this
document... like the requirement that the AVVPAT be printed in both
the language of the voter and the language of the elections officials
(English). What happens if the voter's language is English?
Presumably, they would have to print a second language that was
randomly selected on the ballot (or risk the ballot privacy of all
non-English readers). In addition to the voter confusion this will
undoubtedly cause, I'm not sure this is the best approach...

For example, the requirement to print two languages on the ballot
stems from the fact that if there is only one voter in precinct that
reads, say, Laotian, then given access to the printed ballots (like in
a recount) it would be easy to tell what voter cast the ballot in
Laotian and therefore destroy the idea of a secret ballot. However,
given access to the ballots, doesn't it seem that you could print a
simple distribution of the second-language frequency and be able to
tell, at a minimum, what the most popular second-language was (the
randomly chosen ones should be random).

This would seem to necessitate that the machines should keep track of
all languages used in an election so that the end result, when you
graph a simple distribution of the second-language printed on ballots,
would appear random. Otherwise, the non-randomly chosen
second-languages will just be signal on top of the flat noise of the
randomly chosen ballots.

Am I completely off my rocker? I hope so... A simple solution is to
print all ballots in English only. Then have overlays (simple
transparencies) for each language needed tethered in groups to each
voting machine. Then the voter can pick up the stack of overlays,
find their language and place the overlays over the English ballot.
The overlay would obscure the English and replace it with their own
language... but not obscure their choices. Then, in the OVC system,
the voter would place the ballot in a privacy folder, and still have
the opportunity to verify the barcode.

This is all getting very complex... Ocam's razor be damned!


On Tue, 15 Jun 2004 18:56:27 -0700, Alan Dechert
<> wrote:
> There are some interesting issues with this document. As Joe Hall
> mentioned, it's not clear they apply to OVC at all. The first paragraph
> has, "These standards shall only apply to DRE systems for which an
> electronic record of the vote is created by the DRE and for which that
> electronic record is considered the official record."
> If the OVC system is understood as a ballot printer--not a DRE--then none of
> this would be applicable. On the other hand, why would they be so
> interested in knowing how we will meet these standards if they understood
> they weren't applicable?
> We will find out soon... should be interesting.
> Alan D.
> > It looks like the standards were written for the Avante system.
> >
> > I don't understand the rationale for not letting the voter touch the
> > paper ballot. Also, how do you ensure that the earlier paper audit
> > trails are not visible to the next voter.
> >
> > If the requirement for encasing the ballot so the voter does not
> > touch it were removed, then that would be better for the OVC approach
> > in California.
> >
> > Best regards,
> > Arthur
> >
> > At 4:43 PM -0700 6/15/04, Alan Dechert wrote:
> > >
> > >
> > >
> > >Press release about it
> >
> >
> pdf
> >
> >
> > --
> > --------------------------------------------------------------------------
> -----
> > Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA 94303-4507
> > tel +1(650)424-0202, fax +1(650)424-0424
> >

Joseph Lorenzo Hall
UC Berkeley, SIMS PhD Student
= The content of this message, with the exception of any external 
= quotations under fair use, are released to the Public Domain    
Received on Wed Jun 30 23:17:16 2004

This archive was generated by hypermail 2.1.8 : Wed Jun 30 2004 - 23:17:30 CDT