[Date Prev][Date Next] [Chronological] [Thread] [Top]

Fw: The meaning of auditing, recount, and observation of canvass

Thought you'd all be interested in progress of Calif. Internet Committee.
The more they learn the more scared they get it seems.  Ken, you should
really like this.
-----Original Message-----
From: jefferson@pa.dec.com <jefferson@pa.dec.com>
To: Charles, Alfie <acharles@ss.ca.gov>
Cc: Alfie Charles <acharles@sos_2.ss.ca.gov>; Ben Duchek
<bduchek@wam.umd.edu>; Bernard Soriano <bsoriano@sos_2.ss.ca.gov>; Bob
Yudkin <ryudkin@us.ibm.com>; Brian Gangler <BGangler@sos_2.ss.ca.gov>;
Cameron O'Rourke <corourke@us.oracle.com>; David Jefferson
<jefferson@pa.dec.com>; Dwight Beattie <dwightb@co.sacramento.ca.us>;
Jacquie Canfield <jcanfield@worldnet.att.net>; James Wayman
<jlwayman@aol.com>; Jim Adler <jim@soundcode.com>; Jim Cunneen
<jim.cunneen@asm.ca.gov>; John Mott-Smith <jmsmith@sos_2.ss.ca.gov>;
Jonathan Nagler <nagler@wizard.ucr.edu>; Kaye Caldwell <kaye@ix.netcom.com>;
Kim Alexander <kimalex@calvoter.org>; Larry Sokol <Larry.Sokol@SEN.CA.GOV>;
Linda O. Valenty <Lvalenty@email.sjsu.edu>; Mark Reynolds <mark@ilumin.com>;
Michael Alvarez <rma@crunch.caltech.edu>; Mikel Haas
<mhaasxrv@rov.co.san-diego.ca.us>; Pam Giarrizzo <PGiarriz@sos_2.ss.ca.gov>;
Pete Adlerberg <Pete@soundcode.com>; Peter Pursely <ppursely@earthlink.net>;
Peter Schmidt <peschmid@cisco.com>; Philip Muller <muller@poltech.com>;
Rodota, Joseph D <JoeR@AcmeSoft.com>; Roger Dao
<Roger.Dao@gsa.co.santa-clara.ca.us>; Rom Lopez <romulo.lopez@asm.ca.gov>;
Stacey Morgan <stacey.morgan@asm.ca.gov>; Steve Cunningham
<scunning@cisco.com>; Steve Knecht <skglobal@earthlink.net>; Sylvia Ahern
<Sylvia_Ahern@sterling.com>; Thad Howard <thad@lstreet.com>; Tim Draper
<tim@drapervc.com>; Tom Hill <thill@sos_2.ss.ca.gov>; Warren Slocum
Date: Wednesday, September 01, 1999 4:40 PM
Subject: The meaning of auditing, recount, and observation of canvass

>Folks, I have a problem that I don't think has been addressed in any of
>our committee meetings.  I think this issue straddles both committees,
>so I'd like to hear from the nontechnical committee as well.  Feel free
>to send to the whole mailing list, not just me.
>We all have an expectation that it should be a requirement for Internet
>voting that the process be auditable, and that there should be a procedure
>for recounts, and that the recount procedure should be able to be observed
>by interested parties in some way that is analogous to the same processes
>for paper ballots.
>The problem is:
> What do auditing and recount mean in the case of Internet ballots,
> and what does it mean to observe the process?
>Does "audit" mean that the auditors watch a county employee type Unix
>commands that re-run the authentication, separation, decryption, and
>programs that were alo run the first time?  What can be learned from that?
>Does it mean that a separately-written set of software that does the same
>job as the primary software should be run instead, and its results compared
>to canvass results achieved with the first software?
>Does it mean that the auditors are given the code to read so they can
>check for programming errors?
>And is there any point to a 10% sample in a recount or audit procedure?
>It is probably faster to rerun the entire process on the full 100% than
>it is to generate a truly random 10% sample and then erun the procedures
>on the sample.
>Should we require that the electronic ballots be printed or punched on
>paper so they can be counted by the same mechanical machinery as the paper
>ballots being audited or recounted?  How retro!
>This seems to me to be an example of how Internet voting simply doesn't
>resemble the paper voting process, at least it seems to me.  Auditing
>seems to be meaningless!
>On the other hand, how can we try to sell a process that we claim is not
>This is not just my opinion.  An old friend of mine, a computer scientest
>at Carnegie-Mellon University who is also a lawyer and an election
>certifier for Pennsylvania and Texas, has made similar points.  See
> http://www.cpsr.org/conferences/cfp93/shamos.html
>This is an article on e-voting written in 1993, before the Web; besides
>the discussion of auditing, it is also very interesting for its outspoken
>opinions on other issues of relevance to us.
>Comments please,