Re: Audits

From: Kathy Dopp <kathy_dot_dopp_at_gmail_dot_com>
Date: Wed Jun 28 2006 - 16:14:15 CDT

> Date: Wed, 28 Jun 2006 11:53:01 -0500
> From: Jerry Lobdill <>
> Subject: Re: [OVC-discuss] OVC-discuss Digest, Vol 20, Issue 34
> To:
> Message-ID: <>
> Content-Type: text/plain; charset="us-ascii"
> Kathy,
> Thanks for your response and the references. I printed out and read
> both of them.
> I don't want to put too fine a point on it, but I think a few
> comments are in order.
> First, I didn't mean to recommend any particular law, only to give a
> possible scenario to illustrate a point about the importance of
> examining the implications of any law and the assumptions it implies.

Yes. I agree.

Second, I understand your objection to my credibility concerns, and I
> do agree that ensuring count accuracy is vital. Still, I think there
> is a credibility problem. Else how explain the dismal percentage of
> eligible voters who cast votes? I think that it will probably take
> more than the assurance of a committee of Ph. D.s in mathematics and
> statistics and election officials that the vote count will be
> accurate to convince people that their vote will count.

Well, that is why our audit proposal basically qualifies anyone who wants to
register as a poll worker (and who does not work as a poll worker for that
same election) to become an auditor.

Third, your references have illustrated how difficult it is to be
> sure the selection procedures don't bias the results. As always, the
> devil is in the details. :-)

Yes. We put some gaming people on the vote count audit advisory committee to
get their advice on random selection. Some audits in WA just put the
precinct names in a hat and let reps from two parties draw them out.

Your references prescribe that the vote count of a random selection
> of precincts will be audited. This assumes that all precincts have

No. It leaves that open. Machine or precinct counts may be audited.

equal probability of having discrepancies. It may well be that this
> is not a good assumption, especially if some precincts (or voting
> stations if precincts are being combined in the election) have many
> more votes cast than others (and thus present a more valuable target
> for tampering) or if certain precincts have known tendencies to have
> tampering while others do not. Also, certain races may have a higher
> probability of being tampered with than others, and the distribution
> of votes among precincts or voting stations for these races may be
> different from other races.

Our proposal calls for all races to be audited. I agree with everything you
say here.

Also, given that the central tabulation level presents a more
> valuable target for tampering than individual voting stations might
> it not be possible to tamper at that level? If that were done then
> auditing the voting station results might well not detect the
> tampering. (I looked at Tarrant County, TX Democratic Party primary
> vote data for the March, 2006 election in making these observations.)

You must not have read our proposal thoroughly or misunderstood it slightly
or perhaps it was not detailed enough because it calls for all ballot types
to be audited and for the audit committee to design the entire procedures,
not the election officials, so of course every aspect of the vote counts
will be audited.

I'm impressed with the extensive amount of thought that has gone into
> the generic law you referenced and the statistical analysis provided.
> I hadn't seen this material before.

Thanks. I am pretty sure that it is the best available proposal right now
for vote count audits, after reading several other laws and proposals and
finding their flaws before writing this one.

The topic still needs a lot of work and research like Alan says.


> Jerry Lobdill
> At 10:50 AM 6/27/2006, you wrote:
> > > Suppose the election is conducted under a law that
> > > prescribes a random hand count audit of a certain percentage (X%) of
> > > voting stations (DREs with printer/ballot box). The law says that if
> > > the random audit results are not within Y% of the electronically
> > > tabulated total results then a complete hand count of all votes is
> > > required.
> >
> >Jerry, This is not the law that I would recommend operating under. Here
> is
> >a better, more generic proposal that addresses your concerns:
> >
> >
> >
> >However, there needs to be a 1% minimum, that we may've forgotten to
> >include.
> >
> >Also, I would include, as Alan said to me, a requirement that the number
> of
> >discrepancies found, must indicate that the amount of vote miscount is
> less
> >than the margin, or expanded audit or full recount is required.
> >
> >
> > > indicated margin. As the actual margin decreases, an audit of X% of
> > > the voting stations has a decreasing confidence level
> >
> >
> >The confidence level is determined by the the number of machines (in your
> >scenario) that are considered to be affected by vote miscounts, 'not' the
> >margin. i.e. not the amount of total inaccuracy, but the number of
> minimum
> >machine counts that may be inaccurate. (I'm having trouble making myself
> >clear, obviously - but suffice it to say that it would "not" be
> necessary,
> >do-able, or wise, to base a formula on the official vote count margins as
> >you suggest doing.) If you study what we've written on this issue, I
> think
> >it will become clear to you.
> >
> >
> >We have provided a spreadsheet audit calculator whose URL is mentioned in
> >the footnotes to this mathematical explanation of randomly selected vote
> >count audits:
> >
> >
> >
> >
> >Another problem with the idea of random audits is that this is a
> > > makeshift solution to a credibility issue that requires an
> > > understanding of probability and statistics to feel good about. This
> > > does not satisfy a requirement for transparency very well. DREs and
> >
> >
> >I believe that the core problem is ensuring vote count accuracy, not
> >ensuring credibility.
> >
> >Also, if you take the time to read our proposal, you would see that it
> >provides plenty of transparency. Please take the time to study the two
> >papers above.
> >
> >central tabulation using electronic means as they currently exist
> > > constitute a voting system that is many times more vulnerable to
> > > tampering than the hand counted paper ballot system, and random
> >
> >
> >I disagree with this also, because both systems are extremely vulnerable.
> >The difference is that computerized counts extends the reach of one
> person
> >to tamper with more votes. The ideal system would have a 100% hand count
> >plus a 100% machine count and reconcile them. However, independent
> audits
> >are a close second.
> >
> >You have not been following all the evidence of vote tampering in the
> recent
> >2000 and 2004 elections, if you are unaware of all the paper ballot
> >tampering, paper ballot substitution, paper ballot destruction, and paper
> >ballot ignoring that goes on. One thing I've learned is that there are
> >many more ways to tamper with vote counts than there are voting systems
> and
> >the wisdom of the OVC system is its co-existance of both paper and
> >electronic technologies to prevent, detect, and correct more types of
> >potential tampering than any other system.
> >
> >
> >Kathy Dopp
> (In accordance with Title 17 U.S.C. Section 107, this material is
> distributed without profit to those who have expressed a prior
> interest in receiving the included information for research and
> educational purposes. ProgressiveNews2Use has no affiliation
> whatsoever with the originator of this article nor is
> ProgressiveNews2Use endorsed or sponsored by the originator.)
> "Go to Original" links are provided as a convenience to our readers
> and allow for verification of authenticity. However, as originating
> pages are often updated by their originating host sites, the versions
> posted on ProgressiveNews2Use may not match the versions our readers
> view when clicking the "Go to Original" links.
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL:
> ------------------------------
> _______________________________________________
> OVC-discuss mailing list
> End of OVC-discuss Digest, Vol 20, Issue 35
> *******************************************

Kathy Dopp
National Election Data Archive
Dedicated to Accurately Counting Elections
Subscribe to announcements by emailing
Please donate or volunteer.
"Enlighten the people generally, and tyranny and oppressions of body and
mind will vanish like evil spirits at the dawn of day," wrote Thomas
Jefferson in 1816

OVC-discuss mailing list

= The content of this message, with the exception of any external
= quotations under fair use, are released to the Public Domain
Received on Fri Jun 30 23:17:12 2006

This archive was generated by hypermail 2.1.8 : Fri Jun 30 2006 - 23:17:12 CDT