RE: Barcode Redux

From: Arthur Keller <arthur_at_kellers_dot_org>
Date: Fri Jun 11 2004 - 19:43:27 CDT

At 5:56 PM -0500 6/11/04, John Payson wrote:
> >>
>>If nothing is done to make sure that what the voters put in the ballot
>>box are actually valid ballots, how can election integrity be assured?
>>Suppose that when ballots are counted after 2,000 people are supposed
>>to have voted, there are found to be 995 valid ballots for candidate
>>#1, 990 for candidate #2, and 15 spoiled ballots. Does this mean that
>>some malefactor voided 15 ballots (perhaps votes for candidate #2)? Or
>>does it mean that 15 voters decided to be "cute" or stupid? Is there
>>any means by which anyone can ever tell?
>If 2,000 people voted, and 15 ballots were spoiled, there should have
>been 2,015 electronic votes. Remember, the voter roll also include a
>count of voters.
>I think you misunderstood. 2,000 ballots were cast, of which 15 seem to
>be spoiled. If they were spoiled by the voter before they were cast,
>then they shouldn't be counted. If they were spoiled by an election
>official's sleight-of-hand, then they should be counted (if it can be
>ascertained who they were for) and the election officials investigated.
>With an electronic ballot box that will refuse all spoiled ballots, the
>appearance of a spoiled ballot within the box would suggest that the
>ballot was tampered with after it was cast. But if no such validation
>is done when the ballot is cast, it's impossible to know who's to blame.

You need to understand how spoiled ballots work. If a voter votes a
second time, because the first ballot was wrong, then the first
ballot is spoiled. How do you distinguish between a voter-spoiled
ballot or an official-spoiled ballot? The check is to ensure that
the number of voters = the number of valid ballots, and the number of
cast ballots minus the number of spoiled ballots = the number of
valid ballots.

> >>
>The problem with the printer with a capture/reject assembly is it
>isn't accessible. There has been discussion earlier in this mailing
>list about the benefits and drawbacks of having ballots scanned and
>incrementally tabulated in real time (with results not displayed
>until close out). That's what's done with optical scan ballots in,
>for example, San Mateo County.
>Do you mean "handicapped-accessible"? If so, I fail to see what the
>problem is. If the system provides a voter with the option to have
>votes read back audibly after they are printed but before they're
>escrowed, then while a blind person would be unable to read the paper
>ballot to ensure that it matched what was read back, that would not
>interfere with the process of voting. Since the machine would have
>no way of knowing whether a particular ballot was being printed for
>someone who was in fact able to read it, the machine could not cheat
>without running a substantial risk of detection.

Are they read back from the paper trail, or from the internal record
of the DRE? If they are read back from the internal record of the
DRE, then we do not have a voter-verifiable accessible paper trail.

The DRE sure does know whether the voter can't read, since they voter
uses headphones. Although some voter using the hearing interface
might be able to read, few would do so because the interface is so
time consuming. Plus, the "error" need only happen in a small number
of ballots.

I'm wondering whether you are in consonance with the basic concepts 
of the OVC approach.
Best regards,
Arthur M. Keller, Ph.D., 3881 Corina Way, Palo Alto, CA  94303-4507
tel +1(650)424-0202, fax +1(650)424-0424
= 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 23:17:13 2004

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