Re: Bar Codes

From: Karl Auerbach <karl_at_cavebear_dot_com>
Date: Sun Apr 01 2007 - 20:20:24 CDT

Doug Kellner wrote:

> I do not understand how bar codes on the voter verifiable paper audit
> trail can expedite the manual audit. The auditor would still need to
> verify that the bar code accurately reflects the voterís choice of
> candidates. Can anyone explain what I am overlooking?

I don't think you are overlooking anything but, rather, are hitting on a
point of debate.

As for myself, I do not feel comfortable wit any thing on the ballot
that can not be examined and verified by a person who has no physical
impairments. In other words, I don't like the bar code and would,
instead, prefer that the paper ballots be designed so that machines and
people read the same markings.

Now, in the general case optical reading is somewhat error prone - just
ask anyone who has used a system that tries to reduce a photograph or
fax of a document to a textual representation.

However, on a ballot we have great control over fonts, placements,
paper, and printing quality. And the vocabulary is very, very limited.
  These characteristics combine to make it possible to do a direct read
of a ballot and not have any alternative representation, such as a bar code.

The hardware/software to read bar codes is inexpensive and easy to
deploy; direct reading hardware and software is a rather more involved

My fear is that voters, who are already rather suspicious of electronic
voting systems, are going to look at the bar code and wonder "I wonder
what evil stuff is in there?" And, since one of the goals to which we
ought to be aspiring is the rebuilding of voter confidence, I have to
ask whether the potential gains of bar codes (which by my metric are not
all that large) are offset by the potential of loss of voter confidence.

As for bar code technology - voters are used to seeing simple bar codes
and they may even possess the tools to read the codes. However voters
are less likely to be familiar with 2-dimensional codes (as generated by
recent postal meters) which are required to hold the number of bits
needed to represent the choices on a large (e.g. California
gubernatorial with 134 candidates) contest or one that uses one of the
instant runoff mechanisms.

The concern that Alan mentioned, the ability to read-back a ballot to a
voter with impaired sight, is fairly easily and inexpensively done with
a bar code. It would cost more money to equip each voting place with
the gear to do a direct scan. But in large quantities (where large is
measured in perhaps thousands or tens of thousands) my guess is that the
difference in cost between a bar-code based read-back machine and a
direct-read read-back machine would be mainly the cost of the scanner -
perhaps a few tens of dollars, plus the cost of the software amortized
over the number of units it is deployed on. The per-election
configuration costs ought to be about the same. The failure rate of the
direct-read systems would probably be rather greater than that of a bar
code based system because of the added complexity of the optical scanner.


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 Mon Apr 30 23:17:06 2007

This archive was generated by hypermail 2.1.8 : Mon Apr 30 2007 - 23:17:16 CDT